Exhibit R- (JF-9)

Exhibit R- (JF-9) -

Bellcore

SPECW REPORT SR-TSV-OM275

ISSUE 2, APRIL 1994.

. . .:. ...

BOC Notes’on the LEC Networks - 1994

EXHIBIT R- -(E-9) CASE NO. U-13796 D A P . C I n c 1 L

This document replaces SR-TSV-002275, BOC Notes on the LEC Nerworks - 1990. Issue 1 (Bellcore, March 1991).

For ordering i n f o d o n . see page IV. and the References section of this document.

This donumnt may not be reprodud without the express written permission of Bellcore, and any reprodudon without written authorization, is an infringemnt of Bellcom’s copyright

Copyright0 1994, Bellcore. Au rights reserved.

EXHIBIT R- _(JF-9) CASE NO. U-13796

.* BOC Notea on the LEC Nawork-1994 Nshwlc Design and Conflguratlon

A - b B calk mrou(ed v(a V attempt VB but do not over- fiaw mrwgh me tandem

A B Rwte advanca sequence of VB In orlglneting bafsc table: VB, WE

Route advenca sequence of VB in lnmmlng tafk table: VB

Figure 4-1 8. Restricting DRT to Two-link Paths

4.6 Reliability of Equipment and Systems Local exchange telecommunication networks providiag services such as Plain Old Telephone Service (POTS), Integrated Services Digital Network (ISDN) capabiities, or future telecommunications services must be. highly dependable. The customer has come to expect a high level of network dependability based on the performance levels that have been experienced for POTS, and on the levels that are continuing to be observed for new services such as ISDN.

As new technology, functionality, and services are integrated into the network, objectives have been set that, at a minimum, preserve the expected levels of dependability for these existing services.

A primary characteristic of local exchange network dependability is its avaitability. Availability is strictly defined as the abiity of an item to be in a stak to perform a required function at a given instant of time, or at a desired instant of time within a given time interval. This assumes that the external resources, if required, are provided.

4-43 EXHIBIT R- -(JF-9) CASE NO. U-13796 PAGE 3 OF 16

However, for the inmLATA networks, availability is g e n e d y interpreted as the long- term hct ion of time that the network performs its function as intended (for example, when the network successfXly provides a communications path fiom one customer to another).

There are four major factors influencing the availability of the hhaLATA network as seen by the customer:

Network topology . Equipment architectures

Equipment reliability

Telephone company maintenance practices.

Each of these factors must be considered during equipment and system design, as well as during network planning and design, to ensure that an acceptable level of service dependability is provided to customers.

To ensure adequate equipment and systems reliability, the service depoldability manugemenr process is followed. This process includes the following:

Establishment of service-level dependability objectives . Development of reference network architectures . Allocation of the dependability objectives to various (facility, interoffice, etc.)

Baseline service-level objectives on the availability of a local exchange network have been developed for network engineering purposes and are presented in this section. These end-to-end network objectives are derived from long-estabhhed network- availability objectives, observed network-availabiity performance, and projected availability needs of the network. A reference network architecture is presented and docations to network segments and network elements are given.

These resulting availability objectives for segments, equipment, and systems are primarily used during network architecture studies, product conception engineering studies, design reliability evaluations, and customer reliabiity analyses where reliability modeling techniques are used to estimate performance for comparison to objectives. In addition, they are also used as baseline performance objectives for comparison to actual field performance.

network equipment and systems.

4.6.1 Local Exchange Network Hypothetical Reference Connection The service dependability management process requires a reference network mhitecture, or local exchange network Hypothetical Reference COMeCtiOXl (HRC). The network architecture used as the local exchange network HRC is shown in Figure 4-19.

444 EXHIBIT R- -(JF-9) CASE NO. U-13796 PAGE 4 OF 16

Figure 4-19. Local Exchange Network Hypothetical Reference Connection

Figure 4-19 is a simple model of the exking in!d.ATA POTS network, including the major network segments needed to comect two telephone submibers served by different central office switches. There an: four network segments used in the HRC

Dism'bution - consisting of two distribution network segments related to each subscriber (customer premises equipmen& which is not part of the telephone company network, is not considered here) Switch -the two central office switches

c. -Fac i& Entrance - a f a c i t i t y ~ n t r a n c e P e t w o r k s e g m e n t i n c l ~ ~ ~ ~ u i ~ n ~ ~ c h \ as analog-todigital converters (channel banks), multiplexers, automated digital \ - \terminals, etc.

Interofice - the interoffice transmission facility segment used to transmit calls from - - one central office switch to the other.

4.6.2 End-to-End Network Availability Objective Since it is important from the customer's view that there be a high probability of obtaining a path through the network, a high level of end-bend network availability is desirable. This section addresses only the availability of a path through the network due to failms of the equipment and not the effects of blwking due to traffic congestion. AU segments in the network are indepndent and each network segment contributes directly to the unavailability or downtime of any path. The hypothetical reference circuit shown in Figure 4-20 has a derived end-tmnd availability of 99.93 percent This is approximately 365 minutes per year or one minute per day of unavailability.

. .. .

4-45 EXHIBIT R- -(JF-9) CASENO. U-13796 _. Cr -7 I /

/ 0.01

U

0.01 \

t€- W.W% -j Figure 4-20. End-to-End Network Availability Objective

4.6.3 Network Segment Availability Objectives

Switching and transmission services and equipment availability objectives are traditionally stated in terms of line or channel availability. Each contributes directly to the end-t-nd network path availability. Therefore, these parameters indicate the ability of the equipment in the network segment to perform its function (for example, distribution, switching, facility e n m c e or interoffice transmission) when needed for a customer’s call. Bellcore has investigated the availability of the equipment in each of the network segments shown, and end-bend objectives have been proposed to help e n m the dependability of the services provided over these networks. Availability is affected by many factors, such as equipment failures (for example, resulting from hardware reliabiity failures, software bugs, or procedural enurs). equipment and system architectures, maintainability and repair strategies and uncontrollable factors such as cable cutddig-ups, storms, etc. Each factor must be considered in designing and maintaining dependable equipment and a dependable network. The network segment objectives discussed an based on longstanding performance objectives, on observed field performance, and a summation of the objectives for the equipment in the segment. Each network segment is discussed in detail in the following m p h s .

4.6.3.1 Distribution Network Segment

The distribution network segment, as used here. includes both the feeder and the loop from the switch to the customer’s bondoffice, excluding customer uremises equipment Two distribution segments, one associated with each subscriber, are included in the local 8 4-46

EXHIBIT R- -(E-9) CASENO. U-13796 PAGE 6 OF 16

a - exchange network HRC. In the dismbution network, availability is essential from the telephone customers’ view since they have come to expect access to the network v h i d l y upon demand. The objective on the level of the distribution segment availability that is suggested for engineering existing and new services is 99.99 percent (see TR-NWT- OOO499, Tramport Systems Generic Requirements (TSGR): Common Requirements).’ This would give a maximum unavailability objective of 0.01 percent, which equates to a maximum downtime objective of approximately 53 minutes per year per customer line Figure 4-21). Availability objectives are sometimes expressed in terms of the complement of availability, unavailability, or downtime. Each is discussed in the remainder of this section, which is devoted to the switch netwoxk segment.

4.6.3.2 Circuit Switch The switching systems in the HRC perform interconnection of subscriber lines to uunks to form the end-to-end kmsmission path. The best method known is the total switch downtime objective of 3 minutes per year, which is described in TR-TSY-000512, Reliability, Section 12 (a module of LSSGR).6 In addition, objectives also exist for individual lines (28 minutes per year) and for individual hunks (also 28 minutes per year). For a through transmission path transversing the switch from a line to a trunk, the maximum unavailability objective is the sum of the line and trunk unavailabiities minus 3 minutes per year, since the total outage objective is included in both the line and trunk objectives and need only be considered once on a through path. This results in a through-path objective (line to mmk) of 53 minutes per year or 0.01 percent. Switch unavailability is shown in Figure 4-22.

Additional dependability objectives have been developed for circuit switches in the following areas and are contained in the LSSGR.’

0 Cutoff calls

Ineffective machine attempts

Failure rates

Servicelife.

4.6.3.3 ISDN Switch Reliability objectives for ISDN switching systems have been developed to help ensure that the dependability of ISDN services is similar to that of POTS. ISDN introduces the concept of multiple senices (such as circuit-switched voice and packet-switched data) integrated onto the same physical interface. Thus, the definition of a line failure becomes more difficult. Therefore, parameters that address the numerous partial line outage modes have been developed in addition to objectives of the LSSGR. For Basic Rate Access (BRA), reliability objectives have been developed to be consistent with Figure 4-22 and the USGR. Following are examples of new availability (downtime) parameters:

Figure 4-21. Distribution Network Segment Unavailability Objective

Une +Trunk - 3 minutes per year = Thrwgh Pam (53 minules per year)

Une TNnk 28 minutes per year 28 minutes per yea^

Figure 4-22. Switch Network Segment Unavailability Objective (Through Path) a

EXHIBIT R- -(JF-9) CASE NO. U-13796 " A P.C Q nr: 1 L

BOCNokaontheLECNolwnrlm-1994 Network Design and CarRguratlon

//-

/‘

I 01

I L

.- a Total B-channel cir& mode downtime

Total B-channel packet mode downtime

D-channel packet data downtime

Total ISDN circuit switching capability downtime

rn Total ISDN packet switching capability downtime.

Similar parameters have been developed for Primary Rate Access PRA). The objectives can be found in l’R-NWT-COlC47, ISDN Switching System Reliability Objectives for Basic Rate Access, and its supplement for PRA (ISDN Switching Syqtem Reliability Objectives Supplement for Primary R a e A c c e s s $ - - - - - .

.-.-.. \ - .\ 4.6.3.4 Facility-Entrance Network Segment

tc-analog conversion, framing, digital channel cross-connection, multiplexing,

L.~., \. For purposes of reliability modeling, the facilify-entrance network segment is d e M as’\., the portion of the network that performs functions (such as analog-tdgital and digital-

demultiplexing, etc.) before channels are put onto the interoffice transmission facilities to be sent to the receiving-end centid office. At the receiving-end central office, it is assumed that the facllity-entrance equipment is replicated to perform reverse functions such as demultiplexing. Figure 4-23 illustrates a representative configuration of typical digital terminal equipment contained in the facility-entrance network. The facility- entrance segment & allccated 0.005 percent unavailability a! each end or a total of 0.01 percent for both. Digital terminals in the facility+ntrance network typically have

Asynchronous Digital Multiplexes Requirements and Objectives,’ and TR-NWT-ooo418, Generic Reliability Assurance Requirements for Fiber Optic Transport System,’ for more information. -----------

‘\~

\.\ I ’!

. ’;

.

unavailability objectives of about 5 minutes per year. See TR-TSY-000009, .!

. /

------._ /” .--

4.6.3.5 Interoffice Network Segment The origin of the availability objective for the BOC‘s interofice transmission network is the former “Short-Had Availability Objective,” which has its origin in early microwave radio applications.

The suggested short-haul, 2-way transmission, minimum availability objective for a BOC transmission channel is 99.98 percent (0.9998 availability) at 250 miles. A full statement of the numeric values of the objective is a plot of unavailability linearly prorated by route length. However, only the maximum objective will be used here. Figure 4-24 is a model showing the generic application of the short-haul availability objective to a transmission system in the interoffice transmission segment of the network. The short-haul availability objective has been widely applied to BOC transmission systems in the past (such as fiber-optic system^)?^

4 4 9 EXHIBIT R- -(E-9) CASE NO. U-13796 PAGE 9 OF 16

or 0.01% for both ends combined

Figure 4-23. Facility-Entrance Network Segment Availability

Lhte Terminating Equipment

4 - 0.02% I Figure 4-24. Interoffice Network Segment Unavailability Objective

EXHIBIT R- -(JF-9) CASE NO. U-13796 PAGE 10 OF 16

BOC Notes on UIO LEC Networb -19% Network Design and Configuration

4.6.4 CCS Network Rellability and Unavailability (Downtime) Objectives The integration of out-of-band signaling with intraLATA networks, via Common Channel Signaling (CCS) networks, is an important factor in overall network reliabiity. This is because failures in signaling will prevent completion of calls and will be. viewed by customers as failures of the networlc Following is a brief discussion of CCS network unavailability (downtime) objectives and four potential alternatives to achieve software diversity in the CCS network.

Unavailability (downtime) objectives are intended to control the amount of time a CCS network (or a segment thereof) is unable to perform its rquired signaling functions. They can be represented by a single number equal to the long-term percentage of time. a CCS network, or segments thereof, are expected to be "down." As such, downtime. objectives can significantly influence end-user perception of seMce quality. The expected percentage of downtime for a network element can be interpreted either as

The average downtime over many years for this network element, or as

The average downtime over one year for a population of the network elements.

Accordin to ANSI T1.111, American Narionnl Stmtdard for Telecommunicaiions - SS7 - M V , Section 5.1.2, the Message Transfer Part (MTP) downtime objective for the CCS basic mesh network shown in Figure 4-25 corresponds to (an average of)* no more than 10 minutes downtime per year for the signaling paths between two Signaling End Points (SEPs)** and is broken down as follows:

1%

Each user interface segment should be down (an average of) no more than 3 minutes per Year, Each network access segment should be down (an average of) no more than 2 minutes per year, and The backbone network segment should be down a negligible amount of time (that is, close to 0 minutes downtime per year). Note that downtime for this segment includes failures that prevent use of the backbone segment but do not by themselves disable any other segment(s).

The above allocation assumes an ANSI-based (Recommendation T1.1115 in ANSI T1.111-1988, Seaion 7.2.1)'' reference architecture with two-way diversity for the A- link sets and three-way diversity for the B-ID-link sets. If three-way diversity is not

* ?be original text in the Ameriurn N d o d Srandardfor T e l e c d a r i o n r - SS7 - MTPlo refers to "nominal @mcnts' that have bccn in- in ~ - ~ - W O 2 4 6 . Bell Cavnunicrrrionr Rrrcarch Specifidon of Signaling S y s m N d c r 7," as "avcrage downtime numbers"; thus thc text in the parentheses repmenfs Bellcore's interpretation of this ANSI downtime objective.

** Examples of SEPs are Service Control Points (SCPs) and switches.

4-51 EXHIBIT R- -(JF-9) CASE NO. U-13796

achievable in the backbone segment, the downtime of that segment may no longer be negligible. Hence, the IO-minute end-bend objective may no longer be achievable.

A backbone network segment failure may cause the switches on each Signaling Transfer Point (STP) pair to lose communication with the switches on the remote STP pair, but the switches can stiU communicate with the other switches homed on the same STP pair. It occurs when

The entire B-/D-link set quad fails,

One of the STPs in either mated pair and the B-Dlink set pair of the other STP in the

A-link sets to one of the STPs in either mated pair and its C-link set and the B-Dlink

mated pair fail, or

set pair of the other STP in the mated pair fail.

When the SEps in Figure 4-25 are both CCS Switching Offices (CCSSOs),* the 10- minute end-to-end downtime. objective and the above allocation to network segments correspond to a single trunk p u p with its terminating CCSSOs interconnected using the ANSI-based reference architecture (see Egure 4-26).

A CCSSO is a switch equipped with tbc ISDN Uscr Part (ISUP) of SS7 for eall schlp.


Figure 4-26. Example Case: One Trunk Group with Its Terminating CCSSOs

It also applies when instead of CCSSOs there are Service Switching Points (SSPs)** or SCPS.

Detailed information on CCS network element reliability objectives can be found in TR- NFvT-oMx)82, Signaling TrMsfer Point (XU') Generic Requirements;I2 in TR-NWT- oooO29, Service Control Point Node Generic Requirements f o r 1 ~ 1 ; ' ~ in TR-NWT- 000533, Service Switching Points, FSD 31-01-0000,'4 pius various references!

** An SSP is a switch equippi to halt call progress launch an SS7 puy to obtain additional infomation from an SCP, and mute or tnat the call based on the information d v e d in the S B ' s response. SSPs can be Eod Of6ccs W) or tandem switches. SSPs intaact with databases to provide saviccs and muting.


12.3.3.2 Residence

Total Length Working Length Total Bridged-Tap

Table 12-3 contains summary statistics of lengths of sampled residence working pairs. sampled redence pairs have an average total length of 13.190 ft and an average working lengtb of 11,723 f?. The average bridged-tap length is 1,490 ft. Figures 12-9 through 12-1 1 present cumulative distribntion plots for these statistics.

495 114,838 13,190 245 186 114,103 11,723 236 0 18,374 1,490 44

Table 12-3. 1983 Loop Survey- Residence Length Statistics

him i m n SDM a n on Minirmm Madmum 18374n

Figure 12-9. Bridged-Tap Length Distribution Residence Loops


1983 BOC Lwp Survey

loo00 15ooo 20000 25ooo 3xxn 35ooo

BddeeaTap (8)

Figure 12-1 0. Working-Length Distribution Residence Loops

oh I I I I I 1

4ooo 5ooo 6ow 0 lMxl 2ooo 3wo

Figure 12-1 I. Total Length Distribution Residence Loops

12-15 EXHIBIT R- -(F-9) CASE NO. U- 13796

12.3.3.3 Business

Table 12-4 contains summary statistics of lengths of sampled business working pairs. Business pairs have an average total length of 9,840 ft, an average working length of 8.816 ft, and an average bridged-tap length of 894 ft. Figures 12-12 through 12-14 present Nmula&e distriition plots for these statistics. The average working-loop lengtb for a business service is about 3CLpercent shorter than the average working-loop length for a residence senice.

Total Length 250 , . 100,613 Working Length ux) 99,569 Total Bridged-Tap 0 11,333

Table 124. 1983 Loop Survey- Business Length Statistics

Minimum Maximum Mean SDM ft ft ft ft

9,840 302 8,816 296 894 41

u 1'

0 5000 1 m 15ooo 2oooo 25000 3oMM 3HxxI

TotalLength (11)

Figure 12-12. Bridged-Tap Length Distribution Business ~OOpS


Exhibit R- - (JF-10)

. ... .. ..

Shing Yin & Muayyad Al-Chalabi, RHK Advisory Services, and Telecom Economics July 16, 2003

t Introduction In efforts to differenliate themselves, service providers have issued a barrage of confusing messages to business cuslomers.

Service provider characteristics must be related to customers’ key buying factors, as customer satisfaction is the ultimate goal.

t Surveying the IXC Landscape Evaluated along a set of common metriw, lour lXCs are found to each have their unique strengths and weaknesses. Among the group, AT8T is found to be the strongest overall.

t The Role of RBOCs RBOCs have deep presence in their traditional territories, but their limited reach elsewhere is a barrier to growing in the large- enterprise segment.

t Recommendations There are many potential aspects of differentiation, and no service provider needs to lead in all of them. Plenty of room for improvement exists for all players, lXCs and RBOCs alike.

Discussion 01 nehvork metrics and their relation to setvice provider characteristics.

t What Really Matters?

t Appendix

Executive Summary

in today’s intensely competitive market, service providers are turning to differentiation strategies to push back against price pressures. Their efforts have been met w i th mixed results, as customers struggle to sort through a barrage of messages. Differentiation is meaningful.only if i t impacts customer satisfaction. This report discusses an extensive set of network metrics and haw they relate to key buying factors, focusing on the business market. Evaluated along these metrics, the differences among several large interexchange carriers become evident and significant: AJ&T i s found to b e the best positioned overall, thanks to i t s large network, efficient operations, and demonstrated commitment to continued investment. Other service providers have their unique strengths as well, and plenty of room for improvement remains for all.

Introduction

Three years into the North American telecom crash, today’s post-bubble market continues to be characterized by reduced spending across the value chain: customers curb IT budgets, service providers t r im capital and operating expenditures, and vendors cut back on new product development. Meanwhile, the degree of competition-particularly a t the service provider level-has never been more intense. Rather than lead t o true consolidation, the downturn has instead produced dozens of restructured companies who linger i n the market.’ In addition, new regulatory rules allow service providers to invade each other’s traditional market, effectively increasing competition.

Slow top-line revenue growth exacerbates the situation for al l (see Figure 1). In this stagnant market, service providers can grow only through market share gains-a ta l l order given the competitive environment. With l i t t le differentiation i n service offerings, price competit ion has been the primary strategy so far. Several firms have tr ied to urge a “flight to quality” in l ight of their competitors’ relative financial weakness. Unfortunately, the window of opportunity for any such f l ight i s l imited, since balance sheets and income statements are being systematically cleansed via the bankruptcy process.’

I Companies that have restructured via bankruptcy include: MCI (formerly WorldCom), Global Crossing, XO Communications, 360networks, Wiltel (formerly Williams), McLeod USA, Covad, and a host of others.

asset value-leading to dramatically reduced interest and depreciation expenses. The most prominent example of this i s MCI, whose bankruptcy has wiped out 536 billion in debt and $80 billion in

0 2003 RHK Inc. A l l rights reserved worldwide. 1

EXHIBIT R- -(F-10) CASE NO. U-13796 ”A,.,? 1 /,v<

SeM’ce Provider Metrics Key to Tiue Differentiation July 16, 2003

Figure 1: North American wireline telecom revenues, 2002-2007

$300

$250

2 $200 2003-07 CAGR:

0 .- $150

m $100

$50

2002 2003 2004 2005 2006 2007

Source: RHK Inc

Purely price-based competitive behavior leads prices to fall rapidly-as one example, RHK’s research, presented in earlier Telecom Economics reports, has shown that average revenue per bit for Internet backbone services declined by more than 75 percent over the last two years. Continued price erosion of such magnitude i s unsustainable, as cost declines clearly cannot keep up; To reverse the trend, service providers are now increasingly emphasizing differentiation, persuading customers to see value in the network infrastructure, technologies, and service organizations that they have invested so much to build.

A new era of differentiation The differentiation strategy has been tried before, but it can actually work this time. After three years of curtailing spending, customers are now beginning to take a serious look a t new services and applications. Virtual private networking and security products are receiving serious interest, while emerging Web services and network convergence trends demand even more innovative offerings. As Michael Capellas, chief executive of MCI, remarked in his keynote speech at this year’s CeBIT trade show, “We have had low innovation over the last three years. The game has changed.”’

Capitalizing on this opportunity wil l be a true test of fitness for service providers, requiring leadership in multiple aspects. Customers know the challenges are formidable and that they need to be selective-so service providers are stressing differentiation to win them. This report provides a framework and facts for sorting through their claims, and evaluates four major lXCs using a set of network metrics that RHK believes represents true differentiation.

’ “Reasons to be cheerful - Keynote Speaker - Michael Capellas. Not al l doom and gloom in telecoms,” Financial Times, June 18, 2003.

0 2003 RHK lnc. All rights reserved worldwide. 2

EXHIBIT R- -(JF-lO) CASE NO. U-13796

Sem'ce Provider Metrics Key to True Differentiation July 16, 2003

Confusing messages . In their publicity campaigns targeted at businesses, service providers understandably emphasize characteristics in which they have a seeming advantage. Thus, one service provider will talk of network security, another plays up customer service, and yet another emphasizes geographic reach. Often, two or more firms make conflicting statements on the same metric. For example, one may cite points of presence (POPS), while another refers to capacity metrics, as evidence of network size. Even worse, one may claim to have thousands of POPs, and another claims to have hundreds: yet they each claim to have more than the other. With the volume of messages streaming from service providers today (see Figure 2) , it i s not surprising that customers are confused.

Figure 2: Confusing messages from service providers . . ~ .--... .AT&T's n e l w r k inclUdes a local

sewice infraslruclure -. and unparalleled '~ 'We have he sbnngesf .,,

local lelemm""icatio"s nehwrk in he nation

hanks 10 p a n Of backbone and whally- \ locusedinvesbrrant'

p l C I has) Ihe indushy's most

expansive global IP

, Owned dala nerwrks"

- Verimn .-. -

' S p i n ? ~ ;ndusy leading ".. , service Level

Agreemenls ... guaranlee

---__ . MCI

unmalched Seamless reach. connecting key

business cenwes in 220 , counlrres and tern'lones*

' - Equant Source: RHK inc.

Sorting through and interpreting the messages is critical for all stakeholders, however:

. Customers need to know which service providers are best suited to meet their current and future needs. Vendors need to know which service providers will make the best strategic partners, and what specific solutions they can use. Service providers need to understand where they stand relative to their competitors, so they can tailor marketing efforts.

What Really Matters?

The ultimate goal of any service provider, and the key to success, is to serve i t s customers as well as possible. Therefore, a fair assessment of service providers necessarily begins with a consideration of customers' key buying factors. For customers in the business market (the focus of this report) these can be boiled down to a small group of high-level criteria, as listed in Table 1. RHK believes these criteria, with

3 0 2003 RHK Inc. All rights reserved worldwide.

EXHIBIT R- -(JF-IO) CASE NO. U-13796 PAGF 1 OF 6

SeM'ce Provider Metrics Key to True Differentiation July 16, 2003

appropriate weightings applied, can represent most Customers' current and future telecom needs.

Table 1: Key buying factors for business customers of telecom services

Key buying factor Large enterprises Small and medium enterprises

connectivity Importance; High Imporlance: l o w - The need lo inlerconnect customer - Mulliple locations of different Sizes in - SMEs typically have only a few sites

located in the same regional area: connectivity to outside world usually achieved via ubiquitous public Internet.

sites al geographically diverse locations. either sith dedicated point- to-point links or via a shared network infrastructure.

diverse locations, often international; also require connectivity lo suppliers, partners. customers. elc. via secure extranet.

Security Importance: High Importance: Medium - Protection of a customer's network . High volumes of sensitive daia are - Lower volume of sensitive data

constantly generated and transmitted by critical applications. e.g., financial iransactions. enterprise resource planning, and Web sewices

assets and data against hacking and loss; also inciudes recovery foliowing a disastrous event.

transmission across fifferent locations; Web sites often outsourced.

Reliability . Assurance that the network will Operate

with minimal disruption and downtime: measured in terms of network availability and performance.

Flexibility - Ability to accommodate muitiple

prolocols and applicalions. and to provide custom soiuiions and different service levels (from fully managed to do-it-yourself) according lo customers' unique needs.

Cost - Low total cost of ownershio fTCOI

Imporlance: High . Downtime causes expensive productivity

loss; network is oflen critical to supply chain and other operations.

Importance: High . High value placed on sewice providers'

ability to handle different prolocois, offer bandwidth on demand, and create cusiom integrated networking solutions.

Importance: Medium - Due to the criticalitv of the network

Importance: High - Downtime causes expensive produclivity

loss.

Importance: low . Use of generic applications and few

protocols means only standard network sewices are reouired.

Importance: High - Because onlv standard sewices are . . . ~~ ~

nc La ng costs c l senoes egJ pnent. ma.nlendiice. acm n stial3n. and aoo caions

c.siomers &a y nave larse &ageis an0 :noose senice pro. uers oaseo on qdal ty and caDac.lct es o w once Neienne ess.

reqL.reo. SMEs pace ntgn 'mponance on pnce. ohen nleresled n coS!.sav ng Y C ce caia. ana &'re ess sew ce bm ies

Service - Quality and availabiiib of customer

support staff; ability to quickly resolve customer issues; response lime when lurning onloff new services; and integration of support tools. sales and semice, billing, etc. across product lines.

Stabilitylcommitment

large customers expect lower unit costs due to their high volumes.

Importance; Medium Importance: High . Many large corporations have their own

inlernai telecom support slaffwho can architect and manage their own networks. However, a high level Of Service (fast provisioning, migration support) is crilical when winning new customers.

. SMEs typically have a nm-specialized IT support staff. and thus rely on the Sewice provider to offer the right solutions and resolve problems.

Importance: High Imporlance: Medium Ass.rancB that me sefi ce probcer hi

coniinLe 10 operaie II 're c..stomer s nterests. I e IS 5nancm are sold nieslmenls cant nie to be mace ana 03s nesses h t ncl3e ex led

. 7ne cosls of chang ng senice p~ok~cers are ver, n gn n aoo.L,on to tne a rect COSI of m grat ng sen cos. me ,na iecl cost 01 nlerrLpteo c.s.ness processes s ohen rnmeasuab e Van? cJslomeiS lee o.mea by pie" ods ;rovoers that na\e e x lec prcoJcI mes or shil oonn doceincr

- NO c.sl0mer I kes lo have h e 1 Sew ce nlen.pteu nowever me costs of m 9raton tcr SMEs are ohen sma er an0 more cearaole

%.ice Rnd rc

4 Q 2003 RHK Inc. All rights reserved worldwide.

EXHIBIT R- -(=-lo) CASE NO. U-13796

Service Provider Metria Key to True Differentiation July 16, 2003

Both the large enterprise and small and medium enterprise (SME) segments are significant in size. However, as noted in the table above, differences between these segments affect the emphasis they place on specific factors, and thus influence how they view and evaluate network service offerings. In general, large enterprises have higher demands on the quality and capabilities of the network, and often require customized integrated solutions for their diverse needs. On the other hand, SMEs are more likely to be interested in service bundles that offer simplicity and lower cost. Many service providers actively target one or the other of these segments, depending on their particular strengths and weaknesses.

For a discussion Of the increasing

business and consumer applications, see:

"The Coming Era o f Absolute Availability," RHK Advisory Services, White Paper, May 2003.

importance of network reliability to

Service provider characteristics The ability to satisfy a customer's key buying factors depends on many individual characteristics of a service provider, which collectively differentiate it from i t s competitors. Some of these can be unambiguously quantified and directly compared: others require a more qualitative evaluation. Tables 2 and 3 l is t these characteristics and explain which key buying factors each one addresses. These are the dimensions along which service providers should be evaluated in order to determine their relative position and prospects in the market.

Table 2: Qualitative selvice provider characteristics and relation to customer buying factors Service provider

characteristic Description (why important] Buying factors Impacted (how)

Nstwork integration - The degree to which multiple networks and systems (OSS) are integrated on a common platform. for improved operalional elficiency, service, and support.

- Network Operations - The sophistication of twls used in the network operations center: also the degree of integration and automation.

improve network operations or enable new services,

- Technologylinnovation - The use of advanced technologies lhat

Breadth of services - The completeness of services offered by !he service provider, including basic and value- added services. - Partnerships - Other companies with whom the service provider partners to provide a more complete and convenient solution to the customer.

- Customer relationships - The strength of existing customer relationships. measured by mutual understanding and INsI.

Source: RHK Inc.

. Cosf: More integrated networks typically run more elficienlly.

- Reiiabilily: Integrated networks alsn lend to be mole reliable. since they do not rely on multiple parallel operating systems.

- Reiiabilily: Automated network operalions eliminate human errors. a major source of network failure.

- Reliabiiify: Technologies like MPLS enable quality oi service in iP networks; others enhance network availability.

- Secunly: Technologies such as lPSec allow secure communications on shared networks.

- Fiexibiiily: More service offerings allow the service provider to create a more customized solulion for the cuslomer.

. Fiexlbilily: Partnerships enable more complete snlulions. by leveraging the stengths of partner companies.

- Cost Service providers will often offer pre- negotiated rales for partner services.

. Service: Strong customer relationships are buill lhrough providing beueer service,

0 2003 RHK fnc. All rights r e s e r v e d w o r l d w i d e .

EXHIBIT R- _(JF-10) CASE ~ NO. U-13796

Service Provider Metrks Key to True Differentiation July 16, 2m3

Table 3: Quantitative service provider characteristics and relation to customer buying factors Service provider

characteristic Description (why important) Buying factors impacted (how) - Geographic reach

- Provisioning time

- Network performance

. Networkcontinuily

* Interconnect

Operational efficiency

- Financial health

Source: RHK Inc.

- The number of cities and countries in which services can be provided, preferably via owned facilities, but also through coiiocated and leased facilities, or through partnerships wi!h olher service providers.

. Tne me 'e+ rea 131 pen sen ce IC ~t tumeo .P. mJ0.ng tne ioca. access c ICJJ and . o w na.. iransoon fac liies

- The quantitative and qualitative levels at which a network fulfills its intended function. affecting its ability to provide services satisfactorily.

- The quantitative and qualitative levels of availability of netwoh services. This includes he ability to provide backup links (redundancy), quickly restore !he network afier an outage. and protect data during an oulage or attack

- The degree to which a network is connected lo other networks. extending its ullimale reach. Also, the amount of access capacity available Ihraugh local carriers, affecting Vle bandwidth that can be made available to customers

. The ability 10 utilize resources, especially employees, efuiently and effectively. Employee costs are the largest expense for most service providers. so efficiency direcUy leads to cost savings that can be passed on to cuslomer.

- The general state of a service provider's financial% including its profit and loss performance. balance sheet strength. and committed capital investments.

Connecwryy .ncreasec geog1aph.c leach a. O A S greater connecsriry: nowver. be lype a m qLa dry of conneceviry ,anes-loi exam@e, some 5'1s may have 0n.y 56 Kops access. hh ies others have T1 or nigher Smce. Local presence a Ions sen'ce personne 10 Qe c OSef 10 c.stomers, laut'ral.nq q d w response.

Sen im Shon prov sion ng ume a Ichs new sentces to ce I.rnea ~pfas:er ~ .e i i n~ l i l f Fast pro.is:on ng also a 06 oandn.otn ana sen ces to oe Lined On and 011 on Cemand

Relaonil{ Cons 8:eni.y h.gn performance means we netnor& s mole re,.ao.e

Retiabilily: A highly available network (low downtime) is more reliable. Secunly: The ability lo prevent attacks and outages. and to recover from them. enables greater security lor h e applications and data that reiy on the network.

Conneclivi!~ Higher degree of inlerconnect enables greater connectivitf by leveraging the reach of other networks. Flexibili!y: Large access capacity facilitates nexibility in bandwidlh and services.

Cosl: More efficient use of resources generally means operational cost savings, which can be passed on to the customer. Service: High efficiency implies a well-run organization, which facilitates customer service.

SfabiMy: Better financial health suggests h e Service provider will not only remain in business, but also be able to make the necessaly investments to keep its network and service up-to-date.

Service provider characteristics impact the ability not only t o satisfy current customer needs, but also to meet the more demanding needs of tomorrow. In other words, the ability to roll out new services and applications depends strongly on the overall fitness of the service provider. These new services-such as voice-over-IP and application- aware networks for Web services-are critical to the health of the telecorn market. If service providers do not enable new services, other parties, possibly the customers themseives, will-resulting in a dislocation or disruption of their market assumptions. Vigilance for.new

0 2003 RHK Inc. All rights reserved worldwide. 6

EXHIBIT R- -(JF-IO) CASE NO. U-13796

P . 3

NEWTON'S TELECOM DICTIONARY copyright 0 2000 Harry Newton email: Harry-Newton @Technologylnvestor.com personal web site: www.HarryNewton.com

All rights reserved under International and Pan-American Copyright conventi including the right to reproduce this book or portions thereof in any form whats

Published by CMP Books An Imprint of CMP Media Inc. 12 West 21 Street New York, NY 001 0

ISBN 1-57820-053-9

July, 2000

Sixteenth and a Half Edition, Expanded and Updated

For individual orders, and for information on' special discounts for quantity orders,

. . !i Y % 3

,F3 $ please contact: . . 5

CMP Books 6600 Silacci Way Gilroy, CA 95020 Tel: 800-LIBRARY or 408-848-3854

Email: telecom @ rushorder.com

Distributed to the book trade in the US. and Canada by Publishers Group West 1700 Fourth St., Berkeley, CA 9471 0

F a : 408-848-5784

D. 4

. Dec 18 03 12 :04p P . 5

Exhibit R- (JF-12) -

The Hybrid Cost Proxy Model Customer Location and Loop Design Modules

C. A. Bmh, D. M. Kmnet, J. PrisbrPy and W. W. Sharkey Federal Communications Commission

and

Vaikunth Gupta Panum Telecom. LLC

www.panumtelecom. com

December 15, 1998

EXHIBIT R- -(JF-I2) CASENO. U-13796

Overview of Version 2.6 of the HCPiM

The HCPM has from the beginning been designed to utilize sources of geocoded customer location data. Previous versions of HCPM were designed to also utilize Census block level data as an alternative source of publicly available data. In the current release, a number of small modifications have been made to "fine nine" the model under.the expectation that it will ultimately be used with a source of geocode data. The current release also contains a clustering module. This module incorporates explicit optimization routines as part of the cluster formation process, and also allows the user to choose from, and evaluate, three different clustering algorithms. A new module has been deyeloped to provide an interface between the'cluster and loop design modules. In all other respects, version 2.6 of HCPM is identical to version 2.5, which was publicly released on February 6, 1998.

The Customer Location Module.

The local exchange telephone network must connect every customer who desires service to.a local central office switch. A critical component in the design of such a network is the definition of a "serving area" which consists of a group of customers served from a common remote terminal. Feeder plant connects every serving area to the central office and distribution planfconnects every customer in a given serving area to a "serving area interface." Clustering algorithms group customers to form serving areas on the basis of both a distance constraint, so that no customer is farther from a potential DLC location than is permitted by the maximum copper distance, and on the, basis of the maximum number of customers in a serving area, which depends on the capacity of the largest DLC terminal. because of the multiple-objective nature of the underlying optimization problem, the design of a clustering algorithm is a difficult task, and some algorithms may not perform well in all situations.

The HA1 clustering algorithm IS a "nearest neighbor" algorithm, which forms clusters by joining customer locations to the nearest adjacent locations. Version 2.6 of the HCPM contains three different clustering algorithms, including a version of the nearest neighbor algorithm. Based on both speed and performance considerations. we recommend the use of a "divisive" clustering algorithm in which new clusters are successively split from a main cluster which initially contains all customer locations. Clusters are evaluated on the basis of the relative distance of customers from the line weighted centroid of the new and old clusters, rather than on the basis of distance from a nearest neighbor. After an initial clustering process, two different optimization algorithms look for ways to're-assign customers to clusters, so as to reduce the total distance from the cluster centroids, while satisfying the maximum distance constraints. These optimization procedures significantly enhance the performance of the original algorithms.

After geocoded customer locations have been grouped into clus!~rs, it is necessary to Further process the location data so that tt Can be used'effectively by the ioop design.niodule. Both the BCPM and - HA1 models process these data by creating square or rectangular serving areas without regard to the precise shape of the underlying cluster. Moreover, these models assume that customers within each serving area are uniformly distributed so that a regular pattern of distribution plant can be constructed to serve each customer. This rearrangement of customer locations can, in some circumstances. significantly distort the costs ofbuilding distribution plant to serve the cluster. In version 2.6 of HCPM, as in all previous versions. distribution plant is built. subject to a small margin of error, to exact customer locations. The model accomplishes this task by defining a grid on top ofevery cluster and then subdividing each grid into a large number of microgrid cells. Loop plant can therefore be designed specifically to reacli only populated microgrid cells. For large wire centers. the number of populated microgrid cells will be much less than the number of customer locations. even when a relatively small microgrid size is specified. Therefore, with this approach it is possible to place an upper bound on computing time, while simultaneously placing a bound on the maximum possible error in locating any individual customer.

Desien of Distribulion and Feeder Plant

HCPM 2.6 1 12/15/98

EXHIBIT R- -(JF-12) CASE NO. U-13796 PACK 7 nr: 7

, ..

Distribution plant consists of the set of analog copper cables, structures and other facilities such as network interface devises that are required to connect every customer location to the closest serving area interface (541). Feeder plant consists of the set of fiber. digital copper (TI) or analog copper cables and structures that connect every SAI to the central office. Tbe HCPM considers the possibility that up to four SAIs can be used within each serving area. When there is more than a one SAI, the model identifies a "primary" interface located closest to the central office. Each non-primary interface is connected to the primary interface using T-l (digital) copper cable.

The feeder network, which connects every primary SAI to the central office, is designed using a "minimum cost spanning tree" algorithm modified to take account of the cost of cable and structures rather than simple distance. Beginning at the central office, the algorithm builds a network sequentially by examining both the cable and structure costs involved in attaching new nodes to the network. Lowest cost nodes are attached first. When each new node is attached, the connection is chosen that minimizes the cost of cable and structures that are required to connect that node to the central office using the currently existing network. Distance computation can be done using either rectilinear distance or airline'distance. In addition, "junction nodes" are placed at points due north, south, east and west of the central office along what would be the main feeder routes in a traditional "pine tree" feeder design.

In calculating the cost of the distribution network, HCPM employs two separate algorithms. One algorithiil deploys vertical backbone and horizonta: branching distribution cables from the serving area interface to reach every populated microgrid. Branching cables run along every other microgrid boundary. Each microgrid is subdivided into equal-sized lots. Drop terminals are loc'ated to serve one to four lots and cables are placed on every other lot boundary to connect with the backbone and branching cable leading to the SAI. The second distribution algorithm uses the same minimum cost spanning tree network as was used to design the feed.er network. In this approach, microgrids are divided into lots and drop terminal locations are determined as before. A spanning tree network is then constructed which.connects every drop terminal to its nearest SAI. All distance computation used in constructing the distribution network are based on rectilinear distance.

The HCPh4 incorporates a number of explicit optimization routines in both the distribution and feeder algorithms. It selects the appropriate feeder technology (fiber, digital T-l on copper, or analog copper) on the basis of cost minimization subject to engineering constraints defined by user inputs. The model also selects loop electronics by examining every feasible combination of large and small terminals and selecting the cost minimizing outcome. In the feeder network, the model optimally determines whether to splice two fiber cables or run multiple cables at each junction point. All technology decisions are made on the basis of life cycle costs, based on a table OF technology specific annual cost factors.

HCPM 2.6 2 1211 5/98 EXHIBIT R- -(JF-12) CASE NO. U-13796 PAGF 2 f3F 7

The model assumes the following pro ression of technologies as distance increases: 26-gauge'copper, 24 gauge copper, T I on copper, and fiber." The model treats these thresholds as non-binding constraints in a cost optimization. For example, if relative prices indicate that fiber-is less expensive than T I , fiber will be chosen even if the feeder distance is less than the TI-fiber crossover. HCPM explicitly computes the C Q S ~

of the feeder plant which connects each terminal to the central office, and, subject to the distance constraints described above, selects the cost minimizing technology.

4.1 Distribution Plant Design

The distribution ponion of the loop design module determines the cost of distribution plant for each cluster in isolation (ignoring information from all neighboring clusters). The algorithms described in the following sections compute the cost of all plant that is required to connect each customer within the cluster to the nearest SAI.

4.1.1 Distribution Plant Within a Microgrid

Each microgrid is divided into lots based on microgrid population. The number of lots within a microgrid is determined by the formula lots = H + (BIAverage Business Lines per Location) where the number of lines per household and the average number of business lines per business location are given in Table 16 below.I2 Distribution cable is built to touch every lot in the cell, as illustrated in Figure 2. Backbone cables, which connect cells to the SAI, are assumed to run horizontally and branching cables within a cell are assumed to run vertically.

1, Cable Junction Points

. Dropwire

Vertical Branching Cable

Figure 2: Disiribution Cable Within a Cell

Branching cable is assumed to follow every other vertical lot boundary, beginning at the upper right hand comer of the lower-left most lot as shown in Figure 2. Drop cable is designed to serve groups of four properties whenever possible. If drop wire runs to the center of a lot. its length is equal to one half of the diagonal of the rectangle which defines the lor. This represents the maximum possible drop length in a lot. Alternatively, if the residence is assumed to be located on the midpoint of the lot frontage, the drop length would be equal to one half o f the width of a lot. HCPM assumes that actual drop length is a convex combination of these two extreme possibilities, weighted by a user parameter ?.(set by default equal to X).

TI-fiber crossover equal to the copper-TI crossover, the user can instruct the model to ignore T I technologies in the feeder network. " That is, copper gauge crossover I copperT1 crossover <TI Fiber crossovec

The initial release of the HCPM does not account for multifamily or high rise residential housing units. Census data are available on housing type, and future versions of the model will make appropriate adjustments to take account of i t .

I 2

HCPM 2.6 10 12/15/98 EXHIBIT R- -(JF-12)

PACF A n C 7 CASE NO. U-13796

4.1.2 Connection of Microgrids to the Nearest SAI

In this section. we describe two algorithms that are used to determine the correct amount of cable and structures that are necessary to connect each microgrid to the nearest SAI. When FEEDDIST is run in a fully optimizing mode, it computes the cost of distribution plant for all clusters using both approaches, and selects the approach giving the lower cost!'

Cable Junction Point

Unpopulated Cell

Figure 3: Connection of Cells to the Closest SAI

The first algorithm is most appropriate in densely populated clusters, in which the proportion of populated microgrids to total microgrids is relatively large. Backbone cables run along every other cell boundary, and connect with the distribution plant within a cell at various points as illustrated in Figure 3. The location of the SAI divides the cluster into four quadrants. Beginning with the "southeast" quadrant, backbone cable is tun along the boundary of the first and second rows of cells, running eastward until it ,reaches a point directly below the SAI. Cable from cells in row one is directed upward toward the cell boundary, while cable from cells in row two feeds downward. From the cell boundary point directly below the SAI, vertical cable then completes the connection to the SAI. By continuing along every other cell boundary, but extending cable only to populated cells, it is possible to connect every customer in the quadrant to the SAI. Similar connections apply'to each of the remaining quadrants.

The second algorithm generally gives a more efficient distribution network for clusters with a lower population density. where the number of populated microgrids is smaller. In this case, the construction of an optimal distribution network within a cluster is closely related to the problem of constructing an optimal feeder network for the entire wire center, and we are able to use the same algorithm to provide a solution. In this approach, the algorithm described'in the Feeder Plant Design section below (based on Prim, 1957) is used to connect each drop terminal placed withineach microgrid (as shown in Figure 2) to a network consisting ofal l drop terminal locations and all SAIs.

HCPM always computes a distribution cost using the first algorithm. Since computations involving the Prim algorithm can be time consuming for large clusters. it is recommended that the second approach be used only for lower density clusters. Whenever both algorithms are used, distribution cost is determined by choosing the minimum cost obtained.

4.2 Feeder Plant Design

The model first computes the cost of each possible configuration of primary and secondary SAIs within a cluster, and selects the least cost option. Then. based on the locations of the central office and all

In order to generate approximately optimal results using less computing time, the user has the option of I1

computing distribution costs using both approaches only for the lowest density grids.

HCPM 2.6 11 I?./ I 5/98

EXHIBIT R- -(JF-12) CASE NO. U-13796

primary SAIs within a quadrant. it determines the feeder network which connects each of the primary SAIs to the central office.

4.2.1 Determination of the Optimal Number of Primary and Secondary S A I s within a Cluster

I Figure 4: Connecting Secondary and Primary S A I s

The cost o f copper-based T I terminals and fiber-based DLC (digital loop carrier) terminals are determined as follows. The terminal sizes are a function of the line capacities required. The digital signal hierarchy, typically referred to as the DS-n where n is I , 2 or 3 (and multiples of 3 referred to as OCn), is used to size the terminal requirements. Each DS-I or TI may be used 10 support 1-24 lines. A D S 3 o r T 3 may be used to support 672 lines (28 DS-1's). Finally, an OC-3 (3 DS-3's) may be used to support 2016 lines.

Four terminal sizes with line capacities of 2016, 672, 96 and sub-96 may be used, 'A sub-96 terminal is really a terminal with line capacity of 96 that may be used to support fewer'than 96 lines (i.e. 1-24. 25-48. 49-72 and 72-96) lines by using one, two three or four modules. Each TI module is required to support a line count of 1-24. Thus, if 49 lines need be supported, a sub-96 terminal with three modules and appropriate number of line cards, 4 lines per line card, will be required.

. . , .

Fiber terminals with line capacities of 24, 96, 672 and 2016 are fiber based and will require 4 fibers per terminal. If the line count is greater than a user supplied threshold value, but less than or equal to 672, one terminal of capacity 672 is required. If line count is less than this threshold, but greater than or equal to 96. then multiple 96-line terminals are required. TI terminals, with line capacities of 24 oc 96 reyuire two copper pairs per DSI line: one for transmitting signals from the CO to the primary SAI. and one for receiving signals from the primary SAI at the CO. A user adjustable 4 to I redundancy ratio is assumed in the model, which means [hat 10 copper pairs are required to serve each 96 line TI terminal.

Similar computations apply to the choice of technologies for connecting the primary SAl to its associated secondary SAls. The cost far a given terminal is the cost for input fiber cables extending From the primary SAI to the secondary SAI (4 per terminal) or the cost of copper pair cables (two per T1 plus two for the redundant TI), and a fixed and a variable cost based on the line count.'' For each configuration

The following examples illustrate the above concepts for various line counts: I 4

2017: One 2016-line terminal and one sub-96 terminal with one T l module and one line card; 4 fibers and 4 copper pairs are required. 1344: Two 672-line terminals and 8 fiber pairs are required. 384: One 672-line terminal and 4 fiber cables are required.

HCPM 2.6 12 I211 5198 EXHIBIT R- -(IF- 12) CASENO. U-13796 P A C 0 c nc 7

. ,

of one to four SAIs the model computes the total distribution.cost assuming the cost minimizing use of primary and secondary SAls. Within each cluster, primary’and G c o n d a q SAIs are connected b y T 1 lines using a minimum distance spanning tree network using an algorithm described in the following section. For each configuration of I to A SAIs, the model computes the total cost within the cluster of the distribution plant which connects each customer to the closest primary or secondary SAI; the TI connections which link all primary and secondary SAIs; and the cost o f all associated terminals.

The minimum cost obtained defines the total distribution and internal feeder cost for the cluster, and the optimal number of primary and secondary SAls. At this point the model has defined for every cluster the cost minimizing collection of primary and secondary SAIs, and has computed the total cost of connecting every customer to the closestprimary SAI .

4.2.2 Feeder and Subfeeder Routes

In previous versions of the HCPM, and in other models of the local exchange (e.g. Gabel and Kennet. 1991) feeder plant was deployed in a “pine tree” network in wbich four main feeder routes emanate from the central office along East-West and Notih-South routes. Subfeeder routes perpendicular to the main feeder routes were then used to bring the feeder system closer lo individual SAIs. This design proved to be highly efficient in terms of creating opportunities for the sharing of structure costs among feeder cables serving different SAIs. Lower structure costs made possible by increased sharing; however, came at the expense of longer feeder routes and correspondingly higher cable costs. In order to balance these two opposing tendencies, the HCPM examined a large number of possible feeder systems having different number of subfeeder routes, and chose the configuration giving the lowest cost:

The current version of HCPM uses a variant of an explicit optimization algorithm. discovered by Prim in 1957, to determine the trade-off between structures and cable costs. ” This algorithm is based on some well known niathematical prirLplea of ~ie:\*ork design based or, icchniques of discrc:i mathematics : and graph theory. An abstract network consists of a single “supplier” node, a set of customer nodes, a cost function specifying the cost of connecting any two nodes, and a set of painvise traffic demands between any two nodes. In the application of the Prim algorithm to the feeder network. the supplier node is the central office for a given wire center, and the customer nodes are the remote terminals. or SAIs, that define the interface points between the feeder and distribution portions of the network. The algorithm can also be applied to determine cable routes for distribution networks within a cluster as noted previously. In this case, the supplier node is an SAl within a cluster and the customer nodes represent individual subscribei locations that are to be connected to that SAI.

In both the feeder and distribution portions of the network, the objective of the telecommunications engineer is to minimize the cost of connecting each customer node to the supplier ~rorir. Wiiilc i u gencral this is an extremely difficult problem lo solve, rh&e are sevziai special cases in which efficient algorithms exist which define a fully optimal network solution. One special case of interest is the construction of a “minimum distance spanning tree network“ in which the sole objective is to minimize the aggregate length of communications links within the network. Such a network would be approximately optimal when traffic demands are sufficiently low that the actual cost of each link in the network is largely determined by the cost of structures (whichdepend only on distance).

A minimum distance network can be constricted using the Prim algorithm in the following way. Beginning with a network consisting only of the supplier. find the nearest customer node that is not yet attached to the network and attach it. The network then consists of the supplier and one customer. The

379: Three Yblioe terminals and one sub 96-line terminal i.e. a Ybline terminal with 4 TI modules and 5 line cards; 40 copper pairs are required. l’ See Prim, R.C. (1957). :‘Shortest Connection Xetworks and Some Generalizations.” Bell S,vslem Technical Journal. 36, l3S9-1401 for a description of an efficient algorithm for computing ‘minimum distance networks. A computed coded version of the Prim algorithm, and some extensions. is contained in Cower. J.C. and G.J.S. Ross (1969). “Minimum Spanning Trees and Single Linkage Cluster Analysis.” Applied Siarisrics. 18, 54-64.

HCPM 2.6 13 I UI 5/98

EXHIBIT R- -(JF-I2) r A < F N n 11.1770L


Network Reliability Council (NRC)

Reliability Issues - Changing Technologies Focus Group

New Wireline Access Technologies Subteam Final Report

February 22,1996

Roy Koelbl (chair) Chris Bright Wally Schatzley Jim Fischer Mark Vogel Gary Berkowitz Bill Nelson Keith Williford Gale McNamara K e h Mistry Dan Sills Eric Tollar Glenn Mahony Scott Bachman David Miller Alex Best Bill McDonald

Bellcore ADC ADC Ameritech NME Ameritech AT&T BBT BBT Bellcore Bellcore Bellcore Bellcore BellSouth Cable Labs Clear Communications Cox Communications Fujitsu

Chuck Dougherty William Ray David Large Raja Natarajan Paul Vilmur Bill Weeks Tom Jurus Craig Mead Walt Srode Robb Balsdon Farr Farhan Roy Thompson Tim Wilk Donovan Dillon Duane Elms Chris Barnhouse Jim Haag Earl Manchester

General Instruments Glasgow Electric Media Connections Group Motorola Motorola Next Level Communications NYNEX Optical Sdutions PBNI Rogers Engineering Scientific-Atlanta Scientific-Atlanta Scientific-Atlanta SNET SNET Time Warner Time Warner U S WEST


TABLE OF CONTENTS

. .-

I. E,YECUTIYE SIJX&wRY ............................................................................................................................. 4

2. BACKGROUND .............................................................................................................................................. 5

2. I TEM GOALS ANI) ACTIVITIES ........................................................................................................................... 5 2.2 RECOklMENOATlON AND BEST PRACCICE DEFlNlTlON ......................................................................................... 5

3. SVBTEAV MEMBERSFID AND ORGAYIZATION .................................................................................. 5

4. DATA COLLECTION AND ANALYSIS METHODOLOGY ..................................................................... 7

5. STUDY RESULTS .......................................................................................................................................... 7

5.1 ACCESS NETWORK E V O L ~ ~ O N .................................................................................................................... 5. I . I Telephone Access Network Boundary and Specifications ........................................................................... 7 5.1.2 Telephone Access Network Evolution .................... : .................................................................................. 9 5.1.3 Cable Television ( C A W Network Evolution .......................................................................................... I I 5.1.4 New Wireline Access Network ............................................................................................................... I2

5.2 N E W O R K INTELLIGENCE IMPACT ON R E L l ~ l L l ........................................................................................... 13 5.3 POWERING .................................................................

5.3. I Access Network Powering . The Challenge ................................... 5.3.2 ilccess Meht-arkPowering . Solutions ............................................................ 5.3.3 Porh Forward ................................................................................

5.4.1 Fiber Node Size 5.4.2 AmplFerFailures ......................................................................... 5.4.3 Ingress ................................................................................................................................................... 23 5.4.4 Drop Cable to Customer Premises .......................................................................................................... 24 5 . 4 3 FiberKable Cuts .................................................................................................................................... 26 5.4.6 Other Best Practices ................................................

5.j. I Potential FTTC Reliability Issues ............................................................................

5.4 HYBRID FIBER~COAX .............................. . .

. ................ 5.5 FIBER-TO-THE-CURFI TECHNOLOGY ................................................................................

_ _ >.>.2 Path Fonuard ......................................................................................................................................... 28 5.6 RELIABILITY TEbIPLATE .................................. .................................. ~ C K l T E C r U R E / T E C H N O L O G Y / ~ P E ~ T l O N S ................................................................... I r I s T ~ L . ~ T I C ? l / ? , i . ~ ~ ~ E ~ . ~ ~ C E ...................................................................................... 6. SUMMARY OF RECO>IMXNDATIONS ................................................................................................... 30

7. CONCLUSIONS ........................................................................................................................................... 31

8. ACICVOWLEDG>IENTS ............................................................................................................................. 31

9. REFERENCES .............................................................................................................................................. 32

10. APPENDICES ............................................................................................................................................. 33

APPENDIX A . OVERVIEW OF AN EXMPLE BROADBAND HFc ARCHITECTURE .... ......................................... 3: APPENDIX B . INGRESS RELIr\BILITY ISSUES M HFC ACCESS NETWORKS .............................................

8. I Introduction.. ................................................................................ B.2 Ingress Stirdies ................................................ 8.3 Long- Term lhgress Studies ............................. 8.4 Shorr-Term Ingress .................................................

L

EXHIBIT R- -(JF-13) CASE NO. U- 13196 PAGE 2 OF A6

APPENDIX C - ISSUE ST~TEMEM ........................................................................................................................... 43 APPENDIX D - NEW TECHNOLOGY RELlAslLlTY TEMPLATE .................................................................................... 47

. . . . . . . . . . . .

3 EXHIBIT R- -(JF-13) CASENO. U-13796 PAGE 3 OF 46

1. Executive Summary The Network Reliability Council’s (NRC).Changing Technologies Focus Group established the New Wireline Access Technologies (NWAT) subteam as one of five subteams examining the reliability aspects of key services provided over new network technologies in the Public Switched Network (PSN). The primary objectives of the NWAT subteam were to: ( I ) identify, define, and clarify potential service reliability attributes (i.e., weaknesses and strengths) associated wixh new wireline access technologies, and (2) where possible, identify potential mitigating solutions and provide recommendations for improved reliability. A subteam of more than 30 members representing a cross-section of the industry worked on this assignment from August to December, 1995. Subteam members represented service providers and suppliers from traditional telephone service companies and newcomers, or potential newcomers, to the local telephone service market (e.g,, cable television companies).

The NWAT subteam primarily investigated the reliability of telephony services transported over Hybrid FiberKOax (HFC) and Fiber-To-The-Curb (FTTC) access networks. These technologies were benchmarked and compared with today’s systems to understand potential failure modes and key differences that could improve or degrade reliability. Digital Loop Carrier (DLC) systems and cable television systems were used as benchmarks. Field and system test data was obtained from several sources, including publicly available data, service operators, and suppliers. Based on this data, inputs from subteam members, the following is a high-level summary of subteam findings and recommendations:

The industry’s goal is to provide 99.99% reliability for telephony services provided over HFC and FTTC access networks. Deployment of HFC and FTTC systems is just beginning, therefore, gathering of critical field data is in its early stages. Industry has identified Several key reliability issues and potential mitigating solutions. Operators and suppliers should implement a process to gather field data on systems as they undergo trials and are deployed. Operators and suppliers should institute a process for root cause analysis on outages and deve!op best practices to improve reliability.

Several more detailed findings and recommendations are provided throughout the report.

4 EXHIBIT R- -(m-13) CASE NO. U-13796 PAGE 4 OF 46

2. Background

2.1 Team Goals and Activities Wireline access technologies are rapidly evolving to support the implementation of new, advanced services. These technologies are being deployed (or are planned to be deployed) by a proliferation of emerging service providers. However, it is expected that many of these access networks will also support telephony services such as Plain Old Telephone Service (POTS). The FCC has asked whether the high reliability of telephony services that currently exists in the public switched network is maintained with the deployment of these new access technologies. This report is aimed at investigating the reliability of such wireline access networks.

A copy of the Issue Statement for the focus group is containedin Appendix C. It was the intent o f the New Wireline Access Technologies (NWAT) subteam to identify, define, and clarify potential service reliability attributes (i.e., weaknesses and strengths) associated with new wireline access technologies. Based on these findings, conclusions would be drawn and recommendations made^ where possible. By identifying and clarifying reliability attributes, it was hoped that “myths” and “fears” based on misinformation and/or the lack of information might be mitigated. On the other hand, because limited reliability data typically exists for any new technology, the subteam did not want to foster new concerns due to limited data or information. Consequently, it was decided that a more qualitative than quantitative investigative approach would be taken, with the emphasis on understandkg failure modes and identifying potential mitighting solutions.

Subteam investigations focused on the reliability of HFC and FTTC access networks, because these two technologies are either being deployed now, or are expected to be deployed in the next three years to support POTS service. Fiber-To-The-Home (FTTH) was also considered, but to a much lesser extent.

This report focuses on the reliability aspects of new wireline access technologies and makes no assessment about their cost-effectiveness for providing key telephony services. Reliability issues related to the interconnection of networks based on these technologies to the PSTN are addressed in the Focus Group 11 report on Increased Interconnection.

2.2 Recommendation and Best Practice Definition The terms “recommendation” or “Best Practice” as used in.this report is defined as follows: “recommendations” are those countermeasures (but not the only countermeasures) which go furthest in eliminating the root cause(s) of outages. None of the recommendations are construed to be. mandatory.

Service providers and suppliers are strongly encouraged to study and assess the applicability of all countermeasues for implementation in their company products. It is understood that all countermeasures, may not be applied universally.

. .

3. Subtearn Membership and Organization With ever-increasing competition for the provision of telecommunications services, the industry is beginning to experience a proliferation of new service providers. It is therefore important that any

5 EXHIBIT R- -(n-13) CASE NO. U-13796 D A CF < nF A6

subteam addressing the reliability of new access technologies incorporate Subject Matter Experts (SME) from different segments of the industry that either offer, or plan to offer, POTS service over such networks. Each se-pent of the industry may have a different perspective on the technology and its reliability requirements. Consequently, subteam members include one or more representatives from the Regional Bell Operating Companies W O C ) , cable television companies, and electric utility companies. Both operators and suppliers were represented on the NWAT subtearn. Subteam members and their affiliations are listed below.

Roy Koelbl (chair) Chris Bright Wally Schatzley Jim Fischer Mark Vogel Gary Berkowitz Bill Nelson Keith Williford Gale McNamara Keku Mistry Dan Sills Eric Tollar Glenn h"ahony Scott Bachman David Miller Alex Best Bill McDonald Chuck Dougherty William Ray David Large Raja Natarajan Paul Vilmur Bill Weeks Tom JUIUS Craig Mead Walt Srode Robb Balsdon Farr Farhan Roy Thompson Tim Wik Donovan Dillon Duane Elms Chris Barnhouse Jim Haag Earl Manchester

Bellcore ADC ADC Ameritech NME Ameritech AT&T BBT

Bellcore Bellcore Bellcore Bellcore

Cable Labs Clear Communications Cox Communications

General Instruments Glasgow Electric Media Connections Group iMotorola Motorola Next Level Communications NYNEX Optical Solutions PBNI Rogers Engineering Scientific-Atlanta Scientific-Atlanta Scientific-Atlanta SNET SNET Time Warner Time Warner US West

BBT

Bell South

Fujitsu

The NWAT subtearn was divided into four working groups as follows: ,

6 EXHIBIT R--(JF-13) CASE NO. U-13796 PAGE 6 OF 46

Current Access Networks 0 FTTC Access Networks

HFC Access Networks Impact of Network Intelligence

4. Data Collection and Analysis Methodology

Because of the compressed schedule for completing the team report, a formal industry survey and collection of outage information was not conducted. Instead, the team relied on several other sources of data.

In order to obtain an understanding of how the reliability of the new technologies will compare to existing networks, it is useful to understand the differences between the new and old technologies. Specifically, this includes understanding the current failure modes, if and how they will occur in the new technology, and'what potentially new failure modes will occur. To achieve this understanding, data on failure modes of current access networks (telephony and CATV) was investigated, as was system test data for the new technologies. Data was collected from several sources:

published papers publicly available data

supplier provided data telephone company and cable television company data

Because much of this data is proprietary, detailed data is not included in this report. Instead, a summary of findings will be presented based on this data, as well as consensus views from subteam members. It should be noted that the data gathered for this report was from available sources. No new data was systematically gathered for this study, which made it difficult to properly compare results from different sources. Consequently, detailed recommendations that can be applied globally are limited.

. . . . . .

5. Study Results

5.1 Access Network Evolution

5.1.1 Telephone Access Yetwork Boundary and Specifications The telephone access network connects the local switch in a central office to individual customers terminating at a Network Interface @I) at the customer's location and at the interface with a local switch in the Central Office (CO), as shown in Figure 5.1. It is these two interfaces that bound the new wireline access technologies addressed in this report.


Local Switch Inkrlaee' -Analog - TR-08 -TR-303

Network Interface ' (ANSI T1.401 and Part 68) Central mce

I I 'See text fo( interface dewplion

Figure 5.1 Access Network Boundanes

The NI is the point of connection between the Customer Installation (CI) (i.e., all telecommunication equipment and wiring on the'customer side of the interface) and network facilities. ANSI T1.401-1993, Interface Between Curriers and Customer Installations -Analog Voicegrade Switched Acczss Lines Using Loop-Start and Ground-Start Signaling,"'describes the network interface in terms of the interactions between and electrical characteristics of the network facilities and the Customer Installation (CI). For example, loop-start signaling is typically used for residence and business central office lines and ground-star-signaling for two- way seizure CO hunks. AXSI T1.401 covers only those characteristics of loop-start and ground- start signaling interfaces used by the network'and C1 to establish calls. This document does not' include signals produced by other features, such as call waiting, calling number delivery, AN1 reliability, or performance characteristics of the access network.

ANSI T1.401 is formulated on the premise that the customer equipment and wiring meet the applicable requirements of Part 68 of the FCC Rules and Regulations which sets forth requirements for the registration of CI equipment to protect the network from harm. Subpart F of Part 68 describes jacks that have been standardized through industry agreement that are installed by the telephone company at the NI for the connection of customer equipment and wiring. Part 68 also requires that telephone companies notify, in writing, customers who have registered (or grandfathered) equipment connected to telephone company facilities of any changes in facilities, equipment, operations, or procedures that can render the customer's terminal equipment incompatible with the telephone company facilities. On request, the telephone company will provide interface information.

The interface between the local switch and the access network can be analog or digital (the digital interface has two variations). Traditionally, the access network consisted of a twisted- pair of copper wires connecting a NI to an analog interface on the local switch. Loop carrier systems providing the capability of carrying more than one customers"te1ephone signals over a pair of wires have been in use for about thirty years, with Digital Loop Carrier systems now being the prevalent loop carrier system. Universal Digital Loop Carrier (UDLC) systems were introduced in 'the early 1970's and consist of a Central Office terminal .(COT) located near the switching system, a remote Terminal (RT) located near the customer, and a digital transmission facility connecting the COT and RT. Twisted-pairs of copper wire connect the COT to analog

..


interfaces on the local switch and the RT to Network Interfaces. Bellcore's' TR-NWT-000057, Functional Criteria for Digital Loop Carrier Systems, [" describes the interface between the local switch and the COT and between the RT and the customer. UDLC systems can be used with any local switching system because the interface presented to the local switch by the COT is the same as if the circuit were carried on a twisted pair of copper wires.

The introduction of digital switching made it possible to eliminate the COT by providing Integrated Digital Loop Carrier (IDLC) systems with many of the COT functions integrated into the digital switch and with the RT interfacing directly to the switch. Because there was a large embedded base of SLC8-96 systems at the time of the divestiture of AT&T in 1984,' Bellcore described the SLCO-96 interface in TR-TSY-000008, Digital Interface Between the SLCO-96 Digital Loop Carrier System and a Local Digital Switch'" . This digital interface is referred to as the "TR-08" interface in Figure 5.1. Another digital interface, with expanded capabilities, is described in GR-303-CORE, Integrated Digiilal Loop Carrier System Generic Requirements, Objectives and InterjCace"'and is referred to as the "TR-303" interface in Figure 5.1. The remote terminal for the TR-303 interface is typically referred to as a Remote Digital terminal (RDT). The TRs also defme criteria and requirements for other features and functions, such as those needed to support the high-quality transmission of voice and voiceband data signals; the provisioning of.other services such as call waiting and calling number delivery, and operations.

TR-NWT-00041 S, Generic Reliability Assurance Requirementsfor Fiber Optic Transport Systems"'contains the following service availability objective for the subscriber loop (i t . , the access network):

The two-way service availability objective of a narrowband transmission channel should be a minimum of 99.99% (0.9999 probability) for the subscriber loop.

TR-NM-0000057 and GR-303-CORE have allocations of the loop availability objective to the COT, RT and RDT. The TRs also require that an equipment supplier must, on request, provide hardware reliability predictions for the DLC system based on Method 1 (the "parts count" methodj in the latest issue of TR-NWT-000332, Reliability Prediction Procedurefor Electronic Equipmenl"' or on other methods.

5.1.2 Telephone Access Network Evolution Figure 5.2 illustrates how electronic schemes are evolving into the access network. The first diagram shows an access network consisting entirely of twisted pair copper wires with no electronics. The twisted pair terminates on an analog port on the digital switch denoted as a Subscriber line tnterface (SLI) unit.

. Bellcore's generic requirements provide Bellcore's view of criteria for equipment or systems intended for general use in a Local Exchange Carrier network. A requirement is a feature or a function that, in Bellcore's view, is necessary to satisfy the needs of a typical Bellcore Client Company (BCC). An objective is a feature or function that, in Bellcore's view, is desirable and may be required by a BCC.

SLC is a registered trademark of ATBT.

9 EXHIBIT R- -(JI-13) CASE NO. U-13796 PAGE 9 OF 46

Digital - -

Distribution oigital S Feeder , Swilch L

Anabg Analog

-Digital Digital Digital

Switch

-

. . Digital Transmission

Fiber or Coax CE Switch

Fiber

Transmission Facility

Fiber or Copper

CE: Common Equipment LE: Line Equipment

-Digital

Digiial Digital Switch

- -

. Figure 5.2 Evolution of Electronics into the Access Network

Transmission Facility Fiber or Coax

CE Fiber

The SLI contains the functions needed to interface a digital switch with a twisted pair medium. BORSCHT is an acronym that is sometimes used to illustrate the functions needed in the SLI; with the acronym defined as: B - Battery (power feed); 0 - Overvoltage protection; R - Ringing; S - Signaling and signaling detection; C - Codec (ix., analog to digital and digital to analog conversions); - Hybrid (ix., conversion between two-wire and four-wire transmission); T - Test access. The SLI typically contains the overvoltage protection, signaling detection, codec and hybrid functions and provides a means of applying the ringing and power feed voltages that are typically located in other pieces of equipment.

In the second part of Figure 5.2, the SLI is moved to the access network along with Common Equipment (CE). The CE includes Digital Transmission Facility (DTF) hnctions for interfacing to a DS1 facility connecting to the local digital switch. The DS 1 facility could be transported over twisted pair copper wires, an asynchronous fiber optic system, or a Synchronous Optical Network (SONET) system. The fiber optic or SONET system could be configured as a ring. The CE also typically contains other functions such as power supplies and the ringing generator (each of which are typically implemented with redundant hardware).

In the third diagram, the electronics component has moved to the curb, near the customer's premises: and finally, in the fourth diagram the electronics at the customer's premises. In both diagrams, there might be additional CE for converting between different media ( i t . , fiber and coaxial cable) or between different interfaces to the same media. Note that the Line Equipment (LE) in the fourth diagram is usually dedicated to a single customer (but potentially more than

10 EXHIBIT R- -(JF-I3) CASE NO. U-13496 PAGE 10 OF 46

. .

one line) and is not shared by multiple customers, as is the case with the common equipment in the third diagram.

5.1.3 Cable Television (CATV) Network Evolution Cable television systems provide broadband communication services and emphasize providing subscribers with multiple television channels, primarily for news, entertainment, and other information. Figure 5.3 illustrates the evolution of a cable television system from a tree and branch coaxial cable architecture to a fiber rich architecture with fiber and coaxial cable. In both architectures, a headend is the source of signals. The diagram labeled “CATV.Network” shows a portion of the system having a maximum cascade of sixteen trunk stations.

The diagram labeled “Fiber Rich CATV Network” illustrates a configuration where the signal from the headend is transmitted over several optical fibers to nodes in various places in the previous tree and branch architecture. The cascade is broken in several places on either side of the nodes and the amplifiers on one side of each node are reversed in direction. The fiber rich configuration uses the same number of amplifiers to serve the same number of customers, but, depending on the number of nodes, there will be fewer amplifiers between each customer and the headend, thereby improving reliability and performance.

In an all coaxial network t& amplifiers are powered via the coaxial cable using cable system supplies that are dispersed throughout the system and connected to power utility lines. Because uf the cascade topology, utility power loss near the headend can interrupt service in a major part of the system and a subscriber experiencing the loss of service could still have AC power. For this reason, power supplies might have batteries that provide 2 to 3 hours standby power and that automatically recharge when power is available.

. .

. .

CATV Network

Fiber Rich CATV Netwark

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. Figure 5.3 Evolution to Fiber Rich CATV Network


5.1.4 New Wireline Access Networks Figures 5.4 and 5.5 show an Hybrid Fiber Coax (HFC) network and a Fiber-To-The Curb (FTTC) network used to provide telephone service. For the HFC network, digitally modulated radio frequency (RF) signals are carried over fiber between a digital interface in the CO and a fiber node where optical-to-electrical and electrical-to-optical conversions are made for interfacing with the coaxial cable distribution network. The two-way signals for telephony are carried by the coaxial cable network to a Network Interface Device WID) which converts between the signals on the coaxial cable and the telephone signals at the NI. Although the NID is shown on the side of the house, in some applications,it could also be at the curb near the house with copper wire going from the curb to the NI at the house. The HFC network is typically two-way and has even fewer amplifiers in cascade than the fiber rich CATV network. A more detailed discussion of HFC networks can be found in Appendix A.

I central Ornee

h

optmi Fiber

DSI Fiber Node

. Figure 5.4 Hybrid FiberKOax (HFC) Network

The Fiber-To-The-Curb (FTTC) system [with a Fiber-To-The-Home (FTTH) variation] shown in Figure 5.5 consists of an Host Digital Terminal (HDT), a Passive Optical Distribution (POD) network and Optical Network Units (ONUs). The HDT manages the ONUs and provides the interface to the local switch. The POD physically connects the ONUs to the HDT and contains only passive optical components. The ONUs provide the signal processing needed to convert between the optical signals and the analog telephone signals at the network interface. The ONUs are connzctzd to the nsrwork ifieikx by twisted-pairs-cf coppe; wirc. With the FTTH variation, the fiber extends to the house where the O W is located. Bellcore requirements, including reliability criteria, for the HDT and ONU are in TA-NWT-000909, Generic Requirements and Objectives for Fiber in [he Loop Systems. "'

EXHIBIT R- -(mu) CASE NO. U-13796 PAGE 12 OF 46

: .

F R C

Figure 5.5 Fiber-To-The-Curb (FTTC) [with a Fiber-To-The-Home (FTTH) Variation]

5.2 Network Intelligence Impact on Reliability In comparison to many of the access networks currently deployed, HFC and FTTC access technologies have the potential for significantly improving the customer's perception of service reliability, at least with respect to the access portion of the network. These improvements stem from the ability to monitor the access network andare directly attributable to the "intelligence" being deployed with HFC and FTTC technologies.

In the present network, much of the .customer access portion of the network is unalarmed. Repair of outages in the unalarmed portion of the access network can be started only after a customer detects the outage and notifies the operating company. Customers are therefore aware of mahy outages which would affect them. With the introduction of intelligence, it is possible to remotely detect outages in the network, and possibly effect service restoral before a customer becomes aware that an outage has occurred. An intelligent network can si,gificantly decrease the total actual time that the access network for the cusromer is unavailable, as well as si,&ficantly decrease the number of outages that a customer detects associated with usage of the network.

As naditionally defined, Ketwork Uowntime (XDT) records the time from the initial time the network managers are.aware of an outage to the service restoral for the customer. This is the standard downtime measure used in present telephony. For comparison, let the Customer Downtime (CDT) record the time from the initial time the customer becomes aware of an outage to the service restoral for that customer. If the access network is primarily unalarmed, then clearly the two measures are essentially equal, because outages fxst require detection by a customer. However, if the network is alarmed, CDT could be significantly smaller than NDT, because many outages could go undetected by the customer.

The hypothetical example below illustrates the improvements in the customer's perception of reliability of the access network that could be anticipated with the new wireline access technologies. This example considers a possible access network implementation for telephony, and examines its reliability using the measures of Network Downtime (NDT) and Customer Downtime (CDT).


Example:

Total

Table 5.1 provides the hypothetical NDT reliability associated with an access network, consisting of a Digital Crossconnect System (DCS), Host Digital Terminal (HDT), Optical Network Unit ( O W ) , feeder and drop facilities. It is assumed that only 10% of the DCS, HDT, and ONU failures are silent failures (ie., unalarmed), the fiber feeder is l l l y alarmed, and one third of the unassigned failures are alarmed. This basically follows objectives defined in TR-41 #I.

53.0

. Table 5.1 Hypothetical NDT for Access Network

Component I NetworkDowntime I


. Table 5.2 CDT and NDT for Hypothetical Access Network

Component

DCS HDT ONU

Network Downtime Customer Downtime (min/yr)* (min/yr)

2.0 1.0 10.0 5.5 26.0 13.3

15 EXHIBIT R- -(JF-13) CASE NO. U-13196 n l r r 1 c n r 1,

Feeder

Total Unassigned

6.0 2.9 9.0 7.5

53.0 30.2 .

5.3 Powering To meet the reliability objectives of current telephone service,t backup powering methods are used to ensure the continuation of telephone service during commercial power outages. For traditional telephone service, these methods include the use of batteries, which are charged from commercial power, and secondary backup generators in the CO. These power sources together provide the reliable power needed for the operation of CO equipment and traditional telephone sets. The power is transmitted from the CO to the telephone sets over the same copper wire used for transmitting telephone signals.

A study of the quality of commercial power reported in the INTELEC '93 Proceedings"' illustrates the reliability improvement with backup power sources. The study analyzed a large quantity of power disturbance data from sites throughout the United States that included descriptive information for power disturbance events recorded at a typical 120V AC wall receptacle. The data was analyzed to determine the effect on service availability of Fiber in the Loop (FITL) systems when FITL equipment used commercial power at the customer's location with varying amounts of battery backup capacity. The following table shows reported results of that study. (Note that the unavailability estimates shown below are based on averages and for any given location the actual unavilability could vary.)

Table 5.3 Estimated FITL Service Unavailability vs. Backup Capacity*

I 1 Zero I 4-hr I 8-hr I 12-hr 1 i I Backup I Backup I Back I Average Service Unavailability I 370.2 I 192.4 I 153 I 128.5 I I (in minutes per year) * All unavailability estimates are derived directly from the data, with no adjustment to account for possible secondary backup (e.g., generator) power during long-duration disturbances. All values are calculated based on the combined effect of outage and low-voltage events.

As table 5.3 illustrates, the unavailability for commercial powering is significantly greater than the telephone access network objective of 53 minutes per year.

Power provision becomes an important issue for the new wireline access technologies because as more optical fiber (which does not conduct electricity) and more electronics are deployed deeper (ix., closer to the home) into the access network, it becomes increasingly costly to provide primary and secondary backup power sources. (Cost is an issue because network-powered wireline technologies compete with other technologies, such as wireless, where power for the telephone that transmits and receives the "wireless" signal is provided by the customer.) The remainder of this section describes the challenges associated with providing reliable power for the access network and reports 'findings on how the industry is finding innovative ways, based on local conditions, to meet these challenges.

The availability objective for the access network referenced in Section 4 is 99.99%. The objective can also be stated as an average downtime of 53 minutes per year, which equals the unavailability objective (O.Ol%/yr) limes the number of minutes in a year (525,600).

16 EXHIBIT R- -(JF-13) CASE NO. U-13796 n*nTi I E o r 1,

5.3.1 Access Network Powering - The Challenge Figure 5.6 illustrates how the access network is currently powered today. Batteries provide power for the CO equipment and for transmission over the copper pairs of wire to the telephone sets. The batteries are charged from a commercial power source with secondary backup provided by a permanently available local generator. The cost of the batteries and local generator can be shared by the relatively large number of customers connected to the CO.

Lt-----l Commercial Power

. Figure 5.6 Central Office Powering

Figure 5.7 illustrates the challenge of providing power for new access network teclinologies where a large number of equipment units are located near the customer and fiber optics is used for transmission from the CO. Commercial power at the customer's location could be used to power the network equipment, but battery backup and secondary backup (e.g., generators) would be needed to maktain th: cxisting telephone iletv<.ork :e!isSilit).. Because there are many pieces of equipment, each used by a few or even just one subscriber, providing and maintaining batteries and generators at each piece of equipment is costly.

Power could be provided from the CO, or centralized locations in the access network, but a centralized power supply could require the addition of copper media parallel to the fiber for power distribution. The additional copper media introduce inefficiencies because of the power losses incurred in transmitting the power. The power losses can be reduced by increasing the voltage used to transmit the power.

Fiber or Fiber and Coax and Other Equipment

Switch W

halog ' Wiring Twisted-pair

Digital Transmission Facility tnsiae

17 EXHIBIT R- -(JF-13) CASE NO. U-13796 PAGF 17 nF A 6

,Figure 5.7 Powering for New Wireline Access Technologies

5.3.2 Access Network Powering - Solutions Figure 5.8 depicts a power supply solution used for DLC systems. Batteries with 8 hours (typically) of reserve capacity power the RT and are charged from commercial power. If extended commercial power outages occur (e.g., greater than 8 hours), portable generators are connected to the RT to maintain power for telephone service However, this solution requires that the batteries be routinely monitored and maintained to their reserve capacity and that the number of RT sites be limited to what can be effectively powered with portable generators during widespread power outages.

Cornrnefcial Power

. Figure 5.8 DLC Powering

. Figure 5.9 shows the use of a power node for powering HFC systems with the power transmitted from the fiber node to the electronic equipment over the coaxial cable. In this case, a permanent generator fueled by natural gas is used for secondary power backup. Because there could be a relatively large number of power nodes, the permanent generator removes the need to supply portable generators during long-duration power outages. Also, with a permanent generator, batteries with lower reserve capacity (e.g., 1 to 2 hours) can be used. The power node concept can also be used for DLC. However, the reliability of the generators in an uncontrolled environment could be an issue, and the generators (and batteries) must be monitored and maintained. Practical issues also need to be considered, such as obtaining right-of-way for the power node and dealing with noise resulting from routine starting of the generator to ensure its operation.

18 EXHIBIT R- -(JF-w CASE NO. U-13796 D A C F I P A F d 6

Commercial Power

. Figure 5.9 HFC With Power Node

A number of solutions can also be used for powering FTTC with centralized or distributed power sources. For example, with a centralized power source, metallic condtuctors in parallel with the fiber are needed to transmit power.

Two other examples of solutions that are undergoing trial use or are deployed as follows:

Powering HFC fiber nodes by transmitting up to 450 volts (AC) from the central office power source over power conductors placed around the fiber optic cable. This solution makes use of the central office power and avoids the need to obtain rights-of- way and maintain remote batteries and generators.

Powering FITL equipment at a subscriber’s premises with local battery, solar power backup and equipment designed for low power consumption. This soluti,on helps to provide the advantages of fiber.optic transmission to subscribers in rural areas. Alternatives are being designed to meet local conditions and needs.

5.3.3 Path Forward Although the powering of the new wireline access technologies presents challenges, the industry is finding innovative methods to ensure continuation of telephone service during commercial

.power outages. Because the deployment of these systems is just beginning, additional methods will probably be developed.

<Recommendation 1> The industry should continue to work on innovative, cost-effective solutions for powering new wireline access technologies and should monitor the reliability of the solutions during field trials and early deployment.

EXHIBIT R- - ( ~ - 1 3 ) CASE NO. U-13796 PAGE I9 OF Ah

19

. .. . .

5.4 Hybrid FiberKOax Technology . .

HFC systems are being deployed by telephone,companies and cable television companies. For cable television companies, HFC networks are part of the evolution from tree and branch architectures, to fiber rich networks, to two-way, 750 MHz HFC networks, On the other hand, telephone companies have little or no embedded coax, amplifiers, etc. in the fe.eder and distribution portions of their access networks, but extensively usedfiber-based digital networks. Consequently, the NWAT subteam primarily obtained data from cable television networks and supplier test data to detekine potential failure modes that might occur in HFC networks. In a sense, HFC networks were benchmarked against cable television systems.

As discussed previously, the cable television network has typically been designed with tens of amplifiers in cascade, no power backup, and is designed to provide a broadcast entertainment video service. This service's reliability needs and service requirements are completely different from those of telephony services. In a competitive market, it will become necessary for the cable teievision networks to be upgaded to meet the reliability objectives of the competing telephony .'

service provider if similar services are to be provided. From a reliability perspective, fiber-rich CATV networks differ from traditional cable television networks in that they have:

more fiber

two-way, higher frequency amplifiers. fewer amplifiers in cascade (typically fewerthan about seven in cascade)

In addition, it is anticipated that HFC systems will incorporate one or more of the following:

As an example of the data collected on current cable television failure modes, Fiewe 5.10 summarizes data co1lec:ed from a state office of cable television based on a sampling of 800 trouble reports from 1994 and the first half of 1995. Although the data in Fi&e 5:lO does not provide failure rate information, it does show the types of failure modes thatmay be experienced in today's cable television system. Reporting criteria were based on 50 or more customers affected for 2 or more hours, or 500 or more customers affected for 1 hour or more.

surge protection for power supplies and amplifiers battery backup to keep the telephony services operational during commercial power outages operational support systems that constantly monitor the network and network elements components with higher reliability ( e g , amplifiers).

20 EXHIBIT R- -(JF-13) CASE NO. U-13796 PAGE 20 OF 4h

-g. 7-

. Figure 5.10 Cable Television Sample Trouble Report Data

Based on the data in Figure 5.10, as well as supplier and operator HFC system test data, the following potential failure modes of an HFC access network were categorized:

Commercial power outages Amplifier failures

Drop cable failures

Commercial power outage were addressed in Section 5.3 so are not discussed in this section. The remainder of this section addresses reliability attributes of HFC systems and where applicable, presents potential solutions for failure modes that have been tested, trialed, or deployed.

5.4.1 Fiber Xode Size Selecting the fiber node size (i.e., number of homes passed per fiber node) is a key design decision. In so doing, many factors must considered (e.g., installed first cost and life cycle cost, performance). However, within the constraints of keeping the system cost per subscriber within reasonable economic bounds, reducing the number of homes served from each fiber node can have the following benefits:

I ) Because the number of system amplifiers and line extenders is reduced, the single points of failure are reduced. 2) Failure group sizes are smaller. 3) Oppormnities for ingress are reduced.

In one example, when a fiber node serving area was reduced from 2,000 homes passed to 500 homes passed, the number of amplifiers in the distribution network was reduced fram 100 to 25. The number of amplifiers in a cascade from the fiber node to the most distant subscriber was

i. Ingress and impulse noise

' .

21 EXHIBIT R- -(m-13) CASE NO. U- 13796 PAGE 2 I OF 46

.. :

reduced from 7 to 4 and the maximum failure group for the fiber node was reduced from 2000 to 500.

Reduction of the fiber node serving area must be balanced with the increased infrastructure costs per subscriber that are incurred as the s e w h i area is reduced. Maintenance action rates must also be considered (e.g., increased number of fiber nodes versus decreased number of amplifiers).

5.4.2 Amplifier Failures The amplifiers in the fiber serving area can be distribution amplifiers, which allow a coaxial cable trunk line to extend farther into the serving area or line extender amplifiers, which extend branches €rom the trunk. The number of amplifiers in the fiber node serving area is correlated to the number of homes served. To improve network availability rates, the tendency has been to reduce the number of amplifiers in cascade (e.g., by reducing the number of hornes'served per fiber node). At the same time, improving the reliability of amplifiers used in the distribution plant should be considered. The results published by a major supplier of amplifiers to the CATV industry are shown in Figure 5.1 fully implemented field return monitoring program established in 1989. This figure illusbates that by aggressively correcting root causes of amplifier failures found in the first few years of the program, it is possible to si-gificantly reduce the failure rate. The illustrated amplifier failure rates apply only to actual hardw,are.failures. The amplifier failure rate in a network can be higherdue to factors such as lightning strikes, power surges (e.g., blown fuses), poorly implemented policies and prccedures, and craft error. In addition, these M u r e rates do not include older-generaticr! amplifiers. It is, therefore, recommended that the network operator and the supplier cooperate to identify the root causes of failures and to make the appropriate corrections.

According to the supplier, these failures rates are based on a

<Recommendation 2> HFC network operators and suppliers of CATV amplifiers should work closely together to identify the root causes of failures'of amplifiers and to take the appropriate remedial actions.

EXHIBIT R--(JF-13) CASE NO. U-13796 PAGE 22 OF 46

10% ( . . . ' I . - ' : - TrunkAmps -

Line Amps . - i --..__ \. '.

1- ' I

/ 1..

-. '1 -..I _. ----.

0.1% 1989 1990 1991 1992 !993 1994

Figure 5.11 Reduction History in CATV Amplifier Annual Failure Rate

Y

.

5.1.3 Ingress The existing HFC architecture and that of the near future, uses and will use the 5 MHz to 30/40 MHz band as the return communication band from the subscriber to the end office. This band is subject to ingress interference into the cable distribution plant from short wave broadcasts, amateur radio, CB radio, paging transmitters and maritime radio, as well as impulsive noise interference from home appliances, motors, lightning, and. automobile ignitions. Excessive ingress interference the HFC return plant will limit the capacity of services offered and degrade the quality of services offered. hextreeme cases, it could completely block an offered service.

Characterization shidies of the CATV return plant have been performed that ingress interference in the return plant is a potential impediment to anticipated service offerings requiring the use of the return plant. This is a very important issue because ingress interference directly affects the ability of the HFC network to offer reliable services using the return path.

The summary of available findings reported in Appendix B, suggests that a number of steps that can be taken to either reduce the levels of ingress in the return plant, or mitigate the effects of ingress on the service offered.

These shidies verifo

2 See Appendix 8 far a more detailed discussion of tests and results

23 EXHIBIT R- -(JF- 13) CASE NO. U-13796 PAGE 23 OF 46

1) Incorporate a comprehensive return plant maintenance program. This could include:

a) Tightening all loose connectors b) Replacing corroded connectors c) Sealing the distribution plant from water migration d) Examinating the distribution plant for cracked cable sheath e) Checking for proper bonding and grounding

. .

2) Segregate the.retum plant bandwidth and block that portion ofbandwidth allocated to telephony services against ingress from the home.

3) Provide frequency agility so that the return signal can avoid interfering carriers.

4) Use a robust modulation scheme such as Quadra-Phase-Shift Keying (QPSK), Code Division Multiple Access (CDMA) and Discrete Multitone (DMT), or include Forward Error Correction (FEC) as part of the modulation method.

<Recommendation 3> HFC network operators intending to deploy telephony services should implement a comprehensive return plant maintenance pro-pm before such services are deployed.

. . . . . . , . . .. . (Recocmendation 4> . . . .

Industry continue to focus on developing and implementing methods to alleviate the effects of ingress and impulse noise in HFC networks.

5.4.4 Drop Cable to Customer Premises The coaxial cable drop from the tap (either-in a pedestal or out on a aerial strand) to the customer premises has been identified as a potentially weak link in terms of reliability. This is an issue because of the relativeiy high frequency of trouble call reports attributed to the drop cable by CATV operators, This has been publicly reported, '"'and continues to be an issue, as noted by a recent trouble call breakdown by a major CATV system operator. This breakdown is illustrated in Figure j . i Z . As the data through August of 1995 shows, the cable drop was still the litl'gest contributor to trouble calls for this operator. This is a significant issue that can directly affect the ability of the HFC network to offer reliable services because the cable drop is a single point of failure in the HFC architecture.

24 EXHIBIT R- -(F-l3) CASE NO. U-13796 PAGE 24 OF 46

~

Cable Company "A" Trouble Call Breakdown Aug '94 to Aug '95

16.5%

. Figure 5.12 Recent CATV Trouble Call Breakdown

As noted, studies of the causes of CAT'; iioiibk ~ a i l s " ~ ' prove that t l i t drop cable and k s connectors are the major sources of trouble. Some of the reasons reportedare listed below:

1) Poor installation and maintenance practice including the following factors: Improper Connector Installation -- such as improper stripping of the coax, tightening of the connector shell to its mate, and lack of a weather seal over the connector. . Shallow cable burial - Deeper burial might avoid cuts due to typical lawn care operations.

.Burial without the protection of a containment tube. 2 ) Rodent Damage. 3) Cable Cuts. This can be from digging equipment for buried cables and from tree falls or' pole damage for aerial cables. 3) Tampering with the drop by the home owner.

There appears to be no inherent reason why the coaxial drop cable cannot be mademore reliable. It is believed that proper employee trainins and attention to installation detail will greatly alleviate the problem. Based on reports""'and subteam inputs, some potentially mitigating solutions that might be considered include:

1) Stripping of drop cable for installation of "F" connectors. 2 ) Creating torque specification in tightening the outer shell of the male "F" connector to its mate. 3) Use of nickel plated shells on the male "F" connector to mate with nickel plated screw threads on the tap or NIU.

EXHIBIT R- -(F-13) CASE NO. U-13796 PAGE 25 OF 46

4 ) Using flooded cable for underground installation. 5) Enclosing buried drop cable in conduits. 6 ) Supporting aerial drop cable with an integral steel messenger line. 7) Connecting of the drop cable to a properly installed grounding block at the NIU. 8) Using amored drop cable where rodent damage has been a problem

Some manufacturers of coaxial cable are now supplying a higher quality coaxial cable for drops. This new cable is essentially a scaled down version of the “hard-line” coax used in the distribution plant, which is generally considered to be more reliable than the coax used for drop cables.

<Recommendation 5> It is recommended that the industry should continue to explore improved-reliability coaxial drop connection technologies for HFC networks, given their current major role in subscriber outages.

<Recommendation 6> HFC network operators intending to deploy telephony services shouldestablish an employee t rahng program and policies and procedures that will ensure proper installation and .maintenance of the coaxial drop cable.

5.4.5 FibedCable Cuts . B;?ried 5ber an! coaxial cable is subject to damage from accidentz! dig ~lps. .Fiber/Cable cuts ha-!e contributed to a significant portion of network unavailability time. For example, Bellcore calculated an optical fiber cut rate of 4.39/year/1000 sheath miles!“’ This calculation was for both interoffice and access fiber. In an HFC system, most of the fiber will be access fiber only. Data is not available for the cut rate ofjust access fiber and coax. Using the Bellcore calculation, the annual failure rate due to fiber is 0.44 %/mile. For an average access fiber run of 6 miles and an average mean time to repair (MTTR) of 6 hours, an annual downtime estimate due to fibedcable cuts is 9.5 minutes. Fiber &d cable cuts remain an important issue in overall system reliability and merits review of past recommendations.

Because the ilber depioyment and maintenance practice for HFC will be ‘very similar (0 that used ’

by the standard wireline telephony industry, the discussion and recommendations of Section A. “Fiber Optic Cable Dig Ups: Causes and Cures” hivetwork Reliability: A Report to the Nution “*’ should also apply to optical fiber and coaxial cable used in HFC networks. To further reduce the outage time caused by fiber cuts, some form of alternate path routingmay be necessary. Route diversity in the form of unidirectional and bidirectional self-healing rings has been used in interoffice and long-haul telephone networks. This architecture, as well as other forms of route diversity, may be applicable to HFC access networks.

<Recommendation 7>

’

Recommendations from Section A, “Fiber Optic Cable Dig Ups: Causes and Cures’! in Network Reliabiliry: A Report to the Nurion, should be applied to optical fiber and coaxial cable in access networks.

26


...

5.4.6 Other Best Practices Following the Bellcore Reliability Assurance Practices for components, optoelectronic devices, and hybrid microcircuits (TR-NWT-000357~"' TR-NWT-000468,'"' TR-NWT-000930"'~ can serve as a key guide to improving fiber node and amplifier hardware reliability.

5.5 Fiber-To-The-Curb Technology Several suppliers have developed FTTC systems, several operating companies have had field trials or early deployment of the systems, and at least one operating company has decided to continue deploying the systems. Requirements for FTTC systems were first published in 1990. Siqce then, research and development efforts have addressed many issues, including reliability issues associated with FTTC. Consistent with the learning c w e typically associated with use of new technologies, the field trials and early deployments have identified further needed improvements, many of which were specific to a particular service provider's operations or a particular supplier's product. However, the existing deployments have demonstrated the importance of carefully monitoring field performance during trials or early deployment, doing a root cause analysis of problems, and taking corrective action. The remainder of this sectionidentifies reliability factors that should be considered for FTTC by comparing FTTC with an existing technology, DLC.

5.5.1 Potential FTTC Reliability Issues Similar to the approach used for HFC systems, one method of calculating the reliability of the new FITL technology is to benchmark the new technology to a deployed technology. By doing this, one can evaluate paiential failure modes thai rliight also occur in the new itchuology, as w4l as highlight differences that might improve or degrade the reliability of the new technology. Consequently, the NWAT subteam compared the FTTC architecture with the architecture of existing DLC systems and identified new characteristics of the FTTC architecture, including single points of failure, that could affect reliability. (Section 4 of this report describes DLC and FITL systems.)

Potential new reliability vulnerabilities relative to DLC can be identified by the single points of failure that are introduced in the FITL architecture. Single points of failure are important since traditionally they have the greatest potential of adding to network unavailability. For example, the FTTC architecture introduces unprotected hardware components such as optical components at the HDT and ONU, power supply and controller boards at the ONU, and ringing generators at the O W . The passive distribution network also includes optical splitters thatare a single point of failure. Although the splitters are passive, they are a relatively new technology and could be a source of reliability problems. The reliability of these single points of failure should be monitored during trials and early deployment. If there are high failure rates for any of these single points of failure, service providers should work with their supplier to determine and correct the root cause of the failures.

The ONU environment might be more uncontrolled than that of DLC, and ONUS will be deployed at relatively more locations. For these reasons, the environmental effects on the equipment, including ONU enclosures, should be monitored and any problems corrected before widespread deployment.


..a

..

As discussed in Section 5.3, power supply is an issue because FTTC might be powered differently from DLC, or might have added components in the powering network, such as the copper wire used to distribute power to the ONUS. Because of power loss in the distribution and the power consumption of the added electronics, FTTC could use relatively more power than DLC (e.g., for a given number of customers served), heightening the concemabout maintaining battery reliability. Added ONU equipment for broadband services, beyond telephone service, could also result in added power consumption. Integrating broadband services with telephony services during a power outage may add to the load incurred by the batteries if these services are not separable during a power outage. The effect of powering should be included in monitoring field reliability.

Effective methods, procedures, and training can be developed and tested during trials or early deployment. For example, two areas where procedures and training could increase the reliability of FTTC are fiber splicing, and bonding and grounding at the ONU. FTTC deployment will involve relatively more fiber deployment, and more fiber splices, than DLC deployment. Poor splices could become a reliability issue. Poor bonds and grounds could result in equipment and service failure due to voltage surges caused by lightning. Although splicing, bonding, and grounding can also be necessary for DLC, the relatively greater levels of fiber and equipment deployment can make these factors especially important for FTTC. Reliability problems due to inadequate methods, procedures, and training should be identified and corrected.

5.5.2 Path Fopvard Because of the learning curve associated with the deployment of new technologies, service providers should carefully monitor field reliability during trials or early deployment to identify the largest sources of failures, do a root cause analysis to identify the causes of the failures, and implement corrective action before widespread deployment. A well-designed study should be used to monitor field performance. The study should include a reliability objective for some measure (e.g., downtime or customer trouble reports) that can be used as'a benchmark for the field results. Efforts should also be made to emure that the scope of the field trial is broad enough and of sufficient duration to provide statistically significant results. The trial requires close cooperation between the network operator and the supplier to identify the root causes of failures and to make rhe appropriate corrections.

<Recommendation 8> Network operators, in cooperation with their suppliers, should establish a field reliability monitoring process during trials or early deployment as a means to identify sources of failures, do a root cause analysis, and implement corrective actions before widespread deployment.

5.6 Reliability Template This section presents a checklist of items that should be considered when deploying new wireline access technologies. Generally, the checklist covers the technology-related issues discussed in this report and is not necessarily complete because these technologies are only in the early stages of deployment. The itcms are separated into two broad categories: Architecture/technology/operations and Installation/maintenance.


Architecture/Technology/Operations

Does your company have a reliability objective for the access network? Have you done a reliability analysis of the architecture? Have you identified and addressed disaster recovery requirements (e.g., widespread power outages)? Have you done a reliability analysis of the systems and equipment to be used in the deployment? Have you reviewed the supplier's documentation and does it meet your criteria? What is the reliability of telephone service using this technology (e.g., compared withcurrent technology)? Have you identified the failure modes and the actions required to recover from each failure mode?

InstallationlMaintenance

Have you developed standard equipment configurations? Have you developed installation methods and procedures? Have you documented your installation acceptance procedures? Does your technology have capabilities for generating alarms? Have you developed methods and procedures for routine hardware and software maintenance? Does the technology provide for nonservice affecting s o h a r e change/maintenance ' .

capabilities? Do troubleshooting procedures exist, including procedures for fault visibility, trouble verification and isolation, and recovery and repair? Do methods exist for doing a post-mortem analysis of failures? Is there a process to feedback findings and recommendations to improve future reliability? Are there required craft training courses available in a timeframe consistent with the

Are new test sets required for this technology and will they be available in a time frame consistent with the deployment schedule

d tp lo~xzn: scheliile'? . . .. .

In addition, see Appendix D for a generic New Technology Reliability Template that can be used to assessment of the reliability of any new technology, including new wireline access technologies.

6. Summary of Recommendations

<Recommendation 1> The industry should continue working on innovative, cost-effective solutions for powering new wireline access technologies and should monitor the reliability of the solutions during field trials and early deployment.

<Recommendation 2> HFC network operators and suppliers of CATV amplifiers should work closely to identify the root causes of failures of amplifiers and take the appropriate remedial actions.

<Recommendation 3 2 HFC network operators intending to deploy telephony services should implement a comprehensive return plant maintenance program before such services are deployed:

<Recommendation 4> Industry should continue to fo.cus on developmg and implementing methods to alleviate the effects of ingress and impulse noise in HFC networks.

. .

<Recommendation 9 It is reccz!xended !hat the.indQstq should continue to explore irnproved-reli??.i!iv c 0 4 d drop connection technologies for HFC networks, given their current major role in subscriber outages.

<Recommendation 6> HFC network operators intending to deploy telephony services should establish an employee training program and k i t t e n policies and procedures that will ensure proper installation and maintenance of the coaxial drop cable.

<Recommendation 72 r(ecommenuaiiuns I’rurri Stxiioii A. “Fiber Optic Cable Dig Ups: Causes and Cures” in the Network Reliability: A Report to the Nation, should be applied to optical fiber and coaxial cable in access networks.

<Recommendation 8> Network operators, in cooperation with their suppliers, should establish a field reliability monitoring process during trials or early deployment as a means to identify sources of failures, do a root cause analysis, and implement corrective actions before widespread deployment.

30 EXHIBIT R--(JF-13) CASE NO. U-13796 PAGE 30 OF 46

.

7. Conclusions The new wireline access technologies subteam investigated primarily the reliability of telephony services transported over HFC and FTTC access networks. It was the intent of the subteam to investigate potential service reliability attributes (i.e,, weaknesses and strengths) associated with new wireline access technologies. Based on these investigations, the most significant, high-level findings are as follows:

The most significant, high-level recommendations are as follows:

Industry's goal is to provide 99.99% reliability for telephony services provided over HFC and FTTC access networks Deployment of HFC and FTTC systems is just beginning; therefore, gathering of critical field data is in its early stages. Industry has identified several key reliability issues and potential mitigating solutions.

Operators and suppliers. should implement a process to gather field data on systems as they undergo trials and are deployed. Operators and suppliers should institute a process for root cause analysis on outages and develop best practices to improve,reliability.

. . . Some of the more si-gificant detailed findings and recommendations are as follows:

Commercial powering of HFC and FTTC (as well as DLC) will not meet the current reliability objectives for key services unless backup power is provided. Operators are trialing and deploying several mitigating solutions. Operators and suppliers should continue to focus on evaluating and implementing solutions to reduce the impac't of commercial power outages on network reliability. Ingress and impulse noise can impair the ability of an HFC system to provide reliable, two- way services. Industry has identified several mitigating solutions, including system and network design, and network maintenance. Industry should continue'to focus on developing and implementing methods to alleviate the effects of ingress and impulse noise. Coaxial drop cable failures continue to represent a si-gificant portion of trouble calls in the cable television industry. Service providers that are planning to use coaxial drop cable must continue to consider potential ways of improving drop reliability, some of which are summarized in this report.

Additional detailed finding and recommendations are discussed throughout the text of the report.

8. Acknowledgments The NWAT subteam would like to acknowledge Ken Young's help and thank him for his support and encouragement in developing this report.

3 1 EXHIBIT R- -(JF- 13) CASE NO. U-13796 . PAGE 3 1 OF 46

9. References 1 . ANSI TI .40 1- 1993, fnterface Between Carriers and Customer Installations -Analog

Voicegrade Switched Access Lines Using Loop-Start and Ground-Start Signaling. 2. Bellcore Technical Reference TR-NWT-000057, Functional Criteria for Digital Loop

Carrier Systems, Issue 2, January 1993. 3. Bellcore Technical Reference TR-TSY-000008, Digital Interface Between the SLC@-96

Digital Loop Carrier System and a Local Digital Switch, Issue 2 , August 1987, plus Revision 1, September 1993, plus Bulletin 1 , October 1994.

4. Bellcore~Generic Requirements GR-303-CORE, Integrated Digital Loop Carrier System Generic Requirements, Objectives and Interj$ace, Issue I , September 1995.

5. Bellcore TR-NWT-000418, Generic Reliability Assurance Requirements for Fiber Optic Transport Systems, Issue 2, December 1992.

6. Bellcore Technical Reference TR-NWT-000332, Reliability Prediction Procedure fo r Electronic Equipment, Issue 4, September 1992.

7. Bellcore Technical Reference TA-NWT-000909, Generic Requirements and Objectivesfor Fiber In The Loop Systems, Issue 2, December 1993.

8. Allen L. Black and James L. Spencer, “An Assessme.nt of Commercial AC Power Quality: A Fiber-In-The Loop Perspective,” MTELEC ‘93 Proceedings.

9. Chuck Merk and Walt Srode, “Reliability of CATV Broadband Distribution Networks for Telephony Applications- Is it good enough ?,” 1995 NCTA Technical Papers, pp. 93-107.

10. Brian Bauer, “In-Home Cabling for Digital Services: Future Proofing Signal Quality and Minimizing Signal cju~Ag%,’‘ 1995 SCTE Conrerence un Emerging Techologies, pp: 35-i0G:

11. D. S. Kobayashi and M. Tesfaye, “Availability of Bi-Directional Line Switched Rings”, Report T1 X1 .j/91-070, April 199 I .

12. Network Reliability: A Report to the Nation, June 1993, National Engineering Consortium. 13. Bellcore Technical Reference TR-NWT-000357, Generic Requirements f o r Assuring the

Reliability of Components Used in Telecommunication Systems, Issue 2, October 1993. 14. Bellcore Technical Reference TR-NWT-000468, Reliability Assurance Practices for

Optoelectronic Devices in Central Ofice Applications, Issue 1, December 1991. 15. Bellcore Technical Reference TR-NWT-000930, Generic Requirements for Hybrid

Microcircuits Used in Telecommunications Equipment, Issue 2, September 1993. 16. CableLabsa, “Two-Wai Cable Te.levision System Charactenzation,” Final Report, April 12, ’.

17. Mark Dzuban, “AT&T Ingress Study,” Submitted to IEEE 802.14-941003 November 1994. 1995.

32

EXHIBIT R- -(E-13) CASE NO. U-13796 PAGE 32 OF 46

.: . .

10. Appendices

Appendix A - Overview of an Example Broadband HFC Architecture The Hybrid Fiber Coax (HFC) architecture has many forms and variations. No industry standard, exists but there are basic components common to all HFC networks.

1) An optical fiber feeder from a fiber hub to a fiber serving area (FSA). The fiber feeder terminates in the serving area in a fiber node. 2) The fiber serving area covers from about 100 to 2000 homes served depending on the application. 3) The fiber node feeds the fiber serving area using a coaxial tree and branch type network.

Fiber Feeder

Fiber Hub

Figure A.l Generic HFC Architecture

The fiber hub could be an end office feeding multiple fiber nodes or a remote hub feeding a single fiber node. For a telephony application, the Host Digital Terminal (HDT) could be located in the fiber node, the end office, a remote fiber hub, in a serving office feeding regional hubs, or in an operations center feeding multiple serving offices. A possible high-level depiction of a regional backbone where the HDT could be located in any one of the circles is shown in figure A.2 on the next page. For overall system reliability analysis, the total network between the HDT location and the Fiber Node must be taken into account, but in most implementations the HDT will be in the End Office or Fiber Node.

33 EXHIBIT R- -( JF- 13) CASE NO. U-13796 PAGE 33 OF 46

OC = Operations Center SO = Sewing office €0 = End Office FN = Fiber Node I5O Homes 150 Homes

150 Homes

. Figure A.2 An Example Regional HFC Network

For purposes of analysis, the network can be se,mented into a series connection of functional blocks starting from the insertion of the telephony signal into the HFC network to the W1 I interface at the in-home telephone network.

1) Backbone Network 2) Serving Office 3) Fiber Network and Fiber Node 4) Coaxial Distribution Network 5 ) Network Interface Unit

The backbone network between the Operations Center and the Serving Office can take many forms depending on the total network architecture of a particular system. In some cases the operatinns center will be co-located in the Serving Office so that there. is no backbolie netwoik. If a backbone network exists, the transport path from the Operations Center to the Serving Office must be accounted for in overall system reliability calculations. In most cases, this path will be a part of a regional ring or multiple ring network. These rings are often set up as either unidirectional or bidirectional and “self-healing”; that is, a single break in the path docs not result in a loss of service. The Serving Office may contain the Host Digital Terminal (HDT), passive combiners, amplifiers and optical transmitter.

The fiber network consists of the optical fiber bundie (usually containing 4 to 12 individual optical fibers) between the Serving Office and the Fiber Node. The optical bundle can branch off in the network to serve more than one Fiber Node so that a fiber bundle cut could affect more than one Fiber Node depending on where the cut occurred: The Fiber Node consists of the downstream optical receiver and distribution amplifier as well as the upstream laser transmitter. It is possible for the HDT to also be located at the Fiber Node. A power supply can be co-located with the


. .

Fiber Node to provide power to the Node and to other active elements in the coax distribution plant.

The coax distribution network consists of the coaxial cable, bidirectional amplifiers, any additional power supplies, and passive elements such as splitters, directional couplers, filters and taps. This network today uses the tree and branch construction technique, which is a carryover from all coaxial cable networks. For a large Fiber Serving Area (about 500 homes or greater),this will probably still be the method of choice. Any failure in a non-redundant element of this network between the Fiber Node and the home served will result in a loss of service to one or more homes depending on where the element failure occurred. Smaller Fiber Serving Areas could use a star type network, which could further limit the number of homes affected by a single-point failure.

The Network Interface Unit (MU) provides the interface between the coaxial distribution network and the in-home telephone network. It will usually be placed on the outside of the home but could be inside for multiple dwelling units. To provide lifeline type services, the NIU will often be powered from the network either through the drop cable or through a separate twisted pair that is a part of the dzop cable protective jacket.

The fiber node is a key element in the system. A failure in the fiber node cancut service to aU of the homes served by that node. The fiber node contains the components for downstream and upstream operation. A block dia-mm showing several of the major components in a typical fiber node is shown in Figure A.3. Failures in the optical receiver, return path laser, fiber node power supply, or in amplifiers #1 and $2 affect service to all homes served by the node. A fiber nodewill typically have multiple coaxial ports to feed coaxial hunk lines into the serving area. Failures in any one of the amplifiers feeding these four ports will affect only those homes served from a failed port.

EXHIBIT R- -(F- 13) CASE NO. U-13796 PAGE 35 OF 46

35

Figure A.3 Typical Fiber Node Block Diagram

. . .

36 EXHIBIT R- -(JF-13) CASENO. U-13796 PAGE 36 OF 46

I

Appendix 6 - Ingress Reliability Issues in HFC A c c e s s Networks

B.l Introduction The CATV industry uses the 50 MHz to 450 MHz band for downstream analog video broadcasts. New downstream services are expected use between 550 MHz and 750 MHz in the near term and up to I GHz in the future. The most convenient band for upstream services (from the consumer back to the serving office) is the 5 to 40 MHz spectrum below the downstream band.

One of the issues associated with the HFC architecture is that the return path below 50 MHz is very susceptible to ingress interference from many sources. These include short wave broadcasts, amateur radio, ISM (Industry, Scientific, and Medical) equipment, CB radio, and many possible interference sources from within the home. The entry points of ingress into the return plant are through loose or corroded connectors, improper grounding, improperly sealed network elements (ampiifiers, couplers, taps, etc.), damaged cable drops, and worst of all ingress injected into the return plant from improper or tampered with in-home cabling. By the nature of the coaxial distribution network, all interference sources entering the network in a fiber serving area are funneled into the return fiber. In some cases, the outputs from several fiber serving areas are combined into one return fiber. All of these factors increase the effect of interference signals leaking into the return plant.

B.2 Ingress Studies

Because of the potential of this interference affecting the reliability of upstream communication from the consumer, ingress interference in the cable plant has been studied by CableLabsB, AT&T, and Motorola (among others). The CableLabsB report'"'on characterization of the CATV cable plant is a comprehensive study of five different cable systems accomplishedover about a one year period. The following is a very brief summary of the report:

I) Five separate cable systems were studied. All five systems had significant return plant problems that had to be corrected before meaninghi test3 couid be made. in two of tine systems, return plant problems were never completely fixed.

2) For short-term ingress measurements, the test signal for both upstream and downstream tests was a QPSK carrier modulated with a T1 data stream with no error correction. An HP test set was used to record all the error statistics. All the downstream tests on all five systems met the (3.82 1 end-to-end objectives for Errored Seconds, Severely Errored Seconds, Degraded Minutes and percent availability. The long-term bit error rate in the downstream was in the lo7 area with a few error bursts exceeding 106 during a 48 hour test run. The results in the upstream were much worse than in the downstream. For the most part (3.821 ( i t . , an ITU recommendation for bit error objectives on digital links) objectives were not met. Long-term BER was in the IU5 area with numerous error bursts in the lo-' to 10.' area. Some of these error bursts lasted for many seconds. There was no correlation between these severe error bursts in the upstream and error activity in the downstream. CableLabsB recommendations, based on their report, included the followmg:

37 EXHIBIT R- -(JF-13) CASE NO. U-13796 PAGE 3 1 OF 46

a) Plant Segmentation -- Reduce the serving area. b) Maintenance -- Must have a comprehensive and thorough plant hardening effort and precise return plant gain alignment. Fix loose and corroded connectors, splices and terminators, cracked cable sheath, poor grounding, and water migration. c) Incorporate a status monitoring system d) Control the level of (Impulse Pay-Per-View) IPPV carriers e) Clean up the subscriber drop

The AT&T study ""reported essentially the same ingress levels and included estimates of outage time for QPSK and 16 Q A V modulated camers in the 5 to 40 MHz return band. By current wireline telephony standards, outage time for the system tested was unacceptable for the lower half of this band, even with Forward Error Correction (FEC) used as part of the modulation scheme.

Motorola has also gathered ingress data from 10 different serving areas in 4 different CATV systems. The results of these studies are discussed in the next sections. It is important to note that these are existing CATV systems not upgraded to carry telephony services.

B.3 Long-Term Ingress Studies L m g t e m ingress is charxterized by ingress events th2.t lest frnm seconds to hours. Individual ingress camers are usually confined to a narrow spectral line (25 kHz bandwidth or less) in the 5 to 40 MHz band. The magnitude and frequency of these events vary with the time of day,.the season, and the sun spot cycle. Motorola has also found that the magnitude of these ingress carriers is strongly affected by the return plant maintenance policy of the cable operator. Shown below is the typical ingress levels from two different HFC systems, the first with 959 subscribers in the fiber serving area and the second with 7000 subscribers in the fiber serving area. As can be seen from the graphs in Fi,pres B 1 and B2, a system with a well-maintained return plant has substantially lower ingess levels than the system without such maintenance, even though the fiber node serves seven times the number of subscribers. In both cases, the scans represent incidents of peak aciiviry. . .

38 EXHIBIT R- -(JF-13) CASE NO. U-13796 PAGE 7 % OF 46

CATV System "A".

I 40

5' 30 E

20

p 10

- 0

m V - m U

m l J m

2

.-

-10 0 5 10 15 20 25 30

Frequency (MHz)

Figure B.l An Ingress Scan From a Poorly Maintained Return Plant .

Frequency (MHz)

Figure B.2 An Ingress Scan From a Well-Maintained Return Plant ,

B.4 Short-Term Ingress

Short-term ingress to the cable return plant is characterized by impulsive noise eventsthat are quite short in duration (nanoseconds to milliseconds in time span) but can have very high amplitudes. This could be noise from any number of sources including electrical storms, motors, vehicle ignition noise, and electrical appliances in the home. Motorola has taken short-term

39 EXHIBIT R- -(JF-l3) CASE NO. U-13796 P A r . 0 i n n c A L

ingress measurements from four different CATV systems at three different operating companies and found the characteristics of short-term ingress to be very similar from site to site and system to system. Some of these results are shown in the graphs that follow.

Figure B.3 represents a continuous record of impulse power levels recorded every 2.6 microseconds in a 500 lcHz bandwidth. A return band monitoring frequency was selected that showed no evidence of long-term ingress carriers, so any level recorded above the background noise was assumed to be impulse noise. Recording times varied from 2 hours to 20 minutes. The distribution of impulse power level from four different sites is remarkably similar. Note that high level impulses, although infrequent, were recorded from all sites. For example the curves show that in a one hour period between 2700 and 5400 impulses greater than 30 dBmV. in power level were recorded. This data can be displayed in a way more meaningful to the performance of a data communication system. This is shown in Figure B.4. Here a QPSK return path modulation is assumed with no error correction. The figure shows the expected average BER due to impulse noise versus the return path carrier level. To’stay within the linear dynamic range of return path lasers and amplifiers, the return carrier levels will be limited to the 20 to 30 dBmV area. Looking at Figure B.3, this means that average BER rates in the 3 x lo5 to 3 x lo6 range can be expected without Forward Error Correction (FEC) and can be improved significantly with FEC. (These measurements were made using individual channel bandwidths of 500 Mz. Individual channel frequencies were between 15 Mhz and 28MHz.)

Cumulative Impulse Level Distributions

--30 -20 -10 0 10 20 30 40 50 Impulse Power Level (d8mV)

. Figure B.3 Distribution of Impulse Levels

40 EXHIBIT R- -(m13) CASE NO. U-13796 PAGF An n F AA

. . ..,. ..* 7- . .. . . . . . :..

.... . ...*' . .

Simulated Bit Error Rates for a QPSK Signal usinq a Differential Detector

-1 0 0 10 20 30 40 50 Carrier Level (dErnV)

Figure B.4 Average BER due to Impulse Noise Versus Carrier Level

Figure B.5 shows the cumulative distribution of impulse widths for thee of the sites. A low measurement threshold was used at one site and a fairly high threshold was used at the other two sites These c w e s indicate that 93% to 98 % of recorded impulses were 10 microseconds or less in duration.

41 EXHIBIT R- -(JF-13) CASE NO U-13796 D A P . ~ n i n o A L

impulse Widths above

- Operator B: Threr.-5.5 d8mV

Op. C. Node 2: lhfes.=18 dBmV

- ___. op . C. Node X lhres.=ll d8mV

10 100 1000 Impulse Width (peconds)

1

. Figure 8.5 Cumulative Distribution of Impulse Widths

From it's ingress measurement activities, Motorola concluded the following:

1) A well-planned return path maintenance policy is needed to keep the levels of long-term ingress carriers under control. 2) Most of the severe ingress occurs in the 5 to 15 MHz region of the return spectrum 3) The frequency and duration of impulsive type noise do not vary significantly from system to system.

42 EXHIBIT R- -(JF-13) CASE NO. U-13796 PAGF A? nn A L

Appendix C - Issue Statement Issue Title: Reliability Concerns Arising Out of Changing Technologies Author: Gary Handler

Bellcore

Problem Statementhsue to be Addressed

The national Public Switched Network (PSN) which is truly a network of networks, has the deserved reputation of providing its users highly reliable, survivable and secure end-to-end services. The FCC and its Network Reliability Council (NRC) want to ensure that this remains the standard mode of operation in spite of a dramatic increase in the number of new technologies being deployed, the implementation of advanced new services offered to the public, and the emergence of a proliferation of new service providers. In specific,’the NRC will study a) the reliability aspects of the provision of key services over new network facilities, ( k , broadband hybrid fibedcoaxial cable distribution, SONET and ATM, wireless, and sateliirej, and b) reliability concerns arising out of new technology providing expanded services over new or traditional facilities, i.e., Advanced Intelligent Network (AN) capabilities. The emphasis of this FocusTeam should be on new technology that will be implemented in the public network within the next three years.

Areas of Concern and Problem Quantification

The’ following are the main areas of concern:

I.

. . . : . . . . - . .

Reliability Aspects of Provision of Key Services Over New Network Facilities a) Broadband Network - One concern about new network technologies is how the

reliability of services such as plain old telephone service provided over new broadband networks will compare with that of the same service provided over existing wireline technology. These new systems should be modeled and analyzed for potential reliability risks and possible reliability improvement techniques. Implementation “Best Practices” should be developed and a plan for their dissemination and implementation should be derived. Two specific areas shimid be adckesued:

b) Hybrid Fiber/Coarial Cable Distribution Systems - This technology is expected to be providing telephone service shortly. The reliability issues with this technology need to be defined and addressed.

c) SOiVET Facilities and ATM Technology - SONET transport and ATM technology are rapidly progressing and will be providing new broadband services as well as existing narrowband services over common facilities. The reiidbility issues with these technologies need to be defined and addressed.

d) Wireless Network (Cellular and PCS) - Another example of a concern about new technologies is the role and reliability of cellular facilities in connection with line-based networks. This issue was discussed by the NRC at its September 30, 1992 meeting and in the document Network Reliability; A Report to the Nation. The reliability of the telecommunications services provided over a combination of new technologies has to be reviewed. Customers who rely on cellular technology need service providers to have and follow established “best practices.” These do not now exist. Best practices

. . ’ ’

EXHIBIT R- -(JF-l3) CASE NO. U-13796

43

n. -- 1 “ ^- .I

for Personal Communications Services (PCS) and Networks should also be considered in this study.

e) Satellite Network - Another area of reliability concern is the provision of telephone services over new satellite technology networks such as low earth orbiting satellites. The reliability issues with this technology should also be defined and addressed.

Reliability Concerns Arising Out of New Technology Providing Expanded Services over New or Traditional Facilities, i.e., Advanced Intelligent Network (AIN) Capabilities - Concerns have also been raised regarding the interoperability and reliability of multiple advanced intelligent services with their inherently independently developed software management and control. As John Clendenin stated at the July 6, 1994 NRC meeting “this is not the kind of problem that could be solved (once) and laid aside”. However, to provide a near term objective from which a model or process might be developed, it is suggested that the team focus on the interoperability and reliability concerns in the development of Advanced Intelligent Network Services.

11.

111. Description of Proposed Work

The team working this issue should consider the following total quality process to identify reliability concerns arising out of changing technologies, quantify network vulnerabilities, identify the major reliability issiles and propose problem solutions.

!. Identify the new techqologies being introduce4 i.nto the network

2 . Collect appropriate data from all available industry sources to determine andor c o n f m areas/technologies of greatest criticality and risk, and those with the greatest potential for network reliability improvement potential. (Work with the ATIS Network Reliability Steering Committee (NRSC) and its Network Reliability Performance Committee to coordinate data collection activities).

3. Collect data from the industry concerning thereliability of new technologies if already deployed. (Work with the ATIS Network Reliability Steering Committee (NRSC) and its Ndwork Reliability Fe r ikmnce Committee to coordinate data coilection aciivitiesj

4. Perform sufficient analysis of the data to determine the root cause(s) of the problern(s).

5. From the root cause analysis determine an appropriate action plan to reduce/eliminate the possibility or severity of failures in high risk areas. Also consider ways that recovery procedures may be implemented more quickly or efficiently.

6 . Determine industry “best practices” for dealing with the root cause analysis fmdings and share this information with industry participants as soon as possible. Deployment should consider costbenefit tradeoffs of “best practices.”

7 . Develop a timeline and metrics to measure the effectiveness of the team’s recommendations.

44 EXHIBIT R- (JF- 13) CASE NO. U-13796 DAcn “ * -- .,

. . : f

. .

8. Consider the following tacticsiideas offered by the Steering Team as potential means to supplement the total quality process and address the findings of the root cause analysis. These represent ideas from the Steering Team that we want to share.

A. New Technology Reliability Template - Design a generic template that serves as a reliability screen for assessing the reliability of new network technologies. This could be ,used as a process for the rapid and reliable evolution of the telecommunications networks.

B. Provision of Key Services Over New Network Facilities

1. Broadband Networks (Hybrid FiberKoaxial Cable Distribution and SONET Facilities & A l M Technology), Wireless Networks (CeIIuIar & PCS), and SateIIite Networks.

a) For each technology, determine the scope of the reliability study. Develop a bounded definition of the reliability problem; for example, the provision of basic telecommunications over a new broadband hybrid fiber/coaxial cable distribution network.

b) Construct an order of rna-dtude (major failure modes and vulnerabilities) reliability model of a reference system for each technology.

c) Collect available reliability data (e.g. ciment coaxial cable systems network outage & failure data, current cellular network outage and failure data, current SONET network outage and failure data and ATM switch reliability ), concerns and “best practices” associated with each technology.

d) Analyze data to quantify reliability and determine the most significant problem areas, and the areas with the greatest risks.

e ) Determine applicability of current “best practices” to the new technology and identify any additional “best practices” that describe quality as part of the introduction of new technologies (i.e., “best practices” applicable to hybrid fibedcoauial cable networks, cellular networks, and SONET networks).

t) Recommend implementation strategies for “best practices” and on-going process information for insuring continued quality.

, .

2. Advanced Intelligent Network (AIN) Capabilities 9 ) Determine the reliability issues associated with AIN services (e.g., management

of many different versions of software). h) Identify efforts taken to date to address AIN reliability issues and to ensure

A M service re1,iability. Identify existing “best practices.” i) Identify potential reliability “holes” or problem areas and recommend solutions. j) Identify the role that the IITP process might play as part of an implementation

strategy for interoperability control and as a reliability qualification process for new AIN platforms, services and software. (Coordinate potential overlapping interconnection issues with the Network lntercomection Focus Team)

45 EXHIBIT R- (JF-13) CASE NO. U-13796 n A P Y . . . - - - ~

. . . . . Existing Work Efforts -.

There are several work efforts that have.addressed or are addressing some of these issues. The Fiber Cable Focus Team recommendations in the Network Reliability: A Report to the Nation, the Telecommunication Industry Benchmark Committee (TIBC) Report, Draft Congressional Bills S2101 and HR4394 on one-call legislation, and the ATISMRSCAnnual Report provide significant data from which to begin to address the Provision of Key Services Over New Network Facilities issue. The ATIS Working Group~on Network Survivability Performance, TlA1.2 and the News Release, DA-1343, requesting comments on Joint Petition ‘for Rulemaking on Cable Television Wiring, RM No. 8380, November 15, 1993 provide background on the cellular and coax cable concerns. The Switching Systems (focus on software) Focus Team Recommendations in the Network Reliability: A Report to the Nation as well as ATISMOFIIITP charter and test plans give good background material for addressing the services and software concerns:

Recommended Team Leader Ken Young - Bellcore

46 EXHIBIT R- -(JF-13) CASENO. U-13796 ” A c n n J AP n I


Verizon Network Ready For Mother's Day Call Surge May 8,2003 Previous I Next

The industry and the technology used to transmit calls have changed dramatically, but one thing in telecommunications has remained unchanged for a century: sons and daughters will call Mom on Mother's Day, making it the busiest Sunday of the year on Verizon's network.

Last year, our network handled nearly 1 billion local and long-distance calls on Mother's Day - about 12 percent more than an average Sunday. The normal Sunday volume on the Verizon network is 872 million calls, while average weekday volume is about 1.5 billion calls.

"As we go about the business of making sure our network serves customers 'round the clock with a reliability factor of 99.99 percent, it's nice to stop for one minute and think that what we're doing is making sure these very important calls go through on Mother's Day," said John Bell, Verizon senior vice president-Network Operations. "We're proud that our employees as well as our local and long-distance network are helping sons and daughters stay connected with their mothers."

Bell said that even with the surge in calls on Mother's Day, the Verizon network is engineered and built to handle the added volume, and is supported with backup electrical power if an unexpected interruption occurs at Verizon's central switching offices.

"Since the number of calls we're expecting to handle this Sunday still doesn't approach normal weekday call volumes, a son or daughter who says they tried to call and couldn't get through because of network congestion doesn't have an excuse,'' Bell added. "Whether it's a local call or a Verizon long-distance call on Sunday, we have the capacity to handle it."

Sons or daughters looking for last-minute shopping ideas and local gift delivery can call a helpful Verizon Livesource operator on 41 1 and find a specific florist, candy store or other shop in the same area where their mother lives. In addition, sons and daughters can also visit Verizon SuperPages.com at www.supemages.com and our Verizon Wireless Web site at www.verizonwireless.com to search gift ideas.

EXHIBIT R- (JF-14) CASE NO. U-13796 PAGE 1 OF I


N e w s R e l e a s e

For more information, contact: John Britton 415.537.3360 Phone

SBC PACIFIC BELL CONTRIBUTES $9 BILLION TO CALIFORNIA ECONOMY DURING 2001

Year-End Report Reveals Combined Impact of SBC’s Business Operations, from Employment and Network Investments to Charitable Contributions and Volunteerism

San Francisco, January 9, 2002 - SBC Communication’s Inc.’s family of

companies, including SBC Pacific Bell, has long served California as a dedicated

neighbor, sharing in the community’s challenging and prosperous times alike. Last

year, despite the weak economy, SBC continued its partnership with the state by

investing more than $9 billion in California through a variety of business, economic

development and philanthropic initiatives.

SBC’s commitment to California during the past year includes: = More than $2.4 billion in technology infrastructure improvements to offer a

greater array of services to consumers, including high-speed data;

More than $21 million in community investments, including corporate,

charitable and employee contributions to initiatives focusing primarily on

technology access, economic development and education;

* Nearly $3.4 billion in taxes paid and employee payroll;

9 $3.4 billion in purchases with California businesses;

9 Nearly 975,000 hours of volunteer efforts through the SBC Pioneers.

“SBC’s commitment to supporting the communities in which we live and

operate is more important now than ever before,” said Ray Wilkins Jr., president and

CEO, SBC Pacific Bell. “In this challenging economic climate, we must ensure the

continued development of our communities and delivery of excellent customer

service. We will continue to be there for our customers and neighbors.”

1

-more-

EXHIBIT R- -(JF-15) CASE NO. U-13796 PACF 1 no i

SBC Invests in California/Page 2

Investing in California Residents

Part of SBC's investment throughout the state is in its people. SBC

companies employ nearly 57,000 California residents, with a state payroll of nearly

$3 billion. Furthermore, SBC companies give more than $400 million to critical state

programs through annual taxes paid.

Investing in the California's Technology Future

As part of SBC's ongoing network infrastructure investment program,

designed to enhance service to all residents, the company invested more than $2.4

billion in its central offices and surrounding telecommunications infrastructure in

2001. By investing in regional technology, SBC provides the latest in

communications capabilities, equipping the states and residents for success now

and in the future.

During 2001, the California energy crisis and numerous rolling blackouts

caused the loss of commercial power at more than 50 key central offices.

Redundant, back up power sources engineered into the SBC network functioned

flawlessly, providing California residents and businesses with continuous service.

The company also quickly invested nearly $3 million in back up generators to ensure

that key call centers would remain open for customers, despite any blackouts.

When fast moving backcountry fires ravaged Central California, SBC crews

replaced more than 25 miles of telephone poles, phone cables and fiber optics.

Throughout these, and other weather-related emergencies, SBC continued to

maintain dial tone reliability of greater than 99.99 percent in a network of more than

18.5 million access lines.

Investing in California Communities

SBC gave more than $21 million to charitable organizations and important

community initiatives statewide through the SBC Foundation, corporate

contributions, employee donations and other community investments in 2001.

-more-

EXHIBIT R- (JF-15) CASE NO. UX796 D * P-0 -2- *

SBC Invests in California/Page 3

Through its philanthropic outreach and efforts of SBC's nearly 200,000 strong

volunteer organization, the SBC Pioneers, the company stepped up in 2001 to serve

the region in times of greatest need.

The SBC Pioneers donated nearly 975,000 hours of their time throughout the

.state, equating to roughly $15 million in sweat equity. Pioneer activities included the

Safe Connecfions program, which teaches children the basics of calling 91 1 and

helps them remember the proper way of using the emergency number, and the A

Book About ME! program, which distributed personalized books to kindergartners to

foster self-esteem. The programs reached thousands of students.

Nationally, SBC, through corporate and Foundation giving, employee

donations, and other investments, contributed more than $1 10 million to community

organizations in 2001. More than $65 million of this was through SBC Foundation

grants.

In the aftermath of the Sept. 11 disaster, SBC contributed $1 million to the

New York Times Fund and another $1 million to various relief organizations by

matching dollar-for-dollar the contributions made by the company's more than

390,000 employees and retirees. Through a $250,000 donation to the Veterans of

Foreign Wars Operation Uplink program at the end of 2001, SBC provided more

than 30,000 calling cards to connect current military service members serving in

Operation Enduring Freedom and hospitalized veterans with their families during the

holiday season.

SBC Communications Inc. (www.sbc.com) is one ofthe world's leading data, voice and lnternet services providers. Through its world-class network and its subsidiaries' trusted brands - SBC Southwestern Bell, SBC Ameritech, SBC Pacific Bell, SBC Nevada Bell, SBC SNET and Sterling Commerce - SBC companies provide a full range of voice, data, networking and e-business sewices, as well as directory adverfising and publishing. A Fortune 15 company, America's leading provider of DSL high-speed internet sewice, and one ofthe nation's leading Internet Service Providers, SBC companies currently serve more than 60 million access lines nationwide. In addition, SBC owns 60 percent of America's second largest wireless company -- Cingular Wireless -- which serves more than 21 million wireless customers. internationally, SBC has telecommunications investments in 28 countries.

The SBC Foundation is the charitable giving arm of SBC Communications Inc. and its family of companies. In 2000, SBC, through Foundation and corporate giving, donated more than $95 million to supporl efforts that enrich and strengthen diverse communities nationwide. The Foundation places primary emphasis on supporfing programs that help increase access to information technologies; broaden technology training and professional skills development; and effectively integrate new technologies to enhance education and economic development -- especially for underservedpopulations. SBC has been named among America's Most Generous Comoenies for two consecutive years by Worth magazine (2000 & 2001)

###

EXHIBIT R- (JF-15) CASE NO. U-13796 -

P


The Metamorphosis of the Telephony Network

Executive S u m mary

The deployments ofvoice over Packet (VoP) technologies are quickly reaching an inflection point. The industry has moved from low-scale toll bypass deployments to large-scale competitive carrier deployments. Within the next year, we expect to see large-scale deployment by incumbents worldwide. The purpose of this report is to explain the technologies, the driving force behind deployments, and present a view of the timing of those deployments by market segment. Specifically we show:

. VoP deployments are not just a cost-saving exercise; they can drive revenue growth and retention. Carriers with existing data infrastructure can implement services using this technology and quickly improve their bottom line. . Thus we believe that service providers with existing data infrastructure are the early adopters of packet voice, layering voice services over their network where they already have data presence. We expect this theme to continue for all market segments. . We do not believe that the lower cost of operation of next-generation voice equipment will justify replacement of existing legacy equipment at current service price points. This may change if prices for long distance services decline precipitously. . We believe that a service provider’s entry into offering VoP services is completely determined by the location of its existing data infrastructure and that the market can be seg- mented this way. For example, Incumbent Local Exchange Carriers (ILECs) have existing access data infrastructure4igital subscriber line services. We expect that the first VoP services offered by these carriers will be local voice services. . The most formidable barrier to the deployment of VoP technologies is quality of service. What we call the convergence layer-IP and ATM routers and switches-is primarily responsible for qualirj of service. We believe companies that play in this space will be significant beneficiaries of the deployment of packet voice technologies. . This is expected to be a large market: almost $9.8 billion in 2004. However, it is currently difficult to play with few pure play companies being public. Today’s market leader is clearly Sonus Networks, Inc. (Nasdaq: SONS; Buy-Aggressive; $4.50), with 88% of the large scale, carrier class VoP gateway market in the second quarter of 2001. However, there are many promising private companies targeting this space, and we expect that Sonus will begin to see more competition as more market segments begin to adopt VoP technology.

Michael R. Brown

[email protected] (612) 313.1211

December 10,2001

Stephanie Roscoe (612) 313-1291

[email protected]

EXHIBIT R- -(JF-16) CASE NO. U-13796 ,. . r,T. 1 r,- _.

The Metamorphosis of the Telephony Nehvork December 10,2001

Exhibit 20: Operation of Currently Available Telephone Features . . .

Can Forwarding Sen expbnatory * 72 '73

70 Prevents phone from ringing during Dredefined times

Do Not Disturb

Selective Call Forwarding Sen descriptive

Source. bes t

'63entef ,03 numbers

QoS: The Big Barrier In our view, quality of service is the single most important and difficult barrier facing the adoption ofVoP technology. Anyone who uses both wireless and wireline telephones has personal experience that supports this view. The poor voice quality, network availability, and dropped calls that are routine on wireless networks would be unacceptable to most subscribers of wireline telephone service. The quality of packet voice services on wireline networks must be indistinguishable from the quality of circuit switched services. We divide QoS inro two key components: the availability of the network that provides the service, and the perceived quality of conversation (not including the content of course) once a connection is made. Each component has its own metrics for measuring quality.

The availability of networks is usually measured by the percent of time during a year that the network is available for use. For the Public Switched Telephone Network (PSTN), the minimum acceptable level of uptime is "five nines," or 99.999%. As shown in Exhibit 21, this translates to the network being unavailable for about five minutes per year.

Exhibit 21: Network Availability Levels ----_-_ [ Level of Reilabllllty .'Downtime per Year' I

. . 95.00% ......... I . . ...... 438hours' . . .

. . . . . . . . . . . < . . . . . . . . 99.00% 88 hours

99.99% 53 minutes

99.9999% 32 seconds . . . . . . . . . . . . . . . . . . . . . . . .

. . . . !.' 99.99999% ' ' 3 seconos I. . ........................... I . . . . . ' .

Source: RBC Capital Markets

December 10,2001 The Metamorphosis of the Telephony Network

We believe that nearly everyone who has used the Internet has personal experience that verifies that the telephone network’s availability is far superior to the availability of the Internet and other data networks. Telephone networks seem to always provide a,dial tone, and rarely do they give the fast busy or “all circuits are busy” signals that indicate the network is at capacity.

Contrast this with the Internet and other data networks. The “always on” connections--e.g. DSL or cable access-that have availability superior to that of dial-up Internet access do not come close to delivering the promised “always-on” availability. And once we are connected, the network does not inform us when the backbone has reached capacity: it simply drops our packets, forcing us to try to connect again or to give up. So while the availability of data networks is steadily improving, it must improve further if this infrastructure is to be used to deliver voice services.

The quality of conversation can be divided into two aspects: the clarity of the sound and the cadence of the conversation. The metric most often used to measure the quality of sound transmission once a connection is made is the Mean Opinion Score (MOS). The MOS method specifies that a large number of listeners (greater than 24) rank, on a scale of I to 5 , the quality of specially prepared speech m a t e n a h a controlled environment. The definitions of the level of MOS are given in Exhibit 22. While a subjective measure, research has found that this subjective measure is highly correlated with easily identified and measured characteristics, and thus measurement can be automated.

Exhibit 22: Mean Opinion Scare

Poor Soeecn 6 unOers1anCab.e onw wilh consmefable efforl: hea-enl

Source: IBM

The cadence of conversation is influenced by the delay in transmission of voice signals and the resulting echo. End-to-end delay of less than 100 milliseconds is not noticeable by most people. Delay between 100 milliseconds and 300 milliseconds will seem to most people as if their partner in conversation is slow to respond. Delay above 300 milliseconds is obvious to most people and they will start to back off to prevent interruptions. Exhibit 23 illustrates the relationship among perceived quality ofservice, MOS, and total delay.

Brown 23 EXHIBIT R- -(JF-16) CASENO. U-13796 - . I_


.

, z : , . . . ,. ,

' , !? : . , . . il :L

I :.

T d e ~ ~ a ~ a Services U.nited States .. . . .. ~~

., ,

July 7,2003

Coveraas view: Neutral

The hype surrounding VolP is growing within the teleccm i This iS our first indepth l ook VoiP and why it may change face of communications

Frank J. Governali frank.govemali Bgs.com Portland: 1-207-772-3300 ' .

Gregory Rsgan gregory.regan @gs.com Portland: 1-207-772-3533

Christopher Fine chrisfine @ gs.corn New Yo& 1-212-902-9010

VolP - the enabler of real telecom competition. As quality and reliability concerns diminish, new competitors are using VolP technology. Incumbents will use VolP as a tool to defend market share but it will not likely add incremental revenue. The issue for VolP competitors now is scalability and rollout.

VolP enables the next phase of true competition in the telecom industry Voice over Internet Protocol (VoIP) has made a fairly qulet entrance onto the telecom scene, I

despite the changes that it will Likely bring out in the next few years. VoIP 1s viewed as the springboard for competitors to take market share, and would serve as a defensive tool for incumbents.

The Bells face a three-pronged attack Web-based VOW providers are already beginning to pick away at the retail consumer base. : Cable companies looking to enter the voice-services market quickly can partner with Val? providers to offer telephony service. Last, and Likely most formidably, the Tier-I cable MSO's '

are all testing V o P services using in-house technology. .

Quality concerns for VolP are diminishing -now the issue is scale In the past, Quality of Service (QoS) was a large issue holding back acceptance of VoIP, but in recent years great strides have been made in QoS. On a managed network, quality can be. engineered to PSTN quality or better. When'the Internet is involved in transport, studies have shown VoIP reliability is just under PSTN levels, but eqwl to or better than with mobile networks. Scalability and business models, not quality, are now the focus for competitors.'

'

. '

Stock impact - no near-term impact, but potential to change the industry '

VoIP is not a near-term stock-driving event for the telcos. It is, however, a longer-tern industry-changing technology. The Bells will likely not stan to feel a substantial competitive ' .

impact until the 2W-2005 time frame, when the Web-based providers gain more scale and the cable operators launch telephony service in the majority of their markets.

Goldman Sachs Global Equity Research

FOR REG AC CERTIFICATION, SEE PAGE 36. FOR OTHER IMPORTANT DISCLOSURES, SEE PAGE 39, GO TO http:l/www.gs.com/research/hedge.html, OR CONTACT YOUR INVESTMENT REPRESENTATIVE.EXHIBIT R- -(JF-l7)

CASE NO. U-13796

. . .- .

: .

. . . .


. .

PACIFIC~::BELL~

VICE PRESIDENT'S CORNER , .

H O ~ Y SECURE IS YOUR BUSNES.S'&.YOUR . . CUSTOMERS'? . ,:

These are questions worth serious thought. Especially if you or your' customers have'been through any downsizing. Business Week . ':. recently featured the need for Security Consulting Services in "Revenge of the Downsized Nerds.".:Experts estimated as many as 30% of a company's approved users are no longer around. Executives make the mistake of assuming firewalls will protect them but a third of companies using firewalls say they're stillhacked, into, according to the Computer Security Institute.

Fortunately, SBC Consulting Services has the offerings designed to meet you and your customers' security needs, in specific areas such as Security Policy Development and Penetration Testing. Here's a brief overview of both:

, .

Security Policy Development 1 .This provides customers with the means to understand their organization's

technological security posture and the abilih, to take the correct steps to improve and rein6rce that posture by:

* Providing an inventory and logical description of .the secure technology infrastructure.

* Delivering a comprehensive issue report for the p.olicies and procedures, ranked in oraer of criticality.

* Providina recommendations for remediation of the issues identified. . ..

. Recommending what can be done to improve the policy infrastructure. lconllnued an D a m 21

A-877-722-3755

* DSL Prices & Availability * Place Orders * Check Order S ta tus 9 Billing

* Self-install Help Desk Service Issues

"This number, which can also be remembered as 1-877-SBC-DSL5, is for DSL Internet customers of Pacific Bell, Nevada Bell, Southwestern Bell. Ameritech & SNET.

THE STRATEGIC SHIFT OF CALL CENTERS

. .

Chief Customer Officer: Vice President of. Customer Satisfaction. These are some of the new titles gracing organization charts in your client companies.The advent of these positions signals the rapidly increasing recognition on the part of C- level executives that customer service ' . ,and customer satisfaction are crucial elements to corporate profitability and objectives.

As executive awareness grows, so does the strategic role of the Call Center in leading'organizations: There are three strategic.objectives that call centers can achieve:

1. Acquisition of customers. If this is the company's primary focus, the call center can function in an outbound mode to solicit business, function in an inbound mode tQ fulfill orders from 'new cu.stpmers, or some combination of contacts.

(continued on page 5)

EXHIBIT R- -(JF-18) Our 'Web address is: urww.pacbell.coml?ro~uctsCASE NO, u.13796

.~ , ,_.- . r . -

_ i

VICE PRESIDENTS CORNER (continued) ,. :

.~ 2: Develop best practice. documented up-to-d,ate and

clear organization Information Security Policies by: * Employing industry best practices and

methodologies to assess and document the organization's current security policies from the individual component level to entire systems and networks. - Ensuring that all relevant policy documents have been specified. Make sure the policy analysis focuses on the correct policy infrastructure. Provides'organizations with an unbiased assessment of their current information

.. systems and telecommunications security policies.

3. Ensures the validity of the Secure Technology Policy by: . Analyzing the relevant information systems and

telecommunications security policies. * Identifying and documenting the existing and

potential issues in the secure technology policy infrastructure.

* Assessing the criticality of identified policy issues. * Evaluating and suggesting possible remediation

...

.techniques, tools and methodologies.

Penetrjtion Testing This demonstrates the need for customers. to protect the integrity and confidentiality of the organization's information, computing and telecommunication assets from attacks either outside or inside the organization by:

Establishing how hackers and crackers can gain unauthorized access to the technology and

.'information resources.

'network scanning and war dialing tools and methodologies to accurately simulate a potential hackerkracker attack.

* .Identifying and documenting potential and existing vulnerabilities in the network Derimeter

. .Employing a variety of automated and manual ' '

infrastructure: * Assessing the criticality of identified vulnerabilities. * Analyzing the relevant information systems and

telecommunications security policies. * Evaluating and suggesting remediation

,, company will-use Sterling's application to'more easily identify potential fraud and reduce losses from Check Kiting-a sophisticated scam involving repeated . . ' deposits of bad checks through multiple accounts. By using Sterling's software, one bank reduced annual Kiting losses from $12 million to $10,000. . :

For more information on how SBC can help you and .your customers with Security, please contact your Pacific Bell ConsultanWendor Liaison Manager (see back cover for their e-mail addresses & phone numbers). And watch our next Pacific Bell CVSG N Broadcast on Wed., Dec. f2 from'9-77:30am PSTYou'can either view the broadcast at one of our sites or be a Streamer. Call 1- 888-889-6010 to reserve a spot or see Back-Cover for Streaming information.

. ..

T m b - Kari Watanabe

Consultatittendor Sales Group. Vice President (415) 542,4516 e-mail: [email protected]

DATA WITH DAVID PROM!SE.S OF IP TELEPHONY

Users of traditional Centrex service, PBXs, and Key Systems are currently faced with the challenge of deciding whether they should grow or replace their current service or system. The convergence of voice and data has added new complexities for telecom and IT managers to consider in designing their networks, One of the technology triggers driving network convergence is the pervasive and wjdening growth of lnter.net Protocol as a . ~ . common element in both Local Area Networks (LAN) and Wide Area Networks (WAN). In the past three years, IP telephony has become the major focus of many technology vendors and service providers as users have sought to receive:

techniques, tools and methodologies. * Enhanced features at equal or lower costs than traditional voice Cost savings associated with simplified moves, adds and changes

* Increased productivity from new applications including remote ofice arrangements

* Increased use of data networks for all applications,

Pacific Bell plans to introduce IP Centrex Fourth Quarter 2001 as customers look to this'new technology and seek ways to find less expensive communications solutions. IP Centrex. which is a replacement for traditional circuit switched technology, is an IP telephony application offered over a'managed IP network and not the Internet; a1thoug.h . . IP Centrex will interconnect to.the Internet.

. . Additional SBC Security News 1.SBC Internet Services recently launched an Online

Security Center where DSL and Dial-Up Customers can download and install anti-virus, privacy and firewall software onto their personal computers. SBClS customers receive a 30-day free trial of this service that features McAfee.com products. After the trial period, customers will be charged an annual subscription of $48, if they choose to continue using the McAfee'software.

2.Sterling Commerce Banking System's Division has announced the licensing of its Vector Kite Anti-Fraud Software by a multi-billion bankholding company. The

EXHIBIT R- -(JF-18) CASE , ~ . ,~. NO. - ~ - . U-13796

. . ., . . ~ . ' i . , , i .

3 . . .

.U . _I.... -- By contrast, an alternative to Pacific Bell's first IP telephony service offering is Voice over IP (VolP). VolP is voice-packetized traffic that is transmit.ted over the Internet. This capability is usually deployed, for toll

. . bypass, and because the Internet is not .a managed network and was specifically designed for data, the

. . quality is inconsistent and often extremely poor,. . .

Packet + Centrex = IP Centrex In IP telephony, voice conversations can be digitized and packetized for transmission across the .petwork. IP Centrex consists of Centrex stations that work off a customeis LAN and receive station feature functionality and access from a Pacific Bell IP enabled central office. 1.P Centrex builds on the traditional benefits of Centrex by combining them with the benefits of IP telephony. One of these IP telephony benefits is increased utilization of access capacity. In IP Centrex. a single broadband access facility is used to carry the packetized voice streams for many simultaneous calls. Transport connecting IP Centrex lines to a Pacific Bell Central Office could be over fractional DS-1, DS-I, DS-3 or OC-3 and is a departure from how Centrex is offered today where dedicated pair of wires connects each telephone to the CO. When calls are not active, more bandwidth is available for high speed data sessions over the LAN, like Internet access. This is a much more efficient use of capacity than traditional Centrex. ' ,

.'

. . . .

. .

Network Design EXisting Class 5 DMS and ESS switches support IP . Centrex service in addition to traditional Plain Old Telephone Service (POTS) and ISDN lines. This is accomplished through the introduction of a new network element-the Network Gateway--and a new type of CPE-the IP phone.

The Network Gateway and IP phone signal each other over a managed IP packet network using the H.323 suite of protocols. The packet network connecting the central office and customer premises appears to the DMS or 5ESS switch as an ordinary digital loop carrier (DLC) system, and the telephone sets connected to the local area network appear to the switch as ordinary phone lines. Because the IP Centrex solution is treated as a DLC system by the central office switch, the switch is able to deliver the same features to IP Centrex users that it delivers to analog and ISDN Centrex users. These features include Centrex services, billing. call processing, and access to the public switched telephone network (PSTN). Consequently, an extensive set of features is immediately available to IP Centrex users w.ithout needing to upgrade the Class 5 switch. The central office is equipped with an IP gateway, which allows for interconnection to both the Internet and the PSTN. (See diagram top of next column)

In this example, the customer is using legacy station equipment (analog and ISDN) over traditional copper pairs to serve most users. The user is gradually introducing IP Centrex on a line-by-line basis by converting some users from legacy equipment to IP Phones.

~

. .

:'

I 1 . , ~

. . Benefits IP Centrex provides business customers with some key benefits:

increased Productivity IP Centrex lays the framework for multimedia ' ' ' ' '

communications such as desktop video conferencing, unified messaging, multimedia conferencing, file sharing, . . and white boarding:The timeframe for introduction of these supplemental services is.expected to be in late 2002 or 2003.

One of the primary groups that will benefit from these IP enhancements will be the remote office or "road warrior; workers. With IP Centrex, geographic location is not a factor when defining a Centrex group. Branch offices,,

' 'telecommuters, and even business travelers can be '. cOmbined.into a single Centrex group with the main

'

, , office, even if they are all served by different wire centers. This is possible because packet transport is so inexpensive and is not mileage sensitive. Multilocation. Centrex provides the following advantages to the customer: :'

* Easier for the corporate telecom manager'to

, .

. . administer one large Centrex group rather than several separate Centrex Groups. , . Uniform services and features for all users, regardless of where their office is. . Extend advanced business features to telecommuters' homes without any Foreign Exchange costs. Uniform dialing plan across the company. Abbreviated dialing can be used to call the. office next door or the office across the state. . .

Operational Efficiency Provides toll-grade quality of service implemented on, managed IP networks which are scalable and provide secure, reliable and integrated IP telephony.

Cost Savings For many of our customers, the real value of IP Centrex will be obtained through, lower cost benefits of managing and delivering voice and data together on a converged network-both within the enterprise and on the access connection to the network. Converged networks also improve utilization of'premise wiring. Toll charges'will '

'

.

also be eliminated for calls locations part of the same

-;-

. . . .

. . I rnG i e ' - , ,:

. . Enhanced Feature Set IP Centrex provides complete feature menus that are currently offered today. It provides enhanced customer . control over moves, adds and changes. In the future, IP

(CTI) solutions where telephone functions can be controlled by the user's PC. and where telephone information (e.g.. caller ID information) can be . .

integrated with other-software applications. The'ultimate form of CTI is the softphone, where the PC actually replaces the telephone.

Deployment Introduction of IP Centrex is expected to begin in the . Fourth Quarter of 2001 in Sacramento and Los Angeles, CA; Chicago, IL and Houston, TX. A limited number of Centrex Offices will be.deployed.in these market areas at that time. A final deployment schedule is still under development.

Conclusion IP Centrex also holds out the promise of true convergence. Customers are actively pursuing IP technology as a means to converge their separate voice, video, and. data networks into single unified networks that can handle all their,CQmmUniCatiOn needs. Benefits.of this convergence mean network efficiencies and enhanced features. Since voice will be

.packetized, it means end users can write their.own . applications to a server that sits on the network. Moves and changes become simple: you,simply unplug the phone and move it to another jack resulting in less documentation and cable plant to track. And if the voice traffic is carried on a managed IP network, it runs for free. For work-at-home applications, the employee only. . . needs a DSL.line with voice-over capabilities. and access to an Internet account. IP Centrex is the. natural evolution of migrating from today's circuit switched narrow-band network into the future broad-band networks which offer next generation services without disrupting customer operations, or sacrificing the productivity features inat Centrex users have come to rely on.

'

Centrex will promote Computer Telephony Integration . . .

. .

. .

.

- Tom David Consultant Liaison Manager (949) 855-5055 e-mail. [email protected]

:

4

CENTREX:IS HERE FORYOU, . .

With its many eneigy driven (and high'bandwidth hungry) Industries, such as high tech and entertainment, California. businesses must be uniquely innovative and resourceful, especially during the current'energy crisis. It is critical that businesses and consumers have a p w e r back-up plan in case of rolling blackouts or power outages. At Pacific Bell, our primary goal is to keep businesses' and consumers' telecommunications ' '

systems running with reliable services and products that won't be interrupted by power outages. Our Centrex .. product is the most reliable telecommunications system during any disaster (fire, earthquake and floods) and in particular, during power outages.

During our last CVSG TV Live Broadcast. Donna Marie. . Simmons,,Regional Product Manager for Centrex and VOlP (Voice over IP) Products, provided excellent information to help businesses with a Power Back-up plan that can alleviate your worries and concerns during rolling blackouts. Here are a few key points that Donna Marie noted:

, .

Centrex Staying Power * Powered by the Pacific Bell Central Office , , , ,

* Redundancy * 24 x 7 - Service Availability

. , . . Five 9 s of reliability (999.99 reliable)

Centrex: Telecommunications Back-U,p * Customers are concerned about

telecommunications back-up alternatives . Centrex plays well as a back-up or primary telecommunication solution

* Complements existing telecommunication systems, - PBX w/Cornbination Trunk-Customers who have

an existing PBX that does not have Enhanced Alternate Routing, should order RACF (Remote Access to Call Forwarding). This will allow the customer to'Call Forward their PBX trunk side line . . to their Centrex Back-Up'System. .. .. . .

- .PBX w/PRI- Customers who have an existing PBX trunked with PRI with Enhanced Alternate Routing should be advised to have their Enhanced Alternate Route flow to.their Centrex Back-up System.

* Excellent primary telecommunication system

- Powered by the SBC Central Office - Eliminates the expense associated with

maintaining a switch on the premise

To address the current energy situation and save money and energy, Pacific Bell has a telecommunications Back- Up Plan that is described in the following Centrex Promotion:

Current Centrex Power Up Promotion (Now through Dec 31, 2001) Between 50.100% off of installation charoes deoendtnc

~ - w o n line size and term

$50 off select telephones EXHIBIT R- -(JF-18) CASE NO. U. 13796

. . . : . . , . , .. ... .

. .

* $4.00 off monthly voiceinail box charge $15.00 6ff lacks 2-5

Minimum requirements for this promotion:

Classic Feature Package on each line 1 voicemail box ,.

* 1 Telematrix (two line,phone) 2105 or 2,Telem'atrix (one line phones) 1105's '.

* Pacific Bell Usage Plan 'Note all Centrex tariff requirements must be met for this offer. Power Up with Centrex is a bundle.

Other products that can help in the PBX environment:

* Customer with PBX trunked w.ith PRI with Enhanced Alternate Routing - Enhanced Alternate Route: $47.50 MRC and

- 2 Centrex Lines

$142.49 NRC

: Customer with PBX with Combo trunks - Remote Access to Call Fonvarding (RACF): $2.90 MRC and $9.00 NRC

How to View the Archived CVSG N Broadcast To view the presentation about "Centrex: 'More Than Just Power ... Staying Power" (also "SBC CPE'& Support" & "Video Monitoring of Schools" as well as "Pacific Bell/ SBC News"), type in the following URL. if you encounter problems, call.Techn1cal Support at 1400-266-3373; Press 3; then Press 2.

http://209.247.74.138lse~lefflPConfSe~?ORG_ID=22 &CONFERENCE-ID=47415

If you need. more ideas and information about how to design a power back-up plan for your customers, please call your Liaison Managers at 1-800-552-5299.

- Lowayne Shieh

WEATHER.COM & CINGULAR WIRELESS

Weather.com, the world's leading source of weather on the Web and the official site of The Weather Channel, is the premier weather provider for Cingular Wireless. Cingular customers with Internet capable phones can now view customized weather forecasts from up to three cities-anytime, anywhere. To use this service, users simplysign up by visiting Weather.com's Cingular section located at weattier.com1serviceslphone.html

THE STRATEGIC SHIFT OF CALL CENTERS

2.

3.

. . (continued;

Retention of customers. In this scenario, call centers focus on maintaining customer relationships and customer satisfaction to ensure that the investments made in acquiring customers are not lost. Both inbound and outbound technologies and processes are used' to support this objective. Increased investment by customerr: This objective aims to increase the'revenueper customer, and expand the range and/or frequency of products purchased by customers.

A Call Center may focus exclusively on one of,these objectives, but more typically they'address two or even all three objectives. After all, most businesses do ,not produce just one product or service, and will have different sales; mzrketing. and support strategies based on multiple factors such as positioning of product, volume of sales, or revenue to company. ' I

The strategic objectives of the Call Center must be reflected in the call handling strategies. (Note: in this article the Call Center inclu'des phone, e-mall, and web- based contacts.) For acquisition of .customers, the'most important facet,of a handling strategy may be speed of answer. When an established customer calls and your strategic objective is retention of that customer, caller identification for personalized,service or self-service options may come into play. And when looking to increase the value of a customer, CRM applications and contact options such as web-based self-service are valuable tools.

Call Centers should periodically audit their call handling strategies to ensure thaf these are aligned with corporate objectives. These audits often identity opportunities for consulting services to revamp call handling strategies and accompanying purchases of kchnologies or applications to sbppcit these changes.

Pacific Bell Call Center Solutions offer a complete array of products that help Call Centers meet their . strategic objectives. From ACDs to quality monitoring and workforce management, our Call Center Solutions sales consultants can work with you to select the appropriate products and configurations for your clients.

. . '

. .

.. I '

- Christine.Hertzog . . .. :

For more information about Pacific Bell Call Center Solutions, contact your liaison manager of Christine Hertzog, Regional Sales Director at [email protected]: Hertzog has over I 5 years experience in Call Centers and Computer Telephone lntegration Technologies. 'She has worked in sales,' marketing. product management, and most recently as a consuitant.

EXHIBIT R- -(JF-l8) CASE NO. U-13796 r.* 1 1 1 i I I. - ~-


Pulse Technical Handbook Series - T7 Networking Made Easy

T Qworkin e Easy

. , . .

.. . . . . .. :

. . . : . .

. . . . . . . . . . .. . .~

. . . , , . .. . . ,, . . . .

. . , . ..: . , . . , ., . . .

I _

. .

. . . . i ,~ . . . . ,

0 1998. Pulse, Inc Toll Free “(888) 785-7393 -.-.*.“ Pulse;, (., x, y (,.- Boston (508,660-0340 Baliimore (410) 444-7999 Fon Lauderdale (954) 783-4320 hlinneapol s (612 893-1725 S e a m (425) 348-8003 Los Angeles (909) 597-7461

1 EXHIBIT R- -(JF-l9) CASE NO. U-13796

. , Pulse Technical Handbook Series - T7 Networking Made Easy '' .'

- . ,

. . The 'TI Carrier The T I is what telephone companies have traditionally used to transport '. . ' .

digitized telephone conversations between central offices. As early as the : 19603, a single T I circuit made it possible for a telephone company to'deliver . ' '

24 high quality voice conversations. Since a T I is a fully digital service, there ' :

was no possibility of cross-talk, which. is common in analog carriernetworks where. copper pairs pickup emissions from neighboring pairs. Significant

transmission standard.

Since the early 19803, T I service has been available to private industry throughout the country. This document will discuss the various types of T I services available, how to deploy them, implement them effectively and

. . . .

.~

increases in noise immunity were also achieved by adopting this new digital . .

. .

understand the general guidelines of T I networking.

What Does a T I Look Like?

T7' Bandwidth The bandwidth of a T I is commonly known to be 1.544Mbps:This represents the maximum bit carrying abilihr of a T I . The overhead necessary to frame a T I is 8Kbps. Therefore, the total usable Gandwidth is 1.536Mbps, or the equivalent of 24 DS-0 channels. A single DS-0 has a bandwidth of 64Kbps and is designed:to carry a digitized telephone call. Today, T I technology is being used in private

. .

.~

' .

. ,

' .and public networks to carry both voice and data traffic.

0 1998, Pulse, Inc Toll Free (888) 785-7393

Boston (508) 660-0340 Baltimore (410) 444-7999 Fort Lauderdale (954) 783-4320 Minneapolis (612) 893-1725 Seattle (425) 348-8083 Los Angeles (909) 597-7461

3 EXHIBIT R- -(JF- 19) CASENO U-13796

Pulse Technical Handbook Series - T I Networking Made Easy

I I

Since a T I interface is now being presented to the user, one of the following would be required:

Ensure that the Customer Premises Equipment (CPE) provides a T I interface. Convert the T I into 24 standard analog lines. This would require channel bank.

If the T1 is going to be used in a voice application and needs to be connected to a telephone system, a T I interface is most likely going to be present on the telephone system. In this case all that would be required to terminate the T1 is a Channel Service Unit (CSU). A CSU is required anytime a T I circuit is connected to CPE. It is the device responsible for isolating the public network from the end-user network and also serves as a test point that allows the carrier to run many standard tests. In some cases a CSU is built into the Customer Premises Equipment which provides for direct connection to the T I (no external CSY Trquired).

Telephone T1 Circuit From Carrier rp- System ( P W

Figure 7: T I Circuit Terminated With a CSU

' pu;se. -.- .iY I \ A +& 0 1998, Pulse, Inc. Toll Free'(888) 785-7393

Boston (508) 660-0340 Baltimore (410) 444-7999 Fort Lquderdale (954) 783-4320 Minneapolis (612) 893-1725 Seattle (425) 348-8083 Los Angeles (909) 597-7461

9 EXHIBIT R- -(JF-l9) CASENO U-13796

Pulse,Technical Handbook Series - 71 Networking Made Easy ' :" . ,

. .

If the T1 is to be used in an analog application, a channel bank would be necessary in order to convert the 24 digital voice channels into 24 individual analog voice lines. Some devices that require analog interfaces are old telephone systems and modems. The following diagram shows a channel bank converting a T I into 24 analog circuits for connecting a rack of modems.

I I

Figure 8: T I Channel Bank Application

Summary of PSTN Access With a T I

Accessing the Public Switched Telephone Network with a T I is similar in concept to accessing the PSTN over a standard analog telephone line. In both cases a local loop is deployed between the CO and the customer's premises. In the case of T I access, a T I loop is used and in the case of a single analog line, an analog loop is used. With the connection in place, the CO is responsible for switching all local telephone calls and routing all long distance call to the IXC of choice.

m r h r . w a ' y 0 1998, Pulse, Inc Toll Free (888) 785-7393

Boston (508) 660-0340 Baltimore (410) 444-7999 Fort Lauderdale (954) 783-4320 Minneapolis (612) 893-1725 Seattle (425) 348-8083 Los Angeles (909) 597-7461

10

EXHIBIT R- -(lF-19) CASE NO. U-13796

Y

AM-TR-OAT-000033 DSI Customer Installation: Metallic Interface , -

To, N/A

NIA Priority:

Effective Date: January 1990

Issue Date:

Expires On: NIA

, /

Issue E, January 1990

NIA Training Time:

NIA Related Documents:

Canceled Documents: NIA

Issuing Department: NIA

Distribution:

Business Unit:

Points of Contact: NIA

NA

NIA

Author(s): NIA

Copyright 0 SEC Corporation, 2000 This document is protected by the U.S. Copyright laws.

Any alteration to its text, contents, or presentation format is an infringement of SBC's Copyright rights

i EXHIBIT R- (JF-20) CASE NO. TJ-11796

.. ..

. . ,AM-TR-OAT-000033 '' SBC Practice .,

Issue B, January 1990

3.2. CI

The CI (Customer Installation) is the customer provided equipment and'wiring at the'cKstomer's location on the customer side of the NI.

.

. . . .

3.3. SI

The SI (Service Interface) is that point of termination where all technicallphysical parameters are defined. The SI is located at the NI or may be extended at the customer's request.

4. DSI METALLIC INTERFACE

The ANSI standard provides for a dry interface at the CI. Ameritech;agrees with the standard. Direct Current power normally will not be delivered across the SI in accordance with FCC Re- port and Order 86-423. However, power may be provided to the customer's NCTE (Network Channel Terminating Equipment) under tariff guidelines, consistent with EIA standards, if that equipment requires line power to function. Deregulated regenerators required to implement the extention of the SI may be line powered. The last regenerator or the smart jack will provide power isolation from the SI.

'

, .

i

In accord with the ANSI standard, the loss of the network signal at the SI shall be within'the .. range of 0.0 to 16.5 dB at 772 kHz between 100 ohm terminations.

Caution: To maintain compatibility with existing DS1 service, level coordination must . '

be considered for the Customer Installation signal. Existing end section design guidelines require that the maximum allowable input power level to the line regenerator (typically - 9 dB or less) not be exceeded. The.method for meeting this requirement is to carefully place the last carrier provided regenerator at a proper distance from the SI. This may not be required'for new installations,. . . . .. . .

,

. . . .. . . . . . ..

It is recommended that a line loopback device, commonly called a smart jack, be placed in the end section for ease of maintenance. The smart jack shall be located on the network side of the NI, and shall have an RJ4Wequivalent jack as an integral part of its design. .If a smart jack with a non-regenerative loop-back is used, the maximum loss in the end section design shouid be limited to 15 dB. The smart jack may be used to provide power isolation. Generic requirements for DSI line loopback devices are covered in TR-TSY-000312 (Functional Criteria for DSI Interface Connector). (See reference 1. Section 7.

It is recommended that 100 ohm impedance cable with transmit and receive pairs separately shield be deployed on all customer premises for this service.

. .

Copyright 0 SBC Service, Inc. 2000 This document is protected by the US. Copyright laws.

Any alteration to its text, contents, or presentation format is

EXHIBIT R- -(JF-20) an infringement of SBC's Copyright rights 4 CASE NO. U-13796

. .

SBC Practice . . . AM-TR-OAT-000033 , .

Issue B, January 1990

5. DIAGRAMS

The diagrams shown in the following figures are recommended circuit arrangements for DS1 service to the customer location. These layouts follow the ANSI T1.403-1989 standard for the metallic interface.

Figure 1. REGENERATED LINE - COLLOCATED NllSl

I ........................... CI .. / REGEhTRATOR NCTE

z' ......................... "*TO -16. ................... / '- CUSTOMER SIGNAL PER PL4P.A. 6.2.

CUSTOMER PROWJED \b'IRJ3EQLEF ...--' J......

/ .................................... P , .................................. / ... KCTE Nus1

REGENERATOR R€GEXER+TOR .. +'- ~ , S ! & . R T T I I r CSU~DSU I RJ48 '~- CUSTOhER SIGNAL PER PARA. 6.?. J : ........................ "*TO-16.5dE) .................. /

............. WSTObER pI10yIDm ,;IREIEQuIp ......... i

The arrangements shown in Figure 1 cover a collocated NI/SI. All wiring and equipmenton the customer side of the NI/SI is customer provided. The RJ48 is an integral part of the smart jack

See Section 4 CAUTION statement.

Copyright 0 SBC Service, Inc. 2000 This document is'protected by the US. Copyright laws.

Any alteration to its text, contents, or presentation format is an infringement of SBC's Copyright rights

, . <

. . EXHIBIT R- -(JF-20) CASE NO. U-13796 -

Exhibit R- (JF-21) -

Telecommunications Transmission Engineering

Volume 2 -Facilities

First Edition

Technical Personnel American Telephone and Telegraph Company,

Bell Telephone Companies, and

Bell Telephone Laboratories

Bell System Center for Technical Education

EXHIBIT R- (JF-21) CASE NO. U-13796

Copyright @ 1977 American Telephone and Telegraph Co.

Prepared for Publication by ~ Wes tern Electric Company, Inc.

Technical Publications Winston-Sakm, North Carolina

FIRST EDITION 1977

Printed in the United States of America

i

: . . I: . '

, .'

EXHIBIT R- -(JF-21) CASE NO. U-13796 D A . C E ? ~ V I <

01. 2

the relay ignal ;Iliny )onse from

gen- .stern

and 3-9)

2 the 1 loss luced

mnel x i is pair

s, no gned have

r one ter-

ihich :ions.

per- ntral mote ?atel- kHz. sp to

rable ta or ntral rvard from .ed.

Two digital transmission systems, a Subscriber Loop Carrier 40 (SLC-40) System and the Subscriber Loop Multiplexer (SLM*)

i. System, have been designed to serve long roiite needs. Of the two, the SLC-40 System has proven t o serve telephbne company needs more economically and is more commonly found in service. However, a number of SLM Systems are also in use. Both utilize T1-Carrier System line equipment, discussed in Chapter 22, but the terminal arrangements and system configurations are quite different.

The SLC-40 System. When fully equipped, the SLC-40 System can provide up t o 40 speech channels between a central office and a remote terminal as much as 50 miles away [S]. Channel units provide service to individual (single party) lines, two-party lines with automatic number identification, and a variety of multiparty lines (up t o eight party) with combinations of semi-selective, fully-selective, or coded ringing and automatic or operator number identification. Other applications are being developed to expand further the field of use of this system.

System Layoz~t . Figure 3-10 shows a typical layout of an SLC-40 system. The system is composed of a central office terminal and a single remote terminal interconnected by a T1-type repeatered line. The system provides 40 full-time voice-grade channels as loops between the central office and the remote terminal. Standard voice- frequency distribution facilities are used to extend the loops from the remote terminal t o customer premises.

The length of the repeatered line depends on the type and gauge of cable. For 22-gauge cable, the maximum length is 10 miles for systems powered only from the central office. The length may be increased to 20 miles for systems powered from both the central office and the remote terminal and to 50 miles for systems powered from both ends and from an intermediate power feed point.

Two remote terminal arrangements are available. In one, a weather- proof cabinet that may be pole- or pedestal-mounted is used to house channel units, common circuit units, batteries, battery charger, and a ringing generator. In the other arrangement, the equipment for two SLC-40 systems may be mounted on a seven-foot frame in an equip-

*Trademark of the Western Electric Company.

EXHIBIT R- -(JF-21) CASE NO. U-13796 P A r i R ? n P l <

86 Local Plant Facilities Vol. 2

71 repeaters I

3000 h

Figure 3-10. SLC-40 system layout.

ment hut, in a community dial office building, or a t a customer premises.

Central offee and remote terminal equipment must be synchronized. A 41st channel is assigned to carry timing and maintenance information between terminals. The system can detect loss of synchronization within one millisecond and can correct such a condition within three milliseconds.

As shown in Figure 3-10, the first repeater is placed about 3000 feet from the central office. Other repeaters are spaced about 6000 feet apart, the exact spacing depending on type of cable, gauge, number of systems on the route, and practical problems of land and right of way acquisition. The short spacing a t the central office is provided t o minimize impulse noise impairments that might result from switching transients.

The SLC-40 system is provided with a protection line and automatic protection switching. One protection line may serve t o protect two working lines where the remote terminal equipment is rack mounted. An alternative arrangement is available for rack mounting in which one protection line may protect 5 or 11 lines. In some cases, patching is provided and a single protection line may serve more than two regular lines.

Terminal Equipment. The 40 voice-frequency circuits of an SLC-40 system are each connected a t central office and remote terminals

Chap

throu multi a sin] modE whicl carri face trans

Thc niquc digit: pairs show cont: plexi line : madc conn sign: 57.2 divis the s

e-

1w mc

A : , ,

. . ,.I the :I:. .P J??

01. 2

the relay ignal .lliny Ionse from

gen- stem and

3-9) 3 the I loss /wed

innel 311 is pair

s, no yned have . one

ter- ihich ions.

per- ntral mote zater kHz. 1p to

rable ta or ntral bvard from ed.

h

Chap. 3 Loops and Station Sets 85

Digital Systems

Two digital transmission systems, a Subscriber .Loop Carrier 40 (SLC-40) System and the Subscriber Loop Multiplexer (SLM*) System, have been designed to serve long route needs. Of ~the.two, the SLC-40 System has proven t o serve telepgone company needs more economically and is more commonly found in service. However, a number of SLM Systems are also in. us.e.' Both utilize TI-Carrier System line equipment, discussed in Chapter 22, but the terminal arrangements and system configurations are quite different.

The SLC-40 System. When fully equipped, the SLC-40 System can provide up to 40 speech channels between a.ce.ntral office and a remote te.rmina1 as much as 50 miles away [S]. Channel units provide service t o individual (single party) lines, two-party lines with automatic . .

number identification, and a variety of multiparty lines (up ' to eight party) with combinations of 'semi-selective, fully-selective, or coded ringing and automatic or operator number identification. Other applications are being developed to expand'further the^ field of use of this system.

. System Layoz~t. Figure 3-10 shows a typical layout of an SLC-40 system. The system is composed of a central office .terminal and. a single remote terminal interconnected by a T1-type repeatered line. The system provides 40 full-time voice-grade channels a s lo.ops between the central office and the remote terminal. Standard voice- frequency distribution facilities are used to extend the .loops from the remote terminal t o customer premises.

The length of the repeatered line depends on the type and gauge of cable. Fo r 22-gauge cable, the maximum length is 10 miles for systems powered only from the central office. The length may be increased to 20 miles f o r systems powered fr:om both the central office and the remote terminal and to 50 miles for systems powered from both ends' and from an intermediate power feed point.

Two remote terminal arrangements are available. In one, a weather- proof cabinet that may be pole- or pedestal-mounted 4s used to house channel units, common circuit units, batteries, battery charger, and a ringing generator. In the other arrangement, the equipment for two SLC-40 systems may be mounted on a seven-foot frame in an equip-

, .

. '

*Trademark of the Western Electric Company.

EXHIBIT R- -(JF-21) CASENO. U-13796 n * n_ c AT- 1z

.. ,

86 ' Chap

throu multi a sin! modu whicl carrii face trans

Remole terminal

Figure 3-10. SLC-40 system layout.

ment hut, in a community dial office building, or a t a customer premises.

Central office and remote terminal equipment must be synchronized. A 41st channel is assigned to carry timing and maintenance information between terminals. The system can detect loss of synchronization within one millisecond and can correct such a condition within three milliseconds.

As shown in Figure 3-10, the first repeater is placed about 3000 feet from the centrai office. Other repeaters are spaced about 6000 feet apart, the exact spacing depending on type of cable, gauge, number of systems on the route, and practical problems of land and right of way acquisition. The short spacing a t the central office is provided t o minimize impulse noise impairments that might result from switching transients.

The SLC-40 system is provided with a protection line and automatic protection switching. One protection line may serve to protect two working lines where the remote terminal equipment is rack mounted. An alternative arrangement is available for rack mounting in which one protection line may protect 5 or 11 lines. In some cases, patching is provided and a single protection line may serve more than two regular lines.

Terminal Equipment. The 40 voice-frequency circuits of 'an SLC-40 system are each connected a t central office and remote terminals

Thc niquc digit; pairs show conk plexi line : madt conn sign: 57.2 divk the E

3% m

A the

...~ , . ..

Chap. 10 System Design Features 275

,opriate similar irrents.

Jol. 2

3n be- many post-

.djust- ion to

The main station and terminal buildings used with analog cable transmission systems contain many types of equipment in addition to those described in relation to power, maintenance, and reliability.

red t o How-

lave a

lent is PPIY is id cur- output ividual derive

:onver-

.. ,.

Ma.intenance and Reliability . .

Many features are designed into analog cable systems to provide adequate reliability and to provide access and facilities for maintenance activities. In addition, building. design, cable route selection, and cable placement (aerial, buried, use of'duct.s, etc.) all affect the reliability and maintainability of systems.

Transmission system equipment is distributed over long distances and, of necessity, is often located, in out-of-the-way places which 'are difficult t o reach. For these reasons, it i.s necessary t.o provide each system with specialized equipment designed to facilitate the recognition of trouble conditions, to isolate the trouble to a particular section of line, and then to identify the location of the fault so that repair personnel may be efficiently dispatched to the correct location. In- addition, separate communications facilities for voice and.'or data transmission (order wires) are also provided t.o assist maintenance personnel in their work.

Some transmission systems a re provided with equipment for carrier. group alarm and conditioning. With this equipment, system failure initiates an alarm and conditions affected trunks in various ways. Any connection established over these trunks i s . disconnected in such a manner that time charges are immediately terminated. The trunks are made busy until the system is repaired. They are then automatically tested and restored to service.

'.

central.. . .

3 route ,'

Most transmission systems are operated so that on each route there is at least one fully equipped and fully powered line that does .not carry service. Service can be transferred .to this spare line t o facilitate maintenance work or to restore service in the event of 'failure of a working line. The transfer of service from the working line to the spare line may be accomplished manually or automatically by line protection switching equipment. As previously mentioned, the work- iny and spare lines must be equalized to very nearly the same attenuation/frequency characteristics in order to minimize the transmission performance penalty that might accompany the; transfer of service from one line to another.

~p fea- pment, .iods of power load if

'owered

Term ina I Arrange men ts I mission mffice o r

EXHIBIT R- -(JF-2 1) CASE NO. U-13796 n 1 P.F 7 nT- 1 c

338 Analog Carr ier Systems on Metallic Media . .

Spacing Rule Flezibilitg. All of the spacings previously discussed are established with permissible variations that allow' for geographic anomalies, population distribution, and right-of-way complications. With basic repeaters, some spacings in excess of 1 mile are acceptable. .

Such excess spacings must be compensated by short spacings on both sides of the long section. Short sections that are not being provided as compensation for long sections are built out t o match the nominal 1-mile spacing by the use of repeaters with LBO networks of appropriate values.

'

Main Station Administration Equipment. At every main station, equipment must be provided to interconnect line and terminal equipment. Among the items of terminal equipment are the multiplex terminals, branchin'g filters, line pilot and synchronizing signal administration circuits, protection switching equipment, transmission surveillance and fault location equipment, restoration access arrangements, jumbo- group trunk circuits, and line' connecting equipment. The line connecting equipment is mounted in a transmit-receive bay with trans- .mitting and receiving repeaters. These bays are standard in three arrangements, one for power-feed main stations, one for switching power-feed main stations, and one for terminal stations and terminal main stations. Arrangements for other types of terminal equipment .are unique to the needs of each main station.

Significant differences in detail exist in the main station arrangements for L5 and L5E systems. These differences are due to changes .in the pilot frequencies (near 21 MHz) and the refereme frequencies used for synchronization, and the difference .in the multiplex 'equipment used for the two systems.

Maintenance and Reliability

The L5 system design has many maintenance and reliability features; some are improved versions of similar features previously used ,and some are newly ,developed. Since L5 and L5E have such large circuit capacities, probability of failure must be minimized and outage time must be kept as short as possible.

As in other broadband systems, many circuits are duplicated and provided with automatic switching features. Also, access points are provided for test and emergency broadband restoration purposes. However, features unique to L5 include aspects of line maintenance

~

'01. 2

ussed aphic tions. table. both

jvided minal ropri-

:quip- ment. iinals, ration llance dmbo-

con- trans- three

xhing .minal pment

range- ianges encies equip-

y fea- y used

large outage

?d and its are rposes. enance

Chap. 1 2 Coaxial Ca'rrier Systems 339

and administration, a newly designed protection switchillg system,' a transmission surveillance system with new fault location features, and a new design of four-wire order wires.

"

Line Maintenance and Administration. Many features of L5 line operation are important fo r efficient maintenance and administrat,ion.i Among the most significant are the equalization system and the-pro- c.edures f o r repeater replacement.

Eqzdizat ion System. The individual components of the equalization system, previously described, are the fixed, manually adjustable, and automatically controlled equalizer networks. The principal fixed equalizers are the fixed-gain basic repeaters, the line build-out networks, and the deviation"equa1izers. All of these fixed equalizers are 'installed in new L5 systems according t o carefully defined application rules. The designs are such that residual deviations from ' an ideal (flat) attenuation/frequency characteristic can be corrected by the manually adjustable bump equalizers.

The bump equalizers are distributed along the line for pre- and post-equalization at equalizing and main station repeaters: The characteristics and the method of adjustment are designed t o minimize the mean-squared error in the resulting characteristic after adjustment. The equalizers are adjusted at the time of installation and ocassionaliy thereafter. When later adjustments are required, service must be removed from the' l ine under adjustment by operation of the line protection switching system.

Automatic adjustment of the transmission characteristic is provided by regulating repeaters and by E3' dynamic equalizers. These adjustments compensate primarily for system Ioss/frequency.changes due to temperature variations.

' '

Repeclter Replacement Pi,ocedzlres. Operating dc power is supplied to manhole-mounted repeaters over the center conductors of the coaxial units. Removal of a repeater, fo r test or replacement, opens the dc circuit and special means must be provided t o maintain power continuity. The means are illustrated in Figure 12-9. At the input and output of each manhole-mounted repeater, twin jacks are provided. When a repeater is t o be removed from its apparatus case, the bridging pads are removed from the twin jack assemblies and. a coaxial patch cord is plugged into the vacant jacks. By this action, the repeater is short-circuited and direct current continuity is pro-

EXHIBIT R- (JF-21) CASE NO. U x 1 9 6 9 A c.r n nT1 1 :

Digital Systems

Chapter 22

Digital Transmission Lines

A number of wire-pair and coaxial transmission media are used to transmit the various signals of the digital hierarchy. In each case, regenerative repeaters are installed at specified intervals along the line t o amplify, retime, and regenerate the pulses for transmission t o the next repeater or to a terminal.

The facilities developed f o r this mode of operation have all been designated as T-type systems. Those presently in use include the T1 and Tl/OS, TIC, T2, and T4M Digital Transmission Systems which transmit signals a t the DS1, DSlC, DS2, and DS4 rates respectively.

Most of these systems provide trunk facilities in metropolitan and suburban areas ; there are also some applications in the loop plant. The distances over which these systems can operate etonomically are being increased and some systems now in service extend well beyond metropolitan and suburban areas.

The layout of digital lines is based on repeater spacings established by line loss, interference, bandwidth, and the provision of adequate margins. Line spans that encompass a number of repeater sections are based on operations and maintenance considerations and on the limitations of supplying power to remote repeaters over the transmission conductors. The provision of fault locating facilities, maintenance spare lines, and protection switching arrangements are other aspects of digital line layout that must be considered.

22-1 THE T1 DIGITAL TRANSMISSION SYSTEM The first digital system to find acceptance and the one that is in

most common use is the T1 Digital Transmission System. The field

589

P . EXHIBIT R- -(JF-21) CASE NO. U-13796 :$ '. ..

. ;&; ,. D b . G F l f l n F 1 C

.i"*,i

590 Digital Systems VOl. 2

of application is primarily t o provide switched network, trunks between central offices up to about 50 miles apart [l]. 'Some systems are as short as five miles but, in general, T1 systems are more economical than voice-frequency facilities at distances of 10 miles o r more. The T1 line equipment has also been found economical'for loop applications such as those described for the .SLC-40 system discussed in Chapter 3 [Z].

While T I was initially limited to a maximum length of about 50 miles by operating and maintenance considerations, i t ,may now be used over distances of up to about 200 miles. A maximum of Z O O regenerative repeaters comprising several power spans may be connected in tandem. This increase has been realized by applying more stringent engineering rules to ensure meeting performance objectives and by the addition of operating and maintenance features needed for longer routes. Also, improved features and equipment needed for small installations and more economical central office equipment arrangements are provided. In such applications, the system is called TI Outstate (Tl/'OS) [ 3 ] .

The signal transmitted over a T1 carrier line must have a repeti- tion rate of 1.544 Mb/s, a bipolar, 50-percent duty cycle format, and contain no more than 15 consecutive 0s. These constraints on, signal characteristics must be applied in the system terminal equipment but are necessary in order to assure satisfactory operation of regenerative repeaters located along the line.

. .

. .

Transmission Media

A wide variety of multipair exchange cables may be used for T1 systems. These include polyethylene-insulated conductor (PIC) cables, pulp-insulated copper-pair cables of 19, 22, 24, and 26 gauge, and aluminum conductor cables of 17 and 20 gauge. In most cases, the two directions of transmission can be carried in the same cable provided they are in different cable units, or binder groups, in order t o control crosstalk between systems. Engineering rules specify the manner in which such separation must be accomplished but, even within these rules, a limit is imposed on the number of pairs that can be used as T1 lines.

In some cases, the crosstalk between cable pairs is excessive due to the assignment of t o o many T1 lines per cable. It is then mandatory that the two directions of transmission be carried in separate cables

CI

01

is UC

s> SE

PC m la of

in

ca A be bc th

Ca

tk. B lo a] if

.~ Re

a: ht t c rt rt CC

0 ir

CASE NO. U s P A C E 7 1 nc

(JF-21) i96

. 2

be- ms

or )OP sed

co-

5 0 be

re- on- ore ves fo r for ar- led

eti- ind nal but m-

T1 les, ind the lro-

t o the ven :an

? to ory iles

.. . Chap. 22 Digital Transmission Lines 59 1

or in cables with shielding between binder. groups to pro.vide sufficient '

isolation so that cable fill o f T1 lines is equivalent to that obtained by . '

using separate cables. Such cables are called screened cables. Some systems are operated with the two directions of transmission in the same binder group but with reduced repeater spacing :.

Metropolitan area trunk (MAT) cable was designed to optimize performance for T1 systems and t o minimize costs for all types' of metropolitan area circuits. This cable is most useful in .heavily populated areas requiring large cross-sections: of circuits between central offices. [4].

.

,

:

In addition t o trunk applications, T1 system line equipment is used in the loop plant to carry SLC-40 signals. using standard loop plant cables. Great care must be taken i n all applications to be sure that cable pairs have been checked f o r satisfactory opemting conditions. All bridged taps, load coils, and line build-out ,networks must. have been removed. In many cases, voice-frequency equipment will hav.e been used on the pairs involved for someiprevious service and all of

:

. :

this equipment must also be disconnected. . . . ,

The standard administration of cables 'in the loop plant is not as well suited to T1-system transmission as that in the trunk plant. Loop cables are generally not operated under gas pressure and are therefore more prone to impairment due to moisture. A cable. route from t h e central office to a remote area may be made up of mixed gauges. Binder group integrity may be lost since i t ' i s less important in the loop plant than in the trunk plant. All of .these factors must be. examined carefully and the cable condition upgraded where deficient .

if the T1 line equipment is t o operate satisfactorily.

'

Regenerative Line Repeater

The T1 system utilizes two general types of repeaters called line and central office repeaters. Line repeaters, designed f o r use in man- holes or on telephone poles, have circuits which include two regenerators arranged for either one or two directions of transmission. These repeaters are generally assembled in apparatus cases that house 25 repeaters (50 regenerators). The other type, designed for use in central office repeater bays, contains circuits t o regenerate the signals on only one lixe. Thus, two such regenerators must be used at each intermediate central office repeater point for a complete two-way

EXHIBIT R- (JF-21) CASE NO. U-13796

Chapter 22

Digital Transmission Lines

i

Digital Systems

A number of wire-pair and coaxial transmission media are used to transmit the various signals of the digital hierarchy. In each case, regenerative repeaters are installed at specified intervals along the line to amplify, retime, and regenerate the pulses for transmission t o the next repeater or to a terminal.

The facilities developed for this mode of operation have all been designated as T-type systems. Those presently in use include the T1 and Tl/OS, TlC, T2, and T4M Digital Transmission Systems which transmit signals a t the DS1, DSlC, DS2, and DS4 rates respectively.

Most of these systems provide trunk facilities in metropolitan and suburban areas; there are also some applications in the loop plant. The distances over which these systems can operate economically are being increased and some systems now in service extend well beyond metropolitan and suburban areas.

The layout of digital lines is based on repeater spacings established by line loss, interference, bandwidth, and the provision of adequate margins. Line spans that encompass a number of repeater sections are based on operations and maintenance considerations and on the limitations of supplying power to remote repeaters over the transmission conductors. The provision of fault locating facilities, maintenance spare lines, and protection switching arrangements are other aspects of digital line layout that must be considered.

22-1 THE T1 DIGITAL TRANSMISSION SYSTEM The first digital system to find acceptance and the one that is in

most common use i s the T1 Digital Transmission System. The field

589

EXHIBIT R- -(JF-2 1) CASE NO. U-13796

590 Digital Systems Vol. 2

of application is primarily t o provide switched network trunks between central offices up to about 50 miles apart [l]. Some systems a re as short as five miles but, in general, T1 systems are more economical than voice-frequency facilities a t distances of 10 miles or more. The T1 line equipment has also been found economical for loop applications such as those described for the SLC-40 system discussed in Chapter 3 [2].

While T1 was initially limited to a maximum length of about 50 miles by operating and maintenance considerations, i t may now be used over distances of up to about 200 miles. A maximum of 200 regenerative repeaters comprising several power spans may be connected in tandem. This increase has been realized by applying more stringent engineering rules to ensure meeting performance objectives and by the addition of operating and maintenance features needed fo r longer routes. Also, improved features and equipment needed for small installations and more economical central office equipment arrangements are provided. In such applications, the system is called T1 Outstate (Tl/'OS) [3].

The signal transmitted over a T1 carrier line must have a repeti- tion rate of 1.544 Mb/s, a bipolar, 50-percent duty cycle format, and contain no more than 15 consecutive 0s. These constraints on signal characteristics must be applied in the system terminal equipment but are necessary in order to assure satisfactory operation of regenerative repeaters located along the line.

Transmission Media

A wide variety of multipair exchange cables may be used f o r T1 systems. These include polyethylene-insulated conductor (PIC) cables, pulp-insulated copper-pair cables of 19, 22, 24, and 26 gauge, and aluminum conductor cables of 17 and 20 gauge. In most cases, the two directions of transmission can be carried in the same cable provided they are in different cable units, o r binder groups, in order t o control crosstalk between systems. Engineering rules specify the manner in which such separation must be accomplished but, even within these rules, a limit is imposed on the number of pairs that can be used as T1 lines.

In some cases, the crosstalk between cable pairs is excessive due to the assignment of too many T1 lines per cable. I t is then mandatory that the two directions of transmission be carried in separate cables

E I( a i!

R

2

3e- .m S CO-

o r 'OP jed

50 be re- on- ore ves f o r for ar- led

?ti- ind nal but 'ra-

T1 les, m d the iro- ' t o the ven can

2 to ory bles

Chap. 22 Digital Transmission Lines 59 1

o r in cables with shield,ing between binder groups to provide suffic7ent isolation so that cable fill of T i lines is equivalent to that obtained by using separate Cables. Such cables are called screened c.ables. Some systems are operated with the two directions of transmission in the same binder group but with reduced repeater spacing.

, ,

. .

Metropolitan area trunk (MAT) cable was designed .to.. optimize performance for T1 systems and t o minimize costs for all types of metropolitan area circuits. This cable is most useful in heavily populated areas requiring large cross-sections of circuits between central offices. [4]. . .

In addition t o t runk applications, TI system line equipment is used in the loop plant to carry SLC-40 signals using standard loop plant cables. Great care must be taken in all applications to be sure , tha t ,cable pairs have been checked fo r satisfactory operating conditions. All bridged taps, load coils, and line build-out networks must have been removed. I n many cases, voice-frequency equipment will have been used on the pairs involved for some previous service and all of this equipment must also be disconnected.

. .

The standard administration of cables in the loop plant is not as well suited to T1-system transmission as that in the trunk plant. Loop cabIes are generally not operated under gas pressure and are therefore more prone to impairment due t o moisture. A cable route from the central office t o a remote area may be made up of mixed gauges. Binder group integrity may be lost since it is less important in the loop plant than in the trunk plant. All of these factors must be examined carefully and the cable condition upgraded where deficient if the T1 line equipment is to operate satisfactorily.

Regenerative Line Repeater

The T1 system utilizes two general types of repeaters called line and central office repeaters. Line repeaters, designed fo r use in man- holes or on telephone poles, have circuits which include two regenerators arranged for either one or two directions of transmission. These repeaters a re generally assembled in apparatus cases that house 25 repeaters (50 regenerators). The other type, designed f o r use in central office repeater bays, contains circuits to regenerate the signals on only one line. Thus, two such regenerators must be used at each intermediate central office repeater point f o r a complete two-way

i EXHIBIT R- -(JF-21) CASENO. U-13796