Strategies and technologies for spectrum utilisation and sharing

Page 1 (36)

Project Number: CELTIC / CP5-026

Project Title: Wireless World Initiative New Radio – WINNER+

Document Type: P (Public)

Document Identifier: D3.3

Document Title: Strategies and technologies for spectrum utilisation and sharing aspects of IMT

Source Activity: WP3

Editor: Miia Mustonen

Authors: Matthias Siebert, Albena Mihovska, Eiman Mohyeldin, Pierre Nguyen, Jean-Philippe Desbat, Miia Mustonen and Werner Mohr

Status / Version: 1.0

Date Last changes: 30.03.10

File Name: D3.3_v1.0.doc

Abstract:

This deliverable introduces spectrum utilisation and sharing aspects of IMT as they are considered within WINNER+ project. First, some of the ongoing activities and results regarding spectrum in different regulatory bodies are introduced. Second, one of the key features of LTE-Advanced enabling use of wide bandwidths, namely carrier aggregation is described. Third, sharing options in wireless networks on different levels infrastructure, core network, site, and spectrum sharing are introduced.

Keywords:

spectrum, sharing, ITU-R, CEPT/ECC, ETSI RRS, carrier aggregation, LTE-Advanced, infrastructure sharing, core network sharing, site sharing, cognitive radio system, software defined radio

WINNER+ D 3.3

Version: 1.0 Page 2 (36)

Authors Partner Name Phone / Fax / e-mail CTIF-Aalborg University Albena Mihovska Phone: +45 9940 8639 Fax: +45 9815 1583 e-mail: [email protected] Deutsche Telekom AG Matthias Siebert Phone: +49 228 936 18575 Fax: +49 228 936 881752 e-mail: [email protected] Nokia Siemens Networks Werner Mohr Phone: +49 89 5159 35117 GmbH & Co. KG Fax: +49 89 636 75121 e-mail: [email protected] Nokia Siemens Networks Eiman Mohyeldin Phone: +49 89 5159 39340 GmbH & Co. KG Fax: +49 89 636 75121 e-mail: [email protected]

Orange Labs Jean-Philippe Desbat Phone: +33 1 45 29 48 99 e-mail: [email protected]

Orange Labs Pierre N'Guyen Phone: +33 1 45 29 66 50 e-mail: [email protected]

VTT Miia Mustonen Phone: +358 40 832 4347 (editor) Fax: +358 20 722 2320 e-mail: [email protected]

mailto:[email protected]







WINNER+ D 3.3


List of abbreviations and symbols 3GPP 3rd Generation Partnership Project AC Allocation Constraint ACK Positive Acknowledgment ACLR Adjacent Channel Leakage Power Ratio ACS Adjacent Channel Selectivity BC Bandwidth Constraint BEM Block Edge Mask BS Base Station BSC Base Station Controller CC Component Carrier CCE Control Channel Element CEPT European Conference of Postal and Telecommunications Administrations CL Circular Letter CN Core Network CPC Cognitive Pilot Channel CPG PTA Conference Preparatory Group Project Team A CPM Conference Preparatory Meeting of the ITU-R CQI Channel Quality Indicator CR Cognitive Radio CRRM Common Radio Resource Management CRS Cognitive Radio Systems DB Dual-Band DC Dual-Cell DFT-S-OFDM Discrete Fourier Transform Spread Orthogonal frequency-division multiplexing DL Downlink DRX Discontinuous Reception ECC Electronic Communication Committee ETSI RRS European Telecommunications Standards Institute Reconfigurable Radio Systems

group E-UTRA Evolved Universal Terrestrial Radio Access EVM Error Vector Magnitude FA Functional Architecture FDD Frequency Division Duplexing FRC Fixed Reference Channel GAP General Assignment Problem GMBS General Multi-Band Scheduling GPRS General Packet Radio Service GWCN Gateway Core Network HARQ hybrid automatic repeat request HSDPA High-Speed Downlink Packet Access HSUPA High-Speed Uplink Packet Access HSPA High-Speed Packet Access IMT International Mobile Telecommunications IP Integer Programming ITU International Telecommunication Union ITU-R International Telecommunication Union Radiocommunication Sector

WINNER+ D 3.3


LTE Long Term Evolution MBS Multi-Band Scheduling MCC Mobile Country Code MCS Modulation and Coding Scheme MIMO Multiple-Input and Multiple-Output MME Mobility Management Entity MOCN Multi-Operator Core Network MO-GAP Multiple Objectives General Assignment Problem MPR Maximum Power Reduction MSC Mobile Switching Centre MVNO Mobile Virtual Network Operators NACK Negative Acknowledgment PBCH Physical Broadcast Channel PCFICH Control Format Indicator CHannel PDCCH Physical Downlink Control Channel PDSCH Physical Downlink Shared Channel PDN GW Packet Data Network Gateway PDPC Packet Data Convergence Protocol PF Profit Function PHICH Physical HARQ Indicator Channel PLMN Public Land Mobile Network PRB Physical Resource Block PSS Primary Synchronization Signal PUCCH Physical Uplink Control Channel PUSCH Physical Uplink Shared Channel QoS Quality of Service RA Resource Allocation RACH Random Access Channel RAN Radio Access Network RAT Radio Access Technology RB Resource Block RF Radio Frequency RLC Radio Link Control RNC Radio Network Controller RRC Radio Resource Control RRM Radio Resource Management RSRP Reference Signal Received Power SA Spectrum Aggregation SDR Software Defined Radio SGSN Serving GPRS support node SGW Serving Gateway SE43 Project Team of CEPT – Cognitive radio systems - White spaces (470 - 790 MHz) SEM Spectrum Emissions Mask SLA Service Level Agreement SIR Signal-to-Interference Ratio SSS Secondary Synchronization Signal TDD Time Division Duplexing TR Technical Report

http://en.wikipedia.org/wiki/Random_Access_Channel

WINNER+ D 3.3


TTI Transmission Time Interval UCI Uplink Control Information UE User Equipment UL Uplink VLR Visitor Location Register WG FM Frequency Management Working Group WP 1B Working Party 1B of ITU-R – Spectrum management methodologies and economic

strategies WP 5A Working Party 5A of ITU-R – Land mobile service excluding IMT; amateur and

amateur-satellite service WRC World Radiocommunication Conference of the ITU-R

WINNER+ D 3.3


Table of Contents

1. Introduction............................................................................................ 7

2. Spectrum Regulation............................................................................. 8 2.1 The status of the band plans for bands identified in WRC-07 ................................................ 8

2.1.1 UHF bands Plan in CEPT ................................................................................................ 8 2.2 Activities on Cognitive Radio Systems................................................................................... 9

2.2.1 Activities in ITU-R .......................................................................................................... 9 2.2.2 Activities in CEPT/ECC ................................................................................................ 10 2.2.3 Activities in ETSI RRS.................................................................................................. 10

2.2.3.1 Activities in ETSI RRS under the original mandate (2008-2009)......................... 10 2.2.3.2 Activities in ETSI RRS under the extended mandate (2009+).............................. 11

3. Spectrum and carrier Aggregation..................................................... 12 3.1 Introduction on carrier aggregation....................................................................................... 12

3.1.1 IMT-Advanced Background .......................................................................................... 12 3.1.2 Rationale and Scope....................................................................................................... 13

3.2 Carrier aggregation basics and issues ................................................................................... 13 3.2.1 Carrier Aggregation Principle and scenarios ................................................................. 13 3.2.2 Definition on the component carrier types and channel bandwidth ............................... 16 3.2.3 Channel bandwidth and component carrier characteristics............................................ 17 3.2.4 3GPP Work Item Description: Carrier Aggregation for LTE ........................................ 18 3.2.5 Carrier aggregation issues.............................................................................................. 18

3.2.5.1 RF and radio performances problematic ............................................................... 18 3.2.5.2 Component carrier coverage ................................................................................. 22 3.2.5.3 Initial access, system information and RACH procedures .................................... 22 3.2.5.4 Control signalling design....................................................................................... 23 3.2.5.5 Uplink power control ............................................................................................ 25 3.2.5.6 Discontinuous reception ........................................................................................ 26 3.2.5.7 Component carrier activation/deactivation............................................................ 26 3.2.5.8 Mobility handling.................................................................................................. 26

3.3 Spectrum Aggregation with Multi-Band User Allocation over Two Frequency Bands........ 27 3.3.1 Future work.................................................................................................................... 30

4. Sharing options in mobile wireless networks................................... 31 4.1 Sharing examples .................................................................................................................. 31 4.2 Sharing motivation................................................................................................................ 32 4.3 Spectrum/Carrier Sharing ..................................................................................................... 32

5. References ........................................................................................... 35

WINNER+ D 3.3


1. Introduction This deliverable introduces the strategies and technologies for spectrum utilisation and sharing aspects of International Mobile Telecommunications (IMT) as they are considered within WINNER+ project. In World Radio Conference (WRC) 2007, additional frequency spectrum was identified for mobile and wireless applications on several different bands. However, these bands have very different propagation conditions which also results in different economic impact on systems deployment in these bands. In particular, the support of very wideband carriers for high system throughput is difficult under the additional requirement of competition between operators. With respect to these constraints and the identified frequency bands, new means of spectrum usage are investigated in this deliverable in order to support high throughput, competition and also to enable an economic system deployment. The finalization of the deliverable was delayed by one week in order to accommodate latest information coming from 3GPP WG RAN 1 # 60, RAN 2 #69 and RAN 4 #54 held from 22nd to 26th February 2010 in San Francisco. The rest of the deliverable is organized as follows. In Chapter 2, some of the ongoing activities and results regarding spectrum in different regulatory bodies are introduced. First, the status of the band plans for bands identified in WRC-07 is briefly described and more detailed information is given about the European Conference of Postal and Telecommunications Administrations (CEPT) position on the UHF band. Then the activities regarding cognitive radio systems are explained in three different groups: International Telecommunication Union Radiocommunication Sector (ITU-R), CEPT and European Telecommunications Standards Institute Reconfigurable Radio Systems (ETSI RRS) group. Chapter 3 describes spectrum aggregation which can be considered as one of the key features of Long Term Evolution (LTE) Advanced to support wide bandwidths. First, an introduction is given to the topic and then more detailed description is given about some of the radio layer 1 and 2 aspects of LTE-Advanced that arise from the fact that user equipment (UE) can receive and transmit from/on more than one carrier. Lastly, this chapter describes spectrum aggregation with multi-band user allocation over two frequency bands. Sharing options in wireless networks are introduced in Chapter 4. First, examples on different levels of sharing are given. Concepts of infrastructure, core network, site, and spectrum sharing are introduced. Then some motivation is given to sharing and lastly spectrum sharing is explained in more detail.

WINNER+ D 3.3


2. Spectrum Regulation

2.1 The status of the band plans for bands identified in WRC-07 The World Radio Conference (WRC) 2007 has identified the following new spectrum bands for International Mobile Telecommunications (IMT) systems (which consists of both IMT-2000 and IMT-Advanced), some of which Region1 specific:

• 450-470 MHz globally, • 698-806 MHz in Region 2 and nine countries in Region 3, • 790-862 MHz in Region 3 and part of Region 1 countries, • 2.3-2.4 GHz globally, • 3.4-3.6 GHz in a large number of countries in Regions 1 and 3.

The intention is that the bands previously identified in the Radio Regulations for IMT-2000 are now identified for IMT. The work on the band plans on these bands is ongoing in ITU-R Working Party (WP) 5D on these bands and the finalization of the plans is set to the 10th meeting of ITU-R WP5D in February 2011. In Region 1, the plan for 790-862 MHz band (also known as the UHF band) is already ready and it has been described below in Section 2.1.1.

2.1.1 UHF bands Plan in CEPT WRC-07 allocated on a primary basis the 790 – 862 MHz band to mobile services throughout Region 1. CEPT has developed one preferred harmonized frequency arrangement based on the frequency-division duplexing (FDD) mode, but for Administrations that might wish to deviate from the preferred harmonized frequency arrangement some other approaches to meet specific national circumstances and market demand are developed. CEPT report 31 [CEPT31] provides the technical condition for the use of the band 790-862 MHz and benefits and risk of the different option. In the following is summary on the preferred arrangement option, and the other approaches for individual administrations for the band based on Electronic Communication Committee (ECC) decision (ECC/DEC/(09)03): 1. Preferred harmonized frequency arrangements for the band 790-862 MHz To meet the technical conditions for this band a frequency separation is needed. Both 1 and 2 MHz are viable options for frequency separation at 790 MHz in the context of Base Station compliance with a regulatory Block Edge Mask (BEM) baseline of 0 dBm/(8 MHz), with the 1 MHz option implying larger filters. There is a trade off between increasing the frequency separation at 790 MHz and reducing the duplex gap. In weighing up this trade off it has been decided that the frequency separation should be 1 MHz and the duplex gap 11 MHz. CEPT has concluded that the preferred harmonized frequency arrangement is 2 x 30 MHz with a duplex gap of 11 MHz, based on a block size of 5 MHz, paired and with reverse duplex direction, and a guard band of 1 MHz starting at 790MHz. The FDD downlink starts at 791 MHz and FDD uplink starts at 832 MHz, see Figure 1 below:

790- 791

791-796

796- 801

801-806

806- 811

811-816

816- 821

821- 832

832- 837

837- 842

842- 847

847- 852

852- 857

857- 862

Guard band Downlink Duplex

gap Uplink

1 MHz 30 MHz (6 blocks of 5 MHz) 11 MHz 30 MHz (6 blocks of 5 MHz)

Figure 1: Channeling arrangements for 790-862 MHz band in CEPT (FDD option). 2. Approaches for individual administrations meeting the specific national circumstances and market

demand

1 Region 1: Europe, the Middle East and Africa (EMEA)

Region 2: North- and South America (Americas)

Region 3: Asia Pacific (APAC)

WINNER+ D 3.3


Administrations which do not wish to use the preferred harmonized frequency arrangement or which do not have the full band 790-862 MHz available (e.g. cases, where an Administration cannot make all channels in the band available because they have already been allocated to other services or are not able to coordinate the use of frequencies with neighboring countries), may consider:

− partial implementation of the preferred harmonized frequency arrangements; − the introduction of the time division duplexing (TDD) frequency arrangement in all or part of the

frequency band 790 – 862 MHz, based on a block size of 5 MHz starting at 797 MHz; − a mixed introduction of TDD and FDD frequency arrangements; − implementation of 1 MHz channel raster.

Figure 2 below illustrates this option:

790-797 797-802

802-807

807-812

812-817

817-822

822-827

827-832

832- 837

837- 842

842- 847

847-852

852-857

857-862

Guard band Unpaired

7 MHz 65 MHz (13 blocks of 5 MHz)

Figure 2: Channeling arrangements for individual administrations (TDD option).

2.2 Activities on Cognitive Radio Systems Cognitive radio systems are attracting all the time more attention in the research world as the new promising technology that is able to provide more efficient use of spectrum and to promote the introduction of new services. While introducing the new systems capable for more intelligent and flexible spectrum use, it is also important to preserve the rights of the current spectrum users. Therefore, the introduction of cognitive radio techniques requires acceptance of the regulatory domain. This subsection will give an overview of ongoing studies in different regulatory domains: ITU-R, CEPT/ECC and ETSI RRS, respectively.

2.2.1 Activities in ITU-R As an evidence that the regulatory domain does not maintain ignorant to this new emerging technology in ITU-R WRC-07 an agenda item was developed regarding cognitive radio systems (CRS) for the WRC-12. This agenda item is 1.19 to consider regulatory measures and their relevance, in order to enable the introduction of software-defined radio and cognitive radio systems, based on the results of ITU-R studies, in accordance with Resolution 956 [COM6/18]. The group ITU-R WP 1B “Spectrum management methodologies and economic strategies” has the main responsibility for this agenda item. ITU-R WP 5A “Land mobile service excluding IMT; amateur and amateur-satellite service” is responsible for the technical work regarding this agenda item. In addition, other groups can participate, if they see it necessary. As a first step of the technical work a report SM.2152 from ITU-R WP 5A was published in September 2009 [SM.2152]. This report contains the definitions of Software Defined Radio (SDR) and Cognitive Radio System. A CRS is defined as “a radio system employing technology that allows the system to obtain knowledge of its operational and geographical environment, established policies and its internal state; to dynamically and autonomously adjust its operational parameters and protocols according to its obtained knowledge in order to achieve predefined objectives; and to learn from the results obtained.” To further facilitate the technical work, ITU-R WP 5A is developing an ITU-R report “Cognitive radio systems in the land mobile service”. The purpose of this report is to provide an answer to the questions posed by the ITU Radiocommunication Assembly at 2007. These questions include following aspects of the CRS: definition, closely related radio technologies and their functionalities, key technical characteristics, requirements, performance, benefits, the potential applications, the operational implications, capabilities that facilitate coexistence with existing systems, possible spectrum-sharing techniques and the effect on the efficient use of radio resources. The report is targeted to be finalized by the end of 2010 and it will then be used as a technical guidance by ITU-R WP 1B when drafting the conference preparatory meeting (CPM) text in response to the agenda item 1.19 of the WRC-12. To introduce cognitive radio technologies into the spectrum regulatory framework in the future, it is important to cooperate between the research domain and the regulatory. The research results on cognitive radio systems need to be communicated to the regulatory domain that is in charge of controlling spectrum related matters. New performance indicators are needed to ensure sufficient protection of current spectrum use and evaluation of technical constraints for coexistence cognitive radio and existing systems.

WINNER+ D 3.3


2.2.2 Activities in CEPT/ECC Deployment of CRS is expected to offer additional flexibility and improved efficiency to overall spectrum use in the future. As use of spectrum is further intensifying, it is important to find alternative ways to make the available spectrum more efficiently utilized. Cognitive capabilities like sensing, access to database (in connection with geolocation), use of cognitive pilot channel (CPC), transmission power control, etc. can form a CRS capability toolbox and could facilitate coexistence of CRS with existing radio systems. These capabilities need to be addressed and defined in order to evaluate the possibility and degree of coexistence of CRS with existing systems and with other radiocommunications services. CRS technologies may enable coexistence/sharing in bands where it was previously determined to be not feasible. As CRS is new, it is important to study means to observe the interference that they may generate and which type of performance indicators are needed to describe whether sufficient protection of other users is achieved. Also evaluation of technical parameters for coexistence between CRS and existing services in different/adjacent frequency bands is needed. Monitoring aspect may also need the development of means to observe the interference that they may be generated as it is the case for any innovation in radio environment. Overall spectrum efficiency of CRS would also be a topic for further studies. This may depend, in particular, of the way in which spectrum will be shared between the various CRSs (for example, shall all the CRSs operating in a particular area and in a particular frequency use the same method to access spectrum with the same parameters). Technical approaches of CRS have to be described and considered, too. CEPT/ECC discussed these cognitive radio issues within the ECC structure. Accordingly , ECC has the following groups that are studying cognitive radio and white space:

SE43: is defining the relevant technical and operational requirements to be considered for the operation of cognitive radio systems in the white spaces of the UHF broadcasting band (470-790 MHz. On the basis of the determining the technical requirements, SE43 will investigate the consequential amount of spectrum potentially available as “white space”. It is expected that the outcome from the SE43 work on white space will be frequency dependent, specific to the considered sharing scenarios and to the national deployment scenarios of systems/services to be protected. However, some of the issues considered in SE43 (such as deployment scenarios, general requirements for sensing or considerations related to the use of database) may be applicable for other frequency bands.

Frequency Management Working Group (WG FM): will initiate a work to identify possible candidate bands for CR. WG FM is considering the following aspects:

o The characteristics of foreseen CRS? o Studying cognitive features that ensure compatibility and optimise spectrum o Identify CRS most appropriate bands o Investigate plans or information relating to initial developmental/experimental work for

CRS Conference Preparatory Group Project Team A (CPG PTA): is preparing the draft CEPT brief

on AI1.19. The work of CPG PTA concentrates on general regulatory and technical issues of WRC-12 preparation.

2.2.3 Activities in ETSI RRS European Telecommunications Standards Institute Reconfigurable Radio Systems (ETSI RRS) is a technical committee within the ETSI dealing with software defined radio and cognitive radio (CR) issues. In fact, RRS is considered as the main centre of competence within ETSI with respect to these reconfiguration topics.

2.2.3.1 Activities in ETSI RRS under the original mandate (2008-2009) Being established in early 2008, the original mandate of RRS was expressed in the terms of reference to

to study the feasibility of standardization activities related to Reconfigurable Radio Systems encompassing radio solutions related to SDR and CR research topics;

to collect and define the related Reconfigurable Radio Systems requirements from relevant stakeholders;

to deliver its findings in the form of Technical Reports or ETSI Guides. First version of its findings will be produced within 18-24 months of commencing activity.

Work was hence organized in 4 working groups covering WG1 – System Aspects (NOKIA)

WINNER+ D 3.3


WG2 – Radio Equipment Architecture (Infineon Technologies) WG3 – Functional Architecture and Cognitive Pilot Channel (Alcatel Lucent) WG4 – Public Safety (Copsey Telecommunications)

WG1 was focused on the development of a CR system concept and general CR spectrum aspects. Further, WG1 was meant to maintain the big picture and thus to consolidate efforts and work in other working groups. Potential regulatory aspects of CR and SDR systems have been identified by WG1 and released in a related technical report [TR 102 803]. WG2 concentrated on a terminal SDR reference architecture [TR 102 680] as well as implementations & costs aspects for software defined radio base stations [TR 102 681]. Ongoing work covers a study of flexibility implications on power efficiency and the specification of a multiradio interface for an SDR mobile device architecture and services. WG3 was dedicated to functional architecture and cognitive pilot channel studies. Two respective technical reports have been produced: Functional Architecture (FA) for the Management and Control of Reconfigurable Radio Systems [TR 102 682] and CPC design [TR 102 683]. WG4, finally, was established to consider increased attention of public safety topics with respect to SDR/CR. Among other, national authorities have a vital interest in taking benefit from SDR/CR technology in case of e.g. disaster scenarios. Two respective technical reports “System Aspects for Public Safety” [TR 102 733] and “User Requirements for Public Safety” [TR 102 745] have been produced, accordingly.

2.2.3.2 Activities in ETSI RRS under the extended mandate (2009+) In autumn 2009, ETSI RRS has completed its first main activities on feasibility studies related to standardization of Cognitive Radio & Software Defined Radio as part of its initial Terms of Reference (ToR) mandate. A comprehensive summary on all related activities has been compiled and is available as technical report “Summary of feasibility studies and potential standardization topics” [TR 102 838]. While the initial mandate of RRS was basically limited to feasibility studies, an extended mandate has been given in late 2009 to start with “real” standardization work. The specification of a multiradio interface as already started in WG2 is going in that direction. Further, specification of white space usage in the UHF-TV band is another hot topic currently under discussion. As a general working methodology, ETSI RRS will adopt a horizontal approach in close collaboration with the specific committees. Horizontal approach thereby means to integrate existing dedicated technology solutions as much as possible to enable SDR/CR flexibility and only in exceptional cases RRS will involve itself in defining specific parameters. Examples of such exceptional cases might be (out of band) Cognitive Pilot Channel and White Spaces.

WINNER+ D 3.3


3. Spectrum and carrier Aggregation Spectrum and more recently carrier aggregation have gained a particular importance for the successful adoption of IMT-Advanced systems. The concept of spectrum aggregation consists of exploiting multiple, small spectrum fragments simultaneously to deliver a wider band service (i.e., not otherwise achievable when using a single spectrum fragment. Spectrum aggregation can appear when an operator’s dedicated band is not continuous but is split in two or more parts. In addition, spectrum aggregation can happen in scenarios, in which an operator accesses both a dedicated band, and a spectrum sharing band which is separated in frequencies from the dedicated operator’s band. Spectrum aggregation allows that new high data rate wireless communication systems can coexist while reusing the spectrum of legacy systems. This is also valid for the inter-operator scenario. In this context, it can be very beneficial to explore the scenarios for joint use of spectrum aggregation techniques and radio resource management at higher RAN levels. In similar way, carrier aggregation allows contiguous or non-contiguous aggregation of carriers to achieve support of wideband services in a fragmented spectrum environment. Section 3 explores the benefits of carrier and spectrum aggregation and identifies the research challenges for these areas.

3.1 Introduction on carrier aggregation. Carrier aggregation is one of the key features of LTE-Advanced to support wider bandwidth than supported by Rel-8 LTE. This document will only focus on the carrier aggregation feature in LTE or LTE-Advanced and not in High-Speed Packet Access (HSPA). This document is prior to the technical specification of this functionality intended for release 10. The intent is to understand the problematic by analysing manufacturers input in 3GPP. This means that further update of this document after release 10 completion will be interesting. First, light description of LTE-Advanced is presented here in order to understand the genesis of the Carrier aggregation work item (from section 3.1.2 to 3.1.4). The last section 3.1.5 describes the content of the work item description.

3.1.1 IMT-Advanced Background The 4G succeeds to 3G and 2G standards, with the aim to provide ultra-broadband (gigabit-speed) internet access to mobile. It refers to IMT-Advanced (International Mobile Telecommunications Advanced), as defined by ITU-R. Targets for LTE-Advanced were agreed in [TR 36.913] [RP-080445]. Main points of importance are:

– LTE Advanced shall be an evolution of LTE. LTE terminals shall be supported in LTE-advanced networks. An LTE-Advanced terminal can work in an LTE part of the network.

– Primary focus of LTE-Advanced is low mobility users. – All requirements and targets in [TR 25.913] apply to LTE-Advanced. – LTE-Advanced requirements shall fulfil IMT-Advanced requirements within the ITU-R time plan

(discussions about the detailed LTE-Advanced performance requirements are scheduled). – For LTE-Advanced mobility and interworking:

o Same inter-RAT interworking capability with at least same performance as in LTE Release 8.

o Intra-RAT handover performance shall be same or better than LTE Release 8. The main detailed performance requirements and a comparison to LTE Release 8 and ITU-R requirements are outlined in the Table 1:

Table 1: Performance requirements of LTE-Advanced.

b/s/Hz =

Mbit in 10 MHz per user (10 users per cell)

LTE performance Requirement in ITU-R

3GPP Target

[TR36.913]

DL 15 (4x4)

7.5 (2x2) 15 (4x4) 30 (8x8) Peak data

rate UL 2.5 (1x2, 16 QAM)

3.6 (1x2, 64 QAM) 6.75 (2x4) 15 (4x4)

Cell average DL 1.63 (2x2) 2.2 (8x2)* 2.4 (2x2)

WINNER+ D 3.3


UL 0.86 (1x2) 1.4 (2x8)* 1.2 (1x2)

Control plane ~100 ms less than 100 ms less than 100 ms

Latency User plane ~ 5 ms (one way) < 5ms (FDD)

< 10 ms (TDD) < 5ms (FDD)

< 10 ms (TDD)

Bandwidth up to 20 MHz > 40 MHz

Up to 100 MHz

(Including already carrier aggregation)

* Maximum antenna configuration when beamforming is considered. As it can be seen from Table 1, the main areas where LTE needs to be improved to meet ITU-R requirements are the support of higher bandwidths and higher uplink peak rates (i.e. uplink single user Multiple-Input and Multiple-Output (MIMO)). In order to achieve larger transmission bandwidths (40 MHz and up to 100 MHz), LTE-Advanced user equipment (UE) will have the ability to aggregate two or more carriers, which is called carrier aggregation. Such carriers are referred to as component carriers. RAN4 is now working on the practical scenarios regarding the supported combination of bands and bandwidths.

3.1.2 Rationale and Scope In order to cope with IMT-Advanced criteria, discussions on scenarios came on the RAN 4 table on the second half of 2009. But, quickly, with regard to the complexity of some scenarios, a first question had to be raised: which level of complexity can we accept, and for what? Except that LTE-Advanced has to be IMT-Advanced, many questions occur such as:

- Can we expect any gain in term of spectral efficiency? - Can we expect gain at cell level or only higher rates at user plan level? - Which level of complexity do we expect at UE or Node B radio frequency (RF) side

depending on the number of carriers, different bands, and asymmetric uplink/downlink (UL/DL) scenarios?

This document was then seen as an urgent need to understand the basics of this feature. Hence, the main objectives of this study are:

– To provide a high-level explanation of carrier aggregation basics and issues. Following this introductory section, the second part of this document establishes the carrier aggregation basic definitions and issues. In this part a high level explanation of carrier aggregation is given. In the first section, the carrier aggregation principle is explained and illustrated with RAN4 practical scenarios. The second section provides the definitions needed for a good comprehension of the carrier aggregation. The third section focuses on the component carrier characteristics. The last section provides a high level explanation of the issues raised by carrier aggregation.

3.2 Carrier aggregation basics and issues

3.2.1 Carrier Aggregation Principle and scenarios LTE-Advanced has been motivated by the additional IMT spectrum band identified in WRC-07 and by the support of wider bandwidths (high data rates).

WINNER+ D 3.3


Table 2: 0perating bans for LTE-Advanced (E-UTRA operating bands)

Uplink (UL) operating band BS receive/UE transmit

Downlink (DL) operating band BS transmit /UE receive Operating

Band FUL_low – FUL_high FDL_low – FDL_high

Duplex Mode

1 1920 MHz – 1980 MHz 2110 MHz – 2170 MHz FDD 2 1850 MHz – 1910 MHz 1930 MHz – 1990 MHz FDD 3 1710 MHz – 1785 MHz 1805 MHz – 1880 MHz FDD 4 1710 MHz – 1755 MHz 2110 MHz – 2155 MHz FDD 5 824 MHz – 849 MHz 869 MHz – 894MHz FDD 6 830 MHz- – 840 MHz- 865 MHz – 875 MHz- FDD 7 2500 MHz – 2570 MHz 2620 MHz – 2690 MHz FDD 8 880 MHz – 915 MHz 925 MHz – 960 MHz FDD 9 1749.9 MHz – 1784.9 MHz 1844.9 MHz – 1879.9 MHz FDD

10 1710 MHz – 1770 MHz 2110 MHz – 2170 MHz FDD 11 1427.9 MHz – 1447.9 MHz 1475.9 MHz – 1495.9 MHz FDD 12 698 MHz – 716 MHz 728 MHz – 746 MHz FDD 13 777 MHz – 787 MHz 746 MHz – 756 MHz FDD 14 788 MHz – 798 MHz 758 MHz – 768 MHz FDD 15 Reserved Reserved - 16 Reserved Reserved - 17 704 MHz – 716 MHz 734 MHz – 746 MHz FDD 18 815 MHz – 830 MHz 860 MHz – 875 MHz FDD 19 830 MHz – 845 MHz 875 MHz – 890 MHz FDD 20 832 MHz – 862 MHz 791 MHz – 821 MHz FDD 21 1447.9 MHz – 1462.9 MHz 1495.9 MHz – 1510.9 MHz FDD 22 3410 MHz 3500 MHz 3510 MHz 3600 MHz FDD ... 33 1900 MHz – 1920 MHz 1900 MHz – 1920 MHz TDD 34 2010 MHz – 2025 MHz 2010 MHz – 2025 MHz TDD 35 1850 MHz – 1910 MHz 1850 MHz – 1910 MHz TDD 36 1930 MHz – 1990 MHz 1930 MHz – 1990 MHz TDD 37 1910 MHz – 1930 MHz 1910 MHz – 1930 MHz TDD 38 2570 MHz – 2620 MHz 2570 MHz – 2620 MHz TDD 39 1880 MHz – 1920 MHz 1880 MHz – 1920 MHz TDD 40 2300 MHz – 2400 MHz 2300 MHz – 2400 MHz TDD 41 3400 MHz – 3600 MHz 3400 MHz – 3600 MHz TDD

Note: Frequency arrangement for certain operating bands in

WINNER+ D 3.3 Table 2 may be modified, eg. split into sub-bands, according as the future studies. Introduction of the following other ITU-R IMT bands are not precluded in the future. (a) Possible frequency bands in 3.4-3.8 GHz band (b) Possible frequency bands in 3.4-3.6GHz as well as 3.6-4.2GHz (c) Possible frequency bands in 3.4-3.6 GHz band (d) Possible frequency bands in 450−470 MHz band, (e) Possible frequency bands in 698−862 MHz band (f) Possible frequency bands in 790−862 MHz ban (g) Possible frequency bands in 2.3−2.4 GHz band (h) Possible frequency bands in 4.4-4.99 GHz band However, the IMT spectrum identified at WRC-07 is fragmented and does not allow 40 MHz contiguous bandwidth for 4 operators operating in a same geographical area.

Figure 3: Carrier aggregation example[REV-090003r1].

To meet the ITU requirements and to provide a large system bandwidth, the aggregation of two or more component carriers is required. As can be seen in Figure 4 aggregated component carriers (CCs) can be discontinuous or continuous. Carrier aggregation can also apply to smaller bandwidths, e.g. 10+5 MHz, which allows more flexibility in spectrum usage.

Figure 4: Contiguous and non-contiguous carrier aggregation.[REV-09006]

A terminal may simultaneously receive one or multiple component carriers depending on its capabilities. The practical scenarios (based on operator’s input) on which the RAN 4 is working perfectly illustrate the


WINNER+ D 3.3


supported combination of bands and bandwidths. RAN 4 FDD deployment scenarios considered in [REV-090003r1] The scenarios proposed so far for rel10 are: Intra-band Contiguous CA Ø FDD: UL: 40 MHz, DL: 40 MHz in Band 3 (1800 MHz) Ø TDD: UL/DL: 50 MHz in Band 40 (2300 MHz) Inter-band Non-contiguous CA Ø Region 1 ² 40 MHz UL/DL: 20 MHz CC (Band 7) + 20 MHz CC (Band 20) ² 40 MHz UL/DL: 20 MHz CC (Band 3) + 20 MHz CC (Band 20) ² 40 MHz UL/DL: 20 MHz (Band 7) + 20 MHz CC (Band 3) Ø Region 2 ² 20MHz UL/DL: 10 MHz CC (Band 5) + 10 MHz CC (Band 12), FDD ² 10MHz UL/DL:5MHz CC (Band 17) + 5MHz CC (Band 4), FDD ² TBD Ø Region 3 ² 20 MHz UL/DL: 10 MHz CC (Band 1) + 10 MHz CC (Band 18/19) ² 40MHz UL/DL: 20 MHz CC (Band 38) + 20 MHz CC (Band 40) Intra-band Non-contiguous CA Ø FDD: None Ø TDD: None Furthermore, the following basic concepts for carrier aggregation were proposed in last february 3GPP RAN 4 meeting:. Proposal 1: The signaling and protocol specifications to support carrier aggregation shall be designed in a generic way, and able to support carrier aggregation scenarios that are introduced in later Releases. Proposal 2: Specification of carrier aggregation shall be done in Release independent manner. Proposal 3: Other scenarios which are not treated in Release 10 should be captured in appropriate technical report (TR) for future reference. Proposal 4: The carrier aggregation scenarios to be specified in Release 11 should be discussed again in the future when 3GPP start planning Release 11, also based on the actual spectrum auction outcomes. Proposal 5: New work items should be created when new carrier aggregation scenarios are introduced in future 3GPP specifications. With regard to these scenarios, we see at least two different kind of application of carrier aggregation. The first is to provide higher bandwidth than 20 MHz in a given area. The second is to better use the spectrum by scaling bandwidth accordingly to frequency band plan. The following definitions and characteristics are needed for a good comprehension of the carrier aggregation's issues [R1-092575]

3.2.2 Definition on the component carrier types and channel bandwidth In order to aid the discussions on carrier aggregation, the following definitions have been adopted RAN1 #57bis and are used in all RAN groups. Backwards compatible carrier: A carrier accessible to UEs of all existing LTE releases. A backwards compatible carrier can be operated as a single carrier (stand-alone) or as a part of carrier aggregation. For FDD, backwards compatible carriers always occur in pairs, i.e. DL and UL. Non-backwards compatible carrier: If specified, a carrier not accessible to UEs of earlier LTE releases, but accessible to UEs of the release defining such a carrier. A non-backwards compatible carrier can be operated as a single carrier (stand-alone) if the non-backwards compatibility originates from the duplex distance or otherwise as a part of carrier aggregation. Extension carrier: If specified, a carrier that cannot be operated as a single carrier (stand-alone), but must be a part of a component carrier set where at least one of the carriers in the set is a stand-alone-capable carrier. Therefore, an extension carrier is not backward compatible with Rel-8, but complies with the Transmission bandwidth configuration of Table 3. UE DL Component Carrier Set: The set of DL component carriers configured by dedicated signalling on which a UE may be scheduled to receive the physical downlink shared channel (PDSCH) in the DL.

WINNER+ D 3.3 UE UL Component Carrier Set: UE UL component carrier set is the set of UL component carriers on which a UE may be scheduled to transmit the physical uplink shared channel (PUSCH) in the UL. Whether the definition of the UL CC set will be needed in the specifications, is for further studies.

3.2.3 Channel bandwidth and component carrier characteristics In [TS36.101/104] the following terminology and numerology is defined for the Rel-8 evolved universal terrestrial radio access (E-UTRA) channel bandwidth [TR 36.815]: Channel bandwidth: The RF bandwidth supporting a single E-UTRA RF carrier with the transmission bandwidth configured in the uplink or downlink of a cell. The channel bandwidth is measured in MHz and is used as a reference for transmitter and receiver RF requirements. Transmission bandwidth configuration: The highest transmission bandwidth allowed for uplink or downlink in a given channel bandwidth, measured in Resource Block units. Figure 5 and Table 3 show the relation between the Rel-8 E-UTRA Channel bandwidth (BWChannel) and the transmission bandwidth configuration. The channel edges are defined as the lowest and highest frequencies of the carrier separated by the channel bandwidth, i.e. at FC +/- BWChannel /2.

Figure 5: Definition of Channel Bandwidth and Transmission Bandwidth Configuration for one Rel-8 E-UTRA carrier. [TR 36.815]

Table 3: Transmission bandwidth configuration NRB in Rel-8 E-UTRA channel bandwidths. [TR36.815]

Channel bandwidth [MHz] 1.4 3 5 10 15 20

Transmission bandwidth configuration (NRB)

6 15 25 50 75 100

The channel bandwidth is measured in MHz and is used as a reference for transmitter and receiver RF requirements. The Rel-8 E-UTRA definitions related to channel edges (i.e. FC +/- BWChannel /2) can be re-used for LTE-Advanced with respect to the component carriers at the edges of component carrier aggregation scenarios in order to provide a reference for transmitter and receiver RF requirements. It is also expected that LTE-Advanced component carriers will support the channel configurations of Table 3. To sum up, each CC has the following characteristics:

– Each component carrier can be configured as a Rel-8 carrier. – Each component carrier can have any of the bandwidths supported in Rel-8 (1.4, 3, 5, 10, 15, 20

MHz).


WINNER+ D 3.3


– It is possible to aggregate a (UE-specific) different number of CCs of possibly different bandwidths in the UL and the DL.

– In typical TDD deployments, the number of component carriers and the bandwidth of each component carrier in UL and DL will be the same.

3.2.4 3GPP Work Item Description: Carrier Aggregation for LTE It is important to recall the scope of the work item that encompass carrier aggregations specifications in 2010. The following text is a copy and taste of the work item description. The work item should be based on agreements on carrier/spectrum aggregation taken during the LTE-Advanced study item and fulfil the following objectives:

– Specify carrier aggregation in LTE for the following scenarios – Rel-8/9 backward compatible carriers is the basic building block and should be supported;

non-Rel-8/9-backward compatible component carriers and carrier segment to be considered (subject to agreement in RAN1) if sufficient gain can be shown.

– Component carriers can be adjacent or non-adjacent in frequency to address contiguous as well as fragmented spectrum, subject to restrictions set by RAN4.

– UE-specific asymmetric number of component carriers in DL and UL. – Component carriers can have any of the bandwidths supported in Rel-8. – Terminal complexity should be considered; the number of supported bands and band

combinations per region should be limited. – RAN4 should until RAN #47 conclude on which bands and band combinations that should

be targeted within this WI. – Introduce stage-2 description of carrier aggregation in [TS36.300]. – Introduce support of carrier aggregation in stage-3 specifications, including:

– UL and DL control channel structure. – Clustered discrete fourier transform spread orthogonal frequency-division multiplexing

(DFT-S-OFDM) UL transmission scheme and control-data decoupling (simultaneous physical uplink control channel (PUCCH) and PUSCH transmission).

– L1 procedures. – L2/L3 protocols and procedures. – UE and base station (BS) RF core requirements. – Radio resource management (RRM) core requirements.

The detailed specification shall take other work items addressing LTE-Advanced into account [RP-091440].

3.2.5 Carrier aggregation issues Carrier aggregation is one of the key features of LTE-Advanced to support wider bandwidth than supported by Rel-8 LTE and results in several challenges. The main challenges raised by the carrier aggregation feature, are due to the fact that a UE can receive and transmit from/on more than one carrier. Thus in theory, LTE-Advanced introduces the possibility for a UE transmit and receive from more than one cell. This implies some changes in the radio layer 1 and 2 and adds some complexities at the terminal side and in the E-UTRAN.

3.2.5.1 RF and radio performances problematic Depending upon the type and nature of deployment scenarios as well as the components carriers characterized by factors such as: back-ward compatible, non back-ward compatible, number of carriers, contiguous or non-contiguous, intra-band carriers, inter-band carriers… and the way they are combined, the impacts on the current requirement of RF equipments characteristics are not the same. So, the issues below are listed generally. In fact, there are two options to define the number of resource blocks (RBs) per component carrier in carrier aggregation as follows:

1) Extend the component carrier transmit bandwidth configuration from 100 RBs to 100 … 108 RBs

2) Use 100 RBs component carriers and an additional smaller carrier which could be “extension carrier” or “segment carrier” A new Work Item for “Carrier Aggregation for LTE” was agreed in RAN #46 [RP-091440] For Europe Country the following scenario will be used for the Work Item phase:

Intraband, contiguous: FDD: 2x20 MHz at 1800 MHz (Band 3);

TDD: 50 MHz in 2300 MHz (Band 40);

WINNER+ D 3.3


Interbanb : 20MHz at 800 MHz (Band 20) + 20MHz at 2.6 GHz (Band 7)

20MHz at 800 MHz + 20MHz at 1800 MHz (Band 3)

20MHz at 1800 MHz + 20MHz at 2.6 MHz

Intraband, non-contiguous carrier aggregation is a new concept for RAN4 BS specifications and requires appropriate extension of the transmit (e.g. unwanted emissions, adjacent channel leakage power ratio (ACLR)) and receive (e.g. adjacent channel selectivity (ACS), blocking, ...) RF requirements across the “gaps” between component carriers in order to facilitate co-existence between uncoordinated systems.

Coexistence study requirements: Extend number of RBs over 100: One of the issues for TX/RX RF requirements is that the related frequency offsets from the DC sub-carrier of the aggregation edge component carrier to the nominal channel edge will >10 MHz impacting the mentioned requirements and tests. Furthermore, these frequency offsets (used for start of spectrum emissions mask (SEM), NB blockers, etc) will differ for the various aggregation cases (3x20 MHz, 4x20 MHz, 5x20 MHz, etc) complicating the situation further. For the ACS requirement there may be different rejection needed at a certain offset from the aggressor carrier. For adjacent component carriers that are not co-sited, the relative power difference between the component carriers can be an issue if the gap between the component carriers is filled up with RBs. The narrow-band blocking requirement implies more stringent filter requirements in case of 108 RBs An excessive SEM (exceeding -25 dBm/MHz) for close-in transmission with the current 20 MHz mask in uplink. So, more stringent filter requirement must be applied unless the mask is modified. SEM limits are not always the same between ITU-A regions and ITU-B regions Some relaxations are awaited especially for LTE-Advanced UL Some UE Tx architecture could limit the possibility of UL Carrier Aggregation to Intraband case only RRM requirements: Carrier Aggregation will also change some RRM requirements by considering that a LTE terminal will need to measure several carriers at the same time. The RAN2 main conclusion is that UE is able to measure 2 carriers at the same time in active mode including scenario where one of the two carriers is not configured by the operators to be high-speed downlink packet access (HSDPA) capable. This will then enable to perform inter-frequency mobility compressed mode activation (which will reduce the time for mobility). However, RAN4 highlighted that discontinuous reception (DRX) operations and compressed mode activation should not be totally given up at that stage. Indeed, UE battery consumption could be increased due to the activation of such "several" carriers continuously measured. UE RF required in LTE-A [R4-094674]

Section inTS 36.101

Requirements Comments

5.5 Operating bands Combinations of inter-band carrier aggregation would be captured in this section.

5.6 Channel BW It should be decided in the study item phase whether the current R8 channel bandwidth configuration should be re-used or not. Additional smaller carriers which are being discussed in RAN1/2/4 would also be captured in this section.

Channel raster No change 5.7 TX-RX separation Variable TX-RX separation would be needed in carrier

aggregation. The impacts of variable TX-RX separation to other RF requirements should carefully be investigated. Furthermore, it should also be discussed how to define test cases for variable TX-RX separation, i.e. whether or not relevant RF requirements should be tested for all possible TX-RX separation scenarios.

6.2 UE maximum output power, MPR/A-MPR

MPR/A-MPR in case of UL MIMO and carrier aggregation would be one of the key issues in UE RF area. It is noted that in case that fallback to LTE, UE need to achieve the same coverage as LTE Release 8.

WINNER+ D 3.3


Minimum output power No change Transmit ON/OFF power, ON/OFF time mask

No change 6.3

Power control Power control requirements would be affected due to UL carrier aggregation especially due to contiguous carrier aggregation. The impact of the transmit power imbalance (i.e. imbalance on UL carriers) on PC tolerances needs to be studied.

Frequency error Synchronization in case of inter-band carrier aggregation would be studied.

Error vector magnitude (EVM)

EVM for UL MIMO and 64QAM need to be studied.

IQ component, In-band emissions

Regarding in-band emissions, RAN4 have to study if there is a need to define additional requirements to limit the leakage arising due to UL carrier aggregation (power imbalance scenario).

6.5

Spectrum flatness No change Occupied bandwidth No change SEM/ACLR Co-existence studies would be needed. Based on such

analysis, reasonable requirements which consider both system performance and UE complexity should be specified.

6.6

Spurious emissions Impacts of carrier aggregation need to be studied. Especially, the impact of IM products due to multi-carrier transmissions needs to be studied. Boundary between OOB domain and Spurious domain should also be discussed and decided.

6.7 Transmit Intermodulation Impacts of carrier aggregation need to be studied. 7.3 Reference sensitivity Impacts of carrier aggregation need to be studied. 7.4 Maximum input level No change 7.5, 7.6,7.7, 7.8

ACS/Blocking/Spurious response/Intermodulation characteristics

Impacts of carrier aggregation need to be studied.

7.9 Spurious emissions No change

The UE power consumption is increased with the number of component carrier. Indeed, more carriers imply that the UE will have to monitor more physical downlink control channels (PDCCHs), which are detected by blind decodings in its control channel element (CCE) search space. The number of such blind decodings is directly related to the power consumption of the UE and it is desirable to minimize the number of blind decodings, i.e., the number of component carriers. BS side It was especially agreed that LTE-Advanced requirement could reuse as much as possible current LTE specification but also trying to use the same approach that was use for dual-band dual-cell (DB-DC) HSDPA and DC-HSUPA. Some discussions occurred especially on the introduction of carrier segment and carrier extension within BS requirement. However, some European operators agreed that before starting to specify, RAN4 is awaited RAN1 development work on such aspect before starting to decide if we specify new methodology of test for such kind of features. While the cost /complexity of some hardware components and software components may only depend on the total bandwidth, the cost/complexity of others would scale with the number of carriers. Example, for extension carriers the use case is less overhead and more throughput. There will be a need to define new fixed reference channels (FRC) to reflect this higher throughput. This is also valid for segments. Release-10 will have specific requirements related to the handling of multi-carrier which might trigger new test models. Impacts on BS demodulation requirements [R4-100106]

WINNER+ D 3.3 RAN4 need to wait for RAN1 decision for demodulation performance requirements, similarly to HSDPA and LTE.

TS 36.104 section

Performance requirement

Rel-8 component carriers used when aggregating

Extension carrier used when aggregating

Segments used when aggregating

8.2

Performance requirements for PUSCH

8.2.1

Requirements in multipath fading propagation conditions reuse all per carrier

New FRC with new higher payload size (higher maximum throughput)

New FRC with new higher payload size (higher maximum throughput) New FRC if different single sided bw

8.2.2

Requirements for UL timing adjustment reuse all per carrier



8.2.3

Requirements for high speed train reuse all per carrier



8.2.4

Requirements for HARQ-ACK multiplexed on PUSCH

New requirements for release-10 needed: Non-contiguous resource allocations Uplink control of all CC is transmitted on single UL CC => ACK/NACK and CSI on PUSCH with extended payload

Non-contiguous resource allocations Uplink control of all CC is transmitted on single UL CC à ACK/NACK and CSI on PUSCH with extended payload

Non-contiguous resource allocations Uplink control of all CC is transmitted on single UL CC à ACK/NACK and CSI on PUSCH with extended payload

8.3

Performance requirements for PUCCH

8.3.1 DTX to ACK performance reuse all per carrier N/A N/A

8.3.2

ACK missed detection requirements for single user PUCCH format 1a

New requirements for release-10 needed: Uplink control of all CC is transmitted on single UL CC => ACK/NACK and CSI on PUCCH with extended payload N/A N/A

8.3.3

CQI missed detection requirements for PUCCH format 2

New requirements for release-10 needed: Uplink control of all CC is transmitted on single UL CC => ACK/NACK and CSI on PUCCH with extended payload N/A N/A

8.3.4

ACK missed detection requirements for multi user PUCCH format 1a

New requirements for release-10 needed: Uplink control of all CC is transmitted on single UL CC. ACK/NACK and CSI on PUCCH with extended payload N/A N/A


WINNER+ D 3.3 3.2.5.2 Component carrier coverage In this document, the focus is done on the carrier aggregation scenarios in which all carriers are co-located. There are two types of carrier aggregation: contiguous carrier aggregation and non-contiguous carrier aggregation. Non-contiguous carrier aggregation could be intra-band or inter-band carrier aggregation. These different types of carrier aggregation will form different deployment. For contiguous carrier aggregation and intra-band carrier aggregation, if each component carrier has the same transmit power, then the coverage of each component carrier will be almost the same. It is easier for UEs to use carrier aggregation in the same coverage. For inter-band carrier aggregation, it is more difficult for all aggregated carriers to have same coverage compared with contiguous case. This is mainly due to the large path loss difference in component carriers. Only two downlink component carriers, f1 and f2 (f1 is lower than f2) are shown in Figure 6. In order to support carrier aggregation where multiple layers have different coverage, efficient procedures and measurements are needed to manage usage of component carriers.

f1/f2

Carrier Aggregation

Contiguous carrier aggregation

Non-contiguous carrier aggregation

f1/f2

f1/f2f1/f2

f1/f2

f1/f2f1/f2

f2

If each component carrier has the same transmit power, the coverage of each component carrier will be almost the same.

Intra-band Inter-band

f1

f1f1 f1f2

f1

f1

f1

f2f2

f2f2

f2

Large path loss difference between component carriers, f1＜f2.

Figure 6: Carrier aggregation coverage issue

3.2.5.3 Initial access, system information and RACH procedures To perform initial cell search to camp on a component carrier in LTE-A systems, first a UE tries to detect primary synchronization signal (PSS) on 100 kHz frequency raster. If UE fails in detecting PSS for the current frequency raster, the UE can change the centre frequency to the next 100 kHz raster and repeat the synchronization signal detection procedure. After detecting PSS, secondary synchronization signal (SSS) and the physical broadcast channel (PBCH), the UEs will try to receive system information from the acquired component carrier.


WINNER+ D 3.3

Carrier component

100 kHz channel raster Frequency

Carrier search

Contigurous aggregationNon-contigurous

Figure 7: Component carrier search for the cases with the contiguous and non-contiguous carriers mixed [R1-093389].

At this stage some clarifications on the component carriers' accessibility are needed. Indeed, among the configured component carriers, at least one downlink component carrier should transmit the required information to enable UEs to make an access to the system.

– Do all the component carriers are accessible to all UEs? – Do some component carriers are not accessible to release 8 UEs, but accessible to Rel-10 UEs?

Even UEs with the capability to support multiple component carriers are not aware of the presence of other component carriers and their configuration (carrier frequency, bandwidth) and will detect only one component carrier in the initial access and receive system information from the acquired component carrier. Furthermore, the camped component carrier may not broadcast system information for all non-camped component carriers. If system information is common among aggregated CCs, UE only needs to obtain system information from one component carrier. However, there is also component carrier specific system information. In this case, how do UEs acquire necessary up-to-date system information of multiple component carriers related to the CA transmission? After completing the cell search and acquiring the system information, the UE needs to perform random access in UL to connect to the system. If too many UEs camp on the same downlink component carrier, they shall transmit random access channel (RACH) preamble on the same corresponding uplink component carrier. This could cause RACH congestion. Another concern has been raised by many companies with the asymmetric carrier aggregation case, in which multiple downlink component carriers may be associated with only one uplink component carrier. Under this situation, the configuration between uplink component carrier and downlink component carrier may cause a serious ambiguity problem since the eNB may not know which downlink component carrier is selected by the UE in order to perform the random access (it cannot identify a downlink component carrier to transmit random access response).

3.2.5.4 Control signalling design The design and deployment of control signalling channels for multiple component carriers are crucial for efficient data transmission control and the overall system performance.

3.2.5.4.1 Downlink The downlink control signalling includes scheduling control information and downlink positive/negative acknowledgements (ACK/NACK) associated with uplink data transmission and hybrid automatic repeat request (HARQ) operation. Indeed, the downlink scheduling grant indicating to the UE that it is going to receive data and the transport format that will be used, so that the UE can demodulate the data properly and the uplink scheduling grant indicating to the UE that it can transmit data and the transport format to be used. The UEs must be able to receive downlink control signalling on multiple component carriers. Figure 8 shows the different downlink control signalling information considering a contiguous carrier aggregation scenario.


WINNER+ D 3.3

f1

f2

Multiple uplink scheduling grant Multiple downlink ACK/NACK

Multiple downlink scheduling grant 2 downlink CCs

Figure 8: Downlink control signalling information Three physical channels are defined for control signalling messages transmission: physical control format indicator channel (PCFICH), physical downlink control channel (PDCCH) and physical HARQ indicator channel (PHICH). PDCCH: UEs must be able to decode PDCCH on multiple DL carriers. This means that the eNB must be able to schedule multiple DL/UL CCs to a UE in a sub-frame. As a PDCCH on a CC can assign PDSCH or PUSCH resources in one of multiple CCs, a carrier indicator field is needed to specify to UEs on which component carrier the PDCCH assigns PDSCH or PUSCH resources. Also, if the Rel-8 blind decoding approach is used for carrier aggregation in Rel-10, the blind decoding requirements would scale linearly with the number of carriers. For example, for 5 component carriers, 220 blind decodes would be required. Blind decoding impacts processing requirements and battery life, both of which are important design elements for LTE-Advanced. PCFICH: In a multicarrier LTE-Advanced system, there are several PCFICH design choices and scenarios to consider. Indeed, inability to decode reliably the PCFICH would result in the cross-carrier PDSCH assignment being wasted. When a cross-carrier-scheduled UE succeeds the reception of PDCCH but fails the PCFICH detection of the PDSCH carrier, the UE stores incorrectly received PDSCH in its buffer and sends NACK to the eNodeB. The eNodeB re-transmits PDSCH to the UE with a new redundancy version. As a result, HARQ combining error happens. PHICH: In LTE, UL physical resource block (PRB) index to PHICH resource index mapping is defined under the condition that we have only one pair of DL and UL carriers. However, since the LTE PHICH mapping rule does not always straightforwardly apply in different scenarios of carrier aggregation in LTE-Advanced, there is a need of new PHICH mapping rules in LTE-Advanced (which DL carrier is used for PHICH transmission?).

3.2.5.4.2 Uplink In the uplink direction, the following control information are needed: the channel quality indicator to inform the eNB of the downlink radio quality, the ACK/NACK associated to downlink data transmission and the uplink scheduling request to request resource for data transmission. Figure 9 shows the different uplink control signalling information considering a contiguous carrier aggregation scenario.

f1

f2

Multiple scheduling request

Multiple CQI reports Multiple uplink ACK/NACK

2 uplink CCs

Figure 9: Uplink control signalling information These information are embedded in Uplink Control Information (UCI) messages, which are carried by the physical uplink control channel).


WINNER+ D 3.3 In carrier aggregation, the baseline assumption for downlink component carrier assignment is one transport block (in the absence of spatial multiplexing) and HARQ entity per scheduled component carrier.

Segm.ARQ etc

Multiplexing UE1

Segm.ARQ etc...

Scheduling / Priority Handling

Logical Channels

Transport Channels

MAC

RLC Segm.ARQ etc

Segm.ARQ etc

PDCPROHC ROHC ROHC ROHC

Radio Bearers

Security Security Security Security

...

HARQ HARQ...

Multiplexing UEn

HARQ HARQ...

CC1 CCx... CC1 CCy... Figure 10: Layer 2 Structure for the DL 3GPP TR 36.912

Thus in case of a multiple CC assignment, the UE may have multiple HARQ processes in parallel. This would mean that multiple ACK/NACKs corresponding to the downlink component carrier transport blocks should be transmitted in the uplink. Meanwhile, the UE also shall feedback multiple channel quality indicators (CQIs) reflecting the channel quality of DL carriers. A first important issue is on which uplink component carriers the resources for multiple PUCCHs to a certain UE will be reserved. Another important issue for LTE-Advanced PUCCH design involves the support of asymmetric component carrier aggregation (i.e., having different number of uplink and downlink component carriers).

3.2.5.5 Uplink power control The main scope of power control for PUSCH in LTE-Advanced is very similar to Rel’8. Indeed, it is used to mainly compensate for slow-varying channel conditions (distance-dependent path loss and shadowing) while reducing the interference generated towards neighbouring cells.

Data Interference

Figure 11: Uplink power control Therefore the PUSCH power control concept for LTE-Advanced is expected to have many similarities with the one standardized for Rel’8 LTE. In the uplink, multiple component carriers can be aggregated for one UE in one subframe. So the uplink power control with carrier aggregation should be reconsidered. Indeed, even if the channel propagation condition may be similar in case of contiguous carrier aggregation, the interference conditions are different. That is the reason why LTE-Advanced supports component carrier specific uplink power control for both contiguous and non-contiguous channel aggregation. The following uplink power control aspects remain:

– Which power control parameters are component carriers specific or common to all component carriers?

– Is it possible to derive path loss of several uplink component carriers from the reference signal received power (RSRP) measurement on a single downlink component carrier?

– How to share power between PUSCH and PUCCH in case of power limitation?


WINNER+ D 3.3 3.2.5.6 Discontinuous reception As UEs run on batteries, it is desirable to maximize the standby and talk time so that users do not have to recharge their UEs as often. Discontinuous reception introduced in order to enable UE power savings when no active data transmission of over the air interface takes place. The DRX specify the sub-frames during which the UE shall monitor the PDCCH of the serving cell for potential resource allocations. Power saving becomes more important in multi-carriers system for UE battery. With the introduction of carrier aggregation, the DRX operation has to be extended to multiple carriers in such a way that the UE power saving can still be optimized for a certain targeted QoS and end-user experience. The main problematic is how to design DRX for LTE-Advanced and how to make DRX work effectively with multi-carriers?

3.2.5.7 Component carrier activation/deactivation Although LTE-Advanced UEs will support 100 MHz bandwidth however UE at any given point in time may not transmit/receive in the whole spectrum. Indeed, the component carrier activation/deactivation is needed to:

– Enable the UE to save battery power it would be wise that the UE listens to only some of the component carriers to start with and then the eNB scheduler, based on the activity of the UE can direct it to monitor subset of component carriers.

– Adjust the number of component carrier when the UE is moving in an area. Indeed, in case of inter-band carrier aggregation the coverage of each component carrier will be different as shown in Figure 12.

How to activate/deactivate component carrier?

CC1 CC2

CC3

f1�f2�f3

Figure 12: Inter-band carrier aggregation coverage

3.2.5.8 Mobility handling Concerning carrier aggregation, there are two main issues in relation to mobility that require specific attention. The first is measurements and the second is the actual handover execution.

3.2.5.8.1 Measurements in connected mode With carrier aggregation (CA), as the UE will be served by multiple frequencies in possibly different bands, the definition of measurement will need to be extended to handle the multiple frequencies case. Moreover, there is different measurement requirement between component carrier management within CA cell and handover among CA cells. Measurement related definition and evaluation criteria should be re-considered in CA scenario. How to adapt the radio resource control (RRC) measurement model for CA?

3.2.5.8.2 Handover In Rel-8, every handover leads to packet data convergence protocol (PDCP) / radio link control (RLC) re-establishment and re-keying, regardless of whether the handover is performed within a same eNB or not. Thus, UE in LTE network experiences interruption of data transfer whenever there is a handover.


WINNER+ D 3.3

CC1 CC2 CC3

f1�f2�f3

CC4 & CC5

Figure 13: Mobility handling

In Rel-10 with LTE-A, a UE can be configured with multiple DL/UL carriers. This means that a UE has to prepare for not only Rel-8 cell change procedures but also DL/UL carrier change procedures. If each (re-)configuration of additional carrier is considered as handover, the UE will experience much more data interruption. Unless some smart mechanism is introduced, user perception will be worse in LTE-Advanced than in LTE.

3.3 Spectrum Aggregation with Multi-Band User Allocation over Two Frequency Bands Spectrum access is one of the central issues in next generation communications. Under usage of spectrum resources has been clearly pointed out. The current radio regulation (while preserving the licenses obtained by the operators) is also limiting the network achievable capacity. By supporting additional system capacity and higher data rates through high speed radio access technologies (RATs), such as the IMT-Advanced [CL5], a universally accessible broadband infrastructure will improve the value of these services. An important challenge regards the standardization of IMT-Advanced technologies to operate in the preferred frequency bands [WRC07]. Existing radio spectrum is highly fragmented and does not match the actual demand for transmission and network resources. To overcome this problem, cognitive radio can be a solution achieving very good improvements in terms of capacity and resource exploitation. However, complexity of the solutions proposed and the interaction among primary and secondary systems are still open issues. To counter-fight spectrum fragmentation in a medium term horizon, mobile operators might be forced to aggregate separated parts of spectrum. Spectrum aggregation (SA) can be performed in the same or in different bands, in contiguous and non-contiguous manner. More details can be found in [WIN+D3.1, WIN+D3.2]. In addition, SA is needed in scenarios in which an operator accesses both the dedicated band and a pool of shared frequencies. The common pool shared by operators may be located on the upper part of the available frequency band. SA has been proposed for LTE-Advanced [LTE-A] and has also been also investigated for the WINNER+ air interface, another IMT-Advanced candidate. In a spectrum sharing scenario, the performance might be degraded on one hand by the reassignment of the spectrum to primary and secondary users, and on the other hand, the challenge is to maximize the overall system performance in terms of data rate, power, number of served and satisfied users, etc. [WIN+D3.2], [WIN+D1.5], and [WIN+D5.1] proposed a general multi-band scheduling (GMBS) algorithm for assigning users in the shared band of an operator. The employed resource allocation (RA) allocated the user packets to the available radio resources in order to satisfy the user requirements, and to ensure efficient packet transport to maximize spectral efficiency. The objective is to determine the best user allocation for a single operator over two (or more) frequency bands { }, , ....,b∈ 0 1 m . Two bands are analysed (m = 2). The operator exclusive usage of the 2-GHz band and can access to the shared frequency pool at 5-GHz. The quantity of radio resources available at 5 GHz is determined by spectrum trading (or bargaining, here, a specific methodology has not been considered ) among all the operators that have access to the common frequency pool. The part of the frequency pool assigned to the operator is assumed to be fixed. The performance gains are analyzed in terms of data throughput. The total throughput is a function of the radio channel qualities for each user in the considered bands. The Channel Quality Indicator (CQI) depends on the path loss, which depends on the distance from the BS (eNodeB) but also on the carrier frequency adopted. The operator applying multi-band scheduling (MBS) will have relevant improvements when the UE have heterogeneous spatial distribution in the cell (variable distances from the BS) and different channel qualities in the considered spectrum bands.


WINNER+ D 3.3 Here, we focus on an optimised radio resource allocation of users over two frequency bands. The RA, an entity within the set of RRM algorithms, should have inherent tuning flexibility to maximize the spectral efficiency of the system for any type of traffic Quality of Service (QoS) requirements. The proposed RA maps packets of variable size into variable length radio blocks for transmission over the PHY layer, and the length is dependent on the channel quality. The following events occur: 1. User packets awaiting transmission are prioritized according to the scheduling algorithm criteria; 2. A CQI identifier is selected according to the link adaptation algorithm, using the available CQI options from the PHY layer; 3. An idle ARQ channel is selected to hold and manage the ARQ transmission; 4. The packet is transmitted and received at the UE. Soft retransmissions are combined with previous packet transmissions (chase combining) and the ARQ messages are generated accordingly. These messages are then signalled to the BS, and the ARQ processes are released if the messages are positive acknowledgments (ACKs). The problem of scheduling the users into two bands (2 GHz and part or all of the frequency pool at 5 GHz) can be formulated as a general assignment problem (GAP) optimization problem [KARLOF05]. A Profit Function (PF) is defined and maximized as the total throughput of the operator via a single objective maximization problem. Fairness and QoS requirements of the service classes are not considered here. Multiple objectives can be introduced and implemented in the problem, such as maximizing the total throughput while minimizing the QoS satisfaction indexes for each service class, [MEU08]. Solving Multiple Objectives General Assignment Problem (MO-GAP) can be very difficult and usually the objectives are combined together via a linear combination, called ”scalarization” [KEL05]. The GMBS problem can be solved considering UE with the added capability of transmitting and receiving in multiple frequency simultaneously (multiple transceivers at UE) or when the UEs can just choose one band among all the bands in the network and choosing one of the transceiver configurations available at the UE radio. In both cases, this is an Integer Programming (IP) problem with different constrains formulation. In the GMBS solution, the PF depends on the ratio between the service throughput request and the real goodput available for each user (on each band). The problem is formulated with a load constrain for each band max

bL , as well as a resource constraint based on the available channels. The PF is thus defined considering the ratio between the requested rate by the service flow and the rate available on a single downlink channel. This weight accounts for a real usage of the capacity considering the source traffic generator. The PF is given as:

( ) maxm n

bu bub u

PF W x= =

= ∑∑1 1

(Eq 1)

In each band, the network has multiple data channels. Figure 14 shows multi-code allocation for an HSDPA system. A round robin packet scheduler is assumed in which all the users are satisfied in a three transmission time intervals (TTIs) iteration. Two iterations are presented, each iteration takes 3 TTIs. For example user 1, that is allocated to band 1, X1 1, has the CQI that demands the usage of 5 parallel codes, while for example user 14 allocated to band 2, X2 14 it has 3 codes.

Figure 14: Allocation matrix example over two bands.


WINNER+ D 3.3 Multi-code user allocation over the bands is represented by:

When single code allocation is being considered, bux is equal either to one or zero.

For a multiband UT, with only one active transceiver and single code allocation, the GMBS has two constraints, namely the allocation constraint (AC), which takes care that each user can be allocated only to a single frequency band with a single allocation channel, and the bandwidth constraint (BC), which defines that the total number of users on each band is upper bounded by the maximum load that can be handled in the band, max

bL . The maximum theoretical capacity of the system is defined as the maximum data rate that an UE with the best modulation and coding scheme (MCS) can receive. The objective of the optimization procedure is to obtain the values of bux .

The RA component is responsible for allocating the available radio resources to the user traffic in a cost-effective manner, and includes a scheduling mechanism, link adaptation, code allocation policy, and HARQ scheme to improve the service throughput for users at the cell edge. The CQI is a mapping of the averages of the signal-to-interference ratio (SIR) recorded over time. A common radio resource management (CRRM) entity keeps track of the CQI in both frequencies by making use of the pilot channel. To get the SIR input parameter for the algorithm in both frequencies, several approaches can be followed. The UE can be either in active or passive mode. In active mode the user is continuously measuring the received SIR from both frequencies. In passive mode, the measurements are periodically sent to the gateway entity. The performance of the algorithm is assessed by using the service throughput that is the total number of bits that have been transmitted and correctly received by the all users in the cell:

[ / ]( )

.serv

bits sb pServ thr

k T− = (Eq 2)

where is the number of bits received in given period p, T is the transmit time interval, and k.T is the total simulation time. Users are deployed on the cell with an uniform distribution, within a range of 900m, with overlapping 2-GHz and 5-GHz coverage. The traffic is model with a Poisson distribution and the call duration is exponentially distributed with an average of 180 s. Simulation runs were stopped when a target 95% confidence interval was achieved. The confidence interval arising from the simulations is represented in the graphs by the vertical bars.

( )servb p

Figure 15 shows the throughput results and respective confidence intervals without GMBS. The ”Overall Serv” throughput is the sum of the service throughput in both frequency bands. The number of service requests is equally divided between the two frequency bands. When more than 32 stations are in the system, it cannot satisfy all the UEs.

20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 580

1000

2000

3000

4000

Number of stations

Thr

ough

put [

kbps

]

Traffic requirementsOveral ServServ 2Serv 5

Figure 15: Average throughputs without GMBS

Figure 16 shows the results with the GMBS algorithm.


WINNER+ D 3.3

20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 580

1000

2000

3000

4000

Number of stations

Thr

ough

put [

kbps

]

Traffic requirementsOveral ServServ 2Serv 5

Figure 16: Average throughput with GMBS

As in the 2-GHz band, the SIR is higher and PER is lower than in the 5 GHz, in the beginning of the simulation, the system has resources for users, the constraints are not limitative and the algorithm will choose the mainly the 2 GHz, but the PF will cause some load management interventions, 20-29 users (see Fig 17). When the system gets overloaded, and the constraints become limitative, the algorithm does is job of maximizing the use of resources in the optimal way. Both bands have a higher throughput due to the switching of the user based on their respective channel qualities into the two bands, 50-59 users, in Figure 17. It can be observed that with more users more exchanges are performed and these are independent of the distance. The enhancement provided by the GMBS algorithm is evident when more than 32 stations are in the scenario. We assume that the systems reaches is full capacity when the overall service throughput curve stops growing (around 55-66 users). While, without GMBS, the system reaches its full capacity around 2.5 Mbps, with GMBS, it reaches its full capacity around 3 Mbps. A gain up to 500 kbps may be obtained, i.e., 20% gain. This gain is achieved by dynamically allocating resources to the UE that best suit the overall system. To fully understand where the gain comes from, two input parameters were analyzed in terms of CQI usage and PER.

100 200 300 400 500 600 700 8000

0.05

0.1

0.15

0.2

0.25

Distance to the BS

Num

ber

of e

xcha

nges

/nu

mbe

r of

CR

RM

bal

anci

ng in

terv

entio

ns

20−29 users30−39 users40−49 users50−59 users

Figure 17: Exchanges between systems

Figure 17 shows the average PER as a function of the distance. The use of the GMBS algorithm shows a significant gain in higher frequencies. In lower frequencies, there are cases in which the gain is not very significant and in the case 300 and 400m, a loss occurred. The GMBS is balancing UEs based on PER, i.e, it allocates the UEs with low PER to the 5-GHz band and the other UTs to the 2-GHz band, decreasing the overall PER.

3.3.1 Future work Future work intends to consider the spatial distribution of the users. The GMBS problem, and its optimal solution, will be analyzed to quantify the gap between the presented (and other) suboptimal GMBS algorithm and the best solution possible. Further, the GMBS problem will be solved with use of various types of schedulers proposed for IMT-Advanced systems. Future work will include the QoS requirements into the GMBS formulation via a linear combination of multiple objectives (“scalarization”). The combined solution for the packet scheduler and the spectrum scheduler is foreseen to be able to greatly reduce the bit rate and delay jitters, which are of paramount importance for real time services. Mobility patterns will also be analyzed, showing the effectiveness of GMBS to counter-fight shadowing in support of the aforementioned real time services. The influence of the gap between the two frequency bands can affect the user allocation mechanism. Depending on the capabilities of the UTs, each user could be allocated to a single frequency band or to


WINNER+ D 3.3 both the frequency bands. When the UTs have multi-radio transceivers then these can transmit and receive data on both bands. This requires algorithms for managing the shared band and for the allocation of the users. Furthermore, it is interesting to investigate the problem and the required algorithms for the more complex case of multiple operators. Future work will involve determining the load threshold in an automatic way.

4. Sharing options in mobile wireless networks In general, a shared network is a realization form that mobile operators may choose in implementing their networks whereby elements of the network are realized and used as shared resources. Different motivations apply why actually competing operators may decide to go for such cooperation: Cost aspects, business models, regulative framework, political reasons or lack of alternative options.

4.1 Sharing examples Sharing may be realized on different partly overlapping domains. Figure 18 gives a non-exhaustive overview of different sharing aspects:

SiteSharing

Infrastructuresharing

RAN/CN sharing

Carrier/SpectrumSharing

common backbonejoint transmission

commonnetwork elements

location, air conditionantennas, radio masts

Multiple PLMNsRoaming, MVNO

Figure 18: Sharing aspects in mobile radio networks

Infrastructure sharing is a very generic notation that refers to network architecture and its realization. A common backbone shared by different operators to enable joint transmissions is one example. Logically, mobile radio network architectures basically distinguish between a core network (CN) part and a radio access network (RAN) part, while sharing is possible for either or both parts resulting in CN sharing and/or RAN sharing. Within 3GPP, the latter is defined as “Two or more CN operators share the same RAN, i.e. a RAN node (radio network controller (RNC) or base station controller (BSC)) is connected to multiple CN nodes (serving general packet radio service (GPRS) support node (SGSNs) and mobile switching centre (MSC) / visitor location registers (VLRs)) belonging to different CN operators” [TR21.905]. If sharing applies only to the RAN, the corresponding reference architecture as defined in 3GPP results in a multi-operator core network (MOCN) configuration [TS23.251]. If elements of the CN are also shared, the set up corresponds to a gateway core network (GWCN) configuration. Figure 19 depicts a corresponding setup for 3G. Similar considerations apply for E-UTRAN likewise.

Radio Access NetworkOperator X

CN Operator A

CNOperator B

CNOperator C

RNC

Iu

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

Radio Access NetworkOperator X

SharedMSC/SGSN

SharedMSC/SGSN

SharedMSC/SGSN

RNC

Iu

.........CNOperator A

CNOperator B

CNOperator C

.........

RNC RNC

Figure 19: Multi-Operator Core Network (MOCN) configuration with RAN sharing and partial CN node sharing [TS23.251]

Site sharing addresses the location where RAN equipment and antennas are deployed. The every increasing demand for more capacity results in smaller cells and thus in more sites. However, suitable locations for technology hosting including backbone connectivity, energy supply, or air conditioning are limited. Driven by political and esthetic reasons, local administrations also want to avoid “oceans of antennas” for which existing antenna masts shall be used as efficient as possible by sharing rather than building new masts for each single operator.


WINNER+ D 3.3


Carrier and Spectrum sharing is of particular interest and challenging as well, since this affects directly the interface to the customer. In fact, the ability for a user to function in a serving network different from the home network – better known as roaming [TR 21.905] – can be seen as shared access to a foreign spectrum. From an operator’s perspective, Carrier and Spectrum sharing is particularly interesting in low load scenarios where one operator has a running infrastructure in place but does not fully use its capacity. This leaves room for business models that are based on the sharing idea, e.g. with the constitution of so called mobile virtual network operators (MVNO) that use air interface resources (and infrastructure) from “real” operators for their brand. While the MVNO in fact does not transmit any physical bit himself but relies on transception services of the license holder, other spectrum sharing scenarios result in actual sharing, with two or more physical signals in the same spectrum by different operators. Service Level Agreements (SLAs) are applied to ensure concurrent coexistence. The usage policy may include equal usage rights or primary and secondary access. While today’s regulation in practice is not yet that far, future developments may see sharing based flexible spectrum usage up to the point of spectrum trading and dealing. Whatever sharing principle is applied on the network side, the intention is that UEs basically are served as if they were connected to a conventional network, where RAN and CN belong to one single serving operator.

4.2 Sharing motivation Obviously, one of the key drivers for sharing hardware, network elements or other resources such as spectrum is the ever increasing demand for a cost efficient deployment. This has also been recognized by 3GPP, classifying network sharing as “a way for operators to share the heavy deployment costs for mobile networks, especially in the roll-out phase” [TR 23.251]. Further, spectrum license conditions might drive operators in considering network sharing aspects in their roll-out planning, too. Traditional deployments start in dense urban areas to provide mobile radio services to as many people as possible in a reasonably short time span. Some license requirements, however, as currently associated with the Digital Dividend 800 MHz band in Germany constitute an “onion principle”. The roll-out needs to start in rural areas and only if particular coverage obligations are reached the next layer i.e. sub-urban areas can be opened up, followed by urban areas and hot spots. Since capacity is not a problem in the early phases, initial sharing scenarios might be attractive for boosting coverage and initial service provisioning while a dedicated, operator individual network can be extended out of the shared approach when required. However, it is clearly to say that e.g. site sharing though reducing costs at the infrastructure side might cause negative economic results in others fields: Operators apply sophisticated network planning to achieve nationwide optimized radio conditions. This is a complex process for which expert knowledge and dedicated tools are required. A matched network layout allows operators to develop a unique selling point expressed by superior network performance. If however, site sharing is applied, smart followers such benefit by the intensive pre-studies of their competitors in applying the same cell grid. On the other hand, site sharing may have negative impacts from technology point of view. Usually not all mounting options e.g. on an antenna mast are of equal quality. Feeding cable length may differ, angle selections are limited or sectorization is non-optimal. From strategic point of view, operators may have a preference to exclude network sharing scenarios and to accept higher costs. This is particularly the true in case of first mover strategies or if network sharing is not acceptable due to different QoS targets that operators want to achieve with their deployment. However, there are also technical reasons, why network sharing can be beneficial. This is why the following discussions will not go in the sensitive details of cost/business, strategy or politically motivated reasons for network sharing but focus on the technology options and related benefits, e.g. higher flexibility and better spectral efficiency.

4.3 Spectrum/Carrier Sharing With LTE R8, network sharing with multiple public land mobile network (PLMN) identities on one single carrier is supported. In total, up to 6 PLMNs can share one carrier while there is no restriction to one single country (mobile country code (MCC)). Since correct operator name display is ensured, international roamer can select from all the PLMNs thus no change to PLMN selection principles applies. Being a mandatory feature, the UE subsequently is able to read the PLMN identities and indicates the selected one to the shared eNB that forwards the connection request to a specific service provider. This is how also national roaming can be supported. If required, area restrictions can be configured i.e. no sharing in capacity critical urban city areas or subways.

WINNER+ D 3.3

eNB

PLMN 1PLMN 2PLMN 3PLMN 4PLMN 5PLMN 6

CN PLMN 1

CN PLMN 2

CN PLMN 3

CN PLMN 4

CN PLMN 5

CN PLMN 6

Figure 20: E-UTRAN with Network/Carrier Sharing for up to 6 PLMNs

While the RAN part is subject to a shared deployment, the CN including mobility management entity (MME), serving gateway (SGW) and packet data network gateway (PDN GW) can be operator specific. Biunique signalling between the shared RAN and the dedicated CN is possible due to the LTE inherent S1-flex concept allowing the interconnection of a single eNB to multiple MMEs or SGWs (pooling concept). If the CN elements are not part of a single service provider owned pool but belong to individual providers, the S1-flex concept thus enables RAN sharing, and otherwise S1-flex is a means for support of network redundancy and load sharing of traffic.

Operator A

PLMN A

Operator A

Operator B

PLMN BInternational roamer can select

Operator B

Operator AOperator B

PLMN APLMN B

S1-flex

Figure 21: S1-flex concept as enabler for RAN sharing

Applying carrier sharing has essential benefits. Bandwidth operation as specified for LTE comprises 1.4, 3, 5, 10, 15 and 20 MHz, whereas support 20 MHz and all bandwidths below is mandatory for a UE. Natural “spectrum bundling” hence is possible up to 20 MHz. With LTE-A this will even increase up to 100 MHz. Efficiency of LTE increases with wider bandwidth since the highest overhead is at 1.4 MHz where control channels always use the full bandwidth, see Figure 22. Spectral optimized operation thus should target on large operating bandwidths.


WINNER+ D 3.3


1,4 MHz, 6 RBs

Max 20 MHz , 100 RBs

Control Ch.

Figure 22: Highest overhead for 1.4 MHz operation in LTE

If operators do not dispose of 20 MHz continuous spectrum, e.g. either due to less successful spectrum acquisition or due to not yet completed refarming process, bundling of bandwidth from other operators allows better utilisation, thus resulting in increase data rates, see Figure 23. Trunking gain arises out of this bundling since resources not used by customers of 1st PLMN improve throughput for customers of 2nd PLMN and vice versa due to the wider bandwidth and scheduling gain. It is to say that spectrum bundling in LTE R8/9 is restricted to adjacent parts of the spectrum (no fragmentation). With LTE-Advanced further options will be available. In total, carrier sharing is hence is an interesting means to increase spectral efficiency, preferably to be applied for small bandwidth deployments (e.g. 900 or 1800 MHz).

PLMN1 PLMN2

PLMN 1 + 2

10 MHz 10 MHz

20 MHz

Figure 23: Bundling of resources results in overhead reduction

WINNER+ D 3.3


5. References [CEPT31] CEPT Report 31 “Frequency (channelling) arrangements for the 790-862 MHz band”,

October 2009. [CL5] Circular Letter 5/LCCE/2, Invitation for submission of proposals for candidate radio

interface technologies for the terrestrial components of the radio interface(s) for IMT-Advanced and invitation to participate in their subsequent evaluation, ITU-R, March 2008.

[COM6/18] Resolution COM6/18 Regulatory measures and their relevance to enable the introduction of software-defined radio and cognitive radio systems, 2007.

[KARLOF05] J. K. Karlof, Integer Programming: Theory and Practice, 1st ed. CRC, 2005 [KEL05] H. Kellerer, U. Pferschy and D. Pisinger, Knapsack Problems. Springer Verlag, 2005 [Lee93] William C. Y. Lee, Mobile Comm. Design Fundamentals, 2nd ed. New York, NY,

USA: John Wiley and Sons, 1993. [LTE-A] http://www.3gpp.org/LTE-Advanced [MEU08] F. Meucci, A. Mihovska, B. Anggorojati, and N. R. Prasad, “Joint Resource Allocation

and Admission Control Mechanism for an OFDMA-Based System,” in Proc. The 11th International Symposium on WPMC’08. Lapland, Finland.

[R1-092575] "Summary of email discussion on carrier aggregation terminology", (3GPP meeting #57bis), Nokia.

[R1-093389] 3GPP TSG WG1 RAN Meeting #58 24-28 August 2009 Initial access in LTE-A DL and UL with carrier aggregation, Samsung,

[R4-094674] 3GPP TSG RAN WorkingGroup 4 Meeting #53 Korea 9-13 Nov 2009 NTT DOCOMO [R4-100106] 3GPP TSG RAN Working Group 4 Meeting Adhoc 2010-01 Ericsson [REV-090003r1] 3GPP Mobile World Congress 17-18 Dec 2009, Beijing, China, REV-090003r1 ©

3GPP 2009 IMT-Advanced Evaluation Workshop "LTE-Advanced Physical Layer", Matthew Baker, Alcatel-Lucent.

[RP-091440] 3GPP TSG RAN#46 1-4 December 2009 Work Item Description Nokia [SM.2152] ITU-R Report SM.2152 “Definitions of Software Defined Radio (SDR) and Cognitive

Radio System (CRS)”, September 2009. [TR21.905] V10.1.0 (2009-12), 3rd Generation Partnership Project; Technical Specification Group

Services and System Aspects; Vocabulary for 3GPP Specifications (Release 10) [TR 102 680] ETSI TR 102 680 V1.1.1 (2009-03); Reconfigurable Radio Systems (RRS); SDR

Reference Architecture for Mobile Device [TR 102 681] ETSI TR 102 681 V1.1.1 (2009-06); Reconfigurable Radio Systems (RRS); Radio Base

Station (RBS) Software Defined Radio (SDR) status, implementations and costs aspects, including future possibilities

[TR 102 682] ETSI TR 102 682 V1.1.1 (2009-07); Reconfigurable Radio Systems (RRS); Functional Architecture (FA) for the Management and Control of Reconfigurable Radio Systems

[TR 102 683] ETSI TR 102 683 V1.1.1 (2009-09); Reconfigurable Radio Systems (RRS); Cognitive Pilot Channel (CPC)

[TR 102 733] ETSI TR 102 733 V1.1.1 (2010-03); Reconfigurable Radio Systems (RRS); System Aspects for Public Safety

[TR 102 745] ETSI TR 102 745 Ver. 1.1.1 (2009-10); Reconfigurable Radio Systems (RRS); User Requirements for Public Safety

[TR 102 803] ETSI TR 102 803 Ver. 1.1.1 (2010-03); Reconfigurable Radio Systems (RRS); Potential regulatory aspects of Cognitive Radio and Software Defined Radio systems

[TR 102 838] ETSI TR 102 838 V1.1.1 (2009-10); Reconfigurable Radio Systems (RRS); Summary of feasibility studies and potential standardization topics

[TS23.251] V9.1.0 (2009-12), 3rd Generation Partnership Project; Technical Specification Group Services and System Aspects; Network Sharing; Architecture and functional description (Release 9)

[TR25.913] 3GPP TR 25.913 V8.0.0 3rd Generation Partnership Project, Technical Specification Group Radio Access Network, Requirements for Evolved UTRA (E-UTRA) and Evolved UTRAN (E-UTRAN) Release 8

http://www.3gpp.org/LTE-Advanced

WINNER+ D 3.3


[TR 36.815] 3rd Generation Partnership Project, Technical Specification Group Radio Access Network, Further Advancements for E-UTRA LTE-Advanced feasibility studies in RAN WG4 (Release 9),

[TR36.913] V9.0.0 3rd Generation Partnership Project,Technical Specification Group Radio Access Network, Requirements for Evolved UTRA (E-UTRA) LTE-Advanced. Release 9

[TS36.101] V9.2.0 3rd Generation Partnership Project, Technical Specification Group Radio Access Network, E-UTRA.User Equipment (UE) radio transmission and reception; Release 9

[TS36.104] V9.2.0 3rd Generation Partnership Project, Technical Specification Group Radio Access Network, E-UTRA. Base Station (BS) radio transmission and reception; Release 9

[TS36.300] V9.1.0 3rd Generation Partnership Project, Technical Specification Group Radio Access Network, E-UTRA Overall description Release 9

[WIN+D3.2] (2009, April) EU CELTIC Project WINNER+, Deliverable 3.2, Aspects of Spectrum Preferences, available: http://projects.celtic-initiative.org/winner+

[WIN+D1.5] (2009, October) EU CELTIC Project WINNER+, Deliverable 1.5 Intermediate Report on System Aspect of Advanced RRM, available: http://projects.celtic-initiative.org/winner+

[WIN+D5.1] (2009, October) EU CELTIC Project WINNER+, Deliverable 5.1, WINNER+ Interim Trials Results, available: http://projects.celtic-initiative.org/winner+

[WRC07] Key results of World Radiocommunication Conference (WRC-07). [Online]. Available: http://www.itu.int/dms pub/itut/oth/21/04/T21040000030014PPTE.ppt

Documents

Strategies and technologies for spectrum utilisation and sharing