January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

Frequency Agile Spectrum Access Technologies

This Presentation was originally made to anFCC Workshop on Cognitive Radios

May 19, 2003 by one of Bill’s colleagues.

(Bill was kind enough to share this information with 802.18 SG1 in the interest of promoting discussion.)

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

Agenda

• Requirements

• Spectrum occupancy characteristics– Significant amount of “low hanging fruit”

• Spectrum access methods– Listen-Before Talk

• “TDMA” spectrum

• Broadcast spectrum

– Probe

– Geo-location/database

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

Frequency Agile Radio Requirements

• Create insignificant interference – Secondary operation with minimal requirement for coordination

with primary system licensees– Unlicensed with equipment certifications on a system basis to

assure avoidance of interference

• Operate in multiple bands– Assured capacity

• Offer cost/capacity/link range/deployment benefits– Access more (5 X?) spectrum than any current system– Operate in VHF/UHF TV band– Rapid spectrum agreements for itinerate use

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

Spectrum Occupancy Is Low

• “In many bands, spectrum access is a more significant problem than physical scarcity of spectrum, in large part due to legacy command-and-control regulation that limits the ability of potential spectrum users to obtain such access.”1

• Shared Spectrum’s measurements indicate– Many bands have no detectable occupancy

– Some bands have low occupancy

– Some bands have high occupancy

Note 1: FCC Spectrum Policy Task Force Report, page 3

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

Typical Spectrum Occupancy Measurement

No signals Medium and short duration signals

FCC should conduct and publish spectrum

occupancy measurements to

identify low occupancy bands

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

Initially Harvest the Low Hanging Fruit

• Measurements show a large quantity of long duration, large area spectrum holes

• “Simple” spectrum access methods are sufficient– Minimal coordination between transceivers– Moderate computational costs

• Later evolve algorithms to handle more complex situations– Short duration, small spectrum holes– Optimize frequency assignments for increased capacity

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

Agenda

• Requirements

• Spectrum occupancy characteristics– Significant amount of “low hanging fruit”

• Spectrum access methods– Listen-Before Talk

• “TDMA” spectrum

• Broadcast spectrum

– Probe

– Geo-location/database

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

Adaptive, Receive-Only Spectrum Access Method

• Pmax TX = 10*log10(k * T * B) + PPrimary – Pmeasured - Margin

– Margin = 10 to 20 dB, required for cummulative effects, rapid propagation changes, false alarm minimization

– T – Interference Noise Temperature, in K

– B = signal bandwidth, in Hz

Primary network

Secondary transceiver

1) Assume a transmit power level value (Pprimary)

2) Measure receive power level (Pmeasured)

3) The difference is the propagation loss (Pprimary – Pmeasured)

Pmax TX = Pallowable interference + Pprimary – Pmeasured

Primary network

Secondary transceiver

1) Assume a transmit power level value (Pprimary)

2) Measure receive power level (Pmeasured)

3) The difference is the propagation loss (Pprimary – Pmeasured)

Pmax TX = Pallowable interference + Pprimary – Pmeasured

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

Primary Transceivers

Frequency Agile Transceivers

Interference/Connectivity Limit

Exclusion Zone

Frequency Agile Coverage “Morphs” To Fit Primary Users

Hidden-node problem overcome by each Frequency Agile transceiver listening to all Primary users within range

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

High Sensitivity Receiver Performance

Cyclic Correlation Output SNR

0

10

20

30

40

50

60

70

-40 -30 -20 -10 0 10 20 30 40

Input SNR

Ou

tpu

t S

NR

T=1 sec

T=100 ms

T=10 ms

T=1 ms

T=.1 ms

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

10 ground vehicles spread over 25 km moving at 25

km/hr10 stationary ground vehicles spread over 25 km

•West Virginia location•3 m antenna height

Simulation Example

•Primary users are stationary•XG users are mobile•Omni-directional antennas•420 MHz signal frequency

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

Link closure

Interference

•Longley-Rice Model•420 MHzFree space loss

Actual loss

Propagation Losses

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

Frequency Agile network reduces TX power

automatically

Network needs switch to another frequency at low TX power levels

XG TX Power (dBm)

Target interference level of –100 dBm

Primary Interference Level (dBm)

TX Power and Interference

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

TV Receiver

Frequency Agile Transceiver

TV Transmitter

Region of Potential Interference

PT (kW)

HAAT (m)

Listen-Only Method in the Broadcast Bands

TV Receiver

L1

L2

Differential propagation loss = L1-L2

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

Transmit Power Rule

Pmax TX = Po if Primary signal is not detected

= Transmission prohibited if Primary signal is detected

where, Pmax TX = Frequency Agile transmitter power level, in dBm

Po = specified power value, in dBm

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

TV Receiver with Grade B reception

Frequency Agile Transceiver

TV Transmitter

Terrain Blockage

Building Blockage

LOS Location

LOS to Frequency Agile Transceiver

Multi-Path Effects

Minimal Interference

•Joint probability of three conditions–Agile Receiver doesn’t detect TV signal–Primary user receives TV signal–D/U < 15 dB

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

Minimum Detectable Signal (-121 dBm)

Minimum Signal Strength– “Grade B” (-81 dBm)

Min D/U =15 dB

Acceptable Interference = Thermal Noise (-96 dBm)

Received Power (dBm)

40 dB25 dB

Maximum Differential Propagation Value

Maximum differential propagation value = 40 dB

Maximum practical sensitivity improvement due to special detection processing

Signal level at TV receiver

Signal level at Frequency Agile receiver

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

Simulation of Differential Propagation

Test reception points along a ~ 8 km path

Scenario – Mid-Atlantic Region Elevation contours

TV transmitters

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

Large Change in Propagation Loss over a Short Distance is Rare

Signal from TV station A

Signal from TV station B

~30 dB change in propagation loss over a small distance

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

Low Power Transmitters Have a Small Interference Range

Noise

Obstructed propagation

Free-space propagation

Maximum interference range

of 4 km in free-space

Maximum interference range of 600 m in

obstructed conditions

1 mW transmit power

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

Spectrum Probing Method

75 km to 200 km spacing

Frequency Agile transceiver

Broadcast Receiver

1) Transmit (Pt) at a very low power

3) Measure Pr of the very weak signal using high processing gain

2) Very weak signal doesn’t interfere with primary user

4) Pt minus Pr is the propagation loss

5) Repeat with N nodes to estimate minimum propagation loss into area

Frequency Agile monitors

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

Geo-Location Method

TV Receiver

Frequency Agile Transceiver

TV Transmitter

Region of Potential Interference

PT (kW) HAAT (m)

TV Receiver

Protected Area

Position information

•GPS, telephone

Beacon TXProtected area boundary database information

•Telephone, over-the-air, special beacons, Internet, manually entered

Guard Distance

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

Significant “White Space” Between TV Coverage Areas

Grade B 50% and 90% contoursChannel 5 and channel 54

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

How Large A Guard Distance?

Noise

Maximum interference range of > 100 km in

free-space

Maximum interference

range of 2 km in obstructed

conditions

1 W transmit power

Large guard distances reduce spectrum harvestTV bands: 100 km is too large >> Limit TX power to mW’s

Other bands: Max TX power ?

January 2004

Bill Byrnes, Shared Spectrum Co.

doc.: IEEE 802.18-04-0004-00-000

Submission

Summary• Multiple, robust spectrum access methods

– Listen-Before Talk• “TDMA” spectrum• Broadcast spectrum

– Geo-location/database

• FCC should conduct and publish spectrum occupancy measurements– Many spectrum holes are large and have long duration

• FCC should allow experimental interactive operations– All access methods including Probe– TV and other bands