TSB88 Managing Intersystem Interference for Dissimilar Modulations

1
TSB-88Managing Intersystem Interferencefor
Dissimilar Modulations

• Overview
• ACRR ACPR
• How to estimate Source of Interference to Victim
  Receivers characteristics
• Example Application

2
ACPR Vs. ACRR ACIPR
Receiver Specification
Transmitter Specification
ACRR Reference is a signal generator with
reference modulation
ACPR Reference is Transmitters Power output
ACRR (dB)
Adjacent Channel Rejection Ratio
ACPR (dB)
Adjacent Channel Power Ratio
Cs/N for Reference Sensitivity
Transmitter Noise Power intercepted by the victim
receiver, relative to P0
Receiver Noise Floor
3
Transmitter Requirements, ACP

• FCC 47 C.F.R. 90.543 (12.5 kHz channels) 700 MHz
• -40 dB ACP, 9.375 kHz offset, 6.25 kHz bandwidth
• -60 dB ACP, 15.625 kHz offset, 6.25 kHz bandwidth
• No FCC Requirement for a 12.5 kHz offset
• 12.5 kHz specified in other reference documents
• TIA-102.CAAB-B, 3.2.8 Unwanted Emissions
  Non-Spurious Adjacent Channel Power Ratio
• Same as C.F.R. 47 90.543 at 700 MHz (ACPR)
• Other bands, 67 dB ACPR, 6 kHz BW, 100 Hz RBW
• NCC Planning Guidelines (Now PSWAC)
• Guidelines for 764-776/794-806 Regional Planning
  Committees
• 40 65 dB ACPR, Appendix K V2_0.doc
• 65 dB ACPR for 12.5 kHz offsets

4
TIA 603-C Analog ACPR Requirements 3.2.14

• Requirements vary with Frequency Band
• Higher requirements in bands less than 512 MHz
• Measurement bandwidths vary with Offset Frequency
• Narrower with decreased Offsets

5
Receiver Specifications

• TIA-102.CAAB-B
• 3.1.7 Adjacent Channel Rejection, ACRR (12.5 kHz)
• ACRR is equivalent to ACIPR due to method of
  measurement
• Mobile A/B, 60/60 dB ACRR, Portable A/B, 60/50 dB
  ACRR
• Requires Cs/N be added to determine ACPR
• ACPR ACRR Cs/N, 60/50 Cs/N
• Cs/N 7.6 dB for C4FM (P-25 Phase 1 modulation)
• Maximum degradation, 9 dB per kHz of offset
  frequency C4FM
• Phase 2 modulations not yet finalized
• NCC (now PSWAC) Guidelines
• Derived from TIA
• Assumes 60 dB ACRR, simplified by using 65 dB
  ACPR
• Companion Receiver, considers actual receiver
  ENBW
• FCC
• No specific requirements
• Little potential for setting requirements

6
References 603-C, 102.CAAA/B, TSB-88-B

• Analog Radios
• Companion Rcvr

• Digital P-25 Radios
• 102.CAAA
• 102.CAAB
• Companion Rcvr
• Required ACPR offsets BW

• Recommended methods for all combinations of
  analog and digital radios
• Various Rcvrs, BW and offsets
• Propagation and CATP

7
Area of Influence

• TIA-603/102 - Normative methods requirements
• TSB-88 - Recommended methods only

8
ACRR specifications

• Digital
• 12.5 kHz
• Requires less than 6 kHz Receiver ENBW to
  achieve for Class A
• Analog
• NPSPAC Analog (Special Case) 25 kHz channel, 12.5
  kHz spacing
• 20 dB Offset Channel Selectivity (_at_ 12.5 kHz)
• IF ENBW to meet Analog 25 kHz A 16 kHz
  (B-4-3)

9
Standard Interfering signal

• Analog FM
• Modulate interfering carrier with equal 650 and
  2,200 Hz tones each at 50 of the maximum peak
  deviation values
• Monitor audio output and measure audio signal to
  noise (SINAD), 12 dBS is reference sensitivity
• Project 25
• Use standard Interference Pattern
• Produces a silent output with equal signal
  deviations
• Monitor radios bit error rate, 5 is reference
  sensitivity

10
ACRR - Comparison of Older Analog 400 Hzand
Newer dual tone test modulation
400 Hz old RS204-D
650, 2200 Hz new TIA603
Wider sideband noise than 400 Hz method produces
Full voice deviation
Dual tone method more representative of actual
spectral power distribution. The change in
methodology dramatically changes the measured
ACRR due to the widen sideband energy. Confirm
that the current TIA methods are being used.
11
ACRR
Increase Interfering Carrier w/ Interference
modulation until Reference Sensitivity is regained
Differences between signal generator waveforms
vs. actual transmitters SPD. Very important when
dealing with linear modulations.
ACRR
ACPR
Reference Sensitivity 3 dB
Reference Sensitivity
3 dB
Cs/N
Cs/N
Cs/N
3 dB
12
Comparison of SPDs of 3 Analog and C4FM
Modulations
13
Receiver Modeling

• Need to know the bandwidth of the victim receiver
  (ENBW)
• Need to know the shape of the filtering
• 3 basis filter models
• Square (perfect) filter
• Butterworth (poles-cascaded sections
• 3 different versions based on manufacture
  recommendation
• Root Raised Cosine (RRC)

14
Filter Comparison (20 kHz ENBW)
Which is the correct one for any specific
radio? Manufacturer specified for proprietary,
generic for non proprietary. This slide shows the
difference in roll-off at 20 kHz (arbitrary value)
Worst Case Older Analog Radios
Newer Digital/ Analog radios
Perfect filter
15
Filter Designators
Details are listed in Table 88B and in Annex A
for each type of modulation

• Examples
• 5K50R02 RRC filter 5.50 kHz wide with a 0.2
• 16K0B0403 Butterworth filter 16.0 kHz wide, 4
  poles, 3 cascades

16
Receiver Filter Table

• Footnotes are critical to select the correct
  filter. Details also in Annex A text
• Distinction between 5 kHz and 4 kHz deviation
  impacts IF model due to OCR requirement
• Analog not as stringent as digital
• Butterworth filter primarily for older analog
  receivers and worst case evaluations

17
Footnote Comments

• Narrow analog (12.5 kHz) easily meets 45 dB ACRR
  requirement with a 7K80B0403 filter (7.8 kHz)
• Can get by with lesser number of cascaded stages
• Allowance for frequency drift
• Can achieve 60 dB ACRR with a narrower filter
• Increases the required Cs/N which is offset by
  the lower thermal noise due to reduced bandwidth
• Decreases maximum achievable SINAD
• P-25 Digital requires a narrower filter to meet
  60 dB requirement.
• Radios that support both narrow analog and P-25
  digital probably use a single narrower filter due
  to P-25 requirements

18
Footnote Comments (2)

• Older 5 kHz analog receivers typically use
  16K0B0403 but may be narrower if radio also uses
  NPSPAC channels other than the five I/O channels
• NPSPAC requires 12K6B0403 to achieve 20 dB OCR
• Reduced deviation for NPSPAC channels
• 4 kHz instead of 5 kHz

19
Footnote Comments (3)

• Multi-Mode radios probably use a single,
  compromise, ENBW for both the wide and narrow.
• Compromise ENBW 9.4 kHz B-4-3
• Overly Narrow ENBW for Wide
• 25 kHz ACRR 78 dB vs. 75 dB req mobile
• 30 kHz ACRR 79 dB vs. 75 dB req mobile
• Impacts high signal SINAD performance
• Overly Wide for Narrow (Just makes requirement)
• 12.5 kHz ACRR 45 dB vs. 45 dB req. mobile
• 15.0 kHz ACRR 65 dB vs. 65 dB req. mobile

20
Table A-1 TSB-88B

• Static C/N _at_ reference sensitivity
• Faded C/N for 3 different channel performance
  (DAQs)
• Future changes due to new vocoders
• Definition of DAQs

21
NTIA/Industry Canada Requirements

• Non standard offset frequencies
• Method of determining ACRR for cross border
  situations
• Different interference level Criterion
• IndCan 6 dB below receivers thermal noise
  floor
• Consider a different view of ACPR/ACRR
• TSB-88 application allows generation of different
  views
• Allows looking at offsets other than fixed values
  of TSB-88
• Compares modulation to a family of the
  recommended filters to determine the closest
  value for worst case compared to TIA
  specification
• Provides simple method to estimate receiver ENBW
• Requires converting the ACPR to ACRR
• Can also interpolate the ACPR Tables for the
  fixed offsets

22
Offset Frequency ACPR for various Receiver
ENBW Wide Analog FM 5 kHz Peak Deviation
23
Offset Frequency ACPR for various Receiver ENBW
Narrow Analog FM 2.5 kHz Peak Deviation
24
Offset Frequency ACPR for various Receiver
ENBW C4FM P-25 Modulation
25
Offset Frequency ACPR for various Receiver
ENBWSecurenet DVP (12 kbps)
ACRR gt 60 dB _at_ 25 kHz offset
Other capabilities may require a narrower IF,
such as DVP, in this example requires 11.1 kHz
for 4 dB C/N for 20 dB OCR
26
Estimating ACRR to ENBW

• Red values are TIA requirements, based on TIA
  methods of measurement and B 4-3 filter
  characteristics (conservative)
• ENBW values may be lower than the minimums shown
• Values of ACRR gt75 dB are less useful due to
  limited dynamic range of measured data files

27
TSB-88-B Spreadsheets on CD

• Analog FM 2.5 kHz Peak Deviation (AFM 2.5kHz
  Dev.xls)
• Analog FM 4.0 kHz NPSPAC (AFM 4kHz Dev.xls)
• Analog FM 5.0 kHz Peak Deviation (AFM 5kHz
  Dev.xls)
• C4FM Project 25 (C4FM.xls)
• DIMRS-iDEN (DIMRS-iDEN.xls)
• EDACS Narrow Band (EDACS-NB.xls)
• EDACS Narrow Band (EDACS-NPSPAC.xls)
• EDACS Narrow Band (EDACS-WB.xls)
• F4FM TDMA-2 (F4FM.xls)
• Linear Simulcast Modulation (LSM.xls)
• OpenSky F4GFSK (F4GFSK.xls)
• Securenet, 12 kbits/sec CVSD (Securenet.xls)
• TETRA (Tetra.xls)
• Tetrapol (Tetrapol.xls)
• Astro Widepulse, (Widepulse.xls)
• Additional spreadsheets for Wideband data systems
  are being developed for TSB-88.1-C.
• SAM, IOTA, DataRadio and RD-Lap

28
ACPR Offsets

• 11 combinations
• Tables of results in TSB-88
• Evaluate Source SPD against Victims receiver
• Determine Victims ENBW based on its own
  modulation

29
Spreadsheets Contain

• Calculator for ACP for the different filters
• Square
• Butterworth (P-C)
• 10-4
• 5-4 (special case)
• 4-3
• RRC
• Charts calculator results for each filter
• Graphs the TSB-88 Tables for all offsets
• P-25 C4FM example

30
Sample C4FM TSB-88 Table
31
Sample Spreadsheet Graphic

• 4 graphs
• Small offsets, 25 kHz plans (shown)
• Large offsets, 25 kHz plans
• Small offsets, 30 kHz plans
• Large offsets, 30 kHz plans

32
Conclusions

• Wide receivers are the most susceptible
• Need to evaluate in both directions
• Narrow Analog and P-25 have similar interference
  characteristics
• Minor difference is due to minor different
  sideband energy
• Obtaining non-ambiguous ACRR specifications is
  critical for frequency coordination
• Identify the modulation from emission designator
• ACRR at Offset allows estimating ENBW
• Current TIA methods of measurement
• 7.5 kHz offsets are impractical for interleaving
  with older radios
• Current 800 MHz NPSPAC rebanding provides
  opportunity to create additional spectrum usage
• Repack using worst case actual Source into Victim
  ENBW rather than current adjacent 12.5 kHz 40 dBm
  / 25 dBm

33
Frequency Coordination using ACPR Methodology

• Use Emission Designator to determine the
  modulation
• Determine ENBW of victim based on ACRR or
  recommendations in TSB-88
• Determine the ACPR for the combination to be
  evaluated, source Tx into victim Rx
• Reduce the ERP of the source Tx by the ACPR
• Alternatively reduce the contour criteria by
  adjacent channel ACPR
• Evaluate resultant Talk out service area
  reliability due to co and adjacent channel
  interference sources
• Monte Carlo
• Equivalent Interferer

34
Desired Signal
Numerous Interference Sources
IM Signal(s)
Margin for Reliability
Performance Requirement
Co-Ch, Adj-Ch, OOBE Power
Requirement C/(IN)
Aggregate Noise Interference
Other Noise
Goal is to control the co-channel, adjacent
channel, OOBE, IM power and the receivers own
internal noise to achieve the desired ratio of
desired signal to the composite power of the
undesired signals and their effects for the
desired level of performance.
35
Margins for Mitigating Interference
This margin plus capture is required, for 5 and
7dB capture requires 25.7 dB for s 8 dB.
Public Safety uses 1 NPSPAC band

• NPSPAC Example
• Co-channel 35 dB C/I based on 1 interference at
  contour 7 dB capture 33.4 dB, rounded to 35
  dB.
• Adjacent channel 15 dB C/I 20 dB OCR (25 dB
  ACRR) produces 40 dB effective C/I
• Multiple sources
• Different RPC definitions of Jurisdiction vs.
  Service Area

N
36
Joint Probability, Noise Interference
To achieve 90 in the presence of 5 Interference
requires a C/N reliability of 93.5
37
Monte Carlo Simulations

• Monte Carlo is preferred way to obtain
  statistical factors
• Tested Equivalent Interferer method against Monte
  Carlo method
• Excel spreadsheet
• For 30,000 throws per test, Monte Carlo results
  were typically within 1 of Equivalent Interferer
  Spreadsheet results
• Monte Carlo
• Determine the median signal levels.
• Draw from a Gaussian Distribution with the median
  value at the predicted level for each variable
  (Roll the dice)
• Repeat the process thousands of times for each
  point being calculated
• The probability is then determined by the number
  of successes divided by the number to tests
  (rolls), e.g. 5000 tests, with 4800 successes is
  4800/5000 96 probability of achieving the
  desired criteria
• Additional margins required for confidence level
• Best for Interference and Simulcast predictions

38
Monte Carlo Interference Simulation
Interferer 1
Interferer 2

• Monte Carlo Simulation
• For each draw, pass if C / (S I N) CPCs
  Cf/(IN), otherwise fail
• Accuracy increases with increased number of draws
• Tile Probability Pass/ draws
• Adjacent Channel(s) simulated by reducing source
  ERP by the ACPR value.

I1
I2
Desired
C
Tile
39
Potential NPSPAC Repacking Improvements during
rebanding

• ACPR improvements (assumed 25 dB in original
  deployments) based on 20 dB OCR
• P-25t ? P-25r (use 65 dB) 40 dB
  Improvement
• P-25t ? NPSPAC12.6r (32.5 dB) 7.5 dB
  Improvement
• NPSPACt ? P-255.5 r (57 dB) 30 dB
  Improvement
• NPSPACt ? P-257.8 r (46 dB) 20 dB
  Improvement
• NPSPACt ? NPSPACr (25 dB) No Change
• Allows closer adjacent channel reuse
• Requires RPC acceptance and involvement, could be
  controversial in a particular RPC

40
Additional TSB-88B capabilities

• Propagation models
• Favors those used by most manufacturers
• Acceptance Testing
• Detailed Definitions for testing
• Statistical Theory for determining number and
  size of test locations
• Interference Topics

41
Next Release(s)

• Anticipate sometime in late 2007
• Breaking 88C version into four documents
• TSB-88.1-C Performance Frequency Coordination
• TSB-88.2-C Propagation (Possible new model)
• TSB-88.3-C Acceptance Testing and Interference
  Mitigation
• TSB-88.4-C Broadband (combines elements of the 3
  above exclusive to broadband
• Adding Wide Band elements in 88.1 88.3

42
Contact TSB-88

• TSB88_at_Yahoogroups.com
• Post questions on document
• Answers by
• Bernie Olson (Past chair editor)
• Tom Rubinstein (Current chair editor)
• Technical Appendix to the Best Practices Guide
• http//www.apcointl.org/frequency/project_39/
  downloads/Interference_Technical_Appendix.pdf

43
Questions?
44
Backup Slides
45
Example of Static SINAD vs. C/N for various IF
ENBWs
Narrower IFs require a higher C/N due to
additional distortion
46
Offset Frequency ACPR for various Receiver
ENBW NPSPAC FM 4 kHz Peak Deviation
TIA 603-C 20 dB OCR ACRR