multiband transceivers - [chapter 7] multi-mode/multi-band gsm/gprs/tdma/amps system analysis
TRANSCRIPT
Multiband RF Transceiver SystemChapter 7
Multiband/Multi-modeGSM/GPRS/TDMA/AMPS
RF Transceiver System Analysis
Department of Electronic EngineeringNational Taipei University of Technology
Outline
• GSM/GPRS/TDMA/AMPS Transceiver Architecture
• Frequency Plan and Specifications
• Noise Figure Requirement Calculation
• Linearity Requirement Calculation
• Selectivity and Blocking Performance Evaluation
• Summary
Department of Electronic Engineering, NTUT2/38
Introduction
• A multimode and dual-band superheterodyne transceiverdesign, which actually covers the designs of GSM, TDMA,AMPS and GPRS mobile systems, is discussed.
• We’ll first take a look at the architecture of this multi-modetransceiver. Lately, the specifications of theses applicationswill be given with some tables.
• For the receiver, there are things to be done:
(1) Noise Figure: BER� Eb/N0 � CNR� Sensitivity� NF
(2) IIP3 : IMD requirement� CNR� Phase Noise/NF� IIP3
(3) Selectivity and Blocking
Department of Electronic Engineering, NTUT3/38
GSM/GPRS/TDMA/AMPS Transceiver
• Band: 800 MHz cellular and 1900 MHz PCS dual bands.LNA SW control
Cellular band
PCS band
LNA
LNA
Rx RFSAW
Rx RFSAW
GSMSAW
TMDA/AMPSSAW
IF VGA
I/Q downconverter
BBA
BBABB filter
BB filter
To BB
f/2VHF PLLAt 266.4/
268.04 MHz
Diplexer
Bia
s a
nd c
ont
rol
Ref. Clck
Rx chip
Tx chipCe
llula
r du
plex
er
PC
S d
uple
xer
From LoopLPF
1002– 1028.1MHz VCO
2063– 2124.1MHz VCO
PA
PA
Tx RF SAW
Tx RF SAW
Powerdetector
BB
Driver
Driver
PCS band
Cellular band
AFC VCTCXO
To VCOsLoop LPF
To BB
UHF synthesizer
Ref. Clock
Single side bandUp-converter
130.38 MHzOr
250.76 MHz
VHF PLLAt 521.52 (cell)Or 501.52 (PCS)
MHz
IFVGA
f/2
Σ To BB
Bia
s an
d co
ntr
ol
BB I
BB Q
To UHFsynthesizer
200 kHz
25 kHz
tunable
Department of Electronic Engineering, NTUT4/38
Frequency Plan (I)
• Half-duplex system: GSM, GPRS, and TDMAmode
• Full-duplex system: AMPS mode.
• 70 MHz span
• 1st IF > 70 MHz
� Choose 133.2 MHz for GSM and GPRS RX(200kHz ch spacing)
� Choose 134.04 MHz for TDAM and AMPS RX(30 kHz ch spacing)
• Common reference clockis used
• 19.2 MHz reference clockis popular used in mobilestations.
SystemUplink (MHz)
Downlink (MHz)
Separation(MHz)
Channel Spacing(kHz)
Cellular 824 – 849 869 – 894 20 30 (CDMA)
GSM 900 890 – 915 935 – 960 20 200
E-GSM 900 880 – 915 925 – 960 10 200
DCS 1800 1710 – 1785 1805 – 1889 20 200
PCS 1850 – 1910 1930 – 1990 20 50 (CDMA)
WCDMA 1920 – 1980 2110 – 2170 130 200
802.11b 2400 – 2484 2400 – 2484 – 13000
802.11a5150 – 5350 5725 – 5825
5150 – 53505725 – 5825
––
2000020000
Department of Electronic Engineering, NTUT5/38
Frequency Plan (II)
• UHF synthesizers frequency tuning range (high-side injection):
� 800 MHz cellular band:1002 to 1029 MHz
� 1900 MHz PCS band: 2063 to2125 MHz
GSM and GPRS TDMA and AMPS
VCTCXO 19.2 19.2
Receiver IF (MHz) 133.2 134.04
Transmitter IF (MHz) 178.2 179.04
UHF VCO Tuning (MHz) 2004 – 2058 2063 – 2125
Receiver VHF VCO (MHz) 266.4 268.08
Transmitter VHF VCO (MHz) 356.4 358.08
Use a single VCO with a tuning range 2004 to 2125 MHz and a divide-by-2divider. (2004 to 2125 MHz is about 6% of the VCO operating frequency,and this is a quite reasonable tuning range that still maintains good phasenoise performance)
Department of Electronic Engineering, NTUT6/38
GSM RX Specifications (I)
GSM & GPRS 800/1900 Specifications Note
Frequency Band 869~894 or 1930~1990 MHz
Modulation GMSK
Symbol rate 270.833 ksps
Sensitivity < −102 dBm RBER < 2%
800 MHz band GPRS sensitivity (packet data channel)
< −100 dBm BLER < 10%
1900 MHz band GPRS sensitivity (packet data channel)
< −102 dBm BLER < 10%
800 MHz dynamic range > −15 dBm RBER < 0.1%
1900 MHz dynamic range > −23 dBm RBER < 0.1%
Intermodulation spurious response attenuation > −49 dBmf1: ±800 kHz offset (CW)f2: ±1.6 MHz offset (Mod)
Table A
1
2
3
Department of Electronic Engineering, NTUT8/38
GSM RX Specifications (II)
GSM & GPRS 800/1900 Specifications Note
Adjacent channel selectivity > 9 dBc ±200kHz offset, 2% BER
Atl. adjacent channel selectivity > 41 dBc ±400kHz offset, 2% BER
Blocking characteristic > 49 dBc ±(600 kHz to 1.6MHz) offset, 2% BER
Blocking characteristic > 66 dBc ±(1.6 MHz to 3MHz) offset, 2% BER
800 MHz blocking characteristic > 76 dBc > 3 MHz offset, 2% BER
1900 MHz blocking characteristic > 73 dBc > 3 MHz offset, 2% BER
800 MHz spurious emission <−79 dBm/100 kHz In Receiver Band
1900 MHz spurious emission < −71 dBm/100 kHz In Receiver Band
Spurious emission < −36 dBm/100 kHz In Transmitter Band
Table B
Department of Electronic Engineering, NTUT9/38
TDMA RX Specifications
TDMA 800/1900 Specifications Note
Frequency band 869~894 or 1930~1990 MHz
Modulation π / 4 DQPSK
Symbol rate 24.3 ksps
Sensitivity < −110 dBm BER < 3%
Dynamic range > −25 dBm BER < 3%
Intermodulation spuriousresponse attenuation
> 62 dBcf1: ±120 kHz offset (CW)f2: ±240 kHz offset (CW)
Adjacent channel selectivity > 13 dBc ±30 kHz offset, 3% BER
Atl. Adj. channel selectivity > 42 dBc ±60 kHz offset, 3% BER
Spurious emission < −80 dBm In receiver band
Table C
1
2
Department of Electronic Engineering, NTUT10/38
AMPS RX Specifications
AMPS 800/1900 Specifications Note
Frequency Band 869 – 894 MHz
Modulation FM
Noise Bandwidth ~27 kHz
Sensitivity < −116 dBm SINAD = 12dB
Dynamic range > −25 dBm
Intermodulation spurious response attenuation > 65 dBcf1: ±60 kHz offset (CW)f2: ±120 kHz offset (CW)
Intermodulation spurious response attenuation > 70 dBcf1: ±330 kHz offset (CW)f2: ±660 kHz offset (CW)
Adjacent channel selectivity > 16 dBc ±200 kHz offset, 2% BER
Atl. Adjacent channel selectivity > 60 dBc ±400 kHz offset, 2% BER
Spurious emission < −80 dBm In receiver band
Table D
1
2
3
Department of Electronic Engineering, NTUT11/38
CNR for GSM
• In the receiver systemdesign, we need to first determine thecarrier-to-noise ratio (CNR) for each operation mode at aspecified BER.
• RBER < 2% for GSM speech channel (TCH/FH Class II).
We use BT=0.25 here to evaluate(actually, GSM is 0.3), and get
Eb/N0 = 5 dB
Next step is to find the required CNR.
1(Table A )
Eb/N0 (dB)
Bit
Err
or
Rat
e
Department of Electronic Engineering, NTUT13/38
Impairments Consideration
• Consider impairments:Eb/N0 = 5.6 dB
• Bit rate = 271 kHz and the RXnoise BW = 182 kHz (thechannel selection filterBW), thus theCNR:
• We can useCNRGSM = 8 dB to reserve 0.7 dB margin
Item Specification Eb/N0 Degradation
Total integrated phase noise of two LOs < −25 dBc 0.1
Group delay distortion of channel filters < 2 µsec 0.4
I andQ imbalance in phase and magnitude < 5° and < 0.5 dB 0.1
0
27110log 5.6 10log 7.3 dB
182b b
GSM
E RCNR
N BW= + = + =
0.6 dB degradation from impairments
(see Ch3 slide-24)
�
Department of Electronic Engineering, NTUT14/38
Other Considerations
• However, when the speech channel with AMR (adaptivemultiple rate), the requiredCNR for the same sensitivity of−102 dBmwill be approximately 1.5 dB higher than that of theoriginal speech channels - i.e., (8 + 1.5) = 9.5 dB.
• The requirement on theCNR in the worst case is the channelTCH/AFS5.9 in HT100 propagation condition:
For a −102 dBm sensitivity 9.4 dB CNR is needed evenwithout AMR.
Department of Electronic Engineering, NTUT15/38
CNR for GPRS
• The correspondingCNR for a 10%block error rate (BLER) inpacket data channels (PDCH) is in general approximately8 dBas required by the GSMspeech channels, but in the worst case,it may rise to close to10 dB.
• However, the reference sensitivity of the GPRS in the worstcase (PDCH/CS-4 in 1800 MHz band, for example) is relaxedto −100 dBminstead of−102 dBm.
• For simplicity, the performance evaluation of the GSMandGPRS systemRF receivers later on will be based onCNRmin = 8 dB.
2(Table A )
In the practical system design we should leave enoughmargin to cover the performance in the worst caseincluding GSM speech channel with AMR.
�
Department of Electronic Engineering, NTUT16/38
CNR for TDMA (I)
• In a similar way, we can determine theCNR for performancecalculation of the TDMAreceiver where the signal isπ/4 -DQPSKmodulated.
1(Table C )
Eb/N0 (dB)
Bit
Err
or
Rat
e
Eb/N0 degradation due to ISI from IF (SAW + ceramic) filters
• Eb/N0 for a 3% BER isapproximately equal to 5dB in the case of ISI free.
Department of Electronic Engineering, NTUT17/38
CNR for TDMA (II)
• Assume that the channel filters have a total group delaydistortion 1.5 ps, which causes 0.3 dBEb/N0 increase to keepthe 3% BER, and the other factors such asI and Q channelmismatching and the phase noise of the LOs raiseEb/N0another 0.2 dB.
• Therefore, it needs total 5.5 dB to reach 3% BER.
• ConsideringRb/BW = 2 (2-bits/symbol) in the case ofπ/4-DQPSKmodulation, we obtainCNR to be
• Similar to the GSMsituation, we add 0.5 dB to the aboveCNRTDMA value for performance evaluation of the TDMAmobile receiver:CNRTDMA = 9 dB.
5.5 10log 2 8.5 dBTDMACNR ≅ + = (see Ch3 slide-24)
�
Department of Electronic Engineering, NTUT18/38
CNR for AMPS (Analog Wireless System)
• The SINAD is used in an analog AMPS (FM) to measure thesensitivity and other performance instead of BER.
• The SINAD value defined to measure AMPS receiverperformance is 12 dB, and we useCNRAMPS = 3.0 dB(approximately 0.5 dB higher than the readCNR value).
( )10log dBS N D
SINADN D
+ +=+
( )( )
( ) ( ) ( )
( )
2
2 2
2
32
1 123
11
oC N
C N
f BWC N
B BS NBW f
C N e C NB BW
eπ
−
−
∆ =
∆ + +
−
( ) ( )10log 10log poSNR S N S N G= = +
(Not SNDR)1(Table D )
CNRAMPS (dB)
SIN
AD
(d
B)
�
( )
( )
6 2 20 10
6 2 20 10
1 10 1010log
10 10
p
p
SNR G SNR SNR
SNR G SNR SNRSINAD
− − + − −
− − + − −
+ +=+
Department of Electronic Engineering, NTUT19/38
Required Noise Figure
• The receiver static sensitivity is determined by the noisebandwidth, noise figure, andCNR.
Usually we would like to have a 4 dB margin in the typical case and a 1.5 dBmargin in the worst case.
( )3174 106 10log 182 10 8 7.4 dBGSMNF = − − × − ≅
( )3174 114 10log 27 10 9 6.7 dBTDMANF = − − × − ≅
( )3174 120 10log 27 10 3 6.7 dBAMPSNF = − − × − ≅
(see Ch4 slide-8, we didn’t consider digitalprocessing improvement here)
The NF of this multimode receivershall be 6.7 dB or lower. Themaximum NF should be 9.2 dB orless, and thus the sensitivity still has1.5 dB margin in the worst case.
Spec. GSM/GPRS TDMA AMPS
Noise Floor −174 dBm/Hz −174 dBm/Hz −174 dBm/Hz
Sensitivity (Spec.) −102 dBm −110 dBm −116 dBm
Margin 4 dB 4 dB 4 dB
Sensitivity (Target ) −106 dBm −114 dBm −120 dBm
Noise Figure Requirement 7.4 dB 6.7 dB 6.7 dB
Department of Electronic Engineering, NTUT20/38
Linearity and IIP3
• Receiver linearity is usually measured by theIIP3.
• The linearity requirement is more complicated to determinethan the receiver noise figure.
• The requirement on the overallIIP3 of a wireless mobilereceiver is dominated by the allowed intermodulationdistortion (IMD) or formally referred to as intermodulationspurious attenuation and the phase noise of UHF synthesizerLO.
Department of Electronic Engineering, NTUT22/38
IMD Performance Requirements
• GSM: Desired signal −99 dBm (−102+3)
Minimum interferers −49 dBm
Tone/modulated interferer, offset frequency ±800/±1600 kHz
• TDMA: Desired signal−107 dBm (−110+3)
Minimum interferers 62 dBc higher;
Tone/tone interferer, offset frequency ±120/±240 kHz
• AMPS: Desired signal −−−−113 dBm (−116+3)
Minimum close-spaced interferers 65 dBc higher
Close-spaced tone/tone interferer, offset ±120/±240 kHz
Minimum wide spaced interferers 70 dBc higher
Wide-spaced tone/tone interferer, offset±330/±660 kHz.
3(Table A )
2(Table C )
2(Table D )
3(Table D )
Department of Electronic Engineering, NTUT23/38
You may like to evaluate the desired signal power with thesensitivity (target), e.g., −106 dBm, calculated in slide-20.
Calculate IIP3 by Ignoring Other Influence
• If ignoring other factor influence to the intermodulation:
( )3,min , ,min , min
13
2d i in d iIIP S I S CNR = + − + , min_ 3 dBd i refS S= + : Receiver input desired signal
,mininI : Minimum input interference strength
min_refS
, mind iS CNR−
min3IIP,mininI
( )
( )
,min , minmin , min
, ,min , min
3 32
1 3
2
in d id i
d i in d i
I S CNRIIP S CNR
S I S CNR
− += ⋅ + −
= + − +
3 dB
,d iS
input@Iout,min reaches
minCNR
minCNR,minoutI
Department of Electronic Engineering, NTUT24/38
Estimated IIP3
• If ignoring other factor influence to the intermodulation:
( )3,min , ,min , min
13
2d i in d iIIP S I S CNR = + − +
[ ]3,min
199 3 49 8 21.5 dBm
2GSMIIP = − + × + = −
[ ]3,min
1107 3 62 9 9.5 dBm
2TDMAIIP = − + × + = −
[ ]3,min
1113 3 65 3 14 dBm for close-spaced
2AMPSIIP = − + × + = −
[ ]3,min
1113 3 70 3 6.5 dBm for wide-spaced
2AMPSIIP = − + × + = −
Spec. GSM/GPRS TDMA AMPS
Desired signal −99 dBm −107 dBm −113 dBm
Spurious ResponseAttenuation
49 dBc 62 dBcclose wide
65 dBc 70 dBc
CNRmin 8 dB 9 dB 3 dB 3 dB
Required IIP3 −21.5 dBm −9.5 dBm −14 dBm −6.5 dBm
Department of Electronic Engineering, NTUT25/38
Consider Other Influence
• LO phase noise, spurious, and the receiver noisefigure will also impact the IMDperformance:
The LO phase noise and spurious level, especially at the offset frequenciesequal to those frequencies where the intermodulation test interferers arelocated, should be low enough to ensure that the requested receiver IIP3 forcertain IMD performance is reasonable and feasible.
• The VHF LOphase noise and spurious:Generally have a negligible impact on the IMD performance if the IF channelfilter has good rejection to the interferers.
max, 2 2 2 210 10
3,min ,min , , , ,1 1 1 1
13 10log 10 10
2
nfin ND
in phn j k spu j kj k j k
IIP I P P= = = =
= − − − −
∑∑ ∑∑
Department of Electronic Engineering, NTUT26/38
LO Phase Noise
Cellular Band Synthesizer PCS Band Synthesizer
Frequency Offset (kHz)
Phase Noise (dBc/Hz)
Spurs (dBc)Phase Noise
(dBc/Hz)Spurs (dBc)
±30 kHz −105 −60 −103 −60±60 kHz −117 −85 −114 −85±120 kHz −125 −90 −122 −90±240 kHz −131 −95 −128 −95±330 kHz −134 −95 −131 −95±660 kHz −140 −95 −137 −95±3000 kHz −144 −95 −142 −95
800 MHz Band 1900 MHz Band
Frequency Offset (kHz)
Phase Noise (dBc/Hz)
Spurs (dBc)Phase Noise
(dBc/Hz)Spurs (dBc)
±200 kHz −118 −60 −114 −60±400 kHz −124 −65 −120 −65±600 kHz −127 −70 −123 −70±800 kHz −130 −78 −126 −78±1600 kHz −136 −85 −132 −85
> |3200 k| Hz −141 −90 −137 −90
The phase noise may vary a couple of dB with temperature from room to hot (60°C) or to cold (-30°C).
For
TD
MA
and
AM
PS
For
GS
M
3(Table A )
Table E
Table F
Department of Electronic Engineering, NTUT27/38
Required IIP3 Estimation (I)
• The required IIP3 of 800 MHz band GSMreceiver iscalculated by using phase noise for GSMas
• The required IIP3 of 1900 MHz band GSMreceiver is
• In a similar way, we can obtain IIP3 for other modes andbands based on phase noise and spurious for TDMA/AMPSLO as follows.
3, _1900 14.8 dBmGSMIIP = −
3, _1900 2.8 dBmTDMAIIP = −
3, _800 3.4 dBmTDMAIIP = − 3, _ 12.6 dBmAMPS closeIIP = −
3, _ 9.3 dBmAMPS wideIIP = −
( )3 3 3 85 49 3
1099 8 174 7.4 10log182 10 130 10log182 10 49 3 136 10log182 10 49 3 78 49 3
10 10 10 10 103, _ 800
13 49 3 10log 10 10 10 10 10 10 15 dBm
2GSMIIP− − +− − − + + ⋅ − + ⋅ − + − + ⋅ − + − − +
= − + − − − − − − = −
3(Table A )
3 dB margin
PN@800 kHz PN@1600 kHz
Spurious@800 kHz Spurious@1600 kHz
Use NF=7.4 for GSM
Department of Electronic Engineering, NTUT28/38
Required IIP3 Estimation (II)
• Fromthese results we can conclude that:
TDMA receiver requires the highest linearity, and the receiver operating in the1900 MHz band needs higher IIP3 than when it is running in the 800 MHz bandsince the phase noise of the PCS band LO is worse than that of the cellular bandLO.
• The linearity design of the receiver common path for differentmodes should bebased on the TDMA requirement, but in thecircuit design we should also consider adjustable bias circuitryto change the device bias based on operation modes to save thecurrent consumption.
Department of Electronic Engineering, NTUT29/38
Selectivity and Blocking Performance
• Receiver selectivity and blocking performance are mainlydetermined by :
� Channel filters
� LO phase noise
� Spurious
• The LO phase noise/spurious requirements is also partiallydetermined by the IMDperformance, and therefore we havealready had a basic idea what level phase noise/spurious canbe used in our receiver systemdesign.
Department of Electronic Engineering, NTUT31/38
Channel Selection Filtering
• The channel filter characteristics affect not only the receiverselectivity and blocking performance but also the IMDperformance.
Since no matter what the adjacent/alternate channel interferers are, distanceblocking signals, or intermodulation interference tones/modulated signals will besignificantly attenuated when they pass through the channel filters.
• It is a trial and error procedure to make a tradeoff betweenfiltering requirements and feasibility of implementation.
• Examples of channel filter characteristics for the GSMreceiver and for the TDMAor the AMPS receiver arepresented in Table next slide.
Department of Electronic Engineering, NTUT32/38
Channel Filter Characteristics
GSM Channel Filter TDMA/AMPS Channel Filter
Insertion Loss (dB) Insertion Loss (dB)
Typical Worst Typical Worst
In-band 4.5 5.5 In-band 3.5 4.5
Rejection (dB) Rejection (dB)
Offset frequency
Typical WorstOffset
frequencyTypical Worst
±200 kHz 4 0 ±30 kHz 3 0
±400 kHz 17 12 ±60 kHz 24 20
±600 kHz 27 22 ±120 kHz 40 35
±800 kHz 31 25 ±240 kHz 50 40
±1600 kHz 40 30 ±330 kHz 45 40
±3000 k Hz 40 30 ±660 kHz 43 35
Table G
Department of Electronic Engineering, NTUT33/38
Calculate Selectivity and Blocking
• Utilizing
• GSM adjacent channel selectivity:
,
,1 ,2 ,1 ,2
174 10log
10 10
/ / / / , ,10log 10log
10 10 10 10
10 1010log
10 10 10 10
d i
phn phn IF spu spu IF
S CNR BW NF
adj alt block adj alt block d i d iN BW N R BW N N RS I S S
− − + +
+ −∆ + −∆
− ∆ = − = −
+ + +
3
3 3
99 8 174 10log182 10 7.4
10 10
118 10log182 10 108 4 10log182 10 60 55 4
10 10 10 10
10 1010log 99 46.8 dB
10 10 10 10adjS
− − − + ⋅ +
− + ⋅ − − + ⋅ − − −
− ∆ = + =
+ + + Rejeciton@200 kHz (adjacent ch.)Spurious@200 kHzPN@200 kHz
VHF LO, IF rejection (LO performance not shown here)
UHF LO
Department of Electronic Engineering, NTUT34/38
AMPS Adjacent Channel Selectivity
• The AMPS adjacent channel selectivity is calculated in asimilar way:
In the above selectivity calculations, it is assumed that the VHF LO phase noise atthe corresponding adjacent channel is 10 dB worse than the UHF LO phase noiseand the spurious is 5 dB worse.
• The results showmargins over 37 and 29 dB, respectively, forGSM and AMPS cases.(GSM > 9 dBc, Table B; AMPS > 16 dBc, Table D)
3
3 3
117 3 174 10log27 10 6.710 10
105 10log27 10 95 10log27 10 3 60 55 3
10 10 10 10
10 1010log 120 45.1 dB
10 10 10 10adjS
− − − + ⋅ +
− + ⋅ − + ⋅ − − − −
− ∆ = + =
+ + + Rejeciton@30 kHz (adjacent ch.)Spurious@30 kHzPN@30 kHz
VHF LO, IF rejection (LO performance not shown here)
UHF LO
Department of Electronic Engineering, NTUT35/38
Selectivity and Blocking Performance (I)
• The adjacent/alternate channel selectivity andblocking performance can be estimated with theformula given in previous slide.
• Estimated GSMperformance:
800 MHz Band 1900 MHz Band
GSM mobile receiverLoose LO(Table F)
MarginTight LO(Table E)
MarginLoose LO(Table F)
MarginTight LO(Table E)
Margin
Adjacent channel (dBc) 45.8 36.8 49.4 40.4 45.7 36.7 50.3 41.3
Alternate channel (dBc) 54.0 13.0 56.2 15.2 54.0 13.0 55.9 14.9
Block 0.6 – 1.6 MHz (dB) 59.9 3.9 61.1 5.1 58.3 2.3 60.9 4.9
Block 1.6 – 3.0 MHz (dB) 72.4 6.4 74.9 8.9 69.5 3.5 74.2 8.2
Blocking > 3 MHz 79.4 3.4 82.0 6.0 75.4 2.4 80.1 4.1
Department of Electronic Engineering, NTUT36/38
Selectivity and Blocking Performance (II)
800 MHz Band 1900 MHz Band
TDMA mobile receiver Tight LO Margin Tight LO Margin
Adjacent channel (dBc) 41.2 28.2 40.9 27.9
Alternate channel (dBc) 62.4 20.4 59.6 17.6
AMPS mobile receiver Tight LO Margin Tight LO Margin
Adjacent channel (dBc) 45.1 29.1 NA NA
Alternate channel (dBc) 66.2 6.2 NA NA
• Estimated TDMA/AMPS performance:
Department of Electronic Engineering, NTUT37/38
Summary
• To estimate the required receiver noise figure, onecan start fromthe BERand sensitivity specifications.
BER � Eb/N0 � CNR � Sensitivity � NF
• One can start fromthe IMD requirement, and utilizethe determinedCNR and noise figure to estimate IIP3requirement.
IMD requirement � CNR � Phase Noise/NF � IIP3
• Selectivity and blocking performance needs someinformation, such as filter responses and 1st LO/2nd
LO performance, and sometimes it requires tedious“trial and error” analysis.
Department of Electronic Engineering, NTUT38/38