Transcript

1

CDMA Concepts andCDMA Concepts andApplications in Wireless PCSApplications in Wireless PCS

NetworksNetworks

WFI Technical Training Series:

Developed By: Kamran Etemad

2

Course OutlineCourse Outline

l Introductionl Part 1: CDMA Concepts

» Spreading/Despreading in Time and Frequency Domains» Concept of Multiple Access Using Codes» Spreading Codes (Walsh and Pseudo-Noise Codes)» Rake Receivers and Soft Handoff» CDMA Cell and System Capacity

l Part 2: Applications: IS95 and 3G-CDMA» Physical and Logical Channel Structure in IS95» Forward and Reverse Link Waveforms» Call Processing and Power Control Issues

InformationInformationSourceSource

Source Encoder

Source Encoder

ChannelEncoder

ChannelEncoder

DigitalModulator

DigitalModulator

Information Information DestinationDestination

Source Decoder

Source Decoder

ChannelDecoder

ChannelDecoder

DigitalDemodulator

DigitalDemodulator

AnalogWaveformChannel

Review of a Basic Communication SystemReview of a Basic Communication System

l The function of source coding is datacompression.

l Removing redundancies of thesignals, in its original form, andrepresenting it with minimumnumber of bits.

l Signal compression may be lossy orlossless.

l Question:» How do we compress analog signals such

as voice and music?» Which Applications require lossless

compression?

Source CodingSource Coding

Information Information BitsBits

Channel Channel BitsBits

CompressedInformation

Add Redundancyto Protect Info. bits.

Channel CodingChannel Coding

l Channel Coding: adds redundant bits to information bitsuch that Protects Information Bits against Channel Noiseand Interference by increasing the distance between validcodes.

(Eb/Io)minCapacity

l Using more powerful channelcoding and modulation schemesincrease the tolerance againstnoise and interference.

l This means for a given Bit ErrorRate (BER) coding reduces therequired (Eb/N0).

BER

(Eb/Io)

Without Coding

With Coding

CodingGain

Coding GainCoding Gain

A B C D E F G C E G A F B D Interleaver

A B C D E F G C E G A F B D De-Interleaver

ErrorsErrorsErrorsErrors

InterleavingInterleavingl Conventional FEC schemes work best when the errors are

randomly distributed in time as opposed to being clusteredin bursts.

l In mobile radio channels, however, errors tend to occur inbursts due to fading effects.

l The function of interleaver is randomization of errors intime.

l The bits' order of transmission is altered, so that uponundoing this altering at the receiver, the errors appear tohave random rather than correlated locations

8

Q

IX X

X X

QPSK

X

ωω1

X

X

ωω3

ωω2FSK

Q

IX X X X

ASK

l A Digital Modulator maps a block of L bits toone of 2L Waveforms suitable for transmissionover a physical channel.

l Examples:» ASK (Amplitude Shift Keying)» FSK (Frequency Shift Keying)» PSK (Phase Shift Keying)» QAM (Quadrature Amplitude Modulation)

is a hybrid modulation

Q

IX

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

16QAM

Examples of Digital ModulationsExamples of Digital Modulations

Insecure, Unreliable Digital Fading Channel

Insecure, Unreliable DigitalMemoryless Channel

Information Destination

Source Decoder

Channel Decoder

DemodulatorInsecure AnalogFading Channel

Deinterleaver

Secure, Reliable, DigitalMemoryless Channel

Decryption

Review of FunctionalitiesReview of Functionalities

10

Spread Spectrum IdeaSpread Spectrum Idea

l Based on Shannon’s Capacity equation:» C=W x log(1+S/R)» A spread spectrum communication is designed so that

the system can operate at much lower signal to noiseratio using a much larger bandwidth.

l Starting from a typically narrowband information Signall The energy of the signal is spread over a much larger

bandwidth using:» Direct Sequence Spreading:

– Modulating each information bit by a high rate sequence DirectSequence Spreading

» Frequency Hopping:– Randomly hopping the sub-carrier frequency within a wide

spectrum.

11

SourceEncoder

ChannelDecoder

Modulator

ChannelEncoder

SourceDecoder

Demodulator

PNSource

PNSource

Wideband Wireless Channel

Identical &SynchronizedPseudo-Noise

High RateSignals

TRANSMITTER RECEIVER

Direct Sequence Spread SpectrumDirect Sequence Spread Spectrum

DS-Spread Spectrum FeaturesDS-Spread Spectrum Features

l Interference Rejectionl Anti-jamming Communicationl Frequency Diversity Against Multipath Fadingl Low Probability of Interceptsl Secrecy and Securityl Code Division Multiple Access (CDMA)

Capabilityl Provides high capacity and spectral efficiency in a

cellular network environment.l Provides no advantage over a pure additive

Gaussian Noise Channel.

13

ApplicationsApplications

l Military Based Applicationsl Second Generation Cellular and PCS

Systems (IS95)l Wireless Local loop Systemsl Third Generation/IMT2000 Systems

(CDMA2000, WCDMA,..)l Mobile-Satellite Systems (Global-Star)l ….

BTS

BTS

BTS

BSC

BTS

BTS

BTS

BSC

ISDN

PSTN

AUC

OMC

HLR VLR

MSC

DataNetworks

MS

MS

Base Station Subsystem

Network SwitchingSubsystem

PublicNetworks

Cellular Network ArchitectureCellular Network Architecture

15

Course OutlineCourse Outline

l Introductionl Part 1: CDMA Concepts

» Spreading/Despreading in Time and Frequency Domains» Concept of Multiple Access Using Codes» Spreading Codes (Walsh and Pseudo-Noise Codes)» Rake Receivers and Soft Handoff» CDMA Cell and System Capacity

l Part 2: Applications: IS95 and 3G-CDMA» Physical and Logical Channel Structure in IS95» Forward and Reverse Link Waveforms» Call Processing and Power Control Issues

16

Binary and Bipolar SequencesBinary and Bipolar Sequences

l Example of a binary sequence» 1,1,0,0,1,1,0,0

l Each binary sequence can be represented in abipolar form by mapping the ‘1’s to ’+1’s and ‘0’sto ‘-1’s. Example» 1,1,0,0,1,1,0,0 +1,+1,-1,-1,+1,+1,-1,-1

l For convenience, in our correlation analysis we usebipolar representation of binary sequences

[ 1 -1 -1 1 -1 1 ]

[ 1 1 -1 1 1 -1 ]

0

[ 1 -1 -1 1 -1 1 ]

1

[ 1 -1 -1 1 -1 1 ]

C1 & C2 are Orthogonal

Autocorrelation

Cross-correlationC1C1

C1C1

C2C2

C1C1<C1,C1>=1<C1,C1>=1

<C1,C2>=0<C1,C2>=0

C1 is Normalized

.dt∫

.dt∫

Correlation Between Two SequencesCorrelation Between Two Sequences

[ -1 1 1 -1 1 -1 1 -1-1 1]

Information Bits

Spreading Sequence

After Spreading

“1” “0”

-1

+1

Information Bits

SpreadingSequence

Spreading is achieved whenmultiplying the signal by thespreading sequence.

Spread Spectrum Signal

Spreading WaveformsSpreading Waveforms

Information SignalBefore Spreading

Information SignalAfter Spreading

R

W

Same Power i.e.Same Area

Frequency

2 Spreading Increases the rate of alternations andtherefore Bandwidth.

Effects of Spreading in SpectrumEffects of Spreading in Spectrum

20f

R

W

Processing GainProcessing Gain

l …Therefore, Spreading involves dividing each bit time intoL equal chip times and modulating the bit interval by asequence p(n) of length L.

l As a result of this multiplication/modulation thebandwidth of the transmitted signal increases by about afactor of L, thus the term spread spectrum.

l The ratio of bandwidth after spreading (W) to informationbit rate (R) is called processing gain (GP),» Thus : GP = W/R=L.

l Despreading is accomplished by correlating the receivedwaveform with the the same sequence p(n).

21

[ 1 -1 -1 1 -1 1 ]

[ 1 -1 -1 1 -1 1 ]

C1C1

C1C1

The original low rate “Information bit”

[ 1 -1 -1 1 -1 1 ]

[ 1 1 -1 1 1 -1 ]

C1C1

C2C2

Another high rate signal

freq.

freq.

freq.

freq.

Using theMatchingCode

Using theWrongCode

Despreading: Time/Frequency ViewsDespreading: Time/Frequency Views

1

0

.dt∫

.dt∫

Narrowband Signal

S.S. Interference

l Within the bandwidth of a narrowband system theSS signal looks like a white Gaussian noise.

SS Interference on NB SystemsSS Interference on NB Systems

Narrowband Interference

Narrowband Interference

Desired SS signal

Desired Signal

Before Despreading

After Despreading

Effect of Despreading NB InterferenceEffect of Despreading NB Interference

Despreading (review)Despreading (review)

l As a result of correlating with a user specific code at thereceiver:» The signal from the intended user gets despread.» The additive white gaussian noise remains the same.» The narrowband interference gets spread and appears as AWGN.» Also all unintended S.S. signals, that have been spread using

different codes, remain spread and therefore appear as AWGN.

l This is the basis for a multi-user secure communicationsystem based on spread spectrum idea.

25

Course OutlineCourse Outline

l Introductionl Part 1: CDMA Concepts

» Spreading/Despreading in Time and Frequency Domains» Concept of Multiple Access Using Codes» Spreading Codes (Walsh and Pseudo-Noise Codes)» Rake Receivers and Soft Handoff» CDMA Cell and System Capacity

l Part 2: Applications: IS95 and 3G-CDMA» Physical and Logical Channel Structure in IS95» Forward and Reverse Link Waveforms» Call Processing and Power Control Issues» Link Budget for CDMA Systems» CDMA and 3rd Generation Wireless PCS Systems

Basic CDMA ConceptsBasic CDMA Concepts

l CDMA: assigns one distinct spreading code toeach user

l As long as the codes are orthogonal or almostorthogonal all users can send and receive theirsignal through the same wide band channel.

l Other users’ signals appear like noise.

Frequency

time

f

t

f

t

f

t

f

t

Code

User1: C1

User2: C2

User3: C3

User4: C4Spread Spectrum Channel

Time

27

CDMA and Universal Channel ReuseCDMA and Universal Channel Reusel A CDMA system allows multiple access using

a single CDMA channel.l The same channel can be used in adjacent

cells. Thus CDMA allows a universal reusepattern, or reuse of one.

l The universal reuse implies:» A significant improvement of spectral efficiency

because of increased spectrum available per cell.» A tremendous amount of Co-Channel Interference.

– Because of spread spectrum nature of signals allco-channel interference appear like noise tointended user.

– Since different base stations or users, usedifferent codes with almost zero correlation, thereceivers can reject co-channel interference aspart of despreading.

F1F1

F1

F1F1

F1

F1

Almost Orthogonal Code Sequences

Frequency

FF11 FF22 FF33 FF44

C D M

A

F D M A

PN1

PN2

PN3

PNn-1

PNn

Example: Multiple Access in IS95Example: Multiple Access in IS95

l In IS95, Physical Channels are formed based on acombination of Code Division Multiple Access (CDMA)and Frequency Division Multiple Access (FDMA).

C1

C2

C3

Cn-1

Cn

Uplink Downlink

C’1

C’2

C’3

C’n-1

C’n

Example: IS95 CDMA ChannelsExample: IS95 CDMA Channels

l Each CDMA Frequency Assignment (FA) consists of a pairof 1.23MHz channels for downlink and uplink.

l Within each CDMA RF channel, or FA, signals to and fromvarious users are distinguished using different codes.

l The spreading codes used in forward and reverse link aredifferent.

30

Forward and Reverse Link CDMAForward and Reverse Link CDMA

l In the forward link there are two levels ofspreading:» Each base station uses a different code, so that the

interference from adjacent cells can be rejected at themobile’s receiver..

» Within each cell, the base station uses a set oforthogonal codes for channelization, to separatedifferent users information signals.

– Different users signal are first spread by a distinct code,– Then all spread spectrum signals for all users are added– and the composite signal is spread by the BS stations specific

code.

l In the reverse link each user uses a differentspreading code.

31

A Two Receiver ScenarioA Two Receiver Scenario

l A CDMA base station (BS) intends to send a “1” to user 1.l The BS spread the information bit by code C1.

» User 1 uses C1 for despreading» The other user, User 2, uses a different code C2, which is orthogonal

to C1.

User 1

User 2

[ ]C1 1 1 1 1= + + − −, , ,

[ ]C2 1 1 1 1= + − + −, , ,

[ ] [ ]{ }< >= × + + − − + − + −

= × − + − =

C C1 2 1 4 1 1 1 1 1 1 1 1

1 4 1 1 1 1 0

, / , , , , , , ,

/ { }

Note That C1 andC2 are Orthogonal.

32

[ ]~ , , ,C C n n n n n1 1 1 1 1 11 2 3 4= + = + + − −

Channel Noise

RX= <C1, C1+N> = <C1,C1> + <C1,N> = 1 + ε’

After correlating with the same code

Correlating With the Same CodeCorrelating With the Same Code

[ ]C1 1 1 1 1= + + − −, , ,

[ ]C1 1 1 1 1= + + − −, , ,

33

Channel Noise

RX= <C2, C1+N> = <C2,C1> + <C2,N> = 0 + ε’

After correlating with a different code

Correlating With a Different CodeCorrelating With a Different Code

[ ]C1 1 1 1 1= + + − −, , ,

[ ]C2 1 1 1 1= + − + −, , ,

[ ]~ , , ,C C n n n n n1 1 1 1 1 11 2 3 4= + = + + − −

34

A More Realistic Downlink ScenarioA More Realistic Downlink Scenario

T a C T a C

T a C

T T T T

TX T D

N N N

N

A

1 1 1 2 2 2

1 2

== ====

== ++ ++ ++== ⊗⊗

. , .

. . . .

. . .

R X L TX L T D

R L R X T D

T T D D T

r T C a C a C a C C

r a C C a C a C C

r a a

A

A

A A

N N

N N

N

== ×× == ×× ⊗⊗== ×× ≈≈ ⊗⊗

=<=< ⊗⊗ >=>==<=< >=<>=< ++ ++ ++ >>=<=< >> ++ << ++ ++ >>== ++ ++ ++ == ++

( )

( / *)

( ),

, . . . . . . ,

. , . . . . . ,

( . . . )

1

1 1 1 1 2 2 1

1 1 1 1 2 2 1

1 1 21 1 1

)

εε εε εε

TX:At the BaseStation A

RX:At the MobileStation 1

Spreading with BS’s Code

Same CellChannelization Codes

2nd level despreading

Match Filtering

Despreading withBS’s Code

35

Course OutlineCourse Outline

l Introductionl Part 1: CDMA Concepts

» Spreading/Despreading in Time and Frequency Domains» Concept of Multiple Access Using Codes» Spreading Codes (Walsh and Pseudo-Noise Codes)» Rake Receivers and Soft Handoff» CDMA Cell and System Capacity

l Part 2: Applications: IS95 and 3G-CDMA» Physical and Logical Channel Structure in IS95» Forward and Reverse Link Waveforms» Call Processing and Power Control Issues

36

Spreading CodesSpreading Codes

l To maintain all signal power, after spreading anddespreading, the spreading sequences» Have to be Mutually Orthogonal to each other or» Have noise like characteristics with very small cross correlation.

l Example of such codes are Walsh codes & Pseudo-Noise(PN) codes.

l Walsh Codes are perfectly orthogonal to each other. Theycan be obtained from different rows of Haddamardmatrices.

l The PN codes» have Noise-like characteristics, e.g. Sharp Autocorrelation» Are easily implementable using shift registers» Are Periodic, Long and» Difficult to reconstruct from a short segment

H

H

HH H

H H

HH H

H Hn

n n

n n

1

2

42 2

2 2

2

2 1 2 1

2 1 2 1

0

0 0

0 1

0 0

0 1

0 0

0 10 0

0 1

1 1

1 0

=

=

=

=

=

− −

− −

[ ]

. . . . .

Hadamard Matrices

Rows areOrthogonalto each other.<Ci,Cj>=0

Rows areOrthogonalto each other.<Ci,Cj>=0<Ci,Cj>=0

C0C1C2C3

l In IS95 Walsh Codescorresponding to rows of H64are used

l There are only N orthogonalsequences of length N.

Orthogonal Sequences: Walsh CodesOrthogonal Sequences: Walsh Codes

38

Walsh Codes are Mutually Orthogonal:Walsh Codes are Mutually Orthogonal:

HH H

H H

C C

84 4

4 4

0 0

0 1

0 0

0 10 0

0 1

1 1

1 0

0 0

0 1

0 0

0 10 0

0 1

1 1

1 00 0

0 1

0 0

0 10 0

0 1

1 1

1 0

1 1

1 0

1 1

1 01 1

1 0

0 0

0 1

2 518

1 1 1 1 1 1 1 1

1 1 1 1 1 1 1 1

18

1 1 1 1

==

==

<< >=>= ×× << −− −− ++ ++ −− −− ++ ++

−− ++ −− ++ ++ −− ++ −− >>

== ×× ++ −− −− ++

, ( , , , , , , , ),

( , , , , , , , )

( −− ++ ++ −− ==1 1 1 1 0)

C2

C5

Convert to Bipolar

Pseudo-Noise CodesPseudo-Noise Codes

l The PN codes are pseudo-random sequences» They are deterministic codes which mimic randomness properties

l Randomness or Noise-like characteristics include» Sharp Autocorrelation» “1”s and “0”s appear randomly and independently in a sequence» The number of “1”s and “0” are (almost) the same in any long

segment of the sequence.» Difficult to reconstruct from a short segment

l Additional desirable properties for PN codes are» Easy to Implement» Periodic & Long

l Examples are :» m-Sequences,» Gold and Kasami codes

40

Pseudo-Noise CodesPseudo-Noise Codesl PN codes have very sharp autocorrelation,l It implies that the time-shifted code versions of the same

PN sequence have very small correlation with each other.l For a long periodic PN sequence of length N this

correlation is very small, i.e close to 1/N, so that differenttime-shifted version or offsets of the same pseudo-randomsequence are almost orthogonal to each other.

-1/N

0

RS (t,t+δδ)

1 PN Chip

Period of the Code

Generating PN SequencesGenerating PN Sequences

l Maximal Length Shift Register Sequences, also called m-sequences, aregenerated by an m stage shift register with appropriate linear feedbackconnections defined by prime polynomials with modulo 2 arithmetic.

l The specific feedback configurations, used, ensure than the sequence hasits maximum period, i.e. 2m-1.

l By loading different initial value into the shift register, one cangenerate different offsets of the same sequence.

+

1 2 3 4 . . . . . m-1 mOutput

Feedback

Initial Value

Example: Spreading Codes in IS95Example: Spreading Codes in IS95

l Walsh Codes» 64 Orthogonal Codes (W0-W63)» Each of Length 64 Chips

l Short Codes» A PseudoNoise M-Sequence» Generated by a Maximal Length Shift Register» of Length 215 and period 215 -1 Chips

l Long Codes» A PseudoNoise M-Sequence» Generated by a Maximal Length Shift Register» of Length 242 and period 242 -1 Chips

43

l Orthogonal Walsh Codes:» Walsh codes of length 64 are used in IS95 forward link» There are 64 Walsh codes used to isolate forward link

channels within one cell.» Examples:

l W0: 0000.......................000l W32: 0000....0001111....111

64bits

32bits 32bits

Spreading Sequences in IS95Spreading Sequences in IS95

….010...1110.......10010010011

(215-1) Chips

BS1

BS2 ….10.......10010010011 010...11PN Offset

(i x 64chips)

Same SequenceDifferent Offsets

PN Offset of Short CodesPN Offset of Short Codes

l The Short Code is an “m-sequence” of period 215-1 chips!!l Different BS’s use different offsets of the “short code”.l Each station (or sector) uses only one PN offset.l There are 512 possible offsets, of 64 chips apart, to be

assigned to base stations.

45

Usage of CodesUsage of Codes

Station A

Short Codes: Sa and SbLong Codes:L1 ,L2 and L3Walsh Codes: W1-W63

Sa.W23Sa.W12

Station B Sb.W23

L1 L2

L3

l Short codes are used for spreading as BS’s ID in the forward link.l Long Codes are used for scrambling and spreading as MS’s ID in the

reverse linkl Walsh codes are used for forward link channelization.

Need for SynchronizationNeed for Synchronization

l Spread Spectrum Signals are typically high ratesignals with very sharp autocorrelations

l Correlation based receivers rely heavily on almostperfect synchronization.

l Therefore, maintaining the synchronization has adirect effect on identifying the desired fromundesired signals.

Synch. Out of Synch.

TX

RXRX RXRX

TX

TX

RXRX

l Coarse Synchronization is performed for Code Acquisition.

l Fine Synchronization is performed during code tracking.

Time

Time

Course and Fine SynchronizationCourse and Fine Synchronization

Chip Time

δδδδ

Code Acquisition MethodsCode Acquisition Methods

l Code Acquisition Circuits can be implemented using» a parallel bank of correlators» or a sliding correlator with a feedback

u1

u2

Correlator with p (t-(2Nc-1) Tc) v 2Nc-1

InputSS Signal

Correlator with p (t-2Tc)

Correlator with p (t-nTc)

...

PN Generator

+

-

Adjust n

Correlator with p (t-Tc)

Sliding CorrelatorParallel Bank of Correlator

x

x

x bandpassfilter

envelopedetector

envelopedetector

bandpassfilter

loopfilter

clockVCO

PN generator

-

+

P(t+τ) to Data Demodulator

P(t+Tc/2+τ)

P(t-Tc/2+τ)

Delay Locked Loop

Code Tracking MethodsCode Tracking Methods

l After Code Phase Acquisitionwe need to adapt to timevariations and maintainlocking condition.

l Code tracking circuits operateusing some sort of a feedbackloop. For example» Delay Locked Loop» Tau Dither Loop

50

Course OutlineCourse Outline

l Introductionl Part 1: CDMA Concepts

» Spreading/Despreading in Time and Frequency Domains» Concept of Multiple Access Using Codes» Spreading Codes (Walsh and Pseudo-Noise Codes)» Rake Receivers and Soft Handoff» CDMA Cell and System Capacity

l Part 2: Applications: IS95 and 3G-CDMA» Physical and Logical Channel Structure in IS95» Forward and Reverse Link Waveforms» Call Processing and Power Control Issues» Link Budget for CDMA Systems» CDMA and 3rd Generation Wireless PCS Systems

51

Multipath Effects on NB signalsMultipath Effects on NB signals

l Because of multipath effects, for each transmitted symbol, the receiver,receives a combination of the main symbol and and its echoes.

l In a Narrowband System (e.g. most TDMA based systems) the symbolis relatively wide in time so the main symbol and its echoes overlap intime. This overlap, called InterSymbol Interference (ISI), is not desiredand causes erroneous detection.

l Therefore most NB systems use adaptive equalizers to cancel ISI.l An equalizer in a NB system tries to estimate multipath components

and cancel them.

Transmitted Symbols

Received Symbols

52

Multipath Effects on WB signalsMultipath Effects on WB signals

l Because of multipath effects, for each transmitted symbol, the receiver,receives a combination of the main symbol and and its echoes.

l In a wideband system (e.g. most CDMA based systems) the symbols arerelatively narrow in time so a symbol and its echoes do not overlap intime and therefore they are resolvable.

l Most WB systems use a Rake Receiver to estimate and combine thesignal coming from multipath.

Transmitted Symbols

Received SymbolsMultipaths are Resolvable

Rake Receiver and MultipathRake Receiver and Multipath

T T+ t1 T+ t2

Delay Line &Correlators

Adaptive Combiner

ττ22

r t dtT

( )(. )∫

ττ11

r t dtT

( )(. )∫

ττ33

r t dtT

( )(. )∫

Input Data

MultiPath Components

BS1

Rake Receiver utilizes the spatial diversity.

Station A

Station B

Station A & B

Soft Hand-offSoft Hand-offl The mobile station continuously

scans for pilot signals transmittedby different stations/sectors andestablishes, both uplink and downlink, communication with up to 3stations whose pilot power exceedsa certain threshold.

l This results in a make before breakprocedure for Hand-off, where duringthe transition from one cell toanother the call is served by multiplecells.

l These simultaneous links to multiplebase stations is a form of spatialdiversity which provides a morerobust and smooth Hand-off andimproves capacity and coverageperformance of the system.

Soft HandoffSoft Handoff

Time Margin

Station AStation A

Station BStation B

TADD

TDROP

Signal Margin

Stations A & BStations A & B

Soft Handoff Region

Time/Space

Ec/

I o

Rake Rec. in Soft Handoff (Downlink)Rake Rec. in Soft Handoff (Downlink)

T T+ t1 T+ t2

Delay Line &Correlators

Adaptive Combiner

ττ22

r t dtT

( )(. )∫

ττ11

r t dtT

( )(. )∫

ττ33

r t dtT

( )(. )∫

Input Data

BS1 BS2

SelectionCombining

MSC

Soft Handoff and Spatial DiversitySoft Handoff and Spatial Diversityl There is a diversity gain associated with soft handoff in

both reverse and forward link.l The major gain is in the reverse link due to combiners at

each base station and the selective combining at the MSC.l Rake receivers in both forward and reverse link

contribute to this spatial diversity gain.

RAKE Rec. 1RAKE Rec. 1

RAKE Rec. 2RAKE Rec. 2

DiversityCombiner

RAKE Rec. 1RAKE Rec. 1

RAKE Rec. 2RAKE Rec. 2

DiversityCombiner

Selection DiversityCombining

Vocoder

2 Finger

2 Finger

2 Finger

2 Finger

BS1

BS2

MTSO

To PSTN

Selection Diversity in SHO (Uplink)Selection Diversity in SHO (Uplink)

59

SHO, Power Control and InterferenceSHO, Power Control and Interference

l SHO reduces the average transmit power of mobiles in thehandoff area» A mobile in soft handoff powers up only if all BS’s involved in soft

handoff ask for more power and» it powers down as soon as one of BS’s ask him to power down.

l Therefore statistically mobile’s transmitted power isreduced and so it contributes less to interference level in thesystem.

BS1 BS2

UP only if UP1 and UP2Down if Down1 or Down2

MSCMSCBS1

BS2

BS3

Adv. and Disadv. of Soft Hand-offAdv. and Disadv. of Soft Hand-off

l + Improvements in RF interface» Reduction in Interference» Improvement in Coverage» Increase in Capacity

l + Improvement in Voice Qualityl - Additional overhead to Allocate

» Channel Element» PowerFor users in soft Hand-off.

l In LBA to account for the Soft Hand-offDiversity Gain, for %30-%50 of users in softHand-off region, 2-3 dB is considered. Effectively,this gain is due to a reduction in the fade marginfor the combined signal.

61

Course OutlineCourse Outline

l Introductionl Part 1: CDMA Concepts

» Spreading/Despreading in Time and Frequency Domains» Concept of Multiple Access Using Codes» Spreading Codes (Walsh and Pseudo-Noise Codes)» Rake Receivers and Soft Handoff» CDMA Cell and System Capacity

l Part 2: Applications: IS95 and 3G-CDMA» Physical and Logical Channel Structure in IS95» Forward and Reverse Link Waveforms» Call Processing and Power Control Issues» Link Budget for CDMA Systems» CDMA and 3rd Generation Wireless PCS Systems

Reverse Link InterferenceReverse Link Interference

Same Cell Interference: Isc Other Cell Interference: Ioc

Forward Link InterferenceForward Link Interference

Same Cell Interference: Isc Other Cell Interference: Ioc

Energy per BitEnergy per Bit

S = Received Signal PowerR= Bit Rate=#Bits/SecondEb= Energy per Bit = Signal Power/ (# Bits per Second)

EEbb= S/R= S/R

Assuming Perfect Uplink Power ControlAssuming Perfect Uplink Power Control

P1

P3

P2

R2R1

R3

P1 > P2 > P3P1 > P2 > P3 R1=R2=R3=SR1=R2=R3=SSuch Such ThatThat

Assuming Perfect Power Control

Interference Power Spectral DensityInterference Power Spectral Density

Interference Spectrum

Thermal/BackgroundNoise

Other Users’Interference

NNoo

IISCSC

IItt

No= Noise Spectral DensityIsc = Same Cell Interference Spectral DensityIt =Total Interference Spectral Density

IIscsc = Total Interference/BandwidthNN00= Noise Power/BandWidthTotal Interference Power =IItt.W=(N.W=(N00+I+ISCSC).W).W

Freq.

Total Interference (Single Cell)Total Interference (Single Cell)

PN-1

SSS

S S

(N-1)SP2

PN

P1

Total Interference= Other (N-1) users Interference + Thermal Noise

It W= IscW+NoW = (N-1)S +NoW

IIt t = (N-1)S/W + N= (N-1)S/W + N00

Assuming Perfect Power Control

EEbb/I/Itt

EI

S RN S W N

S RN S W

W RN

b

t

=− +

≈−

=−

/

[( ) ] /

/

[( ) ] /

/

1 1 10

Ignoring thebackground noiseE

IS R

I WE I E N

W RN

E N

NW R

E N

Capacity NW R

E NW R

E N

b

t totalb t b

b

b

b b

= > =

−>

< +

= = + ≈

/

/( / ) ( / )

/( / )

/

( / )

/( / )

/( / )

min min

min

min

maxmin min

0

0

0

0 0

1

1

1

Common Terminology

Coding and Processing GainCoding and Processing Gain

BER

(Eb/Io)

Without Coding

With Coding

CodingGain

(Eb/Io)min

Coding

Capacity

☺.

l Using more powerful channelcoding and modulation schemesincrease the tolerance againstnoise and interference.

l This means for a given Bit ErrorRate (BER) coding reduces therequired (Eb/N0).

l Also note the direct and expliciteffect of Processing Gain on thecapacity.Capacity

☺.

W/R

Speech ActivitySpeech Activity

60%Inactive(Silence)

40%Active

(Speech)

Human’ Speech signal has a duty cycle of about νν=40%

Effect of Speech Activity FactorEffect of Speech Activity Factor

ν νii

N

N

S N S N Seffective

= × − ≈ −=

∑1

1

1 0 4 1( ) . ( )1 24 34

ννi=1 with probability 0.4

ννi=1 with probability 0.6

ννi=1 with probability 0.4

ννi=1 with probability 0.6

N-1 mobiles,Only 0.4(N-1) active

RX=Neff. xS

Effect of Voice Activity on CapacityEffect of Voice Activity on Capacity

N NW R

E N

CapacityW R

E N

effb

b

. maxmin

min

/

( / )

/

( / ).

= ≈

=

ν

ν

0

0

1

Voice Activity Gain

> 1

l The effective capacity increases because of voice activity.l The increase in capacity is achieved without additional

overhead signaling and protocol considerations.

Effect of SectorizationEffect of Sectorization

l For 120o sectored cites, comparedto an Omni-Cite» Almost 1/3rd Interference received in

the uplink» Causes almost 1/3rd Interference in

the uplink

l Reduction in interference resultsin higher capacity in both links.

I I

G

CapacityW R

E NG

Sector Omni

S

bS

= ≈

=

1 3

2 55 3

0

/

.

/( / )

.min

SectorizationGain

Sectorization GainSectorization Gain

l Interference reduction due to directional antenna patternsresults in increase in capacity.

l For three sectors the sectorization gain is close to 3.l There is no loss of trunking efficiency because all sectors of

a cite use a common pool of channels.

Reuse Efficiency FactorReuse Efficiency Factor

Ioc: Other Cells InterferenceIsc: Same Cell Interference = f= f

Total Interference= Isc + Ioc = (1+f ) Isc

EI

S RI I W N

S Rf I W N

EI

S Rf N S W N

W Rf N

b

t sc oc sc

b

t

=+ +

=+ +

=+ − +

≈+ −

/

( ) /

/

( ) /

/

( )[( ) ] /

/

( )( )

0 0

0

1

1 1 1 1ν νReuse Efficiency >1Effect of Other Cells

Ignoring thebackground noise

Ignoring thebackground noise

Effect of Effect of f f on Capacityon Capacity

EI

W Rf N

W Rf N

CapacityW R

E N f

b

t

b

≈≈++ −−

≈≈++

==++

/( )( )

/( )

/( / )

.( )min

1 1 1

110

νν νν

νν

Reduction in Capacitydue to other cellInterference

77

Loading FactorLoading Factor

l The simple capacity equation which ignores the effect ofnoise, called the pole capacity, is theoretical limit toCDMA cell capacity.

l To achieve this limit mobile has to transmit at infinitepower and the system becomes unstable.

l For stable operation of the system a loading factor of %50to %80 is usually considered.

CapacityW R

E N fG L

bS=

+/

( / ).( )

. . .min0

11

1

ν

Pole Capacity

Loading Factor

Big Picture (Reverse Link Cell capacity)Big Picture (Reverse Link Cell capacity)

CapacityW R

E N fG L

bS=

+/

( / ).( )

. . .min0

11

1

ν

Processing Gain

Speech AcivityGain

SectorizationGain

Minimum TechnologyRequirement Frequency

Reuse Efficiency

LoadingFactor

79

ExerciseExercise

l Using this capacity equation and the following practicalassumptions compute the cell capacity for CDMA system.

l Eb/No=7dB=5l Reuse Efficiency Factor f=0.55l Bandwidth W=1.23MHzl Data Rate R=9.6kbpsl Voice activity v=0.4l Sectorization Gain Gs= 2.65l Loading L=60%

.LPC. GSHO

>1<1

CapacityW R

E N fG L

bS=

+/

( / ).( )

. . .min0

11

1

ν

Some Implicit Effects on CapacitySome Implicit Effects on Capacity

l Soft handoff improves Capacity» Because of spatial diversity gains, users in soft handoff region

demand less power from BS and transmit less power therefore theycontribute less to the interference. Reduced Interference meansimprovement in capacity.

l Power Control errors reduces the capacity .» In all of calculations we assumed perfect power control. The effect

of errors in power control is modeled as a factor which is adecreasing function of the error variance.

Soft CapacitySoft Capacity

l In a CDMA system the capacity islimited by a threshold on acontinuous variable i.e. signalquality.

l It is always possible to allow oneextra user by sacrificing some ofthis quality, for all users.

l In FDMA/TDMA systems we havehard capacity, because capacity ishard limited by the number of RFcarriers and number of time slots.

# Active Users

Qu

alit

y In

dex

(E

b/N

o)

Desired Quality

NominalCapacity

More Users &less Quality

82

How About Forward Link Capacity?How About Forward Link Capacity?

l While the reverse link capacity is limited by aggregateinterference effects the forward link is power limited.

l The forward link capacity is defined as the maximumnumber of users, for whom the base station can provide» distinct code channels» enough power

l The code channel limitation practically never dominates.l The power limitation depends on user locations

» the worse case is when all users are far at the cell periphery, inwhich the base can support only few of them.

» The average case, where users are uniformly distributed, in whichcase the forward link capacity estimated using simulation and it isusually higher than reverse link.

l Therefore the CDMA cell capacity is usually determined bythe reverse link.

83

Course OutlineCourse Outline

l Introductionl Part 1: CDMA Concepts

» Spreading/Despreading in Time and Frequency Domains» Concept of Multiple Access Using Codes» Spreading Codes (Walsh and Pseudo-Noise Codes)» Rake Receivers and Soft Handoff» CDMA Cell and System Capacity

l Part 2: Applications: IS95 and 3G-CDMA» Physical and Logical Channel Structure in IS95» Forward and Reverse Link Waveforms» Call Processing and Power Control Issues

1.23MHz

41 x 30KHz AMPS Channels.

. . .

3dB

Downlink or Forward Channel

Uplink orReverse Channel

fN+ 45MHz

Physical ChannelsPhysical Channels

l Each CDMA channel occupies 1.23MHz of spectrum whichis equivalent of 41 AMPS channels.

l All CDMA channels within the Cellular band are organizedbased on AMPS channels.

[825+0.03N] MHz [870+0.03N] MHz for N=1,2,..,799[825+0.03(N-1023)] MHz [870+0.03(N-1023)] MHz for N=990,991,...,1023

Uplink

Downlink

A Band B’A’’ B Band A’

991

1023

1 333

334

666

667

716

717

799

Channel Numbering for Cellular BandChannel Numbering for Cellular Band

BlockDesignator

BandwidthAllocated

(MHz) Uplink Downlink

A (MTA) 30 (15/15) 1850-1865 1930-1945

D (BTA) 10 (5/5) 1865-1870 1945-1950

B (MTA) 30 (15/15) 1870-1885 1950-1965

E (BTA) 10 (5/5) 1885-1890 1965-1970

F (BTA) 10 (5/5) 1890-1895 1970-1975

C (BTA) 30 (15/15) 1895-1910 1975-1990

PCS BlocksPCS Blocks

D CA B E F

[1850+0.05N] MHz N=0,1,..,1199

Uplink Downlink

0 1199 0 1199

[1930+0.05N] MHz N=0,1,..,1199

D CA B E F20MHz

80MHz

1850MHz

CDMA PCS BLOCKSCDMA PCS BLOCKS

l PCS blocks (A, B and C) are 15MHz wide pairs, whereas blocks (D, Eand F) are 5MHz wide pairs.

l PCS spectrum allows up to 1200 center frequencies (and thereforeCDMA channel numbers) of 50KHz separation.

BlockDesignator

Preferred Set of CDMA Channel Numbers

A 25, 50, 75, 100, 125, 150, 175, 200. 225, 250, 275

D 325, 350, 375

B 425, 450, 475, 500, 525, 550, 575, 600, 625, 650, 675

E 725, 750, 775

F 825, 850, 875

C 925, 950, 975, 1000, 1025, 1050, 1075, 1100, 1125, 1175

Preferred Channels for PCS Preferred Channels for PCS

l CDMA carriers are arranged in the middle of preselected channels toallow for sufficient guard bands, the set of these channels is called thePreferred set.

l To access the CDMA system in block A, the mobile station scans thePreferred set of black A until it finds a pilot channel. If no service isfound it may search for Preferred set of another block say block B.

System Information & Paging Information

Access Request

Voice Information & Signaling

Voice Information & Signaling

Control & Voice ChannelsControl & Voice Channels

l Pilot Channell Synch Channell Paging Channell Traffic Channel

» User Traffic Data» Blank & Burst

Signaling» Dim & Burst Signaling» Power Control

l Access Channel

Signaling &Control

IS95 Logical ChannelsIS95 Logical Channels

PilotSynchPaging

TrafficChannels

Forward Link ChannelsForward Link Channels

Pilot ChannelPilot Channel

l Pilot Channel is an unmodulated DSS signal continuouslytransmitted by each CDMA base station used to uniquelyidentify the base station.

l It transmits Walsh-0 (W0) sequence.l It serves as a phase reference for timing, bit

synchronization and coherent demodulation in the downlink

l Also since Pilot is not subject to dynamic power control itprovides a reference for comparing the signal strength ofdifferent base stations.

l Therefore pilot channel plays the major role in determiningbest server and servers in soft hand-off.

Synch ChannelSynch Channel

l Synch channel is demodulated by the mobile right tuning tostrongest pilot.

l It carries some of system ID parameters» System Identification number» Network Identification number

l And some information about timing» Pilot sequence offset index PILOT_PN» Long Code State» System Time» Offset of Local Time» Leap Seconds

l Paging Channel Data Rate.l The data rate of the Synch channel is 1.2kbps.

Paging ChannelPaging Channel

l Paging channel is continuously monitored by the mobileafter reading the information on the Synch channel.

l Some of typical Paging Channel messages are:» System parameters» Access parameters» Page or Slotted Page» Order Messages» SSD update» Data Burst» Authentication» CDMA Channel List» Channel assignment

Forward Traffic ChannelForward Traffic Channel

l Traffic channel carry variable rate voice/data.l In addition to user data/voice, the traffic channels is IS95

also carry some signaling information.» These signaling subchannels are associated to and are

time multiplexed with users data on the traffic channel.» In the Forward Traffic channel the following messages

are sent to the mobile– Order messages– Data Burst– Hand off Direction– In-Traffic System parameters– SSD Update– Power Control Parameters– Neighbor-list Update– MS Registered Message

Pag

ing

Cha

n. 1

Tra

ffic

Cha

n. 1

Pag

ing

Cha

n. 2

Pag

ing

Cha

n. 3

Pag

ing

Cha

n. 4

Pag

ing

Cha

n. 5

Pag

ing

Cha

n. 7

Pag

ing

Cha

n. 6

Tra

ffic

Cha

n. 2

Tra

ffic

Cha

n. 3

Tra

ffic

Cha

n. 2

4

Tra

ffic

Cha

n. 5

4

Tra

ffic

Cha

n. 5

5

Syn

c C

han.

Pilo

t C

han.

W0 W32 W1 W2 W3 W4 W5 W6 W7 W8 W9 W10 .......... W31W33 .................. W62 W63

Tra

ffic

Cha

n. 2

5

Tra

ffic

Cha

n. N

Traffic DataMobile PowerControl Sub-channel

1.23 MHz BandwidthTransmitted by Base Station

1 Pilot1 Sync7 Paging55 Traffic

Total of 64 Walsh CodesTotal of 64 Walsh Codes

CDMA Forward ChannelCDMA Forward Channel

AccessChannel

Mobile’s:OriginationResponds to OrdersPeriodic Reports

Random Access ChannelRandom Access Channel

l Access Channel Signaling consists of messages related to» Access or Call Origination» Respond to a Page» Authentication» Registration» User Generated Data Bursts for the base station.» Other order messages

l It operates based on a variation of Slotted ALOHAProtocol

l Some of the signaling information transmittedover uplink traffic channel are» Authentication Challenge Response» Power Measurement Report» Pilot Strength Measurement» Hand-off Completion» Dual Tone Multi-Frequency (DTMF) Signaling» Order Messages

– Long Code Transition Request and Response, SSD UpdateConfirmation/Rejection, Parameter Update Confirmation,Service Option Control, Base Station Challenge, Mobile StationAcknowledgment, Release (normal and with power-downindication), Local Control, Mobile Station Reject (with andwithout a reason).

Uplink Traffic Channel SignalingUplink Traffic Channel Signaling

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

Acc

ess

Cha

n. 1

Tra

ffic

Cha

n. 1

Acc

ess

Cha

n. 2

Acc

ess

Cha

n. n

Tra

ffic

Cha

n. 2

Tra

ffic

Cha

n. m

Addressed by Long PN CodesAddressed by Long PN Codes

1.23 MHz BandwidthReceived by Base Station

CDMA Reverse ChannelCDMA Reverse Channel

100

Course OutlineCourse Outline

l Introductionl Part 1: CDMA Concepts

» Spreading/Despreading in Time and Frequency Domains» Concept of Multiple Access Using Codes» Spreading Codes (Walsh and Pseudo-Noise Codes)» Rake Receivers and Soft Handoff» CDMA Cell and System Capacity

l Part 2: Applications: IS95 and 3G-CDMA» Physical and Logical Channel Structure in IS95» Forward and Reverse Link Waveforms» Call Processing and Power Control Issues

Variable RateSource EncodingVariable Rate

Source Encoding

Channel CodingChannel Coding

Bit InterleavingBit Interleaving

Long Code ScramblingLong Code Scrambling

Walsh and QuadratureSpreading

Walsh and QuadratureSpreading

Quadrature CarrierModulation

Quadrature CarrierModulation

Variable RateSource DecodingVariable Rate

Source Decoding

Channel DecodingChannel Decoding

Bit De-interleavingBit De-interleaving

Long Code De-scramblingLong Code De-scrambling

Walsh and QuadratureDe-spreading

Walsh and QuadratureDe-spreading

Quadrature CarrierDemodulation

Quadrature CarrierDemodulation

TX RX

WirelessWireless Channel Channel

BasebandProcessing

DownLink ProcessesDownLink Processes

OPERATION CHANNEL RATE

kbps

8 9.6

4 4.8

2 2.4

1 1.2

kbps

Variable Rate

Vocoder

Variable Rate VocoderVariable Rate Vocoder

l There are two rate sets corresponding to» 8 kbps Speech coders (Rate Set I)» 13 kbps Speech coders (Rate Set II)

l At each rate set:» There are 4 possible rates used based on the speech activity.» At lower rates, lower average power is transmitted

LPC filter Coef.

Pitch Parameters (Gain and Lag)

Excitation Parameters (Index and Gain)

MUX

Speech Analysis

Channel Coder

LPC Filter

Excitation

Pitch

Imitation of Vocal Cords

Imitation of Vocal Tract

Speech Waveform

Speech Generation Model

High RateSampled Speech

Low Rate OutputModel parameters

Code Excited Linear Predictive CoderCode Excited Linear Predictive Coder

LPC Filter Coef.Pitch

Codebook Index

LPC Filter Coef.Pitch

Codebook Index

LPC Filter Coef.Pitch

Codebook Index

LPC Filter Coef.Pitch

Codebook Index

40 bits10bits 10 bits 10 bits 10 bits

10b 10b 10b 10b 10b 10b 10b 10b

20 bits

10 bits

10 bits

10bits

10bits

10bits

0 bits

10bits 10bits

10bits 10bits 10bits 10bits

6bits

Rate 1

Rate 1/2

Rate 1/4

Rate 1/8

20 msec frame

171 bits

80 bits

40 bits

16 bits

Encoded Packet

Subframes

Speech Coder Rates (Rate Set 1)Speech Coder Rates (Rate Set 1)

TrafficBlock

Generator

SpeechCoder

CRC

ConvolutionalEncoder++Speech

Blocks

Signaling

Traffic Blocks

ChannelEncoder

Interleaver

Traffic Frames

Mixed Mode Bits

8 Tail Bits

Channel CodingChannel Coding

1/2 Rate Conv. Encoder &

Repeater ( for input rates < 9.6 kps )

1/2 Rate 1/2 Rate Conv. Encoder &Conv. Encoder &

RepeaterRepeater ( for input rates < 9.6 kps )( for input rates < 9.6 kps )

20ms20ms 20ms20ms

Traffic FramesTraffic Frames Coded FramesCoded Frames

Variable RateVariable Rate Fixed Rate Fixed Rate 19.2 kbps19.2 kbps

Convolutional CodingConvolutional Coding

l Traffic data frames are coded using a 1/2 rate convolutional encoder.l The convolutional encoder has a constraint length of 9 and uses an 8

bit shift register that for every input bit generates 2 output bits.l For input rates smaller than 9.6kbps, output bits are repeated to

provide a fixed number of output bits per frame.l So for all input rates there are 384 bits per 20 msec frame which gives

an output rate of 19.2kbps.

Convolutional Encoder&

Repetition Code

BlockInterleaver

Convolutional Encoder&

Repetition Code

BlockInterleaver

Convolutional Encoder&

Repetition Code

BlockInterleaver

Sync ChannelData

Paging ChannelData

Traffic ChannelData

•1.2kbps

•2.4kbps•4.8kbps•9.6kbps

•1.2kbps•2.4kbps•4.8kbps•9.6kbps

4.8kbps

19.2kbps

19.2kbps

Pilot Channel No Data

Forward Link Channel CodingForward Link Channel Coding

Long Code Generator

42 bitLong CodeMask

1/64 Long Code Generator

Scrambling

InterleavedBits 19.2kbpsPaging or Traffic

Encrypted Data 19.2 kbps

Walsh Code Spreading

Long Code ScramblingLong Code Scrambling

l Based on a user specific mask the long code generator, which uses 42 bitshift register, generates a PN sequence of length 242-1 chips of rate1.2288Mbps.

l 1/64 long code generator selects the first chip of every 64 long code chips,and holds it for the duration of 64 chips. This provides a 19.2 kbpssequence.

l The 1/64 long code, used as a MS key, is combined with the input bitstream through XOR operation to encrypts the data.

l The same 1/64 long code is generated at the receiver to undo thescrambling process.

1/64 Long Code Generator

DownlinkPower Control

Algorithm

Power Control Bit position

Power Control Bit

Bit Puncturer

Scrambled TrafficFrames

Walsh SequenceWi

19.2kbps 1.2288Mbps

Power Control Bit PuncturingPower Control Bit Puncturing

l Bit Puncturer replaces 2 consecutive input bits by one power control bitevery 1.25msec.

l The value of this bit is determined by power control algorithms and BSmeasurements from mobile.

l And the position of this bit is determined by long code.

1dB1dB

1.25msecPower Control Group Period

0 0 0 1 0 1 0 1 1 1 1 1 0 0 0 1 0 1 0 1 1 1 1 1

Power Control Bit00

11

timetime

MS PowerMS Power

Change inPower level

+1dB+1dB

-1dB-1dB

Power Control BitsPower Control Bits

Coded & Interleaved

Coded Interleaved

and Scrambled

CodedInterleaved

and Scrambled

W0

+W32

+Wp

+Wi

+

1.22MHz1.22MHz

1.22MHz1.22MHz

1.22MHz1.22MHz

Pilot Channel0kbps

Sync ChannelData

4.8kbps

Paging ChannelData

19.2kbps

Traffic ChannelData

19.2kbps

1.22MHz1.22MHz

Walsh SpreadingWalsh Spreading

Pilot ChannelNo Data, “All 0’s”

Sync ChannelData

4.8kbps

Paging ChannelData

19.2kbps

Traffic ChannelData

19.2kbps

Long CodeGenerator

+

Long CodeGenerator

+

W0

+W32

+Wp

+Wi

+MUX

Power Control Bit

++

++

++

++

Q-Channel Pilot PN Seq.

I-Channel Pilot PN Seq.

Paging Channel Long Code Mask

User i Long Code Mask

1.22MHz1.22MHz

1.22MHz1.22MHz

1.22MHz1.22MHz

Symbol Scrambling

Forward Link WaveformForward Link Waveform

Variable RateSource Encoding

Variable RateSource Encoding

Channel CodingChannel Coding

Bit InterleavingBit Interleaving

Long Code ScramblingLong Code Scrambling

QuadratureSpreading

QuadratureSpreading

Quadrature CarrierModulation

Quadrature CarrierModulation

Variable RateSource Decoding

Variable RateSource Decoding

Channel DecodingChannel Decoding

Bit De-interleavingBit De-interleaving

Long Code De-scramblingLong Code De-scrambling

QuadratureDe-spreadingQuadrature

De-spreading

Quadrature CarrierDemodulation

Quadrature CarrierDemodulation

TX RX

WirelessWireless Channel Channel

BasebandProcessing

UpLink ProcessesUpLink Processes

SpeechCoder

ChannelEncoder

Interleaver

TrafficBlock

Generator

CRC

++Speech Blocks

Signaling

Traffic Blocks

Traffic Frames

Mixed Mode Bits

8 Tail Bits

1/3 Rate Conv. Encoder &

Repeater ( for input rates < 9.6 kps )

1/3 Rate 1/3 Rate Conv. Encoder &Conv. Encoder &

RepeaterRepeater ( for input rates < 9.6 kps )( for input rates < 9.6 kps )

20ms20ms

Coded FramesCoded Frames

Fixed Rate Fixed Rate 28.8kbps28.8kbps

Uses a Different Interleaving Matrix than uplink

Same as downlink

Different Conv. Encoder

Uplink Channel CodingUplink Channel Coding

64ary Walsh Modulator

Coded & Coded & Interleaved bitsInterleaved bits

28.8kbps28.8kbpsTo EncryptorTo Encryptor307.2 kbps307.2 kbps

6 input bits

Each selects one of 64 Walsh Waveformsof 64 bit length.

26=64 combinations

Walsh ModulationWalsh Modulation

l Every block of 6 bits is mapped to one of 64 Walsh sequences of length 64.l This orthogonal modulation improves the error performance of the system.l Note that although Walsh Modulation increases the data rate, it and not

the same as Walsh spreading.

Long CodeLong CodeGeneratorGenerator

307.2kbps307.2kbps 1.2288Mcps1.2288Mcps

Voice PrivacyVoice PrivacyMask GeneratorMask Generator

Long Code Mask42 bits

Walsh ModulatedBits

To QuadratureShort Code

Spreading andModulation

Long Code Spreading/EncryptionLong Code Spreading/Encryption

l The data and signaling on the traffic channel are encrypted/spread witha long code based on user specific long code masks.

l The data on the access channel is not encrypted because the accesschannel long code mask is not private.

Reverse Link WaveformReverse Link Waveform

l No pilot Noncoherent receiversl 1/3 rate convolutional encoder.l 64ary orthogonal Walsh code modulation (not spreading in

down link)l Uplink channels are identified by long PN codes.l Interleaving matrix is different than down link.l Modulation is Offset QPSK.l Message encryption using Long Code Private Mask

IS95 Physical Layer (Rate Set I)IS95 Physical Layer (Rate Set I)

l Multi-access : CDMA combined with FDMAl Bandwidth 1.23MHz per carrierl Voice Circuits up to 55 per carrierl Modulation: Down link :QPSK, Uplink: OQPSKl Speech Coding Variable Rate, QCELPl Channel Coding CRC + Conv. Code + Interleavingl Coding Rate Down link: 1/2, Uplink:1/3l Bit or Chip Rate 1.2288Mcpsl Traffic Rates 9.6, 4.8, 2.4, 1.2 kbps.l Channel Rates Down link: 19.2, Uplink: 28.8

Convulotional Coding Rates:

•Downlink: 3/4

•Uplink: 1/2

Rate Set IIRate Set II

Rate Speech CoderRate

Traffic BlockRate

Channel

Rate1 13.35 14.4 19.2

1/2 6.25 7.2 9.6

1/4 2.75 3.6 4.8

1/8 1.05 1.8 2.4

l Rate Set II is an option that allocates more bits to voice coder and lessto convolutional coder.

l In this set speech and traffic rates are higher but convolutional codingrates are also higher, which adds up to unchanged channel rates.

l For Rate set II» Paging and Access channels are unchanged.» Frame duration, modulation and power control are unchanged.

120

Course OutlineCourse Outline

l Introductionl Part 1: CDMA Concepts

» Spreading/Despreading in Time and Frequency Domains» Concept of Multiple Access Using Codes» Spreading Codes (Walsh and Pseudo-Noise Codes)» Rake Receivers and Soft Handoff» CDMA Cell and System Capacity

l Part 2: Applications: IS95 and 3G-CDMA» Physical and Logical Channel Structure in IS95» Forward and Reverse Link Waveforms» Call Processing and Power Control Issues

Call Processing OverviewCall Processing Overview

l We will discuss the following call processingissues» Mobile Initialization and Access» Registration and Paging» Call Set up and Release (Mobile originated or

terminated)» (Soft) Handoff related Signaling» Power Control

InitializationInitialization

Power UpPower Up

Idle ModeIdle Mode

System AccessSystem Access

Conversation ModeConversation Mode(Traffic/Signaling)(Traffic/Signaling)

?

Mobile Station StatesMobile Station States

Pilot ChannelAcquisition

Synch ChannelAcquisition

Synchronization

System DeterminationAnalog or CDMA

Analog Initialization

To Idle StateTo Idle State

Identify and lock on to the strongest CDMAbase station.

Decodes Synch Channel Messages

Synchronizes it internal timing with the BS.

Analog

CDMA

?

Initialization StateInitialization State

Acquire Primary Paging Channel

Call Origination

Page Response

Registration

Authentication

Idle Hand-off

Update OverheadInformation

Page or Call Origination

No

Access StateAccess State

Idle StateIdle State

System AccessSystem Access

l System Access Mode consists of active or reactive messagetransfers in uplink through access channel.

l These message transactions take place before conversationstage.

l During access mobile BS tells MS about its assigned trafficchannel.

l It consists of the following substates:» MS origination attempt» MS Page Response» Order or message response» Registration Access» Update overhead/configuration information» MS Data Burst message transmission

Mobile Registration TypesMobile Registration Types

l There are various occasion during which MS send aregistration message to BS.» After power on» Before power off» Upon Registration Order» Time based and periodically» Distance based» Zone based» Implicitly during any access» implicitly during conversation on the traffic channel

Page Response

Traf. Ch. Initialization

Waiting for BS’s Order

Waiting for MS’s Answer

Conversation

Call Release

MS verifies the FTC & Transmits on the RTC MS Receives

Release Order

MS Receives Release Order

MS Receives anAlert with Information

MS Answers the Call

MS Receives Release Order or initiates Disconnect

MS Control on the Traffic ChannelMS Control on the Traffic Channel

MSC receivesOriginationfrom PSTN

Null Data (FTC)

Page Message (PC)

Page Response (AC)

Traffic Channel Assignment (PC)

Acknowledgement (FTC)

Traffic Channel Preamble (RTC)

Null Data (RTC)

Service Option Response (FTC)

Alert with Information (FTC)

Connect Order (FTC)

Conversation, Speech Frames (TC)

FTC set up

RTC set up Rec. 2 consecutive valid frames

User Answers, Ring Stops

Starts Ringing

Call flow for Mobile TerminationCall flow for Mobile Termination

Soft Hand-off

Softer Hand-off Soft Softer

Hand-off3-way Hand-off

A1 A2 A3

B1 B2 B3

C1 C2 C3

D1 D2 D3

{D2,C1}

{D1,D3}

{B1,D2,D3}

{A2,B1,D3}

Hand-Off’s in IS95Hand-Off’s in IS95

l Hard Hand-offl Soft Hand-offl Softer Hand-offl Soft Softer Hand-offl Three Way Hand-off

Simultaneous Softer & Soft Hand-off

Channel ElementChannel Element Channel ElementChannel Element

MSC

Signals from the two sectors arecombined locally at the BTS.Only the combined frame is sent toMSC.

Frame selection is performed at MSCbased on the signals received fromthe two sites.

Soft Softer Hand-offSoft Softer Hand-off

EI

PP N W

C

i

i

jj

0 0

==

++∑∑αα

ActiveSet

CandidateSet

Neighbor Set

Remaining Set

Pilot SetsPilot Sets

l MS evaluates each pilot strength based on its powerrelative to the total power received in the forward link.

l Based on their signal strength, the pilots identified by theMS are categorized in Four different sets:

Mobile Station Power ClassesMobile Station Power Classes

MobileStation

EIRP at Maximum Output

Class Minimum Maximum

I -2 dBW (630 mW) 3 dBW (2 W)

II -7 dBW (200 mW) 0 dBW (1 W)

III -12 dBW (63 mW) -3 dBW (0.5 W)

IV -17 dBW (20 mW) -6 dBW (0.25 W)

V -22 dBW (6.3 mW) -9 dBW (0.13 W)

P

S

K.S

P

Interference Spectrum

NNoo

IIoo

IIttEquivalent to K users

Near-Far ProblemNear-Far Problem

Excessive Interference

Poor Signal Quality

Just Enough PowerJust Enough Power

Power ControlPower Control

l The fundamental purpose of power control is to» maintain a satisfactory voice quality subject to» maximizing system capacity and» minimizing power consumption.

l Power Control is applied to:» Mobile Power on initial access» Mobile Power while on the traffic channel» Base Station Power

PilotSynchPaging

TrafficChannels

Within the traffic channel thepower is dynamically allocatedto different users according toto their path loss to maintainthe same voice quality or FERfor all users.

Forward Link: Power AllocationForward Link: Power Allocation

0

5

10

15

20

25

30

35

User1 User2 User3 User4

User1

User4User3

User2

Traffic Channel Power Allocation

Forward Loop Power ControlForward Loop Power Control

∆∆D

∆∆U

PMR shows High FER

N Frames

Forward Link Power Control ProcessForward Link Power Control Process

l MS measures Frame Error Rate (FER), every N frames, on the forwardlink and reports the measurement to the BS.

l The Power Message Report (PMR) contains the number of frames inerror and the total number of frames received during the report timeperiod. The ratio between these two numbers is FER.

l Whenever PMR shows high FER base station powers up by ∆∆Uotherwise it powers down by ∆∆D

Probe 1

1st Attempt 2nd Attempt 3rd Attempt 15th Attempt

RandomTime

Probe 2

Probe 16

Waitingfor ACK

Random Time Access Preamble

1-16 FramesMessage Capsule

3-10 Frames

Time

One Access Channel Slot

Power Increment

Initial Power

Reverse Link Access Power ControlReverse Link Access Power Controll During an access attempt mobile’s power has to be controlled.l Each access attempt consists of the entire process of sending one

message and receiving or failing to receive its acknowledgment.

P

Open Loop: Power control based on mobile measurement of pilot signal strength

Closed Loop: Power control based on BS commandsaccording to its uplink measurements.

Reverse Link Traffic Power ControlReverse Link Traffic Power Control

l Power control in the uplink is done both open loop andclosed loop..

l Open loop PC takes care of slow fading due to shadowingeffects.

l Closed loop PC tries to compensate for multipath fadingeffects.

EI

PP N W

C

t i

i

jj

=

+∑α

0

Open Loop Power ControlOpen Loop Power Control

l The mobile measures the pilot power level from its primarycell along with the total signal received.

l These two measurements, which are also used for Hand-offdecisions, are the basis for open loop power controldecisions.

l The mobile powers up if it receives low Ec/It and powersdown otherwise.

l To avoid too many unnecessary changes in the power dueto fast fading effects on the received signal, the open looppower control has a relatively large response time.

Until Frame Error Exceeds thethreshold.

The set-point value is reducedby a small amount for everyconsecutive frame....

∆∆U∆∆D

Ceiling(10dB)

Floor (10dB)

Time

Eb/No Target or set-point Value

Inner Loop Process

Closed (Outer) Loop Power ControlClosed (Outer) Loop Power Control

dBm Eb

20 msec Frame

1.25 msec

No

1dB

set-point Value From Outer Loop Process

From Outer Loop

Closed (Inner) Loop Power ControlClosed (Inner) Loop Power Control

l BS sends power control bits to the MS to ask it power up or down asneeded to reach the target Eb/No set point determined in the outer loopprocess.

l The power control bits are sent 16 times per 20msec frame.l Each “0” (”1”) bit changes the power level by +1 ( -1) dB.l When set-point is reached the power control bit alternates and

therefore signal level changes +/-1 dB around the set-point.

∆P= ∆POpen + ∆PClosed

Autonomous Slow Large-Scalechanges with largetime constant τ=30msec

Directed Fast changes+/-1dB per 1.25 msec.Dynamic Range 48 dB over 3 frames.

Reverse Link Power Control ( Recap)Reverse Link Power Control ( Recap)

l The faster response time of the closed loop control enablesit to overwrite the open loop commands when it isnecessary.

l The two power control mechanisms are independent andtogether can provide at least 80dB of dynamic range.

Power Control SummaryPower Control Summary

l Objective: Operate BS and MS at optimum power level to» Achieve the minimum FER (e.g. 0.01) to ensure voice quality.» Reduce interference to its minimum and thereby maximize the

operational capacity.» Maximize the battery life of the mobile.

l Process consists of» Mobile Access Power control» Dynamic Allocation of Power among Traffic channels at the BS in

Down Link.» Reverse Link Power Control on the Mobile

– The Autonomous Open Loop control at the MS based on its powermeasurement on the down link.

– The directed or closed loop control based on BS’s Eb/No set points andits power control commands on the traffic channel.

145

Course ReviewCourse Review

4 Introduction4 Part 1: CDMA Concepts

4 Spreading/Despreading in Time and Frequency Domains4 Concept of Multiple Access Using Codes4 Spreading Codes (Walsh and Pseudo-Noise Codes)4 Rake Receivers and Soft Handoff4 CDMA Cell and System Capacity

4 Part 2: Applications: IS95 and 3G-CDMA4 Physical and Logical Channel Structure in IS954 Forward and Reverse Link Waveforms4 Call Processing and Power Control Issues

146

Hope that you enjoyed this course.Hope that you enjoyed this course.

Thank You for Your Participation

147

Useful ReferencesUseful References

l “Applications of CDMA in Wireless/Personal Communications”, VijayK. Garg, Kenneth Smolik and Joseph E. Wilkes, Prentice Hall 1997.

l “CDMA”, Andrew j. Viterbi, Addison-Wesley, 1995l “Wireless Communications, Principles and Practice”, Theodore

Rappaport, Prentice Hall 1996.l “CDMA System Engineering Handbook”, Jhong S. Lee and Leonard E.

Miller, Artech House Publishers, 1998.l “Wideband CDMA for Third Generation Mobile Communications”,

Tero Ojanpera and Ramjee Prasad, Artech House Publishers, 1998.l Magazines:

» IEEE Communications Magazine (Recent Issues)» IEEE Personal Communication Magazines

top related