Principles of Spread Spectrum and CDMA Principles of Spread Spectrum and CDMA

Separation of Overlapping Signals

• Frequency Division Multiplexing

• signals non-overlapping in frequency

freq
Principles of Spread Spectrum • signals overlap in time
f2
and CDMA • separation achieved by filtering f1
- “window” in f-domain time
- convolution with filter impulse response in t-domain
Dr Bhaskar Ramamurthi
Professor • if H 1 ( f ) ↔ h1 ( t ) filters out signal in [f1 , f2]
∞

Department of Electrical Engineering desired_signal = ∫ sum _ signal ( τ )h1 ( t − τ )dτ

−∞

Indian Institute of Technology Madras. ⇒ sliding correlation with h1(-t)

Dr Bhaskar Ramamurthi CDMA_Wireless_Course 1 Dr Bhaskar Ramamurthi CDMA_Wireless_Course 2

Principles of Spread Spectrum and CDMA Principles of Spread Spectrum and CDMA

Separation (contd…) Orthogonality and Separation of Signals

• Time Division Multiplexing • in FDM, signals are orthogonal in frequency

freq

• signals non-overlapping in time ⇒ Signal_1(f) . Signal_2(f) = 0

• signals overlap in frequency • in TDM, signal are orthogonal in time

• separation achieved by windowing ⇒ signal_1(t) . signal_2(t) = 0
t1 t2 tN
− multiplication in t-domain time • signals can be continuous-time or discrete-time
− convolution in f-domain  however in TDM, signal is usual a d-t digital signal
or digitally-modulated c-t signal
• if u(t; t1 , t2) is non-zero in [t1 , t2]
desired_signal = sum_signal . u(t; t1 , t2) • Is any other form of orthogonality possible?

Dr Bhaskar Ramamurthi CDMA_Wireless_Course 3 Dr Bhaskar Ramamurthi CDMA_Wireless_Course 4

 Principles of Spread Spectrum and CDMA Principles of Spread Spectrum and CDMA

Orthogonality: Any Other Way? Orthogonality (contd..)

• How many orthogonal functions ci(t) can we get?
• Yes, if we consider only discrete-time signals
• Landau-Pollak Theorem: N∼ 2WT where
• Let c1(t) and c2(t) both of duration T be such that
W is the bandwidth available
their cross-correlation = 0 when time-aligned
T ⇒ given symbol duration T,
∫ c ( t )c ( t )dt = 0
1 2

N ∼ W / (1/T)

code
0

• Let signal_1 be x1 and signal_2 be x2 in [0,T]

↑

q
fre
• Consider sum_signal(t)= x1c1(t)+ x2c2(t) bandwidth expansion factor
time
⇒ xi can be extracted by correlating with ci(t) • ci(t) have to be time-aligned, i.e. synchronous
T T • signals overlap in time and frequency
∫ sum _ signal( t ).c ( t )dt =xi [ ∫ c ( t )dt]
i 2
i

0 0

Dr Bhaskar Ramamurthi CDMA_Wireless_Course 5 Dr Bhaskar Ramamurthi CDMA_Wireless_Course 6

Principles of Spread Spectrum and CDMA Principles of Spread Spectrum and CDMA

Orthogonal Direct-Sequence Code Division Multiplexing

Orthogonal CDM (contd…)
• c i(t) are binary-valued sequences such that
T
∫ c ( t )c ( t )dt = 0
i j • Example : Walsh codes
0 L c −1

− i.e., ci(t) = ∑ ( 2 c ki − 1 ) h ( t − T c ) where c k = {0,1},

i
k =0

and Tc is the chip period = T/Lc 0 0 0 0 0 0 H 2n-1 H2n-1

0 1 0 1 0 1 H2n-1 H2n-1
Lc −1 0 0 1 1
• equivalently ∑ cki XOR ckj = Lc H2
0 1 1 0
H2 n
k =0
H2 2
• { c } are called codes, hence the terms CDMA and
i
k
• H2 6 used in IS-95
synchronous CDM

Dr Bhaskar Ramamurthi CDMA_Wireless_Course 7 Dr Bhaskar Ramamurthi CDMA_Wireless_Course 8

 Principles of Spread Spectrum and CDMA Principles of Spread Spectrum and CDMA

Asynchronous CDMA Asynchronous CDMA (contd..)

• in orthogonal CDM, codes are synchronised
• even chip alignment will not be there
⇒ difficult to ensure between independent transmitters
at variable distances from receiver (i.e, multiple
access) • if {bk} and {ck} are two truly-random binary sequences
Lc −1
• synchronous orthogonal codes can have large of length Lc, and cross-correlation β = ∑ ( 2bk − 1 )( 2ck − 1 )
k =0
correlation when not synchronised
⇒ not used in CDMA • E[β] = 0 and E[β 2] = Lc
• employ pseudo-random or pseudo-noise (PN) ⇒ interference power from (N-1) signals ∝ (N-1) Lc
sequences x11 x21 • For the signal we desireL to1 extract,
c−
⇒ typically, low but non-zero desired_signal power = [ ∑ ( 2ck − 1 )( 2ck − 1 )] = Lc
2 2

k =0
correlation between any two
sequences x2 2 ⇒ undesired per-user interference suppressed by
x12
factor 1/Lc on the average
⇒ quasi-orthogonal codes
− for large N, variation around average will be less

Dr Bhaskar Ramamurthi CDMA_Wireless_Course 9 Dr Bhaskar Ramamurthi CDMA_Wireless_Course 10

Principles of Spread Spectrum and CDMA Principles of Spread Spectrum and CDMA

Signalling over Fading Channels

• A fading channel has deep nulls in its frequency response Spread Spectrum Signalling and Fading
≡ multiple paths taken by signals from Tx to Rx • c(t) spreads a narrowband signal uniformly
Imp. Freq. over a band T/Tc larger C(f)
Resp Resp
⇒ C(f) must be flat and wide f
τ t 1/2τ f
⇒ Rcc(τ) must be impulse like convolution is uniform
wideband
destructive • multipath signal α1x1 c(t) + α2x1c(t- τ )
Imp. Freq. interference
Resp Resp correlated with c(t)
τ t 1/2τ f ⇒ α1x1 Rcc(0) + α2x1 Rcc(τ) ≈ α1x1Rcc(0) if τ > Tc
• Only a small part of a wideband signal’s spectrum is severely ⇒ delayed signal is suppressed for delay > Tc
distorted by the nulls
• for slow frequency hopping, some bursts of
− in FDM or OFDM, a (narrowband) signal located around symbols falling in spectral nulls are affected
f= 1/2 τ will be lost
⇒ employ coding across bursts to overcome this
• multipath leads to Inter Symbol Interference if τ ~T (symbol
duration)
− in TDM, T is small ⇒ tolerable delay spread is less
Dr Bhaskar Ramamurthi CDMA_Wireless_Course 11 Dr Bhaskar Ramamurthi CDMA_Wireless_Course 12
 Principles of Spread Spectrum and CDMA Principles of Spread Spectrum and CDMA

Spread Spectrum Diversity SS Diversity (contd…)

• estimate path delays greater than Tc in a multipath channel • each finger will give α i Rcc(0) x1
• implement one correlator for every significant path :
⇒ combine to get better decision on x1
RAKE receiver
T

∫ • ISI has been resolved and converted to

0
c1(t) diversity gain
T +τ 1

∫ combine
• if path gains [α i ] are also estimated, can weight
τ1
c1(t-τ1)
each finger proportionately
T +τ 2 path gains
∫ ⇒ “maximal ratio” combining
c1(t- τ2) τ2

else “equal gain” combining

RAKE RECEIVER

Dr Bhaskar Ramamurthi CDMA_Wireless_Course 13 Dr Bhaskar Ramamurthi CDMA_Wireless_Course 14

Principles of Spread Spectrum and CDMA Principles of Spread Spectrum and CDMA

Multi-Transceiver (Macro) SS Diversity Macro Diversity (contd…)

• multiple transmissions on same carrier with different
• The fingers of a RAKE receiver need not be fed by same RF
codes can be combined
front-end
⇒ different transceivers at same base station
c(t) (e.g., adjacent sectors)
T
⇒ different base stations!
∫
α1x1 c (t)+ α2x1b(t - τ)
0
b(t)
combiner
T +τ c(t)
∫ T
τ
RF ∫
0
c(t-τ)
combiner
⇒ signals from two base stations can be combined during handoff T +τ
RF ∫
τ

Dr Bhaskar Ramamurthi CDMA_Wireless_Course 15 Dr Bhaskar Ramamurthi CDMA_Wireless_Course 16

 Principles of Spread Spectrum and CDMA Principles of Spread Spectrum and CDMA

Multi-Access Interference in CDMA Refinement of MAI Computation

• codes employed have low autocorrelation sidelobes, and • effect of bandpass transmission
typically low cross correlation also - since each interferer will have a random phase
offset (actually, small frequency offset),
• assume all signals arrive with equal power MAI will be less
⇒ perfect power control at Tx to counteract path loss and E[cos2ϕ] = 0.5 for uniformly distributed ϕ
fading
• effect of chip mis-alignment
• desired_signal energy at each RAKE finger ∝ Lc2 − if interference is y1 for left alignment and y1
• sum of undesired-signal (interference) energies at each y2 for right alignment,
RAKE finger ∝ NLc actual interference α y1 + (1- α) y2 where α y2
is uniformly distributed in [0,1]
• if MAI is approximated as Gaussian for large N, E{[α y1 + (1- α)y2]}2 = (1/3) [Ey12 + Ey22] = 2Lc/3
⇒ variance of MAI = NLc • Net effect : CIR increases by a factor of 3
⇒ Carrier-to-Interference Ratio (CIR) = Lc / N • For large cells, thermal noise adds to MAI near
cell boundaries : CINR drops
Dr Bhaskar Ramamurthi CDMA_Wireless_Course 17 Dr Bhaskar Ramamurthi CDMA_Wireless_Course 18

Principles of Spread Spectrum and CDMA Principles of Spread Spectrum and CDMA

Sectoring Reduces MAI Power Control in CDMA

• BS employ sectoral antenna
• if an interferer’s power is higher by, say, 10 times
⇒ downlink adjacent cell interference reduced
⇒ uplink uncontrolled interference from ⇒ equivalent to 10 interferers of equal power
adjacent cells reduced ⇒ CIR drops dramatically
• need to control power of each user to ±0.5dB,
• capacity improvement factor vis-à-vis circular cell
= η × 360o/ sectoral angle (η = 2.8 (4.5) for 3 (6) sectors) - use combination of open-loop and closed-loop
control

• sectoral gain improves link budget in noise-limited situation • path loss due to shadowing similar on up and
⇒ no impact in interference-limited case down links even if frequencies are different
• uncontrolled adjacent sector interference reduces capacity by 0.6 ⇒ open-loop control of Tx power based on local Rx
signal strength
⇒ ~ 0.6 x 60 x η (~35 η) users per cell with variable rate coding

Dr Bhaskar Ramamurthi CDMA_Wireless_Course 19 Dr Bhaskar Ramamurthi CDMA_Wireless_Course 20

 Principles of Spread Spectrum and CDMA Principles of Spread Spectrum and CDMA

Power Control in CDMA (contd…) Statistical Multiplexing in CDMA

• individual signal is bursty with activity factor γ
• fading will be different in up and down link (different
frequencies) − bursts occur randomly
⇒ closed-loop control of Tx power based on
feedback from far-end Rx • for large N, total “traffic” at any time t is
- usually a low bitrate binary signal :1/0 → up/down equivalent to γ N non-bursty signals
by 0.5 dB t
• in CDMA, MAI from M bursty signals is not
NPTX/Lc, but γ NPTX/Lc for large N
Open loop Closed loop
• if variable rate voice coder employed
Tx
Tx Tx ⇒ γ ∼ 0.4, N ∼20-30 sufficiently large

feedback
duplexer
control Dplx Dplx • for IP packets, γ ∼ 0.1 or less, but N has to be much
Rx larger due to long-tailed distributions of ON/OFF periods
Rx Rx

Dr Bhaskar Ramamurthi CDMA_Wireless_Course 21 Dr Bhaskar Ramamurthi CDMA_Wireless_Course 22

Principles of Spread Spectrum and CDMA

Pros and Cons of CDMA

• spread spectrum diversity
- multipath, macro
• statistical multiplexing easily exploited
- large number of quasi-orthogonal codes
⇒ large number of bursty users
• easy to support a variety of bit-rates
e.g; if chip rate is 2.048 Mbps service at n kbps, has a
spreading factor of 2048/n
⇒ n can be any power of 2

• strict power control required

- combination of open and closed loop control
• difficult to hand over from one carrier to another
- seamless handoff not possible between carriers, similar to FDM

Dr Bhaskar Ramamurthi CDMA_Wireless_Course 23