MSIT 413: Wireless Technologies

N O R T H W E S T E R N MSIT | Master of Science in Information Technology

U N I V E R S I T Y

MSIT 413: Wireless Technologies Week 6

Michael L. Honig Department of EECS

Northwestern University

November 2014


U N I V E R S I T Y

Outline

•  Finish diversity

•  Multiple Access techniques –  FDMA, TDMA –  CDMA (3G, 802.11b) –  OFDMA (4G, WiMax)


U N I V E R S I T Y

Diversity •  Idea: Obtain multiple independent copies of the received signal.

–  Improves the chances that at least one is not faded.

•  Macroscopic (space): copies of signal are received over distances spanning many wavelengths.

•  Microscopic (space): copies of signal are received over distances spanning a fraction of a wavelength

•  Different types…


U N I V E R S I T Y

Macroscopic Diversity

Copies of signal are separated by many wavelengths.

MSO


U N I V E R S I T Y

Microscopic Space Diversity

•  Want signals s1 and s2 to experience independent fading (why?). –  distance between antennas should be ≥ ½ wavelength. –  Ex: 900 MHz, λ = c/f ≈ 1/3 meter

2 GHz, λ ≈ 0.15 meter

Antenna 1 Antenna 2

s1 s2


U N I V E R S I T Y

Frequency Diversity

frequency

channel gain

Wideband signals exploit frequency diversity. Spreading power across many coherence bands reduces the chances of severe fading. Wideband signals are distorted by the channel fading (distortion causes intersymbol interference).

coherence bandwidth Bc

f1 f2

Frequencies far outside the coherence bandwidth are affected differently by multipath.

signal power (wideband)


U N I V E R S I T Y

Time Diversity

•  Transmit multiple copies of the signal in time. –  Error control coding: add redundant bits

•  Problem: slow fading –  Combine with power control


U N I V E R S I T Y

Path Diversity

•  Called a “RAKE” receiver, since it “rakes” up (combines) the energy in the different paths. –  Can substantially increase the S/I.

•  An important component of CDMA receivers. –  Each branch in the Rake is typically referred to as a “finger”.

Delay τ2 - τ1

adjust phase

adjust phase

+ τ1 τ2

received signal


U N I V E R S I T Y

Multiuser Diversity


U N I V E R S I T Y

Multiuser Diversity

d1

d2 > d1

Received power user 1

user 2

time transmit to user 1

transmit to user 2

transmit to user 2

transmit to user 1

The BST can choose to transmit to the user with the best channel. Exploits variations in signal strength across users.


U N I V E R S I T Y

Selection Diversity

Received power

antenna 2

time select ant. 1

select ant. 2

Choose the best signal (highest instantaneous SNR). Easy to implement (antenna switch).

Antenna 1 Antenna 2

s1 s2

antenna 1

select ant. 2

select ant. 1


U N I V E R S I T Y

Benefit of Selection Diversity (Example) •  Suppose that the signal on each antenna experiences

independent Rayleigh fading.

•  Determine the probability that the received signal is faded: Recall Rayleigh fading formula: Probability that the signal power is less than a x P0 (average received power) = 1 – e-a Hence the probability that the signals on both antennas are less than a x P0 = (1 – e-a)2 Without diversity, probability of a signal fade = 1 – e-1 = 0.63 With 2-branch diversity, probability of a signal fade = 0.632 = 0.39


U N I V E R S I T Y

Benefit of Selection Diversity (cont.) •  Suppose that there are N copies of the signal

(e.g., N antennas, paths, coherence bands, etc.)

Probability that the signal power is less than a x P0 (average received power) = 1 – e-a Hence the probability that all N signals are less than a x P0 = (1 – e-a)N Without diversity, probability of a signal fade = 1 – e-1 = 0.63 With 4-branch diversity, probability of a signal fade = 0.634 = 0.16 Without diversity, Prob(signal is faded by more than 10 dB) = 1 – e-0.1 ≈ 0.1 With diversity this probability is (1 – e-0.1)4 ≈ 0.0001 !


U N I V E R S I T Y

Coherent Combining

•  “Coherent” means that the phases of the two signals are estimated at the receiver and aligned.

•  Performs better than selection combining (why?).

•  Example: RAKE receiver

•  Can weight the combined signals to maximize the received SNR. (How should the weights depend on the signal levels?)

adjust phase

adjust phase

+

S1 (ant. 1)

S2 (ant. 2)


U N I V E R S I T Y

Probability of Error with Fading

add diversity

•  Diversity can transform a fading channel back to a non-fading (additive noise) channel. •  Essential for mobile wireless communications.


U N I V E R S I T Y

Multiple Antennas: Multi-Input/Multi-Output (MIMO) Channel

Multi-Channel Detector

Transmitted

Data (multiple data streams)

Estimated Data

•  Multiple (M) antennas at receiver and transmitter •  Channel has multiple inputs and multiple outputs.

16


U N I V E R S I T Y

Single Transmit Antenna


Transmitted

Data (single stream)

Estimated Data

•  Multiple receiver antennas provides spatial diversity •  Lowers error rate

•  “Single-Input/Multiple-Output (SIMO)” channel

17


U N I V E R S I T Y

Multi-Input/Single Output (MISO) Channel

Single-Channel Detector

Transmitted

Data (single or multiple streams)

Estimated Data

•  Transmitting the same symbol from all transmitters provides transmit spatial diversity (e.g., select the best antenna, turn the others off).

•  Practical for cellular downlink. 18


U N I V E R S I T Y

Downlink Beamforming

Different beams can use the same frequency!

Narrow “beam” focused on one user

19

•  M antennas at the base station (single or multiple antennas at mobiles)

•  Can support up to M data streams. •  Multi-user MIMO: multiple users on the same channel

–  Introduced in LTE, 802.11ac


U N I V E R S I T Y

OFDM Signal

Orthogonal Frequency Division Multiplexing (OFDM)

Split into M substreams source

bits

Modulate Carrier f1

+

Modulate Carrier f2

Modulate Carrier fM

substream 1

substream 2

substream M …


U N I V E R S I T Y

Multiple Antennas: Multi-Input/Multi-Output (MIMO)Channel

•  Multiple (M) antennas at receiver and transmitter.


Transmitted Data Estimated

Data

21


U N I V E R S I T Y


•  Multiple (M) antennas at receiver and transmitter. •  Transmitted data is divided into M substreams, one for each antenna.

•  Transmit antennas are used to multiplex multiple data streams.


Substream 1

Estimated Data Substream M

22


U N I V E R S I T Y



•  Transmit antennas are used to multiplex multiple data streams. •  Multiple receiver antennas (plus signal processing) are used to

remove interference from the different antennas.


Substream 1

Estimated Data Substream M

23


U N I V E R S I T Y



•  Transmit antennas are used to multiplex multiple data streams. •  Multiple receiver antennas (plus signal processing) are used to

remove interference from the different antennas. •  Data rate (Shannon capacity) is proportional to M!


Estimated Data

Substream 1

Substream M

24


U N I V E R S I T Y

WiFi Evolution: 802.11n

•  Technology based on OFDM with multiple antennas at the transmitter and receivers

•  Supports data rates up to 540 Mbps –  4 spatial streams, 40 MHz bandwidth –  Can replace USB 2.0 connections.

•  Also important part of 802.11ac (multi-user MIMO)

25


U N I V E R S I T Y

The Multiple Access Problem

•  Frequency-Division (AMPS)

•  Time-Division (GSM)

•  Code-Division (3G, Bluetooth) Direct Sequence/Frequency-Hopped

•  Orthogonal Frequency Division Multiple Access (OFDMA)

•  Random Access (Wireless Data)

How can multiple mobiles “access” (communicate with) the same base station?


U N I V E R S I T Y

Duplexing (Two-way calls)

Frequency-Division Duplex (FDD)

Time-Division Duplex (TDD)

Channel 1

Channel 2

Time slot (frame) 1

Time slot (frame) 2


U N I V E R S I T Y

Combinations

•  FDMA/FDD (AMPS) •  TDMA/FDD (GSM) •  TDMA/TDD (IS-136 or 2G in the U.S.) •  CDMA/FDD (IS-95, CDMA2000) •  CDMA/TDD (3G/UMTS) •  Frequency-Hopped CDMA/TDD (Bluetooth) •  OFDMA/TDD and FDD (WiMax, 4G)


U N I V E R S I T Y



•  Time-Division (IS-136, GSM)

•  Code-Division (IS-95, 3G) Direct Sequence/Frequency-Hopped



How can multiple mobiles access (communicate with) the same base station?


U N I V E R S I T Y

Cellular Spectrum (50 MHz)

A* A B A* B*

uplink

downlink

824

869 870 880 890 891.5 894

825 835 845 846.5 849

AMPS (1G): 30 kHz Channels

416 FDD Channels (requires 12.5 MHz): •  395 FDD voice channels •  21 FDD control channels


U N I V E R S I T Y

Properties of FDMA •  Can be analog or digital (AMPS is analog).

•  Narrowband: channel contained within coherence

bandwidth – undergoes flat fading.

•  Low capacity

•  Best for circuit-switched (dedicated) connections.

•  Requires guard channels for adjacent channel interference.


U N I V E R S I T Y



•  Time-Division (GSM)

•  Code-Division (IS-95, 3G) Direct Sequence/Frequency-Hopped





U N I V E R S I T Y

Frame

H 1 2 N H . . .

N time slots H: Frame Header

Time Division Multiple Access

Time slots

H 1 2 N H . . .

H 1 2 N H . . .

Channel f1

Channel f2

Channel fK

. . .


U N I V E R S I T Y

Frame

H 1 2 N H . . . N time slots H: Frame Header

Data to or from user K + control information

Time Slot

Preamble and synch

Guard time

Time Slot


U N I V E R S I T Y

H 1 2 H

TDMA/Time-Division Duplex

3 4 5 6

Uplink time slots Downlink time slots { {


U N I V E R S I T Y

Properties of TDMA •  Data transmission occurs in bursts.

–  Must ensure small delays for speech. –  High peak to average power on reverse link.

•  Can measure signal strength in idle time slots (e.g., for handoff).

•  Can assign multiple time slots for higher data rates.

•  Significant overhead/complexity for synchronization. –  Guard times needed between time slots for delay spread.

•  May require an equalizer to mitigate intersymbol interference.


U N I V E R S I T Y

Global System for Mobile Communications (GSM)

•  Originated in Europe –  Main objective: allow roaming across countries –  Incompatible with 1G systems

•  More than an air-interface standard: specifies wireline interfaces/functions

•  TDMA/FDMA, FDD •  Dynamic frequency assignment •  200 kHz channels •  270.833 kbps


U N I V E R S I T Y

GSM Frame Structure

•  200 kHz FDD channels divided into 8 time slots per frame •  Total number of available channels = (12.5 MHz – 2 X Guard Band)/200 kHz

100 kHz guard bands è 62 channels •  Total number of traffic channels = 8 X 62 = 496 •  Channel data rate = 270.839 kbps •  Without overhead, data rate/user = 24.7 kbps

TS0

TS1

TS2

TS3

TS4

TS5

TS6

TS7

4.615 ms

156.25 bits 576.92 µs

T0 T1 T2 T22 T23 T24 I/S T10 T11 T12 S T13 T14 T15 ... … 120 ms

Tn: nth TCH frame S: Slow Associated Control Channel frame I: Idle frame

Speech Multiframe = 26 TDMA frames

Frame


U N I V E R S I T Y

GSM Frame Structure


100 kHz guard bands è 62 channels Total number of traffic channels = 8 X 62 = 496 Channel data rate = 270.839 kbps Without overhead, data rate/user = 24.7 kbps

TS0

TS1

TS2

TS3

TS4

TS5

TS6

TS7

4.615 ms

156.25 bits 576.92 µs




Frame


U N I V E R S I T Y

GSM Frame Structure


100 kHz guard bands è 62 channels •  Total number of traffic channels = 8 X 62 = 496 •  Channel data rate = 270.839 kbps •  Without overhead, data rate/user = 24.7 kbps

TS0

TS1

TS2

TS3

TS4

TS5

TS6

TS7

4.615 ms

156.25 bits 576.92 µs




Frame


U N I V E R S I T Y

GSM Time Slots

0 1 2 3 4 5 6 7

4.615 ms

120 ms

6.12 s

3 57 1 26 1 57 3 8.25

576.92 µs

“Normal Burst” Traffic Channel (TCH)

•  148 bits/time slot •  114 coded information bits •  Frame efficiency 74% (total bits – overhead bits)/(total bits)

Superframe

Multiframe

Frame

Time slot

Hyperframe = 2048 superframes lasts ~3 hrs 28 min 54 sec

≈


U N I V E R S I T Y

GSM Capacity

•  Total bandwidth = 12.5 MHz, 200 kHz channels è 62 channels –  With cell cluster size N=3 (typical),

capacity is (62/3) x 8 ~ 165 users/cell


U N I V E R S I T Y


•  Use different frequencies (FDMA)

•  Use different time slots (TDMA)

•  Use different pulse shapes (CDMA)



U N I V E R S I T Y

Code Division Multiple Access •  Users transmit simultaneously over

the same frequency band •  Performance limited by interference


U N I V E R S I T Y

Two-User Example

time User 1:

T T/2

time T

T/2

User 2:

T 2T 3T 4T 5T

s2(t) 1 0 1 1 0

T 2T 3T 4T 5T s1(t)

1 1 0 1 0 bits:

T 2T 4T 5T

received signal r(t)= s1(t)+s2(t)

How to recover each users’ bits? 3T

2

-2

1

-1


U N I V E R S I T Y

Chip Sequence

time T

T/2

User 2:

T 2T 3T 4T 5T

s2(t) 1 0 1 1 0

chips

bits:

1 -1 1 1 -1 Transmitted chips: -1 1 -1 -1 1

chip duration Tc

symbol duration T=2Tc

User 2’s chip sequence (1, -1) is called a signature.


U N I V E R S I T Y

Chip Sequence

time T T/2

User 1:

chips

1 1 -1 1 -1 Transmitted chips: 1 1 -1 1 -1

chip duration Tc

symbol duration T=2Tc

T 2T 3T 4T 5T s1(t)

1 1 0 1 0 bits: 1

-1

User 1’s signature is (1, 1).


U N I V E R S I T Y

Two-User Example

T 2T 3T 4T 5T

s2(t)

T 2T 3T 4T 5T s1(t)

T 2T 4T 5T

r(t)= s1(t)+s2(t)

3T

2

-2

1

-1

1 -1 1 1 -1 Transmitted chips: -1 1 -1 -1 1

2 0 0 2 -2 Received chips: 0 2 -2 0 0

1 1 -1 1 -1 Transmitted chips: 1 1 -1 1 -1

1 0 1 1 0

1 1 0 1 0


U N I V E R S I T Y

Correlation

Given two sequences: a1, a2, a3, …, aN

b1, b2, b3, …, bN

The correlation between the sequences is defined as:

(a1 x b1) + (a2 x b2) + (a3 x b3) + … + (aN x bN)

Examples: 1 1 1 1 1 correlated with 1 1 1 1 1 = 5 1 1 1 1 1 correlated with 1 -1 1 -1 1 = 1 2 4 1 3 1 correlated with -2 1 8 2 0 = -4 + 4 + 8 +6 + 0 = 14

If the correlation between two sequences is zero, they are said to be orthogonal.


U N I V E R S I T Y

Correlator Receiver

r(t) Sample Chips

Correlate with desired user’s signature

Bit Decision < 0 à 0 > 0 à 1

estimated bits


U N I V E R S I T Y

Why Does This Work?

Correlate with User 1’s signature

A1 s1 + A2 s2

signature (1,1) amplitude


A2 s2 0

The user signatures are orthogonal.

Now observe that:


A1 s1

2A1

2A1


U N I V E R S I T Y

Correlator, or “Matched Filter” Receiver


A1s1+ A2s2 User 1’s bits


User 2’s bits

The correlator is “matched” to user 1’s signature s1, and rejects s2 (and vice versa).




U N I V E R S I T Y

Observations •  Users transmit simultaneously (not TDMA). •  Users overlap in frequency (not FDMA).

frequency

Spectrum: User 1

0

signal bandwidth is roughly 1/T

frequency 0

signal bandwidth is roughly 1/Tc = 2/T Spectrum: User 2

Bandwidth expansion (factor of 2) è “spread spectrum” signaling.


U N I V E R S I T Y

Users and Bandwidth Expansion To guarantee orthogonal signatures (no interference), the length of the signatures must be ≥ the number of users.

Example (4 users):

time T/2 T/4 User 3: 3T/4

T time T/2 T/4 User 4: 3T/4

T

time T/2 T/4 User 2: 3T/4

T time T/2 T/4 User 1: 3T/4

T

The chip rate is 4 times the symbol rate, hence the bandwidth expansion is a factor of 4.

signature: 1 1 1 1 signature: 1 1 -1 -1

signature: 1 -1 -1 1 signature: 1 -1 1 -1


U N I V E R S I T Y

Correlator Receiver (4 users)


s1+ s2 + s3 + s4 User 1’s bits


User 2’s bits




User 3’s bits Bit Decision < 0 à 0 > 0 à 1


User 4’s bits Bit Decision < 0 à 0 > 0 à 1


U N I V E R S I T Y

Processing Gain

•  The PG is essentially the bandwidth expansion factor, given by (1/Tc)/(1/T) = T/Tc (chips per symbol), which is the length of the signature.

•  The signature (sequence of chip values) is also called a spreading code. The signature may be randomly generated, in which case it is called a pseudo-noise (PN) sequence.

•  “Direct-Sequence” CDMA uses a spread spectrum signalling scheme in which the signal is spread by transmitting a sequence of chips at a rate faster than the symbol rate.

RateSymbolSignal Spectrum Spread of Bandwidth)PG( Gain Processing ≡


U N I V E R S I T Y

DS-CDMA Transmitter

Spreader (generate chips)

Modulator (e.g., QPSK)

Source bits chips RF signal

Ex: 100 source symbols, processing gain = 10 è 1000 chips

X

sin 2πfct

Baseband signal Passband (RF) signal

time

Nyquist chip shape

Tc


U N I V E R S I T Y

Orthogonality and Asynchronous Users

•  Orthogonality among users requires: –  Synchronous transmissions –  No multipath

T 2T 3T 4T 5T s1(t)

1

-1

1 1 0 1 0

T 2T 3T 4T 5T

s2(t) 1 0 1 1 0

time

Asynchronous users can start transmissions at different times.

Chips are misaligned è signatures are no longer orthogonal!


U N I V E R S I T Y

Correlator, or “Matched Filter” Receiver


s1(t) + s2(t-τ)


The multiple access interference adds to the background noise and can cause errors. For this reason, CDMA is said to be interference-limited. Because CDMA users are typically asynchronous, and because of multipath, it is difficult to maintain orthogonal signatures at the receiver. Consequently, in practice, the signatures at the transmitter are randomly generated.



delay Signal 1 + multiple acess interference (MAI) From user 2

Signal 2 + multiple acess interference (MAI) From user 1


U N I V E R S I T Y

Correlation and PG

Example: PG=4 s1: 1 -1 -1 1 s2: 1 1 1 -1 Correlation = -2

Energy in s1 (or s2) is 12 + (-1)2 + (-1)2 + 12 = 4 Normalized correlation = correlation/energy = -2/4 = -1/2

Example: PG=10

Conclusion: On average, the correlation between signatures decreases as the signature length (PG) increases.


U N I V E R S I T Y

Correlation and Bandwidth

T 2T 3T 4T 5T

s2(t) 1 0 1 1 0

0 1

T 2T 3T 4T s2(t)

1 1 0

5T

frequency 0

frequency 0 Increasing the PG increases bandwidth, but decreases the correlation between user signatures.


A2 s2

correlation between s1 and s2 à multiple access interference

Increasing the PG decreases multiple access interference. Bandwidth expansion therefore provides “immunity” to interference (all kinds: analog, multiple access, multipath, narrowband, etc).


U N I V E R S I T Y

Example

•  IS-95 (2G CDMA) –  Total bandwidth = 1.25 MHz –  Data rate = 9.6 kbps –  PG ≈ 130

•  3G/CDMA2000

–  Total bandwidth = 1.25 MHz –  Data rate varies between 14.4. kbps (voice)

up to 2 Mbps (1X-DO) –  PG varies from 1 to 130


U N I V E R S I T Y

Properties of CDMA •  Robust with respect to interference

•  No frequency assignments (eases frequency planning)

•  Asynchronous

•  High capacity with power control –  Power control needed to solve near-far problem.

•  Wideband: benefits from frequency/path diversity.

•  Benefits from voice inactivity and sectorization. •  No loss in trunking efficiency.

•  Soft capacity: performance degrades gradually as more users are added.

•  Soft handoff


U N I V E R S I T Y

Near-Far Problem SO… THEN THE THIRD

TIME I CALLED CUSTOMER SERVICE, I SAID &%$#%^…


U N I V E R S I T Y

Near-Far Problem


A1 s1(t)+A2 s2(t) User 1’s bits Bit Decision < 0 à 0 > 0 à 1

User 1

User 2

A1+A2 × (correlation of s1 and s2)

amplitude A1

amplitude A2

Output of correlator receiver is signal + interference. As the interferer moves closer to the base station, the interference increases. In practice, power variations can be up to 80 dB!

Conclusion: User 1’s signal is overwhelmed by interference from user 2!


U N I V E R S I T Y

Closed-Loop Power Control

User 1

User 2

•  Base station gives explicit instructions to mobiles to raise/lower power. •  Needed to solve near-far problem (equalizes received powers). •  Introduced by Qualcomm in the late 80’s. •  Requires “closed-loop” feedback.

•  BST controls powers through feedback channel. •  Why “closed-loop”?

raise power

lower power


U N I V E R S I T Y

Closed-Loop Power Control

User 1

User 2

•  Base station gives explicit instructions to mobiles to raise/lower power. •  Needed to solve near-far problem (equalizes received powers). •  Introduced by Qualcomm in the late 80’s. •  Requires “closed-loop” feedback.

–  “Open-loop” power control (no feedback) is inadequate due to frequency-selective fading.

raise power

lower power


U N I V E R S I T Y

Closed-Loop Power Control: Properties

User 1

User 2

•  Crucial part of CDMA cellular systems (IS-95, 3G). •  Minimizes battery drain. •  Complicated (increases cost) •  Requires overhead: control bits in feedback channel to tell transmitter

to lower/raise power. •  Cannot compensate for fast fading.

raise power

lower power


U N I V E R S I T Y


•  No frequency assignments (eases RF planning).

•  Asynchronous

•  High capacity with power control. –  Power control needed to solve near-far problem.


•  Benefits from voice inactivity and sectorization. –  No loss in trunking efficiency.


•  Soft handoff


U N I V E R S I T Y

Bandwidth and Multipath Resolution

direct path (path 1)

reflection (path 2)

signal pulse

Narrow bandwidth è low resolution Receiver cannot distinguish the two paths.

signal pulse

Wide bandwidth è high resolution Receiver can clearly distinguish two paths.

multipath components are resolvable

τ (delay spread)

τ

T

T < τ T > τ


U N I V E R S I T Y

CDMA and Path Diversity

•  CDMA uses wideband signals (chips are very narrow pulses), so that multipath is resolvable.

•  A “RAKE” receiver collects (“rakes up”) the energy in the paths:

power delay profile

τ

delay τ + adjust phase

received signal

received signal with combined multipath


U N I V E R S I T Y



•  Asynchronous





•  Soft handoff


U N I V E R S I T Y

CDMA Capacity Performance depends on

Let S= Transmitted power (per user), R= information rate (bits/sec), W= Bandwidth, K= Number of users Eb= S/R (energy per second / bits per second) N0= (Number of interferers x S)/W = ((K-1) x S)/W Therefore Eb/N0 = (W/R)/(K-1) = (Processing Gain)/(K-1) For a target Eb/N0, the number of users that can be supported is K = (Processing Gain)/(Eb/N0) + 1

bandwidthunit per power Noise ceInterferenbitper Energy

0 +NEb ≡


U N I V E R S I T Y

CDMA Capacity: Example

•  For IS-95, want Eb/N0 ≥ 7 dB •  For 3G, want Eb/N0 ≥ 3 to 5 dB •  Suppose W=1.25 MHz (single-duplex), R= 14.4 kbps,

target Eb/N0 = 7 dB: K= 1 + [(1.25 × 106)/(14.4 × 103)]/5.01 ≈ 18

•  Compare with GSM, cluster size N=3: K= 8 (users/channel) × (# of 200 kHz channels) = 8 × 1.25 × 106 / (200 × 103 × 3) ≈ 16


U N I V E R S I T Y

Increasing CDMA Capacity •  Must reduce interference

•  Antenna sectorization

–  Interference reduced by 1/3 –  Trunking efficiency is not a major

issue (no channels/time slots).

•  Voice inactivity automatically increases the capacity relative to TDMA with dedicated time slots.

•  CDMA has a “soft” capacity: each additional user marginally degrades performance for all users.

other-cell interference


U N I V E R S I T Y



•  Asynchronous





•  Soft handoff


U N I V E R S I T Y

Interference and CDMA Capacity If interference is reduced by a factor 1/g, then the number of interferers can be increased by g (N0 is replaced by g x N0): If W/R is large, then reducing interference by 1/g (approximately) increases the capacity by a factor of g.

)Ng)(E(RW+=Kb 0//1/1

Previous example: voice activity of 1/3 combined with 120o sectors increases capacity by a factor of 9!


U N I V E R S I T Y

Refining the Capacity Estimate

•  Capacity for previous example is 9 × 18 ≈ 162 •  Have not accounted for:

–  Other-cell interference •  Approximately 1/3 to 1/2 of total interference power

K à 1/(1+1/2) × K ≈ 108

–  Multipath / fading •  Some multipath is combined by the Rake receiver, the rest is

interference

–  Power control inaccuracy

Precise capacity predictions become difficult, best to rely on field trials…


U N I V E R S I T Y



•  Asynchronous




•  Soft capacity: performance degrades gradually as more users are

added.

•  Soft handoff


U N I V E R S I T Y

Soft Handoff (CDMA) ”Make before break”

MSC

BSC BSC

MSC

BSC BSC

MSC

BSC BSC

Hard Handoff (TDMA)

MSC

BSC BSC

MSC

BSC BSC

MSC

BSC BSC

BEFORE DURING AFTER

MSIT 413: Wireless Technologies

Documents

Transcript of MSIT 413: Wireless Technologies