Adaptive Antenna Systems

All Rights Reserved © Alcatel-Lucent 2007

Adaptive Antenna Systems

Tibor ASZTALOS, NE/WiMax; Andrei OANA, NE/REMS

May 2007

All Rights Reserved © Alcatel-Lucent 20072 | AAS | May 2007

Agenda

1. Adaptive Antenna Concept

2. AAS Algorithms

3. AAS in WiMAX

4. [AAS implementation in A9155 V6.6], not covered


Adaptive Antenna Concept


Adaptive Antennas are:

Arrays of elementary antennas

With a precise spacing between elements (i.e. λ/2)

Amplitude and phase control on each element

1. Adaptive Antenna ConceptWhat is an Adaptive Antenna ?!


1. Adaptive Antenna ConceptWhat is the purpose ?!

Main capabilities of adaptive arrays

Range Extensionsteering of the direction of maximum transmit powerAAS gain due to multiple antennas (N elements)

– Gain of 20·log(N) on DL– Gain of 10·log(N) on UL

Capacity Enhancementsteering of the direction of maximum transmit powerreduction of interference by ‘null steering’Data rate increase due to higher modulation (due to C/I)


Beam Forming by Adaptive Antenna Systems (AAS)天线阵列单元要求信号相关requires coherent signals at the array elements

antenna spacing must be smaller than coherence distance (typically λ/2)

performance degradation in strong multi-path environment

Diversity gain• antenna spacing must be larger than coherence distance• Requires uncorrelated signals for highest gain

Space Time Coding (STC) through a Multiple Input Multiple Output(MIMO) System

requires propagation of the signal through independently fading channels

antenna spacing must be larger than coherence distance

most benefit in strong multi-path environment

1. Adaptive Antenna ConceptOther techniques…


1. Adaptive Antenna ConceptBeam Forming with AAS

Beamforming:

Element spacing λ/2

Beam is formed by compensating phase differences

Sidelobe control by amplitude tapering

Possibility to insert nulls

w1

w2

wM

x (k)M

y(k)

Direction FindingBeam Forming

x (k)2

x (k)1

AntennaElements

All Rights Reserved © Alcatel-Lucent 2007

User

strong interfererPreferred Application Scenarios

Coherent signals at the antenna arrayRestricted angular spread of multi-pathsTypical scenario: BS significantly higherthan surrounding reflectors

Macro-cellular: rural, sub-urban and urban

Antenna System requirements:Antenna spacing ~ λ/2: compactnessOnly BS side

1. Adaptive Antenna ConceptBeam Forming with AAS


1. Adaptive Antenna ConceptDiversity gain

Diversity gain:

Element spacing 10 .. 20λUsed to combat fading by exploiting decorrelation of amplitudes

Coherent combining by additional phase correction

Cophasingand Summing

Cophasingand Summing

a1 a2 aM

Equal GainCombining

Maximal-RatioCombining


1. Adaptive Antenna ConceptMIMO

STC with MIMO:

Element spacing 10 .. 20λUsed to transmit and receive with spatial multiplexing

Signals have to be uncorrelated

N

Rx

Radio channel

Tx

M

1 1

N- Tx antennas M – Rx antennas


AAS Algorithms


Conventioanl beam steering

steers the maximum beam towards the wanted user

uses only the phase to control the weights

Null steering

steers a maximum towards the wanted user

steers a null towards the strongest (N-1) interferers

uses phase and amplitude in the complex weights

Minumum Mean Square Error

gives the best C/I under heavy interference conditions

uses a reference signal and minimizes the error (y(t)-r(t))

2. AAS Algorithms


Beam Patterns for Up-LinkAdaptive Antenna Processing

Main Lobe Steering (Blue)

lobe suppression (15dB over ± 40°scan range) through amplitudetapering

No cancellation of individualinterference

Interference Cancellation (Red)

Mitigation even of interferencebeing ‘close in direction’

Single Antenna Patterns (Green)

HPBW: 90°

Signal - 111 °, Interference - 91 °

2. AAS AlgorithmsBeamforming example


2. AAS AlgorithmsBeamforming mechanism

The following functions are performed for Beamforming

Direction of Arrival estimation for all incoming signals

Identify desired user signal

The beam is steered with the weights in the direction of wanted user

User is tracked while moving


2. AAS AlgorithmsBeamforming Mathematical model

Steering vector (S)Ψi – is the delay of the signal arriving at antenna element i


2. AAS AlgorithmsBeamforming algorithm

The weights are selected to be the conjugate of the steering vector

wH·S=1

Advantage:• Very simple algorithm• Provides maximum SNR if noise is uncorrelated

Is used only on DL in Alcatel-Lucent WiMAX implementation (W3)


Antenna Array Power Patterns

-50

-45

-40

-35

-30

-25

-20

-15

-10

-5

0

-180 -150 -120 -90 -60 -30 0 30 60 90 120 150 180

Angle [deg]

Pow

er in

Rad

iatio

n Pa

tter [

dB re

lativ

e to

m

ain

lobe

at b

ores

ight

]

140° 130° 120°110° 100° 90°80° 70° 60°50° 40°

2. AAS AlgorithmsBeamforming patterns


2. AAS AlgorithmsNull steering algorithm

The weights are computed such as to:

Steer the main beam towards the wanted user (S0)

Steer nulls towards the k=N-1 interfering users (Si)

Drawback:

Does not provide the maximum SINR


2. AAS AlgorithmsMinimum Mean Square Error algorithm

A known reference signal is needed

The weights are determined such as to minimize the error (y(t)-r(t))

No direction of arrival estimation needed

Advantage

Works very good in high interference conditions

Is used on the UL in the Alcatel-Lucent WiMAX implementation (W2.1)


2. AAS AlgorithmsBroadcast pattern

provides a uniform coverage of the preamble and broadcast information within the cell

Constant amplitude and phase used during the broadcast

~2dBi gain compared to elementary antenna gain


Array AntennaAndrew, 2.5 GHz (2.3 GHz – 2.7 GHz)Mounted in open space environmentA rotor allows to turn the antenna for pattern measurements

Base StationNeMo HW3 PHY Layer W2.1 with adjustable pattern weightsMAC SD7 / SD9 in demo mode4 FEUs

Mobile StationsFixed antennas at various positions, approx 15 m air linkZyXel/Runcom MSS connected by cableAlternatively Laptop with MSS can be carried through the field

Application3 Videos running in parallel in DL

2. AAS AlgorithmsAntenna pattern measurement: test platform


2. AAS AlgorithmsAntenna pattern measurement: Antenna Element Patterns


2. AAS AlgorithmsAntenna pattern measurement: Broadcast Patterns


2. AAS AlgorithmsAntenna pattern measurement: User #1 Individual Pattern


2. AAS AlgorithmsAntenna pattern measurement: User’s #1, #2 and #3 Individual Patterns


2. AAS AlgorithmsAntenna pattern measurement: Pattern’s Max Limit vs. Angle of Arrival


AAS in WiMAX


Typical AAS parameters;

Frequency: 2500 / 3500 MHz

Number of elements: 4

Gain of one element: 17dBi

HPBW of one element: 90deg horizontal, 5 deg vertical

Boresight gain: 23 dBi

HPBW Boresight steering: 25 deg

Length: 1.35m (2500MHz), 1m (3500MHz)

Antenna manufacturers:Andrew: APW425-12014 (2500 MHz),

APW435-12014 (3500 MHz)

RFS: W425-90ANV (2500 MHz),

W435-90ANV (3500 MHz)

3. AAS in WiMAXAntenna parameters and manufacturers


System requirements to support beamforming

Dedicated pilot mode fordown-link PUSC zone, down-link AMC 2x3 zone.

Feedback of physical CINR on pilots ofdown-link PUSC zone,down-link AMC 2x3 zoneprovided through the Channel Quality Indicator Channel (CQICH)

Up-link soundingprovide training for interference cancellation in up-link as well as in down-link operation.

Provisioning of up-link permutations for beamforming operation:up-link PUSC without sub-channel rotationup-link AMC 2x3.

3. AAS in WiMAXAAS and WiMAX PHY

All Rights Reserved © Alcatel-Lucent 200730 | AAS | May 2007 U

IUC

12

Com

pressed DL M

ap & C

ompressed UL M

ap

Com

pressed DL M

ap & C

ompressed U

L Map

DL Burst #1

DL Burst #3

DL Burst #2

DL Burst #4

DL Burst #5

DL Burst #6

DL Burst #7

UIU

C 0

Preamble

#k #k+1

#k+2

.... #k+2

×n

.... .... .... .... .... .... .... .... .... .... .... .... .... .... .... .... .... .... .... .... .... .... .... .... .... .... .... #k+2

×m

Logical Sub-Channels

FCH

#k+2

×m+1

.... #k+2

m+3

×p

.... .... .... .... .... #k+2

m+3

×q

.... .... .... .... .... #k+4

7

TTG

UL Burst #1

UL Burst #2

UL Burst #3

UL Burst #4

UL Burst #5

UL Burst #6

UL Burst #7

UL Burst #8

UL Burst #9

3. AAS in WiMAXUL beamforming (W2.1)

Beamforming applied only on the UL PUSC zone

No changes in the frame structure required (same as single antenna)

UIUC 0 - Fast feedback channel carrying the CQICH (CINR)

UIUC 12 – CDMA based ranging


3. AAS in WiMAXDL Beamforming (W3)

Broadcast pattern

Adaptive pattern

Each user has his own adaptive beamforming weights


3. AAS in WiMAXUL Beamforming (W3)

UL PUSC

ULSounding

UL AMC2x3

#k+2×m+3×p

#k+2×m+1

........

........

........

........

........

#k+47

UL Burst#1

UL B

urst #3

UL B

urst #4

UL B

urst #5

UL B

urst #6

UL B

urst #7

UL B

urst #8

UIUC 13

UL Burst#2

#k+2×m+3×p+1

#k+2×m+3×p+3

UIUC 12 initial / handoverranging

UIUC 12 periodic ranging

UIUC 0

(e.g. CQICH)

Broadcast pattern

Adaptive pattern

UIUC 0 - Fast feedback channel carrying the CQICH (CINR)

UIUC 12 – CDMA based ranging

UIUC 13 – UL sounding zone, training pilots for UL interference cancellation


3. AAS in WiMAXUL Sounding Zone example


PHY layer stores the individual weight vectors for each subscriber

The association weight vector – subscriber is done through MAC layer

A handle is exchanged between MAC and PHY as entry in the table of UL/DL weights (during UL allocations)

The handle is updated by the MAC layer

3. AAS in WiMAXInterface between MAC and PHY layer


Thank you!

www.alcatel-lucent.com

Adaptive Antenna Systems

Documents

rights reserved

compressed

adaptive antenna

adaptive antenna

link pusc

wimax ul beamforming

lucent wimax

wanted user