Smart Antennas

2nd Feb 2007

Overview
 Current Networks
 What are Smart Antennas
 Digital Beamforming Basics
 Switched Beam
 Adaptive Arrays
 Comparison
 Trial Results
 Simulation
Current UMTS Networks

 Issues
 Interference Limited / Capacity
 Limited Spectrum

 Usual Solutions
 Optimization
 New Carrier
What if we could narrow the beam to the UE?

Such System would:

Ability ■ Decrease DL interference since Tx

to power is concentrated in a narrow beam
Steer
■ Increase UL gain since Rx signal is
confined in a narrow beam at the direction
of arrival

-> Capacity Increase

-> Better Ec/Io
What is a Smart Antenna (System)?

 SDMA
Origins
 World War II rapidly-steerable Radar system to aid in
the landing of airplanes in England
 More recent military examples:

Ballistic Missile Early

Warning System (Alaska)
PATRIOT
Phased Array TRacking to
Intercept Of Target
Array Geometries

y z y y
z
x x x x

180º 360º 180º 360º

Array Examples

Smart Antenna IC
Motia, California

8-element array at 1.8 GHz

Telia Research AB, Sweden

900MHz 900 MHz

Ericsson Ericsson&Mannesmann
Digital Beamforming Basics
Incident Wave

α
■ The received signal phase is
0(α,0) different for each element (antenna).
d
1(α,d)
Receiver

■ The received signal phase

depends on the incident wave DoA*
2(α,2d) and the distance d between the
elements

3(α,3d)

* DoA: Direction of Arrival

Digital Beamforming Basics (2)

S
■ Gain and phase (delay) of the sinal
of each element is changed through an
S
array of “complex weights”
S

S ■ This way the system can

effectively create a narrow beam that
is looking at the DoA

■ The beam width depends on the

number of elements
Angle of Arrival (AoA)

AoA of UE moving at 120 km/h near a NodeB

Usual Smart Antenna Systems Types
 Switched Beam  Adaptive Array
 Fixed beam patterns  Adaptive Beamforming
 Static (No Tracking)  Dynamic (UE Tracking)
90°0dB 90° 0dB
120° 60° 120° 60°
-4dB -4dB

-8dB -8dB
150° 30° 150° 30°
-12dB -12dB

-16dB -16dB

180° 0° 180° 0°

210° 330° 210° 330°

Conventional
Antenna
240° 300° 240° 300°
270° 270°
 Switched Beam
 A number of antennas* with narrow beams are covering sector area
 There is either a baseband processing circuit or an RF measurement circuit
that senses which antenna is receiving more power (DoA**)
 The antenna that is chosen is switched as the active antenna for the receiver
 The monitoring system decides if switching to a new antenna is necessary

User 1
User 2

User 3

* Antennas can also be implemented as complex weights

** DOA: Direction Of Arrival
Switched Beam
Example: Antenna Diagrams for 3 Subsectors
9030
120 60
Antenna Gains:
20
150 30 12.65dB (Pilot antenna)
10 16.86dB
17.42dB
180 0 16.86dB

Front-to-Back-Ratio:
210 330
20 dB

240 300
270
Adaptive Array
 An intelligent algorithm is utilised to “steer” the
beam to follow the user
 Additionally the algorithm can be made so that it
reduces interference by creating nulls in the AoA
of the interferers
UE
Interferer
Interferer
Adaptive Array
Some Algorithm Parameters

UE 5

UE 4 UE
UE G intracell interferer
intercell interferer

UE 1 G: gain in active UE direction

N: Number of nulls (max interferers)
D D: depth of a null

UE 2

UE 3
Adaptive Array
Algorithm Performance
 The performance of the adaptive algorithm is measured in
terms of:
 Convergence speed
 Stability

 Weights of adaptive algorithm have two basic variables

that are related to the algorithm performance:
 Initial Values
 Step size
Comparison of the two systems
Switched beams Adaptive Array
Simplicity, price High gain
Usually maximizes RSCP Flexibility and ability to operate in all
environments and traffic conditions
Lower gain than Adaptive Array
Usually maximizes RSCP or Ec/Io
Complexity, price
Comparison of the two systems (2)
•Unloaded Network •Loaded Network
Implementations
Ka Kb Kc Ka Kb Kc

PA / PA / PA /
Level Comparison
Equal Equal Equal
Signal Distribution

Comparison of BER
PA /
C & I Separation
Equal

carrier n carrier n
Channel carrier 2 carrier 2 Channel
(De)Coder carrier 1 carrier 1 (De)Coder
Pros and Cons of Smart Antenna Systems
Ec/Io
High capacity gains
Improved coverage (13-17 dB div gain*)
Improved quality
Cost effective at a later network stage
Lower UE/MS power consumption
Cost
Complex systems that might have high maintenance costs
Introduce modest delays in the receiving path
Downlink beamforming for FDD difficult
Effectiveness for WCDMA still to be proven

* Performance of ArrayComm Intellicell

Smart Antennas for UE
 Still under feasibility study
 Texas Instruments: Gains of 7-9 dB have been achieved with antenna
separations of 0.15 λ or more (≥ 2.2 cm@2100MHz) *

MPRG Antenna
Array Testbed

 Philips: Significant BER improvements (10x). Minimum 0.5 λ (7.2

cm@2100MHz) – IC available for TD-SCDMA

* From a 2-years joint research report from Texas Instruments, MPRG (Mobile Portable Research
Group) and VTAG (Virginia Tech Antenna Group)
Trial Results
 Shanghai, China (Huawei)
 WCDMA
 Capacity increased by 90% and coverage by 30%

 Bristol, UK (Tsunami II Project, Orange/Bosch Telekom/Univ of Bristol

 Rural coverage extension of 54 %, 30 dB interference reduction

 Ottawa, Canada – Nortel Networks & Sprint

 Voice capacity of live network increased by a factor of 2

 ArrayComm Intellicell (GSM results)

 SNR Improvement 10 dB  DL Tx Power Reduction 90%
 Range (Area) Increase 1.9 (3.7)  Capacity Increase >2

 Giessen, Germany (Mannesmann/Ericsson)

 GSM 900 MHz, 3 BTS with smart antennas in a commercial network
 Up to 120% of capacity increase for tighter frequency reuse
Trial Results
 Metawave Communications
 Technology: CDMA IS-95
 Objective: Load Balancing

ORIGINAL AZ

OPTIMAL AZ
 Example
Adaptive Array Simulation
Questions and Discussion