Adaptive array smart antennas system for wireless communications

Dissertant

Mansur, Bashshar Mundhir

Thesis advisor

Jamal, Thamir M.

University

University of Technology

Faculty

-

Department

Department of Electrical Engineering

University Country

Iraq

Degree

Master

Degree Date

2013

English Abstract

A smart antenna is actually combination of an array of individual antenna elements and dedicated signal processing algorithm.

Six adaptive algorithms are proposed in this work.

The suggested algorithms fall into four approaches and offer a significantly improved solution to reduce interference levels and improve system capacity.

They provide better range or coverage by focusing the energy sent out into the cell, multi-path rejection by minimizing fading and other undesirable effects of multi-path propagation.

To evaluate the performance of proposed algorithms for mobile communication systems, an Additive White Gaussian Noise (AWGN) and realistic Rayleigh fading multipath channels are used in the simulation.

The first proposed algorithm is called Error Control-Variable Step Size (ECVSS) algorithm.

Through simulation results of smart antennas for AWGN channel, this algorithm achieves about 5 dB and 13 dB improvements in deeper null compared with LMS and NLMS algorithms respectively.

The weight stability occurs after 22 iterations in ECVSS, while weight stability starts after 100 and 30 iterations in LMS and NLMS respectively.

Moreover, when this algorithm is applied for Rayleigh fading channel, it achieves about 5 dB and 15 dB improvements compared with LMS and NLMS algorithms respectively.

The second proposed algorithm is called LMS-CGM algorithm.

Through simulation results of smart antennas for AWGN channel and Rayleigh fading channel, this algorithm achieves about 1 dB improvement compared with CGM algorithm, and starts to converge after 4 iterations while CGM starts to converge after 6 iterations.

The third proposed algorithm is NLMS-CGM algorithm.

Simulation results using AWGN channel showed that this algorithm achieves about 3 dB improvements compared with CGM algorithm and starts to converge after 2 iterations.

Moreover, when this algorithm is applied for Rayleigh fading channel, it achieves about 2 dB improvement compared with CGM algorithm and starts to converge after 4 iterations.

The fourth adaptive algorithm is ECVSS-CGM algorithm.

Using AWGN channel, this algorithm achieves about 10 dB improvement compared with CGM algorithm and starts to converge after the iteration number 2, and achieves about 4 dB improvements and starts to converge after 4 iterations when Rayleigh fading channel is used.

The fifth proposed algorithm is SMI-ECVSS, this algorithm achieves about 14 dB and 18 dB improvement in interference suppression compared with LMS and NLMS algorithms respectively.

While the LMS and NLMS algorithms start to converge from 65 and 25 iterations, the SMI-ECCVSS starts to converge from the initial iteration.

Moreover, when this algorithm applied for Rayleigh fading channel, it provides about 7 dB improvements compared with LMS and NLMS algorithms and starts to converge from the initial iteration.

Finally, the sixth proposed algorithm is the Fast Euclidean Direction Search (FEDS) algorithm.

Through simulation results for AWGN channel, the RLS and FEDS start to converge after the iteration number 7 and 8 respectively, whereas, the LMS and NLMS start to converge after 60 and 12 iterations respectively.

These algorithms generate a deeper null of about (-26 dB for LMS), (-22 dB for NLMS) and (-30 dB for both RLS and FEDS).

In the presence of Rayleigh fading channel, the LMS, and NLMS algorithms converge after 70 and 30 iterations respectively, while the RLS converge from the initial iteration and FEDS converge after the iteration number 20.

The deeper null generated is (-25 dB for LMS), (-16 dB for NLMS), (-23 dB for RLS) and (-30 dB for FEDS).

Main Subjects

Electronic engineering

Topics

• Adaptive antennas

• APA
• MLA
• AMA

Language

English

Data Type

Arab Theses

Record ID

BIM-417645