Date of Award

Fall 2011

Document Type

Dissertation

Degree Name

Doctor of Philosophy in Electrical Engineering - (Ph.D.)

Department

Electrical and Computer Engineering

First Advisor

Yeheskel Bar-Ness

Second Advisor

Alexander Haimovich

Third Advisor

Ali Abdi

Fourth Advisor

Osvaldo Simeone

Fifth Advisor

Laurence B. Milstein

Abstract

The multipath nature of the wireless channel, results in a superposition of the signals of each path at the receiver. This can lead to either constructive or destructive interference. Strong destructive interference is frequently referred to as deep fade and may result in temporary failure of communication due to the severe drop in the channel's signal-to-noise ratio (SNR). To avoid this situation, signal diversity might be introduced. When having more than one antenna at the transmitter and / or receiver, forming a Multiple-Input Multiple-Output (MIMO) channel, spatial diversity can be employed to overcome the fading problem. Space time block codes (STBC) have been shown to be used well with the MIMO channel. Each type of STBC is designed to optimize a different criteria such as rate and diversity, while other characteristics of the code are its error performance and decoding computational complexity. The Orthogonal STBC (OSTBC) family of codes is known to achieve full diversity as well as very simple implementation of the Maximum Likelihood (ML) decoder. However, it was proven that, with complex symbol constellation one cannot achieve a full rate code when the number of transmitting antennas is larger than two. Quasi OSTBC are codes with full rate but with the penalty of more complex decoding, and in general does not achieve full diversity.

In this work, new techniques for OSTBC transmission / decoding are explored, such that a full rate code can be transmitted and decoded with linear complexity. The Row Elimination Method (REM) for OSTBC transmission is introduced, which basically involves the transmission of only part of the original OSTBC codeword, resulting in a full rate code termed Semi-Orthogonal STBC (SSTBC). Novel decoding scheme is presented, such that the SSTBC decoding computational complexity remains linear although the transmitted codeword is not orthogonal anymore. A new OSTBC, that complies with the new scheme's requirements, is presented for any number of transmit antennas. The performance of the new scheme is studied under various settings, such as system with limited feedback and multiple antennas at the receiver.

The general decoding techniques presented for STBC, assume perfect channel knowledge at the receiver. It was shown, that the performance of any STBC system is severely degraded due to partial channel state information, results from imperfect channel estimation. To minimize the performance loss, one may lengthen the training sequences used for the channel estimation which, inevitably, results in some rate loss. In addition, complex decoding schemes can be used at the receiver to jointly decode the data while enhancing the channel estimation. It is suggested in this work to apply adaptive techniques to mitigate the performance loss without the penalty of additional rate loss or complex decoding. Namely, the bootstrap algorithm is used to further refine the received signals, resulting in better effective rate and performance in the presence of channel estimation errors. Modified implementations for the bootstrap's weights calculation method are also presented, to improve the convergence rate of the algorithm, as well as to maintain a very low computational burden.

Share

COinS