Date of Award

Fall 1999

Document Type

Dissertation

Degree Name

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

Department

Electrical and Computer Engineering

First Advisor

Hongya Ge

Second Advisor

Yeheskel Bar-Ness

Third Advisor

Joseph Frank

Fourth Advisor

Alexander Haimovich

Fifth Advisor

Roy D. Yates

Abstract

Wireless cellular communication is witnessing a rapid growth in market, technology and range of services. Current and future demands for wireless communication services motivate the need for handling multi-media traffic types. In a multimedia communication system, users with different and even time-varying rates and quality of services (QoS) requirements, such as voice, image and data, must be accommodated. The use of Spread Spectrum modulation with Code Division Multiple Access (CDMA) technology is an attractive approach for economical spectrally efficient and high quality cellular and personal communication services. This dissertation explores the technologies of applying different interference cancellation techniques to multi-rate CDMA systems that serve users with different QoS.

Multiple Access Interference (MAI) and multipath propagation are the major issues in wireless communication systems. It is also true for multi-rate CDMA systems. Multi-user detection has been shown to be effective in combating the near-far problem and providing superior performance over conventional detection method. In this dissertation, we combine both linear minimum mean squared error (LMMSE) detector, nonlinear decision feedback detector, with other signal processing techniques, such as array processing and multipath combining, to create effective near-far resistant detectors for multi-rate CDMA systems.

Firstly, we propose MMSE receivers for synchronous multi-rate CDMA system and compare the performance with the corresponding multi-rate decorrelating detectors. The multi-rate decorrelating detector is optimally near-far resistant and easy to implement. The proposed linear MMSE multi-rate receiver can be adaptively implemented only with the knowledge of the desired user. Due to the fact that MMSE detector offers best trade-off between the MAI cancellation and noise variance enhancement, it is shown that multi-rate MMSE receiver can offer better performance than the multi-rate decorrelator when the interfering users' Signal to Noise Ratio (SNR) is relatively low comparing to the desired user's SNR.

Secondly, the asynchronous multi-rate CDMA system, we propose multi-rate multi-shoot decorrelating detectors and multi-rate multi-shot MMSE detectors. The performance of multi-shot detectors can be improved monotonically with increasing the number of stacked bits, but a great computational complexity is going to be introduced in order to get better performance. A debiasing method is introduced to multi-rate multi-shot linear detectors. Debiasing method optimizes multi-rate detectors based on the multi-rate multi-shot model. Debiasing multi-shot MMSE detector for multi-rate signals can offer performance than the corresponding debiasing multi-shot decorrelating detector.

Thirdly, we propose linear space-time receivers for multi-rate CDMA systems. The minimum mean-squared error criteria is used. We perform a comparative study on the multi-rate receiver which uses either multipath (temporal) processing or array (spatial) processing, and the one which uses both array and multipath (space-time) processing. The space-time receiver for the multi-rate CDMA signals give us the potential of improving the capacity of multi-rate systems. The space-time processing combined with multiuser detection have the advantages of combating multipath fading through temporal processing, reducing MAI through MMSE method and provide antenna or diversity gain through spatial processing and increasing the capacity of the multi-rate CDMA systems.

Lastly, the group-wise interference cancellation methods are proposed for multi-rate CDMA signals. The non-linear decision feedback detection (DFD) schemes are used in the proposed receivers. The proposed interference cancellation schemes benefit from the nature of the unequal received amplitudes for multi-rate CDMA signals. Users with same data rate are grouped together. Users with the highest data-rate are detected first. Interference between the groups is cancelled in a successive order. The results show that the group-wise MMSE DFD yields better performance than multi-rate linear MMSE detector and multi-rate decorrelating detector, especially for highly loaded CDMA systems.

Share

COinS