Adaptive Signal Processing Techniques for Robust, High Capacity Spread- Spectrum Multiple Access

摘要

This project is concerned with signal processing and coding techniques, which can improve the performance of spread-spectrum multiple-access systems. Specific topics investigated during the course of this project include: (1) Reduced-rank interference suppression; (2) Combined coding and interference suppression, (3) Joint transmitter-receiver optimization in the presence of multiple access interference; and (4) The effect of limited feedback on the performance of joint transmitter-receiver optimization schemes. The theory of large random matrices has been used to analyze the performance of both interference suppression and interference avoidance schemes. For example, we have used these techniques to analyze the performance of adaptive reduced- and full-rank least squares filtering for interference suppression with limited training. This analysis shows the effects of algorithm parameters, which determine the initialization and data windowing, along with system load and noise level. Other contributions include optimization of the ratio of pilot-to- data power and code rate with adaptive linear interference suppression, and signature optimization for combined interference avoidance and pre-equalization of multi-path. Transmitter optimization with limited feedback has also been considered, and bounds on the achievable performance as a function of feedback bits per dimension have been obtained. This project has resulted in a patent application for an adaptive reduced-rank filter, a paper award (for work on reduced-rank filtering), and numerous journal and conference publications on the preceding topics.