Multiuser detection for CDMA systems

Abstract

Spread spectrum-based code division multiple access (CDMA), has taken on a significant role in cellular and personal communications. We concentrate on direct sequence CDMA (DS/CDMA). We show that there is a natural modification of the present systems that is potentially capable of significant capacity increases. By "natural modification" we mean a modification that can be made conceptually clear, not that it is easy to implement. Indeed, the optimal multiuser detector is much too complex and most of the present research addresses the problem of simplifying multiuser detection for implementation. The objective of the article is to make the basic idea intuitive and then show how investigators are trying to reduce the idea to practice. We also indicate multiuser receiver structures with potentially acceptable levels of complexity and address potential obstacles for achieving theoretically predicted performance in practice. As a result of these investigations, an answer to the following question is expected. Is there a suboptimal multiuser detector that is cost effective to build with significant enough performance advantage over present day systems? A definitive answer is not yet available.>