Multiuser detection for CDMA systems

Abstract

Present day direct sequence code division multiple access (DS/CDMA) systems treat each user separately as a signal, with all ofthe other users representing interference or noise. Potentially significant capacity improvements can be achieved ifthe negative effect that each user has on the others is cancelled. In other words, joint detection (multiuser detection), rather than separate detection, is beneficial. Optimal multiuser detection is, however, too complex. Thus, suboptimal approaches have been investigated. The basic principles of optimal and suboptimal multiuser detection are well understood now and effort is underway on investigating practical and economic feasibility. • A comprehensive reference set on spreadspectrum until 1985 is [SI85].

Decision feedback multiuser detection: a systematic approach

“…Using (3) then, we have . Moreover, since this bound is valid for any by choosing the best among the upper bounds, we have (8) Since the sum of single-event probabilities is an upper (union) bound on , we can obtain a lower bound on by substituting the union bound in place of in (6). In the computation of the limit of the ratio of this union bound to , the term in the union bound that corresponds to (asymptotically) the slowest rate of exponential (in SNR) decay will dominate the union bound so that it alone determines the lower bound.…”

“…Verdú [51], Duel-Hallen [8], and Moshavi [23]). Our brief survey here offers another perspective with a discussion of a few key results while focusing on linear modulation and the Gaussian multiuser channel.…”

Data Chunking in Quasi-Synchronous DS-CDMA

“…A common rake receiver architecture designed to recover a single message is shown in figure 11 [24].…”

A method to partially suppress ISI and MAI for DS SS CDMA wireless networks

“…Problems of ISI and in particular MAI can be fought by the use of advanced multiuser detection [6]. This, however, requires considerable complexity of receivers, and is rather unlikely to be implemented for such high data rates soon.…”

“…Within the 50 m coverage area those differences may be in the order of a few code symbols (chips) depending on the data rate. This effect, referred to as Multiple Access Interference (MAI) [6], is particularly severe for very short spreading sequences, like 16-bits Walsh-Rademacher functions and 7-bits Gold codes studied in [5].…”