Linear multiuser receivers for multipath fading channels

Abstract

I n t h i s p a p e r we s t u d y t h e impact of channel estimation e r r o r s o n t h e performance of line a r multiuser receivers, as well as t h e channel estimation problem itself. Of particular interest a r e t h e scalability p r o p e r t i e s of t h e channel a n d data estimation algorithms: w h a t h a p p e n s to t h e performance as t h e system b a n d w i d t h a n d t h e n u m b e r of users (and hence channels to estimate) grows ? O u r main results involve a s y m p t o t i c expressions for t h e signalto-interference r a t i o of linear multiuser receivers in t h e limit of large processing gain, w i t h t h e n u m b e r of users divided b y t h e processing gain held constant.

Large-system performance of iterative multiuser decision-feedback detection

“…In subsequent iterations , the S-DFD is replaced by the IS-DFD, where the order of the users is reversed at each iteration. The corresponding feedback error term for a finite system is (25) and depends on . As (26) The two integrals on the right represent the contribution of users and to the feedback error variance, respectively, assuming the new ordering.…”

“…The effect of multipath can be taken into account for a particular user by conditioning on the user's channel, as in [24]. With small delay spreads (i.e., ignoring the intersymbol interference), the variance of the large-system multiple-access interference can be estimated as in [24] and [25].…”

Asymptotic analysis of improved linear receivers for BPSK-CDMA subject to fading

“…Notice that when is equal to zero; indeed solving the problem (7) for is equivalent to solving (4) and (8) coincides with the classical decorrelator detector given by (4). A block scheme is given in Fig.…”

“…The proof can be obtained by applying some standard combinatorial techniques [9], but the general proof for every is very long and tedious. Another proof relies upon an application of the following Lemma from free-probability theory [15], [4] as applied to random matrices Lemma 1: Let and be a set of -dimensional Hermitian random matrices with independent of such that and then, as , for any…”

Design of MMSE multiuser detectors using random matrix techniques

“…Asymptotic analysis of DS-CDMA systems with random spreading has received much attention because it helps to understand the behavior of multiuser receivers in large (in the sense that the processing gain and the number of users are large) DS-CDMA systems [1], [2], [11]. Particular attention has been paid to the analysis of the linear minimum mean-squared error (MMSE) receiver.…”