Joint active users’ identification and multiuser detection in cell DS-CDMA system

Abstract

In this paper, the problem of joint Bayesian active users' identification and multiuser detection in an uncoded cell DS-CDMA system with unknown colored noise is considered. The number of active users follows the Poisson distribution and the observation vectors are considered as the compound Poisson process vitiated by the colored noise. According to the actual condition of users' calling and being called in cells, it defines five kinds of active users' state parameter space movements. We show the conditional posterior probability density function and the equation of conditional posterior likelihood ratio of multiuser detector by introducing subspace algorithm and Bayesian inference. Two-layer nesting iteration Markov Chain Monte Carlo (MCMC) method of Reversible Jump MCMC (RJMCMC) and Gibbs Sampler is used parameters estimation and multiuser detection jointly. Simulation results support the effectiveness of nesting iteration methods.