Joint Bayesian the Number of Active Users Estimation and the Signatures Detection in DS-CDMA System

The estimation of the number of users and detection of the signatures in the unknown colored noise is considered. We exploit the Reversible Jump Markov Chain Monte Carlo (RJMCMC) method to solve the problem. Through introducing subspace algorithm and converting condition state spaces of the posterior distribution, the method based on Bayesian inference achieves the joint posterior probability density function of desired parameters by integrating out the nuisance. Simulation results support the effectiveness of the algorithm.