Blind signal detection and parameter estimation for an asynchronous CDMA system with time-varying number of transmission paths

In this paper, we propose algorithms to estimate blindly parameters and input sequences for an asynchronous multipath and multi-user code division multiple access (CDMA) system. We model the communication system as a frequency selective fading channel with multipath propagation. The number of active propagation channels is unknown and time-varying. Estimates of input sequences and unknown parameters, such as, for example, time delays and the number of multipaths, are computed via a variant of the expectation maximization (EM) algorithm, known as the space-alternating generalized expectation maximization (SAGE) algorithm. The proposed iterative scheme has a computational cost that grows linearly with the number of users and propagation paths. Computer simulations illustrate the performance of our detection and estimation method. The errors of the channel estimates are compared to the Cramer Rao lower bounds (CRLB).