Adaptive detection algorithms in DS-CDMA

In direct sequence (DS) code division multiple access (CDMA) systems, multiuser detection offers important performance gains compared to conventional detection. The optimum multiuser detection is computationally very complex and requires knowledge, e.g. of the signal energies and cross-correlations. In this paper two adaptive suboptimal multiuser detection algorithms are investigated in the symbol-synchronous case and tested by simulations. The algorithms are based on the classical projection theorem of Hilbert spaces and the idea is to minimize the error vector between the matched filter output and the detected information bit. Both the algorithms have computational complexity linearly dependent on the number of users. The simulations show that it is possible to achieve the multistage detector performance based only on knowledge of the matched filter output.