Multicode Sparse-Sequence CDMA: Approach to Optimum Performance by Linearly Complex WSLAS Detectors.

This paper investigates the performance-complexity tradeoff of the wide-sense likelihood ascent search (WSLAS) detectors in large multicode sparse-sequence CDMA. It is illustrated that when each sequence has sparsely only 16 nonzero chips, in a channel load up to 1.05 bits/s/Hz and a broad SNR region, the linearly complex WSLAS detectors can achieve the benchmark optimum BER while the complexity is significantly reduced from 0.5 times bit number to a constant less than 30 additions per bit by the sequence sparsity. The evaluation result of multiuse efficiency also shows that the sparse sequences of 16 nonzero chips can already provide a sufficient degree of freedom. [ABSTRACT FROM AUTHOR]