Non-average performance of chaos-based DS-CDMA: a tensor approach to analytical any-order correlation of spreading sequences

This paper deals with the mathematical tools needed to optimize the performance of a chaos-based DS-CDMA system. The optimization exploits the degrees of freedom offered by the choice of a spreading sequence for each user. The impact of this choice depends on the map adopted and can be estimated by computing high-order autocorrelation functions of the spreading sequences, i.e. of the quantized version of chaotic trajectories. This computation can be effectively formulated and finalized within the tensor algebra framework proposed in this paper.