A chaotic spreading code and its application to blind channel estimation

The proposed system introduces a new approach to chaotic-based signal transmission and blind channel estimation. It is now well known that channel estimators based on FIR-digital adaptive-filters (FIR-DAF) approach the upper bound to their performance when driven by samples of chaotic signals. A fixed chaotic-system parameter is assigned to every m-valued information symbol and thus after symbol detection the chaotic sequence can be reconstructed. Assuming proper symbol synchronisation, the complete transmitted sequence (which is in general deteriorated by noise) can be reconstructed and employed as drive signal for the blind channel estimation algorithm. The performance of the proposed system is evaluated for a channel with low pass characteristics and a 10 MHz bandwidth. Channel estimation is carried out using the finite correlation least squares (FCSL) algorithm recently introduced by the authors