Adaptive equalization in hybrid DS-TH CDMA system using State-Space RLS with Adaptive Memory

Modern wireless communication networks need to support a variety of data intensive applications in various medium/channel scenarios. It requires high network throughput as well as robustness against Multiuser Interference (MUI), multipath fading and other channel effects. This is achieved by using the hybrid Direct Sequence Time Hopped CDMA (DS-TH CDMA) systems. In this paper, a novel usage of State-Space Recursive Least Squares with Adaptive Memory (SSRLSWAM) with non-coherent detection is proposed for adaptive equalization of DS-TH CDMA system. The proposed algorithm is able to mitigate multipath fading effects through adaptive equalization by effectively initializing the SSRLSWAM and tuning its forgetting factor. The choice to use non-coherent detection is made because of the fact that non-coherent detectors do not utilize the phase information at the receiver side, resulting in lowering computational complexity. The BER performance of the proposed algorithm is compared with the correlating detector and Minimum Mean Square Error (MMSE) detector. It is shown that SSRLSWAM results in BER performance better than both the correlating and MMSE.