Comparison of adaptive algorithms performances for a multistage Recursive Successive Interference Cancellation (RSIC) multiuser detection strategy

Horizontal-linked wireless acoustic communication is hampered by many environmental factors. Apart from coping with the problems of Doppler effect, multipath propagation fading, phase fluctuations and intersymbol interference (ISI), the receiver has the added task of mitigating the effect of multiple access interference (MAI) from other users in Shallow Water Acoustic Networks (SWAN) communication. Recently, a novel technique of multiuser detection strategy based on Recursive Successive Interference Cancellation (RSIC) was proposed and successfully implemented in several sea-trials. The adopted adaptive algorithm is a driving force that is of particular importance. This paper compares the performances of the Variable Step-size Least Mean Square (VSLMS) and the Improved Recursive Least Square (IRLS) algorithms over the conventional LMS and RLS algorithms for adaptation by the RSIC receiver structure. The algorithms were accessed with trial experimental data for their rate of convergence, tracking performance and computational complexity.