An improved multistage parallel interference cancellation for CDMA in multipath fading channels

This paper introduces an improved multistage parallel interference cancellation (PIC) technique in which H_∞ filtering is employed to estimate MAI and ISI at the initial stage. Because the statistics of MAI and ISI are unavailable for multipath fading channels, the estimations of the MAI and ISI, especially in the earlier stages, contain significant errors. A poor estimation of MAI and ISI due to the relatively high decision bit error rate (BER) may lead to a poor cancellation or even a higher BER at the following decision stages. The H∞ filtering approach is a robust optimization from the minimax point of view, and it tends to be more effective even in the case that the co-channel interference (CCI) statistics are less valid. It minimizes the effect of the worst disturbance (including co-channel interference and additive noise). Computer simulation demonstrates that a better transmission performance can be achieved by using the H_∞ filtering algorithm at the initial stage of PIC