New Parallel Interference Cancellation/Decoding for Convolutionally Coded CDMA Systems over Fading Channel

In this paper, based on soft computation of tentative values, a new parallel interference cancellation and decoding (soft PIC/decoding) is proposed for convolutionally coded DS-CDMA systems. Two forms of soft PIC/decoding are analyzed, and compared over Rayleigh fading channel. It shows that the soft PIC/decoding achieves the dominant performance gain if the tentative decisions of coded symbols are computed in hyperbolic tangent function in terms of the sum of initial a priori log-likelihood rate (LLR) and updated a priori LLR, while a significant performance loss will occur if the tentative decisions of coded symbols are computed in hyperbolic tangent function in terms of the updated a priori LLR only. Besides, we compare the soft PIC/decoding with matched-filter (MF) receiver and conventional PIC/decoding, which shows that the soft PIC/decoding dominantly outperforms the MF receiver and conventional PIC/decodings.