Decision-feedback soft-input/soft-output multiuser detector foriterative decoding of coded CDMA

摘要

The optimal decoding scheme for a code-division multiple-accessn(CDMA) system that employs convolutional codes results in a prohibitivencomputational complexity. To reduce the computational complexity, anniterative receiver structure was proposed for decoding multiuser data inna convolutional coded CDMA system. At each iteration, extrinsicninformation is exchanged between a soft-input/soft-output (SISO)nmultiuser detector and a bank of single-user SISO channel decoders.nHowever, a direct implementation of the full-complexity SISO multiuserndetector also has the exponential computational complexity in terms ofnthe number of users. This paper proposes a low-complexity SISO multiuserndetector based on tentative hard decisions that are made and fed backnfrom the channel decoders in the previous iteration. The computationalncomplexity of the proposed detector is linear in terms of the number ofnusers and can be adjusted according to the complexity/performancentradeoff. Simulation results show that even with this simple feedbacknscheme, the performance of the coded multiuser system approaches that ofnthe single-user system for moderate to high signal-to-noise ratiosn(SNRs)