Iterative multiuser detection and decoding for highly correlated narrow-band systems and heavily loaded CDMA systems

When a soft-in soft-out (SISO) iterative multiuser detector cooperates with a bank of SISO single user decoders, the multiuser system performance can be shown to converge to that of the single user system. In this paper, we propose a novel SISO iterative detector which employs a decorrelator on the output of soft interference cancellation. By making use of advantages of decorrelating detection, the performance of our proposed system is improved only with a little complexity increase compared with pure soft interference cancellation. Performance improvement is reflected in lower bit error rates at low signal-to-noise ratios and the higher convergence speed. Therefore, our iterative detector is especially suitable for highly correlated narrowband systems and heavily loaded CDMA systems. Both performance analysis and simulation results are provided to show this improvement. Finally, the computational complexity of our detector is analyzed.