Reduced complexity maximum likelihood estimator for high speed wireless access-TCS-MLSCE

This paper proposes a received signal estimator with excellent compensation capability against multipath distortion, referred to the estimator as tap coefficient shift maximum likelihood symbol candidate estimator (TCS-MLSCE). The estimator adopts symbol-by-symbol maximum likelihood criterion based on both symbol candidates and decision feedback data for pre- and post-cursor estimates, respectively. It is shown that the structure can dramatically reduce the hardware size and the amount of calculation, compared to those of maximum likelihood sequence type estimators. By introducing a tap coefficient shift algorithm, the estimator improves the performance, especially in a case of a non-minimum phase. In the simulation, the degradation against DDFSE ranges less than 2.5 dB of E_b/E₀