On the various LMS Algorithms for FS-DFE with Low Hardware Complexity suitable for the High Order QAM Cable Modem

Abstract

In this paper, the performance and convergence time comparisons of various LMS algorithms that is used for coefficient update of channel equalizer and adaptive filter, are presented. We choose the optimum LMS algorithm suitable for the low complexity, high performance and high order QAM applications. The presented Fractionally Spaced Decision Feedback Equalizer accommodates 64 and 256QAM formats, and incorporates T/2-spaced feed-forward filter(FFF), feedback filter(FBF), a carrier recovery loop and an error monitor. The FFF and FBF have both16-tap filters based on various LMS(LMS, data signed LMS, error signed LMS, signed signed LMS) coefficient updating algorithms. To achieve area efficiency, time-multiplexed multiplication and tap-sharing techniques are employed. We simulated the design using SPW in AWGN channel and also severe multipath channel.