Reduced complexity LC-LMS joint MMSE equalization and square root raised-cosine matched filter processing via constraint windowing

Recently, a new class of fractional-spaced equalizers has been developed that perform joint MMSE channel equalization and square-root raised-cosine (RRC) matched filter (MF) processing via the linearly constrained least-mean-square (LC-LMS) algorithm. While the time-multiplexing scheme employed by this joint process equalizer permits a single bank of multipliers to service the demands of both the channel inversion and MF processing so as to conserve the RRC MF FPGA real estate, the demands of an additional responsibility imposed upon the equalizer's update effectively doubles its computational workload. We present a novel modification to the LC-LMS equalizer that drastically reduces the workload required to develop the RRC MF within the digital equalizer.