NLMS-based algorithm for decision feedback equalisation

A novel block-based normalised least mean square (NLMS) algorithm with set-membership filtering criteria is proposed for symbol rate decision feedback equalisation. The introduced algorithm provides substantial reduction in filter weight updating operations and low bit error rate (BER) while keeping the convergence speed almost the same as existing algorithms for the same purpose. It is shown in the results that the proposed algorithm yields better BER performance in the presence of co-channel interference.