Bayesian distributed blind equalization based on density-sum filters

This paper introduces a new Bayesian algorithm for blind equalization of frequency-selective channels in a setup where a single transmitter broadcasts to multiple remote receivers. The algorithm approximates some posterior probability functions, which are density mixtures with an exponentially growing number of terms, by mixtures with constant term count via a moment-matching technique. We verify via numerical simulations that the proposed algorithms exhibit bit error rate (BER) performances similar to that of particle-filtering-based algorithms while incurring in reduced internode communication cost.