A Particle Swarm Optimization based Training Algorithm for MCMA Blind Adaptive Equalizer

An equalizer is a device that performs the reverse operation of the fading channel to mitigate the effect of inter-symbol interference (ISI) and nonlinearity. In contrast to conventional adaptive equalizers, blind equalizers do not require transmission of training symbols. Constant modulus algorithms (CMA) are a class of blind adaptive equalizers which equalize the received symbols based on the property that their magnitude (modulus) belongs to some finite number of discrete values. In this paper, a particle swarm optimization (PSO) based learning algorithm for modified constant modulus algorithm (MCMA) digital channel equalizer is proposed. Proposed equalizer does not produce any phase ambiguity as is inherent in conventional CMA and does not get stuck in local optima. Performance of the proposed equalizer is evaluated for transmission of 4-QAM signal over complex channels and compared to other blind equalizers in literature.