A comparative study of bees colony algorithm for blind source separation

This paper proposes a novel method for blindly separating unobservable independent component signals based on the use of a bee colony optimization algorithm (BCO). It is intended for its application to the problem of blind source separation (BSS) on linear instantaneous mixtures. In this work, results obtained by BCO algorithm for solving BSS problem based on a set of cost functions are compared. These cost functions based on the fusion of two important paradigms, higher order statistics and information theory are established to measure the statistical dependence of the outputs of the demixing system. This paper demonstrates the possible benefits offered by BCO in combination with BSS, such as robustness against local minima and a high degree of flexibility in the evaluation function. Results show that the performance of the BCO is better than or similar to other evolutionary algorithms such as particle swarm optimization (PSO) with applying mutual information in combination with kurtosis on its own cost function.