Particle Swarm Optimization for Antenna Selection in MIMO System

This paper investigates the receive antenna selection problem to maximize capacity in wireless MIMO communication system, which can be formulated as an integer programming optimization problem and can not be directly solved because of its non-convex characteristics caused by the discrete binary antenna selection factor. To deal with this challenge, a computationally efficient approach, particle swarm optimization(PSO) algorithm is introduced, in which the particle is defined as the discrete binary antenna selection factor and the objective function is associated with the capacity corresponding to the specified antenna subsection represented by the particle. Furthermore, in order to meet the condition that the number of selected antennas should keep fixed, the particle elements are relaxed to change between [0 1] and the position of the higher elements are taken as the index of the antenna subsection to be activated. Then the best antenna subset can be found by seeking the global optimal particle in PSO. Numerical results reveal that PSO algorithm exhibits a promising performance when applied to both the classical benchmark function and our antenna selection scenario.