Particle swarm optimization for MIMO receivers

MIMO receivers need to have good knowledge about channel. To get good channel estimation needs to good pilot sequences and good channel estimator. The good pilot sequences can ensure the channel estimator converge the channel parameters by short length of pilot sequences, while good channel estimator can employ fewer iteration to estimate channel parameters. To reach the first goal, this paper proposes a design of pilot sequences for MIMO channels, only with ML training symbols; to reach the second goal, this paper proposes a PSO channel estimation algorithm. Different from the known PSO algorithms, in the proposed PSO estimator algorithm, the ‘particle positions’ are virtual, they represent the MIMO weights matrices, and the ‘particle velocity’ is also virtual, it represent the updating increment of MIMO weights matrices, which simplifies the PSO algorithm's implementation. In the iteration of the estimator, the proposed PSO algorithm finds desired MIMO weights matrix through the interaction of individuals in a population of weights matrices. Simulations show that the BER performance of the proposed PSO MIMO equalizer is better than those of conventional adaptive equalizers.