MULTIPLE STAGE ANT COLONY OPTIMIZATION ALGORITHM FOR NEAR-OPTIMAL LARGE-MIMO DETECTION

In this paper, we propose a multiple stage ant colony optimization (MSACO) algorithm for symbol vector detection in large multiple-input multiple-output (MIMO) systems. The proposed algorithm uses minimum mean squared error (MMSE) solution as an initial solution in every stage, and produces a set of solutions by using the ant colony optimization (ACO) based MIMO detection. Finally, a best solution from the generated solution set is selected using the maximum likelihood (ML) metric. Simulation results show that the proposed algorithm significantly outperforms the existing ACO algorithm and some of the other MEMO detection algorithms in terms of bit error rate (BER) performance and achieves near ML performance. Furthermore, the BER performance of the proposed algorithm shifts towards single input single output (SISO) additive white Gaussian noise (AWGN) performance with increase in number of antennas which adds to the importance of MSACO algorithm for detection in large MEMO systems.