A Novel LS-LMMSE Channel Parameter Tuning Approach using Particle Swarm Optimization in MIMO-OFDM

The performance of Multi Input Multi Output-Orthogonal Frequency Division Multiplexing system solely depends on the estimate of the channel. Least square (LS) and least minimum mean squared error (LMMSE) channel estimation techniques are widely used for channel estimation. Least square estimate suffers from inherent additive Gaussian noise and Inter Carrier Interference and least minimum mean squared error estimate had higher computational complexity. Due to multi antennas and channel delays the performance of channel estimation algorithm degrades. For large MIMO systems channel coefficients increases and the problem becomes multimodal. Particle swarm optimization has capability to solve problems when numbers of variables are more. In this paper, parameters of least square estimate & least minimum mean squared error estimate are efficiently used for tuning particle swarm initial parameters to optimize the performance of \pmb2x\pmb2 MIMO system. The minimum bit error rate at signal to noise ratio from LS estimate & LMSSE estimate is used as a target for swarm particles for finding optimized channel matrix. Results showed that the swarm was able to lower the bit error rate at all signal to noise ratios during multiple runs.