Joint order and channel estimation with evolutionary algorithm

The classic adaptive channel estimation algorithms exhibit high dependence on the prior knowledge of channel order. In this paper, a joint order and channel estimation algorithm is proposed. By exploiting the excellent global optimization performance of the evolutionary algorithm, the channel order and channel impulse response are evolved at the same time with the MSE as fitness function. The experimental results on simulated channel show that the proposed algorithm outperforms the traditional LMS estimation approach in initial convergence rate as well as the tolerance upon order underestimation.