A joint-optimization NLMS algorithm with linear function approximation penalty for sparse channel estimation

In this paper, a linear function approximation penalty is introduced into the joint-optimization normalized least mean square (JO-NLMS) algorithm to utilize the sparseness of the multi-path channels in the hilly terrain communication environment. The proposed algorithm is denoted as linear function approximation (LFA) penalty JO-NLMS algorithm (LFAJO-NLMS), which is realized by integrating the LFA into the JO-NLMS's cost function. The proposed LFAJO-NLMS algorithm is mathematically given in detail and its channel estimation behaviors are confirmed over a multi-path model. The computer simulation results gotten from the sparse channel estimation show that our developed LFAJO-NLMS algorithm performs better than the early reported NLMS and its corresponding sparse versions.