Reweighted l1-VFF Modified RLS-based Channel Estimator for OFDM-IDMA Systems

In recent years, efforts have been directed towards estimation of channel state information (CSI) for Orthogonal Frequency Division Multiplexing (OFDM)-based systems. However, despite the fact that the OFDM channel is generally sparse in nature, most of the research efforts did not take advantage of the sparse nature of OFDM channel in the development of the channel estimation techniques for the system. In this paper, the sparse nature of the OFDM channel is exploited in developing an estimation scheme for OFDM-Interleave Division Multiple Access (OFDM-IDMA) wireless communication systems, a candidate for 5G wireless communication networks. First and foremost, a variable forgetting factor (VFF) mechanism is derived from general recursive least square (RLS) algorithm's a posteriori error that has been subjected to a constraint in a similar way to some previous works. Thereafter, a log-sum penalty, tagged reweighted l1 norm penalty, is applied to the conventional RLS's cost function to obtain the reweighted regularized version of modified RLS algorithm. The resulting channel estimator is named reweighted l1 variable forgetting factor modified RLS (Reweighted l1-VFFMRLS)-based estimator. The results obtained from simulations demonstrate that the reweighted l1-VFFMRLS estimator shows improved performances with negligibly high computational complexity when compared with other closely performing estimators considered in this paper.