ZA-LMS-based sparse channel estimator in multi-carrier VLC system

Visible light communication (VLC) is an affordable green technology that utilizes visible light as a medium for high-speed wireless data transmission. However, performance of a realistic VLC system is limited by ambient light, user mobility and multipath between the receiver and the transmitter. Inter-symbol-interference and lowering of the instantaneous signal-to-noise ratio caused by a frequency domain spreading due to multipath and the user mobility, respectively, can be largely mitigated using recently proposed orthogonal time frequency space (OTFS) modulation. Since the delay-Doppler representation of a time-varying channel by OTFS modulation is sparse in nature, this study presents a zero-attracting least mean square (ZA-LMS) algorithm for channel estimation to exploit this inherent sparsity. In this paper, we present a formal analysis of the convergence and bit-error rate of the proposed ZA-LMS algorithm, along with supporting simulations. We compare performance of the proposed algorithm with the traditional least mean square (LMS) and orthogonal matching pursuit (OMP) algorithm. From the simulations conducted over realistic mobile random-way point VLC channel, superior mean square deviation and bit error performance of ZA-LMS-based estimator are observed over classical LMS and OMP estimator.