Joint iterative channel estimation and guard interval selection for adaptive powerline communication systems

The classical in-house electrical wires are identified as an attractive medium to enable broadband power line communications (PLC). In PLC systems an orthogonal frequency division multiplexing (OFDM) has been adopted as physical access scheme to cope with the multipath channel reflections at the receiver end. However, for coherent detection in OFDM the accurate channel estimates are required. Currently, the channel estimation (CE) schemes based on time division multiplex (TDM) and frequency division multiplex (FDM) pilots are foreseen, where the channel delay-time domain windowing is used for noise reduction (denoising) based on the pre-defined guard interval length. In this paper, we present and evaluate a mechanism for efficient CE with adaptive denoising algorithm designed based on mean square error criteria. The method is designed to reduce the noise in the system which is exploited by channel equalizer to improve the system's performance. The adaptive algorithm is used as a driver for the transceiver with adaptive guard interval based on the channel conditions. The algorithm performance is evaluated by computer simulation to illustrate efficiency of the adaptive PLC system.