Abstract
Power Line Communication (PLC) offers a convenient and inexpensive medium for high-speed data transmission; however, this technology still faces difficult challenges. In particular, the Power Line (PL) channel suffers from multipath propagation accompanied with frequency-selective fading. This paper presents a method for constructing Low Density Parity-Check (LDPC) codes for Forward Error-Correction (FEC) on PL channel combined with Chaotic Interleaving (CI) and linear equalization to compensate for burst errors and multipath effects. First, the PL channel characteristics are introduced. A new way of noise modeling, which takes into account both background and impulse noises, is presented. Unlike Middletonʼs class A noise model, the proposed model is suitable for PLC channel. Noise bursts last no longer than 0.1 ms and they are characterized by Poisson distribution with an impulse arrival rate of 0⩽λ⩽5×10−3. In addition, we present a new algorithm for detection and removal of the impulse noise, by which a gain of 1.5 dB could be obtained. Concerning the system design, Zimmermannʼs model for PLC channel is considered, and then the proposed LDPC code construction method is presented. A Modified Shortest-Path (MSP) algorithm of low complexity is applied to obtain a code graph of the desired structure. The parity-check matrix of the code is constructed column-by-column subject to bit-degree and girth constraints. Both high rate and high girth codes have been constructed, e.g., code rates of 0.94, 0.95, 0.96, and 0.97. For better decoding of LDPC codes, an accurate estimation of Signal-to-Noise Ratio (SNR) is required. Therefore, we introduce a new method for SNR estimation based on higher-order signal statistics. The proposed method is especially beneficial at low SNRs. Furthermore, to achieve better immunity to burst and impulse noise, a novel interleaving scheme based on chaotic Baker map is proposed. Eventually, transmission of images on PL channel has been carried out using computer simulations. Clear images were reconstructed and the proposed algorithms were shown to perform well over PLC channel.