Improved DPTE technique for impulsive noise mitigation over power-line communication channels

Signal blanking is a simple and efficient method commonly used to reduce the impact of impulsive noise (IN) over power-lines. There are two main ways to implement this method, namely, (a) the unmodified scheme and (b) the dynamic peak-based threshold estimation (DPTE) technique. Concerning the first, in order to optimally blank IN the noise characteristics must be made available at the receiver otherwise the system performance will degrade dramatically. Whereas in the DPTE case, only estimates of the signal peaks are required to achieve best performance. In this paper, however, we propose to enhance the capability of the conventional DPTE technique by preprocessing the signal at the transmitter side. To evaluate system performance, we consider the probability of blanking error (P_b), probability of missed blanking (P_m) and probability of successful detection (P_s). In light of this, closed-form analytical expressions for the three probabilities are derived which are then validated with simulations. The results reveal that the proposed DPTE technique can significantly minimize both P_b and P_m and maximize P_s. It is also shown that the proposed system is able to attain up to 3.5 dB and 1 dB SNR enhancement relative to the unmodified and the conventional DPTE techniques, respectively, as well as improving the symbol error rate performance.