Preprocessing-based impulsive noise reduction for power-line communications

Signal blanking is a common technique for mitigating impulsive noise (IN) in power-line communications. When signal samples unaffected by IN are erroneously blanked, part of the useful signal will be lost and performance will degrade. In this paper, we show that the performance of this technique is sensitive not only to the blanking threshold but also to the signal's peak-to-average power ratio (PAPR). We thus propose enhancing the capability of the conventional blanking technique by preprocessing the signal at the transmitter. With this in mind, a closed-form analytical expression for the probability of blanking error is then derived and the problem of blanking threshold optimization is addressed. The results reveal that the proposed technique is able to minimize the probability of blanking error dramatically and can provide up to 3.5-dB signal-to-noise ratio (SNR) improvement relative to the conventional technique. Furthermore, it will be shown that if the transmitted signal's PAPR is maintained below a certain threshold, then not only can a considerable SNR enhancement be achieved but it is also possible to completely alleviate the need for any prior knowledge about the IN characteristics in order to optimally blank it.