Improving Soft Decoding by Spectral Leakage Reduction in Presence of Narrow Band Interference in PLC

Communications over power lines (PLC) is a promising technology for a variety of applications. The use of a single network for power and data transmission with PLC provides weight, space, and cost savings for vehicular communications. Current work focuses on broadband PLC (BPL) systems compliant with the IEEE 1901 standard. Given that transmissions occur on a physical medium not designed for data communication, the channel and noise characteristics are mostly unfavorable. To deal with these conditions, turbo codes have been widely employed due to their high performance. However, their performance strongly depends on the accuracy of the noise estimation. This work addresses the estimation of the noise power spectral density (PSD) to perform soft-decoding. A general background noise (GBN) model and a narrowband interference (NBI) model is considered. To the best of our knowledge, this is the first work that analyzes the effects of the NBI leakage on noise estimation when it is performed during the reception of the data frame. We propose the use of the frame control (FC) symbol to reduce interference leakage on noise estimation. In addition, we investigate the use of payload symbols to further reduce spectral leakage. Finally, we show that system performance considerably improves by using the FC-based estimation in soft-decoding procedures.