A novel approach of canceling cyclostationary noise in low-voltage power line communications

Narrowband noise (NBN) and periodic impulsive noise (PIN) are two of the most disturbing interferers in low-voltage (LV) power line communication (PLC). They exist commonly in different scenarios and degrade the performance of PLC systems significantly. Accurate estimation of the NBN and the PIN are important for the power line channel characterization and can help improve the quality of service (QoS). In contrast to other noise classes, both NBN and PIN may exhibit strong cyclic behaviors. In this work, the cyclic properties of both of them are investigated. Based on the results, an universal algorithm is developed to detect and cancel narrowband and periodic impulsive components from PLC signals. The proposed method is executed purely on the physical layer and does not require specialized hardware. Its performance is verified through simulations and real measurements.