Design of least-mean square (LMS) adaptive impulse noise canceller for intrabuilding powerline communications system

The study presents the analytical and experimental results for designing an adaptive system that cancels the periodic impulse noise exhibited by an intrabuilding powerline channel. The study uses a least-mean square (LMS) based algorithm for the design of finite-impulse response (FIR) adaptive filter that converges to the characteristics of the intrabuilding powerline channel. The characteristics of the intrabuilding powerline impulse noise are investigated both in time and frequency domain under binary-phase shift keying (BPSK) demodulation process. Results of the study show that the harmonic noise of demodulated powerline noise using two different carrier frequencies generated similar sinusoidal components thus; real-time adaptive impulse noise cancellation can be realized. The adaptive filter response provides a significant average intrabuilding powerline noise reduction of up to 12.8dB. Experiments show that the proposed system provides a significant improvement of biterror counts of up to 11000 bits compared to a similar powerline communication system without using adaptive impulse noise cancellation especially when the system uses higher transmission rates and the powerline channel exhibits high attenuation