Research on noise reduction of Internet of things based on power line using independent component analysis

摘要

In order to reduce the influence of noise in internet of things based on electric power line, an algorithm of improved robust independent component analysis (robustICA) in blind source signal processing was used. The receiver of communication adopts multiplex signals. The correlation of multiple signals was reduced by pre-whitening matrix. Multiple signals were separated by separation matrix estimated by Based on the relaxation factor of robust independent component analysis algorithm. In the end, the data signal was found by the decision threshold. The experiment of simulation of power line channel noise shows that the algorithm can remove noise effectively, SNR get up 2–3 times and the optimized algorithms improve the convergence.