The Restraining Method Research Of Power Line Communication System

With the deeply extending of the smart grid in the nationwide promotion, so power distribution network automation system is very important. Communication is regarded as a key technology in power distribution network automation system, and it seems crucial. Because PLC technology is an advanced technique that is not only low cost, but also convenient operation. But noise exists in PLC system will affect the communication quality. In order to realize reliable communication, we need to study noise in PLC system. And we need to put forward effective attenuation noise method.In view of the OFDM system based on PLC, we firstly analyze background noise and random impulse noise in this paper. According to the statistical parameters that are obtained from analysis, noise is modeled and simulated by MATLAB. Moreover, in order to improve the reliability of PLC system, it is proposed that the decision feedback estimation method of noise based on SBL in this paper. This algorithm is different from other methods. Because the noise statistical model in this algorithm isn’t assumed, it needn’t to be trained. This algorithm has three steps. Firstly, the noise of the PLC system is projected to null and pilot tones. Secondly, we add the information in the data tones to perform joint noise estimation and symbol detection. Finally, we use decision feedback from the decoder to enhance the accuracy of noise estimation. Noise estimation is subtracted from received signal, we will obtain reliable transmit information.The noise mitigation effect of this algorithm is verified based on OFDM system of PLC by MATLAB. We can compare the proposed methods with conventional mitigation noise method based SBL that only uses null tones. According to the simulation result, the proposed methods achieve SNR of up to 8dB in coded systems for random impulsive noise. And it obtains SNR gains of up to 3dB in coded system for background noise. In addition, we also analyze BRE of OFDM system, thus we prove the feasibility and effectiveness of this algorithm.