Research Of Speech Discontinuous Transmission Based On PLC In Low SNR Environments

In recent years, modern communication technology has developed rapidly, with the advent of the era of digitalization and informatization, and the construction of the smart grid, as information carrier, power line has a huge market competitiveness with the advantages of its wide distribution, cheap price and mature system. Power Line carrier Communication(PLC) uses the power line has been laid to transmit the signal, because of the low-voltage power line network covers a wide range and is necessary in people’s life, low-voltage Power Line carrier Communication(L-PLC) has become the research focus of scientific research workers at home and abroad. But because the power line network is designed to transmit electricity, and is connected to all kinds of electric and electrical equipments, most of the transmission environment is relatively poor, such as background noise and multipath fading channel, resulting in low signal noise ratio, so to research a kind of practical and effective methods of signal processing and transmission to ensure the reliable transmission of the power line data signals is very important.In this paper, in view of the serious channel noise environment, the discontinuous transmission of speech signal based on L-PLC is researched. Firstly expounds the principle of carrier communication technology and main parts of speech signal transmission system of L-PLC, the technology of modulation and demodulation, and discontinuous transmission, etc. Then the simulation analysis of low-voltage power line channel and all kinds of noises is made, and their models are established. Secondly introduces a kind of methods based on Fast Independent Component analysis(FastICA) and Voice Activity Detection(VAD) to reduce the channel noise and improve the anti-jamming ability, using the FastICA to process the noise speech signal with strong noise interference, and then do voice activity detection for the processed speech signal, and achieve the speech segments meets quality requirements.The Matlab simulation experimental results show that in low SNR environment, can achieve the conditional continuous transmission of speech signal by the above method, and improving to a certain extent the quality of the voice transmission. Finally, by the emulator try to make the algorithm to realize in DSP, the processed waveforms in the CCS environment are basically as same as the Matlab simulation results, shows the speech signal processing method is correct.