Research On Channel Estimation Optimization Algorithm Of Low Voltage Power Line Carrier Communication Based On OFDM

With the development of smart power grid and information industry in China,low-voltage power line carrier communication makes full use of the existing power line network and has advantages such as no need to redistribute network lines,convenient network formation and low cost.However,due to noise interference,multipath effect and frequency selective fading when signal is transmitted over low-voltage power line channel,there are still some problems to be solved.In this paper,an improved LS-GPR channel estimation algorithm is proposed based on the OFDM low-voltage power line carrier communication system.Based on the traditional LS algorithm combined with Gaussian process regression,the windowed filter function is introduced to improve the traditional algorithm,and the pulse interference is further eliminated by nonlinear iteration.Compared with the traditional algorithm,the bit error rate in the low SNR environment is reduced to a large extent,and the interference of impulse noise is eliminated relatively fully,and the performance of channel estimation is improved.Specific research contents are as follows:(1)The key methods of OFDM technology are studied and analyzed,and the mathematical model of low-voltage power line communication system based on OFDM is established by combining the power line communication standard and OFDM technology.(2)Establishment of low voltage power line channel model.Analyze and verify the impedance,attenuation and noise characteristics of power line channel;Three kinds of background noise models and two kinds of impulse noise models are built and simulated.The modeling method of low-voltage power line channel from top down and bottom up is studied.It is further verified that the environment of power line channel is complex and changeable,and the noise has the characteristics of randomness and uncertainty,which leads to the multipath fading of the signal.(3)A comparative study of low voltage power line channel estimation algorithms based on pilot frequency.LS,MMSE,LMMSE and LMMSE algorithm based on Neumann series expansion are simulated and compared in terms of bit error rate and mean square error.Experiments show that although the influence of channel noise is not taken into account,the LS algorithm has a low computational complexity and is easy to be improved.(4)The LS algorithm based on nonlinear GPR is improved,and the threshold zero criterion is introduced into the nonlinear iterative process to optimize the algorithm twice.Gaussian process regression is used as the training model of LS algorithm data to obtain LS-GPR algorithm,and windowing filter is introduced to improve the algorithm.Aiming at the problem of pulse interference in the channel,nonlinear iteration experiments of the improved algorithm are carried out by combining the threshold zero method.The results show that the improved LS-GPR algorithm can fully eliminate the pulse interference,has a lower bit error rate than the traditional algorithm in the low SNR environment,and greatly improves the performance of channel estimation,which proves the feasibility of the proposed method.