Research On The Channel Estimation Of Power Line Communication Based On OFDM

Power line carrier communication refers to the communication through signal transmission over the power line network,has the advantages of multiple users,widely distribution,long transmission distance,high reliability,no need of rewiring and power grid construction synchronization.In order to make avoid multi-path fading effect of the power line communication,achieve the best effect of the performance of the system,we need to be able to track the changes in the response of the channel,channel estimation seems to be an ideal method.The performance of the system is directly affected by the accuracy of the channel estimation method..In this paper,the power line carrier communication channel estimation method based on OFDM is studied.Firstly,we do the research of the background and significance on power line carrier communication channel estimation.First introduces the development status and channel estimation of PLC at home and abroad,and then introduces the basic principle of power line carrier communication,analyzes the characteristics of the power line channel,including time-varying characteristics,attenuation,impedance and noise characteristics.Then,the power line carrier channel is modeled,and the channel model with 4 and 15 channels are simulated with MATLAB.The basic channel estimation method based on OFDM is introduced,and the reasoning and simulation results of classical LS algorithm,LMMSE algorithm and SVD algorithm are given.This paper introduces the basic definition of Gauss process(GP)and regression(GPR).The power line channel estimation process based on LS-GPR in [48] is given and the results are obtained by MATLAB simulation.On the basis of the algorithm using LS does not take into account the influence of the relationship between the received signal and the pilot signal and noise so as to improve the system performance of channel estimation,the MMSE is used to reduce the influence of noise.On this basis,we know that the white noise of noise subspace satisfies the zero mean Gauss distribution,ICI also satisfies Gauss distribution with zero mean,the impulse response of the useful information can be seen change more slowly than the noise component of the transformation.Here we can make full use of the IFFT transform to separate the useful information and the noise.And use MATLAB to achieve the theoretical analysis,the simulation results proved the effectiveness of the method..Based on the above,this paper uses the least squares Gauss regression process to estimate the channel.Gauss regression process using probabilistic method in the Bayesian criterion of learning data,has very strong nonlinear fitting ability and learning ability of small samples,can be made by a process of automatically obtaining the optimal hyper parameters optimization.In this paper,two improved methods are proposed,and the results show the effectiveness of the method.