Font Size: a A A

Nonlinear Distortion Compensation For LED In Visible Light Communication

Posted on:2019-10-05Degree:MasterType:Thesis
Country:ChinaCandidate:X D XuFull Text:PDF
GTID:2428330566999264Subject:Electronic and communication engineering
Abstract/Summary:PDF Full Text Request
As the key technology of the next generation of mobile communication network,visible light communication technology solves the problem of limited spectrum in the traditional mobile communication system,and the data rate can reach hundred-mega-level.However,similar to RF(ratio frequency)communication technology,nonlinear distortion also exists in visible light communication system.Because of the nonlinearity distortion,the relationship between the transmitted signal and the received signal is non-linear,so that it will have a greater error rate at the received side.In this thesis,new methods based on machine learning are proposed to compensate for the nonlinear distortion of LED,the details are as follows:Firstly,the model of visible light communication system is introduced.MQAM,orthogonal frequency division multiplexing(OFDM)technology,two commonly used methods of clipping are introduced,especially.Besides,according to the nonlinear characteristics of LED,several classical LED mathematical models are given.Secondly,based on the summarization of the existing LED post-distortion compensation techniques,a post-distortion compensation technique based on BP neural network and a post-distortion compensation technique based on SVM are proposed.The post-distortion compensation technique compensate for the received signal by adding a post-distortion module at the receiver side and fitting the reverse characteristic curve of the LED so that the compensated signal is linear with the signal which will be transferred by LED,and reducing the bit error rate of decoding effectively.Simulation results show that the above two machine learning methods can effectively compensate for the nonlinear distortion of the LED.Finally,on the basis of studying the post-distortion technique and summarizing the previous methods,a pre-distortion compensation technique based on BP neural network is proposed.The difference with the post-distortion technology is that the object of the pre-distortion compensation is the signal which will be transferred by LED,that is,a non-linear distortion module is added at the transfer side to compensate the signal.Similarly,the simulation results show that this method can also effectively inhibit the nonlinear distortion of the LED,thus reducing the bit error rate.
Keywords/Search Tags:visible light communication system, LED, machine learning, BP neural network, SVM, post-distortion, pre-distortion
PDF Full Text Request
Related items