Research On Channel Modeling Of Low Power Line Communication Based On Chaotic And Multifractal Theory

The low-voltage power line communication (L-PLC) means transferring data withlow-voltage power line. Due to its convenience and economy, L-PLC becomes a majorconcern. However, low-voltage power line is a complicated nonlinear finite-dimensionaldynamical system. So it is difficult to build its channel model.There are some limitations in current researches. Firstly, there is no comprehensive indexto assess channel of L-PLC. Secondly, the basic theory of L-PLC analysis is linear theorywhich cannot reflect variability and non-linear of L-PLC channel.In this dissertation，chaotic and multifractal theory are applied to researching thecharacteristics of L-PLC channel. The main work of this dissertation is as follow.1. Lots of experimental data are collected by data measurement system.2. Through chaotic and multifractal analysis of L-PLC time series with differentconditions, this paper proves that L-PLC channel has characteristics of chaos and multifractal.Moreover, this paper indicates that maximum Lyapunove exponent and multifractaldimension are important indexes to describe L-PLC channel.3. Models generated by GP algorithm are often too complicated and instable. Therefore,GP algorithm is improved through simplifying the new generation. By experiments theimproved GP algorithm is proved to be more suitable for L-PLC modeling.4. The improved GP algorithm is applied to modeling L-PLC channel. The optimalmodel is chosen by maximum Lyapunov exponent. And through analyzing the optimal modelof L-PLC channel we find that relation of high dimensions is close.5. By studying increments probability distribution, and statistical self-similarity ofL-PLC channel, this paper reveals the inherent relationship between the chaotic channel andFractional Brownian Motion (FBM) model. And the multifractal model of chaotic channelbased on FBM is established and applied to reconstructing received signals.