Research On Key Techniques Of Low-Voltage Power Line High Rate Data Communication

Power line high rate date communication exploits the existing low-voltage distribution network as the communication channel to transfer data signals, which is applied to the broadband access, smart grid, smart home networking and enterprise internal networking. However, like all other technologies, power line communication (PLC) also faces its own set of obstacles and technical challenges. Power line is suited to convey electrical power, not information. Due to load fluctuations in the power system, PLC channels experience time-varying, frequency-selective fading in multipath propagation and represent a non-linear behavior. Moreover, transfer function of the power lines depends on the location, the topology network, the connected loads, and is impaired by the colored and impulsive noise.PLC systems by applying the technologies of orthogonal frequency division multiplexing (OFDM) can be more effective to deal with the poor PLC environment, so can realize the high rate data communication. This paper establishes channel model and noise model based on actual measurement of power line communications channel transmission characteristics, and then researches dynamic channel estimation technology and multiuser adaptive dynamic resource allocation algorithm.The main research tasks and innovations are as follows:1) A novel method for measurement power line channel transmission characteristics based on PN sequence’s excellent periodic autocorrelation properties is presented, and then the empirical multipath channel model of power line is presented from the measured results. Matching pursuit (MP) algorithm is proposed for multi-parameter identification of power line channe multi-path model. The over-complete atom dictionary is designed based on typical multi-path channel model. In order to reduce the computation complexity of MP algorithm, genetic algorithm (GA) is applied to effectively search for the best atom at each step of MP in the dictionary of atoms. The multi-parameter identification result demonstrates that the combined algorithm based on MP and GA can acquire more evaluation precision of PLC channel.2) The low-voltage power line channel noise is measured. Then the statistical analysis in time and frequency domain respectively is performed for the measurement channel noise. The basic noises model for power line channel is studied. The characteristics of the background noise and the impulse noise are researched, and then the simulations have carried on for the measured noise.3) The OFDM principle and technical characteristics is studied. The system state-space model for the measured power line channel is established. Time-varying channel estimation schemes respectively based on a joint PF algorithm and decision feedback method, and combined PF algorithm and linear interpolation method are proposed. Comparisons have also been made with the channel estimation scheme respectively based on least square (LS), Kalman filtering (KF), unscented Kalman filtering (UKF) and the proposed algorithm. For LS algorithm, the influence of noise is ignored during channel estimation process. KF and UKF algorithm are lack of the capacity for dealing with the common nonlinear and non-Gaussian problems, while PF algorithm has the powerful capacity for these problems. Therefore, the performance of the proposed power line channel estimation scheme is much better than that of the other schemes.4) The PSO_GA hybrid algorithm based on particle swarm optimization algorithm and genetic algorithm is proposed for PLC multiuser OFDM system dynamic resource allocation. The hybrid algorithm effectively fuses GA’s strong global search ability and PSO’s fast convergence property, so can improve the system’s convergence speed.