Impulsive Noise Suppression In Power Communication Systems

With the development of smart grid and the popularity of electrical equipment, impulse noise effect on the electric power communication system has become increasingly serious. Though impulsive noise in power communication has been deeply studied, the processing mechanism of impulsive noise suppression is still immature. IEEE 802.22 wireless Regional Area network(WRAN) and IEEE 1901.2 are two kinds of representative standards in power communication. We concentrate on processing method of impulsive noise in this thesis.OFDM is adopted as the modulation scheme in the IEEE 802.22 and IEEE 1901.2 physical layers. This thesis focuses on the OFDM-based power communication system. The interference of impulsive noise on power communication system is deeply analyzed, and the schemes of impulsive noise mitigation are designed. The main contributions of this thesis are summarized as follows:Firstly, the characteristics of impulsive noise in power line communication and WRAN communication are researched. Three types of impulsive noise models, i.e. Bernoulli-Gaussian, Middleton class A and KATA model based on power line, are studied. The mathematical expression and parametric characteristics of the model are analyzed in detail. The time-domain sequence and the statistical distribution of the noise model are simulated. The noise model is transplanted into the channel module of simulation platform.Secondly, aiming to solve the problem of waveform distortion caused by nonlinear processing in time domain in WRAN communication, the impulsive noise suppression algorithm based on Blanking nonlinear combined with hard decision feedback compensation is proposed. First of all, the traditional suppression of Blanking and Clipping methods are introduced. The SNR gain of these two kinds of methods in Bernoulli-Gaussian model is deduced and modified in detail. Therefore the optimal threshold of nonlinear processing is presented. According to the technical characteristics of OFDM and theoretical analysis of the Blanking processing, the time-frequency joint suppression algorithm is proposed. The most prominent feature of this algorithm is that it can suppress the impulsive noise of different power, and can compensate the waveform distortion caused by Blanking, which greatly improves the system’s anti-interference ability.Thirdly, the noise processing method based on interleaving technique in PLC system is studied. On the basis of IEEE 1901.2, the characteristics of FFT parallel data processing are used to improve the performance of system in two aspects. First, in order to prevent impulsive noise from spreading into the system via FFT and interleaving, the Clipping algorithm, which is implemented before the demodulation of receiving signal, can reduce the influence of impulsive noise. Secondly, on the basis of the frequency domain interleaving in the IEEE 1901.2 standard, the time domain interleaving algorithm is added to form a double interleaving, which can fully whiten the impulse noise, so as to realize the performance optimization.