Research On Parameter Estimation And Mitigation Methods Of Impulsive Noise For OFDM Systems

In some communication systems such as smart grid communications,power line communications,industrial internet of things systems,digital subscriber loop systems,etc.,impulsive noise has been considered as a kind of interference that cannot be ignored.It is generated by electronic or electrical devices and presents non-gaussian characteristics,which greatly impairs the performance of communication systems.Orthogonal Frequency Division Multiplexing(OFDM)technology is widely used in various communication systems due to its good resistance to multipath fading.However,when disturbed by impulsive noise,OFDM spreads impulses to all subcarriers,which greatly degrades the system performance.In this paper,a fractional order absolute moment-based estimator is proposed for BernoulliGaussian noise,which is used to estimate its parameters.The impact of the order is also analyzed.Finally,the performance of the fractional order absolute moment-based estimator is compared with that of traditional 1,2,3-order moment-based estimator and maximum likelihood estimator through simulations.The results show that the fractional order absolute moment-based estimator has smaller normalized mean square error than the 1,2,3-order moment-based estimator and lower computational complexity than the maximum likelihood estimator.Using the noise parameters estimated by fractional order absolute moment-based estimator,this paper continues to study the OFDM demodulation technique in the simultaneous presence of multipath fading and Bernoulli-Gaussian noise,and proposes a joint channel equalization and impulse noise mitigation method.Firstly,the state vector of Bernoulli-Gaussian noise is estimated based on the bit flipping algorithm.Then the minimum mean square error equalization is performed under the condition of the estimated state vector.Finally,the simulation compares the proposed method with some existing impulsive noise mitigation methods.The results show that the performance of the proposed method is significantly better than other impulsive noise mitigation methods when the probability of impulse occurrence is greater than or equal to 0.01.For example,when the impulse occurrence probability is 0.1,the proposed method has a BER performance improvement of 3～8 d B compared with the sparse Bayesian learning method.