Intelligent Power Estimation of Communication Interference in the Presence of Alpha-Stable Noise

In the present age, wireless communication systems are affected by various communication interference signals from the outside world. By estimating the parameters of communication interference signals, the prior information can be provided for the anti-interference of the system. In this paper, a neoteric intelligent power estimation method of communication interference in the presence of alpha-stable noise is proposed. First, the weighted myriad filter is used to suppress non-Gaussian noise in the obtained communication interference signal. Second, the covariance matrix is computed to obtain the network training label and network output. Then, compose the network training data set and train the deep residual network (Resnet) until it converges. Finally, input the test data and use the output of the trained network to estimate the power of the communication interference signal. Simulation results demonstrate that under low generalized jamming noise ratio (GJNR), the proposed method has better estimation performance under different alpha-stable noise characteristic parameters, which is also better than the existing method and basic network structures.