Interference Identification Method for UAV Based on Improved Support Vector Machine

Abstract

To improve the interference suppression ability of the UAV data link system, it is necessary to accurately identify various interference signals in complex electromagnetic environments. Support vector machine (SVM) is widely used in the field of interference identification, but its classification performance is greatly affected by parameter selection. Therefore, in order to solve the problem of low interference identification accuracy in the traditional SVM algorithm, an improved SVM algorithm based on Sparrow Search Algorithm (SSA) is proposed. The method optimizes the penalty factor and kernel function parameters of the SVM algorithm through SSA, which improves the classification accuracy and efficiency of the algorithm. Simulation results show that the identification performance of the proposed SSA-SVM algorithm is better than the traditional SVM and random forest algorithms. When the interference-to-noise ratio is 0dB, six kinds of UAV interference signals can be accurately identified. In addition, the identification accuracy is also significantly improved at low signal-to-noise ratios. For compound interference, the SSA-SVM algorithm also shows good identification performance. When the interference-to-noise ratio is greater than 2dB, 15 types of compound interference signals can be accurately identified.