Identification of Frequency-Hopping Spread Spectrum Signals Using SVMs with Wavelet Kernels

This paper proposes a novel method of using wavelet kernel functions in Support Vector Machines (SVMs), and this method is applied to identification of individual communication transmitter which works in frequency-hopping spread spectrum modulation. The adoption of kernel function can improve the classification rate. The experimental results show how the recognition rates change with the parameters of wavelet kernel function. In a certain specific range, the classification rates maintain at a high level.