Study on time-frequency distribution based on Randon transformation

A new algorithm is proposed in the paper to analysis the linear multi-frequency signal. When processing multi-frequency signal, cross terms often interference with self terms. Firstly, by using cosh(t) function as kernel function to restrain the cross terms, and also introducing parameter α to control the degree of the cross terms restraining. Secondly, Radon transform is used after time-frequency analysis to filter the cross terms. The simulation and test results show the algorithm not only has higher time-frequency resolution, also can restrain the cross terms. Therefore the proposed algorithm is fit to process the linear multi-frequency signals.