Abstract: An algorithm to detect frequency hopping signals and estimate parameters of frequency hopping signals is presented. The proposed algorithm is based on atomic decomposition and information theoretic criterion. The stopping criteria of atomic decomposition is supplied by intercomparing the information theoretic criterion measure of each iteration to find the first local minimum. The number of valid iterations which correlates with the first local minimum equals to the number of signals in the sample. Detection and parameter estimation are performed after atomic decomposition. The proposed algorithm can be performed without any prior knowledge about signals. Numerical experiments have shown the efficiency of the proposed algorithm.