A joint algorithm of parameters estimation for frequency-hopping signal based on sparse recovery

Parameters estimation is a crucial and challenging component for Frequency-Hopping (FH) communication. Time-frequency analysis is a valid signal processing tool to estimate parameters of FH signal. However, the existing Time-Frequency analysis methods have several shortcomings such as weak suppression noise interference and feeble concentration of Time-Frequency, resulting in inaccurate parameter estimation. To tackle these challenges, we propose a novel joint algorithm by incorporating the approximation of L0 norm (AL0) and sparse linear regression (SLR). In detail, AL0 is used to obtain accurate frequency sets from the Time-Frequency representation, and SLR is used to estimate the hop period, combined with the frequency sets got from AL0 algorithm. Finally, an accurate estimation of the frequency hopping pattern is achieved. Simulation results demonstrate that the joint algorithm can effectively obtain the frequencies and hop period. Besides, the joint algorithm outperforms the SLR algorithm in estimating the period with low Signal Noise Ratio (SNR). Finally accurately recover the FH pattern.