本文针对通信对抗中的跳频信号参数盲估计问题,提出了一种基于原子分解算法的跳频信号参数盲估计方法。根据跳频信号内部结构的特点,以加窗正弦函数为基函数构造了一种与跳频信号匹配的3-参数时频原子字典,将跳频信号分解为有限个3-参数时频原子的组合,最终根据这些时频原子的参数值来估计跳频信号的参数。仿真实验表明,该方法能够在未知任何先验知识的情况下,对跳频信号的跳周期、跳变时刻和跳频频率进行有效估计。
A novel algorithm is presented for blind parameter estimation of frequency hopping signals. According to the structure of frequency hopping signals, the three-parameter frequency-time atom is obtained from the unit windowed sine function. On the base of these,a atomic decomposition based parameter estimation algorithm of frequency hopping signal is developed. Using this algorithm, the parameter of frequency hopping signals i. e. as hop duration, hop timing and hopping frequencies can be estimated without any priori knowledge. Simulation results demonstrate that the method is effective and efficient.