为提高信号频率估计精度和扩展已有方法的适用范围,提出一种基于相位积累的多段异频等长信号的频率估计算法。首先生成异频分析参数矩阵,克服分段信号频率不等问题;其次设计相位差补偿因子,克服分段信号相位不连续问题;最后搜索相位累积频谱最大值,获得信号频率估计值。为验证算法的正确性,给出了相应的数学推导证明。仿真实验结果表明,该算法具有较好的频率估计精度,频率估计误差约为现有方法的1/2,较为接近克拉美罗下限。
To improve the frequency estimation precision of signals and extend the applicable range of existing methods, a frequency estimation algorithm of multi-section signals with the same length and different frequencies based on phase accumulation is proposed. In the algorithm, firstly, different-frequency analysis parameter matrix is generated to solve the problem of unequal frequencies among various sections of signals. Secondly, phase-difference compensation factor is designed to overcome the problem of discontinuous phases of signals. Finally, the maximum of phase accumulation spectra is searched to gain the frequency estimation. Besides, corresponding mathematical justifications are given to verify the algorithm in the paper. Simulated experimental results demonstrate that the proposed algorithm has better estimation precision, and its root mean square error is about 1/2 of existing method, which approaches Cramer-Rao lower bound.