一种低复杂度线性调频信号参数估计算法

摘要

A quadratic estimation algorithm is proposed to reduce the complexity of accurate Linear Frequency Modulation (LFM) parameter estimation. First, the frequency rate and initial frequency are estimated coarsely by short time coherent integral and incoherent accumulation. Then, the parallel Partial Matched Filters combined with FFT (PMF-FFT) and quadratic interpolation are utilized to estimate the residuals of the frequency rate and initial frequency. Last, the final estimated values are obtained by synthesizing the results of both estimations. Simulation shows that the proposed algorithm has a low SNR threshold, and the accuracy is close to Cramer-Rao Lower Bound (CRLB). The complexity and hardware consumption of the proposed algorithm are much less than the frequency rate test algorithm and joint estimation algorithm based on interpolation.%为降低线性调频(LFM)信号参数估计的复杂度，该文提出一种二次估计算法。首先通过短时相干积分与非相干累加对频率斜升和起始频率进行预估计；然后以预估计结果为中心，利用多路并行部分匹配滤波快速傅里叶变换(PMF-FFT)和二次插值对参数进行精确估计；最后综合预估计和精确估计结果得到参数的最终估计值。仿真结果表明，该算法信噪比门限较低，估计精度接近克拉美罗下界，其计算复杂度和资源消耗均远低于频率斜升试探算法和插值联合估计算法。