| 摘要: |
| 针对时差估计（TDE）易受噪声影响产生模糊的问题，推导了单频信号、线性调频（LFM）信号以及二进制相移键控（BPSK）信号的互相关（CC）函数表达式，以最小二乘（LS）曲线拟合细化CC函数的幅度谱，获取精确的时差估计。首先对CC函数的幅度谱进行谱峰搜索，得到时差的粗估计，然后对CC函数的幅度谱以时差的粗估计为界，根据脉内调制类型，选择相应的一次或二次曲线，进行LS曲线拟合，降低噪声项对信号项的影响，可有效排除伪峰和局部最大，获得时差的精确估计。仿真结果显示，在信噪比大于6 dB时，BPSK信号的时差估计均方根误差接近零。 |
| 关键词: 时差估计 互相关函数 最小二乘曲线拟合 |
| DOI： |
|
| 基金项目: |
|
| Time delay estimation by least squares curve fitting |
| ZENG Xiaodong |
| () |
| Abstract: |
| According to the problem of blurring in time delay estimation(TDE),a cross-correlation(CC) derivation of single frequency signal,linear frequency modulation(LFM）signal and binary phase shift keying(BPSK）signal is studied. Moreover,an algorithm by least squares(LS) curve fitting is proposed to obtain accurate TDE with the thinning CC amplitude spectrum. First,the spectral peak is searched and the rough estimation of time delay is got. Then,the order one curve or conic is chosen on the basis of intra-pulse modulation type and LS curve fitting is performed beside the rough estimation. This algorithm can eliminate false peaks and local maxima to achieve the accurate TDE. The simulation results show that when the signal-to-noise ratio(SNR) is more than 6 dB,the root mean squared error(RMSE) of BPSK TDE is close to zero. |
| Key words: time delay estimation cross-correlation function least squares curve fitting |
