基于主成分分析的UWB信号的TOA估计算法

TOA estimation of UWB based on PCA

提出了利用主成分分析法估计信号在超宽带(UWB)系统中的波达时间(TOA)的算法。算法先对带有噪声的接收信号的相关矩阵进行主成分分析,求出信号特征值对应的特征向量,然后把接收信号分别投影到上述的各个信号特征向量上,求投影之和。通过设置复杂度和采样频率较低的阈值,比较了主成分分析法和MEP算法的时延估计精度。仿真结果表明,主成分分析法时延估计误差小于12 cm的概率大于90%,而MEP算法的概率小于45%。

An algorithm based on PCA（ Principal Component Analysis） was presented for estimation of signal arrival time in UWB system. This algorithm analyzed the signal eigenvectors of the correlation matrix for the received signal associated with noise components, and then the received signal was projected to every signal eigenvector to get the sum of the projections, Selecting a threshold with low complexity and sampling rate, estimate precision between Multiscale Energy Products （MEP） and the proposed method was compared. Simulation results show that the probability of estimated mean absolute error which is less than 12cm exceeds 90% in the proposed method and that of MEP is lower than 45%.