摘要
为提高估计精度,提出一种基于低复杂度的矩阵束方法(MPM)的OFDM信号超分辨率TOA/AOA二维定位方法。首先,对信道频率响应(CFR)采用MPM方法进行时延估计,得出首径(FAP)时延,即OFDM信号的TOA估计;然后,对天线接收信号的频域阵列信号使用同样的算法进行AOA/DOA估计;最后,联合TOA、AOA估计结果在二维平面内确定目标的位置。由于载波间的相关性和相干多径的影响,使AOA估计的RMSE偏差比TOA估计大,所以对TOA估计过程中的各径幅值,设置权值对非首径的幅值进行抑制,减少对后续首径AOA估计的影响。利用OFDM信号进行仿真,其结果表明,非首径幅值抑制方法能明显提高AOA估计精度和无偏性,使整体定位误差明显改善,满足单基站或接入点室内定位的需要。
For improving the prediction accuracy ot the ＇IDA and AOA parameter, this paper proposed a low complexity lVtatnx Pencil algorithm for OFDM signal＇ s super-resolution TOA/AOA two-dimensional positioning. First, the first-path delay, that was, it obtained the OFDM signal＇s TOA estimation through using MPM method to channel frequency response parameters. Then,it also gained the AOA/DOA estimation by using the same MPM method to parameters received by antenna array in fre- quency domain. Finally, the location within 2-D plane could be acquired by joining the TOA estimate and AOA estimate. Be- cause of the correlation between the carriers and the impact of coherent muhipath, the AOA estimation RMSE was much bigger than TOA＇ s. So,it added the different weights to each path＇ s amplitude in order to strengthen the FAP amplitude and decrease the other paths＇ amplitude, to lower the impact on the follow AOA estimation process. Simulations presented here employ the OFDM signal to illustrate the efficiency of the proposed algorithms. The results show that restrain to non-FAP amplitude can im- prove RMSE performance and unbiasedness of the AOA estimation, also, can improve the whole positioning process＇ accuracy significantly.
出处
《计算机应用研究》
CSCD
北大核心
2013年第2期534-536,540,共4页
Application Research of Computers
关键词
矩阵束
到达时间
到达角度
无线定位
OFDM
Matrix Pencil
time of arrivals（TOA）
angel of arrivals（AOA）
Wireless positioning
OFDM
