摘要
研究了Ad Hoc网络中移动节点的一种快速DOA估计算法,此算法并非独立估计全部时刻的DOA,而利用曲线拟合方法,从前面时刻的DOA拟合成的曲线来获得后面时刻的DOA信息,因此节省了计算时间。算法中曲线拟合的精度决定了DOA估计的准确度和分辨率,采用了代数多项式、拉格朗日插值法和最小二乘法,对3种曲线拟合方法进行了误差分析。计算机仿真结果表明,最小二乘法的精度比较高。所得出的结论对这种快速算法在Ad Hoc网络中移动节点的估计有一定的参考价值。
A fast DOA estimation algorithm of moving nodes in Ad Hoc network has been studied in this paper. Instead of using the independent estimation at different time of conditional DOA algorithms, based on curve fitting the algorithm emploies the curve deriving from the previous DOA to provide the latter information on the DOA of the nodes. The algorithem reduces the computational time. As the accuracy of curve fitting affects DOA estimation, three methods of curve fitting, Algebra polynomial, Lagrange The error analysis shows that the precision of Least Mean interpolating and Least Mean Square are evaluated Square is comparatively higher. The conclusions could be useful for the fast DOA estimation of moving nodes in Ad Hoc network .
出处
《南京邮电大学学报(自然科学版)》
EI
2007年第4期49-53,共5页
Journal of Nanjing University of Posts and Telecommunications:Natural Science Edition
基金
国家自然科学基金(60402005)资助项目