Application of Improved LLL Lattice Reduction in BDS Ambiguity Decorrelation
作者单位：Geospatial Information InstituteInformation Engineering University
会议名称：《第五届中国卫星导航学术年会》
会议日期：2014年
学科分类：08[工学] 080401[工学-精密仪器及机械] 081105[工学-导航、制导与控制] 0804[工学-仪器科学与技术] 0825[工学-航空宇航科学与技术] 0811[工学-控制科学与工程]
关 键 词：Ambiguity Resolution Decorrelation LLL Lattice Reduction, Gram‐Schmidt Orthogonal Transformation QR Decomposition.
摘 要：In the process of BDS navigation and positioning, fast and precise ambiguity resolution plays a key role in improving the efficiency and precision of positioning. In cases of highly correlated ambiguities, since the searching space is too large, it takes a long time to fix the ambiguities. Therefore, it is very necessary to deccorelate the covariance matrix. In essence, the decorrelation process is the closest vector problem in Lattice Theory. On the basis of proving the equality of lattice reduction and decorrelation, the LLL decorrelation algorithm is analyzed and it is improved on the basis of QR decomposition. BDS triple‐frequency data is adopted to carry the experiment. The decorrelation effect of the LLL and the improved LLL algorithm are analyzed from aspects of spectral condition number, average correlation coefficient and reduction time. It is proved that the improved LLL algorithm has better decorrelation effect although it requires longer reduction time.