An Improved United Compass Ambiguity Decorrelation Algorithm

Abstract

An improved unity ambiguity decorrelation algorithm of Beidou was proposed due to the drawback of unity ambiguity algorithm in high-dimensional calculations. The algorithm will transform vector matrix K through rounding one by one, and have a descending order according to the size of the diagonal elements after the end of iteration. Comparing the new algorithm with the standard unity decorrelation algorithm through simulation, the results show that the new algorithm can achieve smaller condition number and larger decorrelation number no matter in low-dimensional or high-dimensional calculations, which can solve the problem of ability decreased of decorrelation in the original algorithm. At the same time, compared with LDLT algorithm based on sorting, the effect of decorrelation of the new algorithm is smaller and less time consuming.