Data Fusion Algorithm Based on Fuzzy Similarity Weighted Least Square for Positioning with the Global Positioning System

The Global Positioning System, GPS customs solutions to determine the coordinates of the GPS receiver location and the receiver clock offset from data extracted from at least four pseudoranges. The constancy and accuracy are essential requirements in positioning calculation. The Least Squares, LS estimate has been widely used for solving GPS positioning problems. Aside its valuable properties, the LS estimate can be affected by outliers which reflect to its performance in terms of accuracy. In this paper, a new approach is applied to LS estimate to increase its accuracy and reliability. Assuming six or more satellites are observed. First, several sets of measurements are formed by making all possible combinations of observed satellites at least five satellites in each set. Second, the LS estimate approach is applied for each set of measurement to estimate the receiver position. A cluster of each set of measurements is obtained and its statistical properties mean and standard deviation are computed. Grubbs’s outlier algorithm is applied to all clusters to find the outlier measurements. The fusion of position data set is based on the fuzzy similarity between the sets of cluster position where the importance weight of each set of data is extracted. According to the proposed algorithm, software is developed using MATLAB. The proposed algorithm is tested, and the position accuracy is improved. Moreover, it reflects the efficiency and feasibility to real-time data processing and monitoring