Weighted combination for multiple fixes
Xinghu Zhang, Ng Hang Sun, Sharon Tan Sihong · International Conference on Information Fusion · 2012
Fusion of multiple fixes is an important in determining signal source location. Various fusion algorithms for multiple fixes and fix with bearing have been developed. However, the result from the commonly used fusion algorithm, which is normalized by error covariance matrix, is largely affected by the size and orientation of the respective error ellipse. In some applications, this could deduce an unreasonable estimation as the fused point may be largely biased from the centroid of these fixes. This paper first presents a unified fusion algorithm with weighted combination for multiple fixes, and then by taking various typical types of weights, including matrix or scale weights, we specify the formulations in these cases for different scenarios and purposes. Fusion of fix with bearing is also brought up as it is an important case to show the necessity for weighted fusion. Testing using a high-fidelity simulation system demonstrated that the algorithm well performed to realize different expectations.