Average Linear Least Squares Positioning: Preliminary Results

In this paper, a linear least squares (LLS) estimator to solve the range-based positioning problem is investigated. We show that the LLS estimation can use a subset of or the whole available information. The former yields many single set (SS) solutions and the later yields the full set (FS) solution. We also show that the FS solution can refine the SS solutions. An experiment with phase of arrival (PoA) measurements demonstrates that the FS and SS solutions have different accuracies. The FS solution corresponds to the average of the SS solutions. The FS solution was utilized to reduce the SS positioning errors by about 8-28%.