Impact of APs removal on accuracy of fingerprinting localization algorithms

In this paper the impact of the Access Points (APs) removal on localization accuracy of Rank Based Fingerprinting (RBF) will be investigated using simulations. The RBF algorithm was previously proposed by the authors and it seems to improve performance of the traditional fingerprinting algorithms and reduce impact of the device change on the accuracy of the positioning process. In the first part of the paper the impact of APs with different Received Signal Strength values on the performance of the positioning using RBF algorithm will be investigated. Based on results achieved in the simulation, algorithm for APs removal will be proposed and tested in simulations and real world experiments. Simulations will be performed in the model created in Matlab environment. Results achieved by the RBF algorithm during will be compared to the results achieved by the traditional Nearest Neighbour family algorithms.