Optimization of rank based fingerprinting localization algorithm

This paper deals with optimization of the Rank Based Fingerprinting (RBF) algorithm, which was previously proposed by the author. Results achieved in the real world experiments shown that RBF algorithm can achieve more accurate position estimate compared to some of traditional fingerprinting algorithms. Accuracy of the RBF algorithm seems to be less affected by change of device and small signal fluctuations since algorithm is based on assumption that bias and scale of measured RSS data will not affect rank of the APs. Optimization algorithm that significantly reduces computational complexity of the RBF algorithm will be introduced. Function of the algorithm was investigated in the simulations and real world experiments. Achieved results show that proposed optimization algorithm can significantly decrease computation complexity, especially when radio map database is high enough. Achieved results show that proposed optimization algorithm allows use of the RBF algorithm in applications, where the position estimate must be calculated in a very short time e.g. tracking and navigation applications.