A segmentation-based matching algorithm for magnetic field indoor positioning

Abstract

Magnetic field-based location fingerprinting techniques are emerging technologies used in indoor navigation that take advantage of magnetic field anomalies. k Nearest Neighbours (kNN) is one of the general matching algorithms that is widely used in fingerprint-based indoor positioning systems to estimate the location of users. However, the standard kNN algorithm always visits all the data in a database in order to take the appropriate nearest k neighbours into account while calculating the estimated location. One of the key disadvantages associated with kNN is the fact that computational complexity is quite large. In order to deal with this issue and improve the precision of this method, this paper proposes the use of a new method called Segmentation-based kNN algorithm. This approach conducts suitable selection and partitioning on the target positioning area before calculating the kNN. We have calculated the accuracy rate of the proposed algorithm and compared it with standard kNN algorithm, and the results show that the proposed algorithm performs better than the kNN algorithm with an improvement of 9.24% in average accuracy.