Indoor Localization Using Cluster Analysis.

One of the key requirements of context based systems and intelligent environments is a user’s location. Numerous indoor localization solutions have been proposed. In this paper, we propose an enhancement to an already implemented indoor localization algorithm that utilizes the JUDOCA operator to linearly find a match to an input image within a geo-tagged dataset of pre-stored images. The proposed approach is based on k-medoids cluster analysis, which is used to compare distances calculated with the same JUDOCA operator used in the original algorithm in an attempt to enhance its execution time. The results showed that the proposed approach introduced an enhancement in the execution speed of around 10 times compared to the original approach.