Instability Live Signal of Access Points in Indoor Positioning Using Particle Swarm Optimization and K-Nearest Neighbor (PSO-KNN)

Nowadays, the technology of object position estimation has developed significantly, with the existence of the fingerprint technique as a method of estimating position. This study discusses the object position estimation in the room with a value RSSI as an indicator and Access Point as a research parameter. The algorithm used to determine RSS Fingerprint in this study is Particle Swarm Optimization and K-Nearest Neighbor. Position estimation is conducted at building of Faculty of Computer Science, Universitas Sriwijaya. The estimated position results are obtained by comparing training data with testing data. The results of this study using the PSO-KNN algorithm have an accuracy rate of 70% with using 3 Access Point and 65% with using 6 Access Point. So, the reduction of parameters can become the solution of accuracy. Using of 6 Access Points which is not completely stable to check instability live signal of those combining method using Particle Swarm Optimization (PSO) and K-Nearest Neighbor (KNN).