KNN-Based Visitor Positioning For Museum Guide System

Abstract

This study focuses on designing and implementing a Positioning System (PS) that is addressed as a component of the Location-aware Museum Guide System (GMS). The level of accuracy of the Positioning System (PS) is an important aspect in determining the suitability of the information received by the visitors. The design flow of the system begins by identifying the location of implementation. After that, choose the components to build the system. The principle used in this study is the use of existing infrastructure to reduce the cost of system development. A recent study was completed in the museum gathered information about the museum environment to assist with the design process. The system design is proposed by using WLAN technology with RSSI-based fingerprinting techniques. The algorithm used for this fingerprint technique is KNN. The addition of Access Point (AP) and AP filtering methods were also applied to improve the system performance. The test results showed that there were significant differences on accuracy level of PS among three times trial tested to the expectation of accuracy level at 1.2 meter. First trial was without additional support the existing infrastructure in the Museum is unable to provide an accurate estimating position. It was only 3.75 m. The second was by adding five APs from 6 to 11 APs, the accuracy level was 2.55 m. The last was to implement the AP filtering. It can provide improvement to 1.83 m.