Optimization of access points deployment for enhancing indoor localization in a Wi-Fienvironment

Abstract

In the recent era, the indoor environment based navigation system has high demand. The main reason behind this is unviability of GPS based system for the indoor environment. To handle this issue, many techniques and technologies have been explored. Many types of research have been done for the reduction in positioning error in existing infrastructure, and many optimized solutions have been proposed. In these days Wi-Fi coverage is very easily available because mobile phones can easily interact with existing indoor Wi-Fi-based systems. With the optimized deployment of access points and with the help of hybrid techniques to select APs for localization, localization accuracy significantly increased and it also reduces the system overall cast, power consumption and dimensionality issue. These positioning systems need no supplementary arrangement, so they require very little amendment. In this research work, first of all, a recent development for indoor navigation systems and issues related to them are explored. Then major work starts, in which Motley-Keenan multi-wall radio propagation model has been used along with the free space propagation model. To implement the proposed model, image processing techniques like Hough transformation has been applied to the architectural floor plan for obtaining the locations and thickness of the walls. Here we have defined various wall types with different attenuating characteristics. Deployment of access points are manual, and then it has been verified with a color map. For the deployment of additional AP in an existing Wi-Fi Indoor Positioning System, Geometric Dilution of Precision (GDOP) has been used. GDOP is an important concept that has been derived from GNSS Satellite Navigation, which has been assessed in the error of a receiver's geographical position. To move forward with the selection of three required APs primarily based on Received Signal Strength Indicator (RSSI) and User-to-AP Position Vectors. So the positioning method is based on the status of received signal strength indicator (RSSI) of Wi-Fi signals and position vectors. Localization steps include the formation of an offline RSSI database, deployment of APs using Genetic Algorithm and GDOP, and an online phase consisting of a comparison of the database with RSSI measurements recorded at the test locations in order to determine their coordinates. The receivers have used RSSI and Position Vectors, that has been determined to choose an AP for their location in case of failure of one or more APs. For the improvement of accuracy and to reduce the positioning error, motley Keenan model and free space model along with RSSI and position vectors have been used. Experimental measurements have been taken out to determine the performance of the proposed algorithm.