A preliminary study on indoor positioning by angle-of-arrival approach using existing Wireless Access Points

The Global Navigation Satellite System (GNSS), such as GPS and GLONASS, directly delivers 3D coordinates of almost any points on Earth, has stimulated the rapid growth of many location based services (LBS). The weakness of the GNSS is that the radio signal transmitted from satellites is too weak to penetrate the roof and wall. Therefore, the GNSS is not applicable for indoor environments, which makes indoor positioning become a very popular research topic in recent years. Some of the image-based researches rely on the scene recognition and analysis. Due to the variety of indoor scene, these image based methods usually requires setting up specific pattern, such as QR Code, in advance. So the user must find where the pattern is before positioning. Some other researches rely on matching characteristics of the magnetic or gravity field. This approach requires the whole field map, which means the user has to walk throughout the whole place before positioning. Therefore, most of the researches look for other radio wave sources to replace GNSS, such as LTE, Wireless LAN, Bluetooth, Infra-red, i-Beacon, or ZigBee. Since Wireless Access Points (APs) has been widely deployed in many public area and each has reasonable coverage range up to 100m, it is very suitable as the positioning source. There has been several positioning algorithms based on radio wave, such as TOA (Time of Arrival), TDOA (Time difference of Arrival), RSS (Received Signal Strength), and AOA (Angle of Arrival). Both of the TOA and TDOA algorithms calculate the distance based on the travel time of the radio signal. It requires synchronous clocks on all equipment and transmitting time stamp. The Wireless APs cannot meet the requirement. The RSS algorithm compares received signal strength to existing fingerprint map. The accuracy and the stability of the fingerprint map is the key to positioning. However, the disturbance come from other sources will affect the received signal strength. Furthermore, some APs are designed to automatically adjust power according the clients, which makes the fingerprint map no longer represent the real scenario. Some research try to deploy more reference APs to produce real-time fingerprint map, but it will increase the cost and the signal traffic. To achieve indoor positioning without extra cost or modification to existing equipment, this research apply the AOA algorithm based on the space resection of the nearby APs. The direction of each nearby AP is determined while it's signal reaches strongest during the rotation of user's smartphone. The experiments are carried out in 3 different site, a classroom, a corridor, and a stadium. The results shows the potential of the proposed method, but there is still some issues need to be worked out.