Performance Evaluation Considering Real Usage in Positioning and Direction Finding System Using Radio Signals

Abstract

This study evaluates the Bluetooth low-energy (BLE) positioning and direction-finding system. The evaluation conditions are set closer to real usage; considering the number of devices, type of devices, and time interval between measurements of the database and measurements of evaluation. Subsequently, positioning and direction-finding performance evaluation experiments are performed in a real environment. The results of the comparison with a previous method show that the hybrid method including both positioning and direction-finding performs better under the conditions considering real usage. The analysis using confusion matrices reveals the trends of the direction errors. Furthermore, the hybrid method maintains the positioning and direction-finding performance and reduces the number of BLE beacon installations. Consequently, the effectiveness of the hybrid method under the evaluation conditions considering real usage is demonstrated and the importance of performance evaluation closer to real usage is shown.