Effortless 3D radio maps generation for fingerprinting-based indoor positioning system

Abstract

The estimation of user or object location in an indoor environment is in high demand nowadays. Indoor Navigation Systems (INS) have utmost importance in large shopping malls, hospitals, universities, airports, etc. Many Indoor Positioning Systems (IPS) have been developed nowadays to enable several types of location-based services. One of the most prominent techniques is the fingerprinting technique, which plays a significant role in estimating the user’s indoor location. In this technique, the fingerprinting database is generated by human experts involving fingerprint collections from indoor places, thereby making it a labor-intensive task. As a result, the calibration cost of the database becomes very high, and it affects the adaptability of this technique. To minimize the survey cost, many techniques have been proposed by researchers, such as the automated generation of a fingerprinting database. These techniques mainly revolve around 2D-based path loss heat maps and thus ignore the height information, such as the z-axis. We have proposed a systematic approach to estimate the indoor location of an object/user by generating 3D path loss heat maps. These 3D-based path loss Received Signal Strength (RSS) values incorporate height parameters as well to estimate the user/object location in an indoor 3D environment. Extensive experiments are performed, and the results outperform the manual and existing techniques by estimating the indoor position more accurately in the 3D indoor environment. Moreover, the 3D path loss RSSI heat maps have been generated without much effort by using the 3D path-loss technique, which has better accuracy than the manual technique. Using the proposed approach, the 1-meter error is obtained for 20% of the cases, and the 2-meter error in the cases 40%, which is better than the manually gathered database.