Method for Reliable Detection of Vehicles with Location Information Errors in Spatio-Temporal Data Retention System

Some types of Internet of Things data depend on the time and location at which they were generated. We refer to such data as spatio-temporal data (STD). To effectively utilize STD, we previously proposed an STD retention system called STD-RS that uses vehicles to retain STD within a specific area. In STD-RS, vehicles autonomously operate based on their location information. However, location information errors at vehicle nodes may lead to STD being distributed outside the target area. This could lead to an increase in packet loss and pose the risk of information leakage. Therefore, this study proposes a method for detecting vehicles with location information errors based on the attenuation of signal strength with distance. Specifically, the distance to vehicle nodes and the received signal strength obtained during STD distribution are collected and stored on multi-access edge computing servers. Machine learning is then applied to this information for detection. Simulations demonstrate that vehicles with location information errors can be detected with an accuracy of approximately 80%.