Reliability-aware Belief Functions Combination in Intelligent Transportation Systems

Intelligent Transportation Systems (ITS) play a pivotal role in modern traffic management, with road anomaly detection standing out as a critical aspect for ensuring road safety. This paper addresses the challenge of enhancing accuracy and robustness in road anomaly detection by integrating sources’ reliability within the Dempster-Shafer framework. The central contribution lies in evaluating different combination rules, including Dempster’s rule, MCR rule, and PCR6 rule, within this framework. To illustrate practical implications, we employ simulated examples and a real-life dataset. These examples serve to demonstrate the significance of considering sources’ reliability during the data fusion process. The obtained results not only emphasize the importance of reliable road anomaly detection systems in ITS but also contribute to advancing road safety by establishing a concrete connection between considering sources’ reliability and improved system efficacy.