Cooperative Localization and Target Tracking With Data Exception Based on Classification

This article considers the problem of cooperative localization and target tracking (CLTT) with data exception from compromised/faulty sensors. An efficient classification approach is used to resist abnormal data from neighbors and then improve fusion performance. Unlike traditional fault-diagnosis-based methods, the proposed classification-based CLTT can be deployed and executed online without condition monitoring, achieving high real-time capability and reliability. To validate the effectiveness of the proposed algorithms, simulation examples and real-world experiments are provided.