A Composite Trust and Behavior-based Secure Routing Strategy for MANETs using Network Path Inference and Layered Threat Detection

Mobile Ad Hoc Networks (MANETs) are of vital significance in infrastructure-less and quickly changing environments, i.e., military missions, disaster relief, and automotive-communication. Their dynamic and decentralized nature renders them extremely vulnerable to numerous security threats like blackhole, wormhole, and Sybil attacks. This paper presents a new framework, Trustworthy-Route-Identification and Attack-Mitigation (T-RIAM), inspired by Network-Path-Inference-Theory (NPIT), for strengthening secure routing in MANETs. T-RIAM has a three-stage architecture: (i) Network Path Inference and Trust Evaluation, where routes and nodes are rated by dynamic trust parameters like Node-Reputation-Index (NRI) and Path-Trust-Score (PTS); (ii) Attack-Identification and Route Validation, where outliers are detected by methods such as Time-of-Flight analysis, Behavioral Pattern Matching, and Multiple-Route -Verification; and (iii) Attack-Removal and Safe-Route-Selection, where safe paths are selected with adaptive route scoring and redundant routing schemes. Simulation results illustrate that T-RIAM considerably outshines current protocols such as SAODV and THR with regards to Path Trust Score (PTS), Attack Detection-Rate (ADR), Attack-Mitigation-Success-Rate (AMSR), and Packet Delivery Ratio (PDR), particularly in big network and dynamic network environments. Through the integration of trust-based assessment and real-time anomaly detection, T-RIAM presents an effective and scalable solution for secure and robust communication in MANETs.