Multi-dimensional dynamic trust routing for multi-attack underwater acoustic sensor networks

Underwater Acoustic Sensor Networks (UASNs) are indispensable for marine environmental monitoring, resource exploration, and maritime defense. However, when deployed in open and hostile environments, they are highly susceptible to concurrent multi-attacks that can severely compromise routing reliability and network integrity. Trust models, as an effective line of defense, play a crucial role in identifying malicious behaviors and enhancing network resilience. Yet, most existing models are tailored for single-type attacks and remain decoupled from routing protocols, leading to delayed or inadequate responses.To overcome these limitations, we propose a Multi-dimensional Dynamic Trust-based Opportunistic Routing protocol (MDTBOR) tailored for underwater multi-attack scenarios. MDTBOR adopts a reverse-analysis-based behavioral evidence extraction method to monitor node behaviors and generate real-time direct trust scores. These are combined with indirect trust from spatially distributed neighbors and historical evaluations across different temporal stages, forming a comprehensive multi-dimensional trust model. A dynamic threshold adjustment rule, integrated with Weighted Linear Combination (WLC) and Dempster-Shafer Theory (DST), ensures continuous trust updates and drives secure relay selection.Simulation results show that MDTBOR achieves high trust evaluation accuracy, with MDTM maintaining a false alarm rate below 5%. Under multi-attack conditions, MDTBOR improves the packet delivery ratio by over 20% compared with T-SAPR, while achieving lower end-to-end delay and energy consumption than other benchmark protocols. These results confirm that MDTBOR provides a robust and efficient trust-driven routing framework, significantly enhancing both security and performance in UASNs.