A Collaborative System for Reputation-Based Security in Underwater Acoustic Networks

Although widely employed in mission-critical applications, Underwater Acoustic Networks (UANs) are intrinsically susceptible to Denial of Service (DoS) attacks, due to the broadcast nature of the underwater acoustic channel. These security breaches can severely compromise the UAN’s operational integrity and disrupt the overall network performance, requiring adequate safeguards to be put in place. Reputation-based trust systems represent one of the most effective countermeasures against these threats, protecting from a wide range of attacks by predicting the reliability of nodes’ future behavior based on historical evidence such as their level of participation in network communication. However, this type of communication-based trust evidence, collected by nodes about their 1-hop neighbors, often leads to wrong trust assessments caused by poor link quality that could be avoided by taking channel conditions into account and/or incorporating third-party reputation information when evaluating trustworthiness. In [1], a channel-based reputation system weighing observations based on channel state is presented. Work presented in [2] expands such system by employing Trust- Related Data (TRD) dissemination to make any node in a multihop UAN aware of the trustworthiness of each other. This paper is a substantial extension of [2], introducing first-hand and second-hand TRD integration and rendering the system suitable for deployment by minimizing the amount of transmitted data. Additionally, we present a more comprehensive simulationbased evaluation of the proposed security framework, including a performance analysis in a scenario involving a mobile node, tests against attacks targeting the reputation system itself (such as slandering and spoofing) and a real-world validation conducted through a dedicated sea trial.