PROJECT TITLE :

A Trust Update Mechanism Based on Reinforcement Learning in Underwater Acoustic Sensor Networks

ABSTRACT:

Underwater acoustic sensor networks, also known as UASNs, have found widespread use in a variety of marine contexts, including marine military operations, auxiliary navigation, and offshore exploration. Traditional security mechanisms are not applicable to underwater sensor nodes (UASNs) because of the limitations that underwater sensor nodes have in the areas of communication, computation, and storage. Recently, a number of different trust models have been researched for their potential as useful tools in the process of making UASNs more secure. However, the existing trust models do not have flexible trust update rules. This is especially problematic when considering the unavoidable dynamic fluctuations that occur in the underwater environment as well as the extensive variety of possible attack modes. The purpose of this investigation is to propose a novel trust update mechanism for UASNs that is based on reinforcement learning (TUMRL). The plan is formulated through a series of three stages. First, an environment model is created to quantify the impact of underwater fluctuations in the sensor data, which helps in updating the trust scores. This is done by using the data from the model. Then, the definition of key degree is presented; during the process of updating trust, nodes with a higher key degree react more sensitively to malicious attacks, thereby improving the network's ability to protect important nodes. In conclusion, a brand new trust update mechanism that is based on reinforcement learning is presented. This mechanism is designed to resist shifting attack modes while simultaneously achieving efficient trust update. The findings of the experiments demonstrate that the proposed method has a performance level that is adequate for enhancing the effectiveness of trust update and the safety of the network.