A Deep Q Network Optimization Algorithm for DoS Attack in WSN

Wireless Sensor Networks (WSN) are frequently used to gather information based on the environmental and physical factors. Security is still a significant challenge in the design of WSN. A key security issue observed in WSN is Denial of Service (DoS) attacks. In order to classify attacks in WSN, this paper introduces an artificial Deep Q Network (DQN) method with functional link neural network (ADQN-FLNN) model. Data preprocessing is first applied to the proposed ADQN-FLNN model to make the data more understandable. Secondly, the FLNN is efficiently used to recognize and categorize the WSN intrusions. The performance of FLNN model is improved by utilizing the ADQN to tune its parameters in the best possible way. An extensive experimentation evaluation has been conducted on test data to show how the ADQN-FLNN model has produced superior results, and the findings highlighted the enhanced outcomes.