Intelligent Cyber Security Model for Network Level and Node Level in WSN Using Adaptive Neural Network

Interconnected sensor nodes in wireless sensor networks (WSN) make data exchange easier. WSN requires smooth coordination and communication between different nodes at the network level to enable efficient data routing, scalability and network optimization. Different sensors help with data collection, processing and transmission at the node level while taking care of certain tasks. In this research, our goal is to improve an innovative cyber security model for securing the network level and node level in WSN. The research proposed the Giraffe Kicking Optimized Deep Belief Network (GK-DBN) methodology for enhancing security practices in WSN environment. Initially, we collected a dataset containing a variety of network traffic, which provides an effective structure for assessing cybersecurity solutions. The data cleaning process was conducted to pre-process the gathered raw data; it enhances the quality and mitigates the redundancy. Giraffe Kicking Optimization (GKO) is employed to improve the crucial characteristics of the proposed DBN architecture. The suggested security system aims to improve the cyber security system with network level and node level in WSN scenarios. The result analysis phase is performed with various metrics including encryption time and decryption time. The comparison analysis is conducted with various existing approaches to determine the effectiveness of the proposed security model. The experimental findings demonstrate that the recommended model performed better than other conventional approaches in terms of recall (97.70%), precision (99.3%), accuracy (99.1%) and f1-score (98.70%), for providing an enhanced cyber security system to secure WSNs.