Advancing wireless sensor network security and efficiency through CNN-GSWO, and stochastic gradient descent intrusion detection
S. Murugan, R. Suganya, R. Ramesh · 2025
A comprehensive approach to bolstering security and efficiency in wireless sensor networks (WSNs) by amalgamating cutting-edge techniques. By using Convolutional Neural Network with Group Search Wolf Optimization (CNN-GSWO) and Stochastic Gradient Descent (SGD) for intrusion detection, this framework aims to tackle the evolving challenges of WSN security. CNN-GSWO enhances the network&s;s intrusion detection capabilities by effectively extracting features from sensor data and optimizing parameters to improve detection accuracy. Additionally, the integration of SGD provides a robust mechanism for detecting anomalies in WSN traffic, further strengthening the network&s;s resilience against potential threats. The integration of SGD further strengthens the system by refining the detection model in real-time, allowing it to quickly adapt to new types of threats and anomalies in WSN traffic. By harnessing the power of machine learning and optimization techniques, the proposed framework enables WSNs to detect and mitigate intrusions more effectively while minimizing false positives. Furthermore, the integration of GSWO ensures that the intrusion detection system is optimized to adapt to changing network conditions and attack scenarios. An experimental result achieves 98% of accuracy. Overall, this research presents a promising avenue for addressing the pressing security concerns facing wireless sensor networks and advancing their capabilities in safeguarding critical infrastructure and data.