Recent Advances in Dynamic Path-Controllable Deep Unfolding Network to Predict the K-Barriers for Intrusion Detection Using a Wireless Sensor Network: A Systematic Review

Abstract

Wireless Sensor Networks (WSNs) play a vital role in applications such as surveillance, border security, and intrusion detection, where maintaining reliable coverage is essential. A key challenge in WSN-based security is the prediction and maintenance of k-barriers, which ensure that intruders cannot traverse the monitored region undetected. Barrier coverage is closely associated with network resilience, aiming to minimize vulnerable gaps in detection. Recent advancements in Artificial Intelligence (AI), particularly deep learning and deep unfolding networks, have significantly improved the accuracy and efficiency of k-barrier prediction. Techniques such as Artificial Neural Networks (ANN), Convolutional Neural Networks (CNN), and Long Short-Term Memory (LSTM) networks effectively model complex relationships among parameters like sensing range, node density, and communication characteristics. Furthermore, dynamic path-controllable deep unfolding networks combine optimization principles with data-driven learning to enable adaptive intrusion detection and efficient resource utilization. Hybrid approaches, including CNN-LSTM and fuzzy neural networks, further enhance performance by capturing both spatial and temporal dependencies. This review analyses recent studies, identifies emerging trends, and highlights challenges such as scalability, computational cost, and real-time implementation, emphasizing the potential of AI-driven models for next-generation WSN security systems.