Predicting k-Barriers for Intrusion Detection Through LSTM-Based Deep Learning in Wireless Sensor Networks

This paper presents a deep learning-based mechanism to predict k-barriers in intrusion detection systems (IDSs) for wireless sensor networks (WSNs) using long short-term memory (LSTM) networks. This study targets intrusion analysis and preprocessing in WSNs, feature extraction, and the construction of a full set of features to enhance detection performance. We propose an LSTM model to predict k-barriers; this also helps in reducing false alarms and improving detection accuracy. One of the major contributions of this study is the comparative testing with other classic intrusion detection models of the proposed Deep BarrierNet based on LSTM. The proposed method achieves better and more reliable accuracy compared to traditional methods. To accelerate the training process, a correlation-based feature selection model is incorporated to identify the most relevant features for LSTM-based intrusion detection. This indicates that LSTM-like networks can efficiently predict k-barriers and will contribute to the final performance of intrusion detection for WSNs. This paper highlights the prospects of using LSTM networks for securing WSNs and proposes a scalable, reliable IDS mechanism to address current security challenges in wireless communication systems. Received: 21 April 2025 | Revised: 20 August 2025 | Accepted: 9 September 2025 Conflicts of Interest The authors declare that they have no conflicts of interest to this work. Data Availability Statement Data are available on request from the corresponding author upon reasonable request. Author Contribution Statement Ganga Bhavani Thsaliki: Methodology, Software, Formal analysis, Investigation, Resources, Data curation, Writing – original draft. Ali Altalbe: Writing – review & editing. Prasanna Kumar Rangarajan: Conceptualization, Validation, Visualization, Supervision, Project administration.