Comparative Evaluation of Hybrid Deep Learning Approaches for Intrusion Detection in Iot Based Wireless Sensor Networks: Traditional vs. Advanced Techniques

View at publisher Summarise with VIRAT
Authors: K Dhivakar., M Jayaprakash
Cited by: 0
References: 10

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

The main goal of the research is to conduct a comparative analysis of hybrid deep learning methods of intrusion detection in the sphere of Wireless Sensor Networks (WSNs) with the aim to compare conventional methods with highly advanced and modern ones. The research will evaluate the efficiency of such strategies when applied to enhance the accuracy of detection, decrease the occurrence of false positives, and improve security in WSN settings. The paper deploys a hybrid deep learning system of convolutional neural networks (CNN) and recurrent neural networks (RNN) along with conventional artificial intelligence models like Support Vector Machines (SVM) and Decision Trees (DT). The training and evaluation are performed on datasets of the real-time traffic data of WSNs. Simulation of intrusion detection is done under various conditions, including sinkhole attacks, Sybil attacks, and selective forwarding to fully analyze the methods. The efficiency of the traditional and advanced techniques is compared by using performance metrics such as accuracy, precision, recall, and$\mathbf{F 1}$-score. The findings indicate that the sophisticated hybrid deep learning solutions are better than the conventional methods to identify sophisticated intrusion patterns in WSNs. The CNNs combined with the RNNs demonstrate a substantial uptick in the detection accuracy whereby a drop in false positives is observed. Although more effective in dealing with some categories of attacks, the traditional approaches are inadequate in dealing with the dynamic nature of the modern intrusion attempts. The paper draws a conclusion that the higher hybrid deep learning methods are more beneficial than the older ones in terms of detecting intrusion in WSNs. CNNs and RNNs offer greater accuracy, overall generalization to different types of intrusion and provide greater security to the WSN applications. These observations indicate that incorporating the highly developed machine learning algorithms within the WSN security systems can greatly enhance the defense systems against upcoming threats.

Fields of Study & Subjects

Security in Wireless Sensor Networks Network Security and Intrusion Detection Energy Efficient Wireless Sensor Networks

Publication details

Journal / source
International Conference on Communication, Computing and Electronics Systems (ICCCES)
Pages
734–739
Publication date
21 January 2026
Language
EN