Skip to content
Conference Open access

A Unified Deep Learning-Based Intrusion Detection and Prevention System for Wireless IoT Networks

Sep 2026 · 2026 International Conference on Intelligent Computing and Sustainable Innovations in Technology (IC-SIT) · pp. 1-6 · 0 citations · 12 references

Abstract

As an emerging network paradigm, the wireless Internet of Things (IoT) network are facing mounting cybersecurity threats due to its characteristic of distributed topography with open wireless communications and limited device resources. Therefore, traditional rule-based IDSs are unable to recognize novel and more sophisticated threats. To address the aforementioned challenges, in this paper, we propose a machine intelligence assisted security enforcement framework comprising of an end-to-end deep learning based intrusion detection and prevention system (IDPS) for wireless IoT networks. The network traffic is cleaned and optimized by using an autoencoder to remove redundancy in features while retaining meaningful information. We assume a CNN-LSTM for detecting both spatial and temporal attack behavior, to detect attacks that are short-lived as well as complicated. Anomaly detection: An Isolation Forest-based method is used to check for undisclosed attacks. Furthermore, a bypass prevention device automatically warns of and blocks attack traffic by confidence against an attacker attack and the real time network status can be seen through an LED system controlled by ESP32.Experiments show that the proposed CNN-LSTM model outperforms classical machine learning algorithms in terms of accuracy and precision, verifying the effectiveness of enhancing wireless IoT network security.

Read PDF

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.