A Comprehensive Approach to Intrusion Detection in IoT Environments Using Hybrid Feature Selection and Multi-Stage Classification Techniques
The rapid expansion of Internet of Things (IoT) devices has led to an increasingly complex threat landscape, challenging traditional Intrusion Detection Systems (IDS) to effectively handle the vast and diverse data generated by IoT networks. This paper presents a novel IDS that integrates Quantum-In...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10851274/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1823857127950647296 |
|---|---|
| author | G. Logeswari J. Deepika Roselind K. Tamilarasi V. Nivethitha |
| author_facet | G. Logeswari J. Deepika Roselind K. Tamilarasi V. Nivethitha |
| author_sort | G. Logeswari |
| collection | DOAJ |
| description | The rapid expansion of Internet of Things (IoT) devices has led to an increasingly complex threat landscape, challenging traditional Intrusion Detection Systems (IDS) to effectively handle the vast and diverse data generated by IoT networks. This paper presents a novel IDS that integrates Quantum-Inspired Particle Swarm Optimization (QIPSO) with Adaptive Neuro-Fuzzy Inference System (ANFIS) for optimized feature selection, followed by a multi-stage classification pipeline using Capsule Networks (CapsNets) and Attention Augmented-Recurrent Neural Networks (RNNs). The proposed approach addresses the limitations of existing methods by capturing both hierarchical and temporal dependencies in network traffic, improving the detection of subtle and advanced attacks. Evaluation on the TON-IoT and BOT-IoT datasets demonstrates that the proposed method significantly outperforms state-of-the-art techniques. Specifically, it achieves 98.83% accuracy, 98.56% precision, and 98.65% F-Measure on the TON-IoT dataset, and 98.6% accuracy, 98.5% precision, and 98.94% F-Measure on the BOT-IoT dataset, showcasing its superior performance over existing IDS models. This paper’s key contribution lies in the integration of feature selection and classification techniques tailored for IoT environments, filling a critical gap in the state of the art and offering a more adaptive and efficient solution for real-time intrusion detection. |
| format | Article |
| id | doaj-art-527112fa50034a3189d7c2bf45d3f81f |
| institution | Kabale University |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-527112fa50034a3189d7c2bf45d3f81f2025-02-12T00:01:42ZengIEEEIEEE Access2169-35362025-01-0113249702498710.1109/ACCESS.2025.353289510851274A Comprehensive Approach to Intrusion Detection in IoT Environments Using Hybrid Feature Selection and Multi-Stage Classification TechniquesG. Logeswari0https://orcid.org/0000-0002-7534-5527J. Deepika Roselind1https://orcid.org/0000-0002-3610-2979K. Tamilarasi2https://orcid.org/0000-0003-0664-984XV. Nivethitha3School of Computer Science and Engineering, Vellore Institute of Technology, Chennai, Tamil Nadu, IndiaSchool of Computer Science and Engineering, Vellore Institute of Technology, Chennai, Tamil Nadu, IndiaSchool of Computer Science and Engineering, Vellore Institute of Technology, Chennai, Tamil Nadu, IndiaSchool of Computer Science and Engineering, Vellore Institute of Technology, Chennai, Tamil Nadu, IndiaThe rapid expansion of Internet of Things (IoT) devices has led to an increasingly complex threat landscape, challenging traditional Intrusion Detection Systems (IDS) to effectively handle the vast and diverse data generated by IoT networks. This paper presents a novel IDS that integrates Quantum-Inspired Particle Swarm Optimization (QIPSO) with Adaptive Neuro-Fuzzy Inference System (ANFIS) for optimized feature selection, followed by a multi-stage classification pipeline using Capsule Networks (CapsNets) and Attention Augmented-Recurrent Neural Networks (RNNs). The proposed approach addresses the limitations of existing methods by capturing both hierarchical and temporal dependencies in network traffic, improving the detection of subtle and advanced attacks. Evaluation on the TON-IoT and BOT-IoT datasets demonstrates that the proposed method significantly outperforms state-of-the-art techniques. Specifically, it achieves 98.83% accuracy, 98.56% precision, and 98.65% F-Measure on the TON-IoT dataset, and 98.6% accuracy, 98.5% precision, and 98.94% F-Measure on the BOT-IoT dataset, showcasing its superior performance over existing IDS models. This paper’s key contribution lies in the integration of feature selection and classification techniques tailored for IoT environments, filling a critical gap in the state of the art and offering a more adaptive and efficient solution for real-time intrusion detection.https://ieeexplore.ieee.org/document/10851274/Enhanced recurrent neural networkadaptive neuro-fuzzy inference systemcapsule networksInternet of Thingsintrusion detection systemquantum-inspired particle swarm optimization |
| spellingShingle | G. Logeswari J. Deepika Roselind K. Tamilarasi V. Nivethitha A Comprehensive Approach to Intrusion Detection in IoT Environments Using Hybrid Feature Selection and Multi-Stage Classification Techniques IEEE Access Enhanced recurrent neural network adaptive neuro-fuzzy inference system capsule networks Internet of Things intrusion detection system quantum-inspired particle swarm optimization |
| title | A Comprehensive Approach to Intrusion Detection in IoT Environments Using Hybrid Feature Selection and Multi-Stage Classification Techniques |
| title_full | A Comprehensive Approach to Intrusion Detection in IoT Environments Using Hybrid Feature Selection and Multi-Stage Classification Techniques |
| title_fullStr | A Comprehensive Approach to Intrusion Detection in IoT Environments Using Hybrid Feature Selection and Multi-Stage Classification Techniques |
| title_full_unstemmed | A Comprehensive Approach to Intrusion Detection in IoT Environments Using Hybrid Feature Selection and Multi-Stage Classification Techniques |
| title_short | A Comprehensive Approach to Intrusion Detection in IoT Environments Using Hybrid Feature Selection and Multi-Stage Classification Techniques |
| title_sort | comprehensive approach to intrusion detection in iot environments using hybrid feature selection and multi stage classification techniques |
| topic | Enhanced recurrent neural network adaptive neuro-fuzzy inference system capsule networks Internet of Things intrusion detection system quantum-inspired particle swarm optimization |
| url | https://ieeexplore.ieee.org/document/10851274/ |
| work_keys_str_mv | AT glogeswari acomprehensiveapproachtointrusiondetectioniniotenvironmentsusinghybridfeatureselectionandmultistageclassificationtechniques AT jdeepikaroselind acomprehensiveapproachtointrusiondetectioniniotenvironmentsusinghybridfeatureselectionandmultistageclassificationtechniques AT ktamilarasi acomprehensiveapproachtointrusiondetectioniniotenvironmentsusinghybridfeatureselectionandmultistageclassificationtechniques AT vnivethitha acomprehensiveapproachtointrusiondetectioniniotenvironmentsusinghybridfeatureselectionandmultistageclassificationtechniques AT glogeswari comprehensiveapproachtointrusiondetectioniniotenvironmentsusinghybridfeatureselectionandmultistageclassificationtechniques AT jdeepikaroselind comprehensiveapproachtointrusiondetectioniniotenvironmentsusinghybridfeatureselectionandmultistageclassificationtechniques AT ktamilarasi comprehensiveapproachtointrusiondetectioniniotenvironmentsusinghybridfeatureselectionandmultistageclassificationtechniques AT vnivethitha comprehensiveapproachtointrusiondetectioniniotenvironmentsusinghybridfeatureselectionandmultistageclassificationtechniques |