Lightweight Intrusion Detection in MQTT Based Sensor Network
- Resource Type
- Conference
- Authors
- Jaafar, Fehmi; Malik, Yasir; Serre, Johan; Wang, Haoyu; Wang, Tianqi
- Source
- 2022 International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME) Electrical, Computer, Communications and Mechatronics Engineering (ICECCME), 2022 International Conference on. :1-9 Nov, 2022
- Subject
- Bioengineering
Communication, Networking and Broadcast Technologies
Components, Circuits, Devices and Systems
Computing and Processing
Engineered Materials, Dielectrics and Plasmas
Fields, Waves and Electromagnetics
General Topics for Engineers
Nuclear Engineering
Photonics and Electrooptics
Power, Energy and Industry Applications
Robotics and Control Systems
Signal Processing and Analysis
Transportation
Runtime
Protocols
Intrusion detection
Support vector machine classification
Telecommunication traffic
Feature extraction
Internet of Things
IoT
MQTT
Intrusion Detection
Cyberattacks
- Language
The rapid development of IoT devices improves the quality of life and efficiency of industries. However, it also brings the attention of cybercriminals. Many IoT devices do not have enough resources to deploy a standard intrusion detection system and thus there is a need to design an IDS that is lightweight and has a narrower focus on the types of attacks that are specific to IoT devices. Previous studies have suggested a few ways to improve the efficiency of the IDS. However, most of them use datasets that only contain general network traffic but not IoT protocol-related traffic. Our proposed approach focuses on improving the efficiency and performance of a machine learning-based intrusion detection system using a support vector machine that targets attacks within an IoT system that uses Message Queuing Telemetry Transport (MQTT) protocol. We use feature selection techniques to reduce the complexity of the model and evaluate the result using multiple metrics.