Machine Learning based Intrusion Detection for Cyber-Security in IoT Networks

Abstract

IoT network is a promising technology, IoT implementation is growing rapidly but cybersecurity is still a loophole, detection of attacks in IOT infrastructures is a growing concern in the field of IoT. With the increased use of Internet of Things in different areas, cyber-attacks are also increasing proportionately and can cause failures in the system. IDS becomes the leading security solution. Anomaly based network intrusion detection (IDS) detection plays a major role in protecting networks against various malicious activities. Improving the security of loT networks has become one of the most critical issues. This is due to the large-scale development and deployment of loT devices and the insufficiency of Intrusion Detection Systems (IDS) to be deployed for the use of special purpose networks. In this article, the performance of several machine learning models has been compared to accurately predict attacks on IoT systems, the case of imbalanced classes was subsequently treated using the SMOTE technique. The Nystrom based kernel SVM is the first time used to detect attacks in the IoT network and the results are promising. The evaluation metrics used in the performance comparison are accuracy, precision, recall, f1 score, and auc-roc curve.

Anomaly Detection IDS for Detecting DoS Attacks in IoT Networks Based on Machine Learning Algorithms

“…Khatib et al [ 6 ] presented ML solutions that detect and protect systems from abnormal states. Furthermore, several ML classifiers were used to analyze the effect of data oversampling on ML models.…”

“… Author Dataset Performance Metrics Feature Selection Approach Average Detection Rate or Accuracy Aversano et al [ 1 ] Integrated dataset based on public IoT traffic traces Accuracy and F-measure Autoencoder and a hyperparameter optimization analysis The best accuracy is obtained (0:9989 for the top hyperparameter permutation). Verma et al [ 3 ] CIDDS-001, UNSW-NB15, and NSL-KDD Accuracy, specificity, sensitivity, false positive rate, and area under the receiver operating characteristic curve Random search algorithms 96.74% By using CART Khatib et al [ 6 ] UNSW-NB15 Accuracy, recall, F1 score, and ROC AUC curve Not mentioned 95% accuracy with DT, RF, and Nystrom-SVM Naung Soe et al [ 8 ] UNSW-NB 15 Processing time and detection accuracy CFS Not mentioned Khammassi et al [ 12 ] KDD99 dataset and the UNSW-NB15 dataset Accuracy, Recall, DR, FAR Wrapper, GA, and LR With a subset of only 18 features, the KDD99 dataset showed 99.90% accuracy of classification, 99.81% DR, and 0.105% FAR. Mukherjee et al [ 13 ] From the ML data repository Kaggle, which was provided by Xavier.…”

“…Based on the studies reviewed in the previous sections, we found the following: Most papers targeted the IoT network environment when designing their systems (e.g., [ 6 , 8 , 13 , 18 , 22 , 23 , 28 , 33 ]). Few papers targeted other environments, such as wireless medical sensor networks (e.g., [ 20 ]).…”

“…In turn, node heterogeneity has raised security concerns, one of the most challenging aspects of the IoT [ 5 ]. However, IoT device security cannot be achieved by employing security methods like IDS, access control, and authentication [ 6 ].…”

Anomaly-based Intrusion Detection System for IoT Application

“…We deployed a supervised ML methodology for designing and developing an anomaly-based IDS. This work used binary and multiclass classification on the datasets, including classifiers LR, LDA, KNN, CART, and SVM [103]. LR (Logistic Regression) estimates the probability of an event based on the previous data provided.…”