Artificial Neural Network Model for Decreased Rank Attack Detection in RPL Based on IoT Networks

Musa Osman ； Jingsha He ； Fawaz Mahiuob Mohammed Mokbal ； Nafei Zhu

《International Journal of Network Security》 23卷3期 (2021/05) Pp. 496-503

https://doi.org/10.6633/IJNS.202105_23(3).15

摘要

Internet of Things (IoT) cyber-attacks are growing day by day because of the constrained nature of the IoT devices and the lack of effective security countermeasures. These attacks have small variants in their behavior and properties, implying that the traditional solutions cannot detect the small mutant variations. Therefore, a robust detection method becomes necessary. One of the common attacks is routing protocol for low power and lossy network attacks, which has not been well investigated in the literature. In this paper, we propose an artificial neural network (ANN) model for detecting decreased rank attacks, which includes three phases: Data pre-processing, Feature extraction using random forest classifier, and an artificial neural network model for the detection. The proposed model has been tested in multi and binary detection scenarios using the IRAD dataset. The results obtained are promising with accuracy, precision, false-positive rate, and AUC-ROC scores of 97.14%, 97.03%, 0.36%, and 98%, respectively. The proposed approach is efficient and outperforms previous methods of precision, recall, and F-score metrics.

關鍵字

6LoWPAN ； Attacks ； Detection Technique ； IoT ； RPL ； Security