We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Deep learning model-based detection of jamming attacks in low-power and lossy wireless networks.
- Authors
Jayabalan, E.; Pugazendi, R.
- Abstract
Communication in the wireless sensor networks could be disturbed by the jammer and this jamming attacks could be identified as a distinctive type of Denial of Service attacks. This might result in the degradation of the network which could be recognized. Due to the jamming signals, the packet transmission might not be proper. These kinds of attacks are destructive in low power and lossy wireless network due to the attributes such as disruption in communication and rapid trench in the batteries. To detect the attack and to perform the secure data transmission, the illegitimate nodes have to be eradicated. Machine learning and deep learning methods are used for network analysis of intrusion detection and security. Hence, the deep learning model is proposed to identify the attacks which result in secured data transmission. Also, the logistic regression is applied to classify the behaviour of the node so that the deviation could be determined. Deep learning is applied so it can adaptively learn the attacks and classify with higher accuracy. It is efficient as it is adaptive in learning and has improved precision. The performance of the model could be proved with simulations by varying the nodes. The secured data transmission could be proved by analysing several metrics such as packet drop rate, normalized routing overhead, and jitter. The packet drop is decreased to 50%, jitter is decreased to 6% and throughput shows a significant increase by 6% when compared with the Dodge-Jam, a lightweight Mitigating Stealthy Jamming Attacks (MJSA) technique.
- Subjects
RADAR interference; DEEP learning; DENIAL of service attacks; WIRELESS sensor networks; DATA transmission systems; MACHINE learning; WIRELESS communications
- Publication
Soft Computing - A Fusion of Foundations, Methodologies & Applications, 2022, Vol 26, Issue 23, p12893
- ISSN
1432-7643
- Publication type
Article
- DOI
10.1007/s00500-021-06111-7