Research Area: Internet of Things
Software-defined industrial network has emer-ged as an autonomous ecosystem where the network control relies on a centralized controller to provide seamless data transfer. However, the reliance on a centralized controller can lead to several challenges, such as single point of failure. An adversary can initiate a denial of service attack and limit the availability of the controller by projecting malicious or uncontrolled traffic flows. To overcome this, in this article, a deep-learning-based blockchain framework is designed for providing secure software-defined industrial network. In this framework, a blockchain mechanism is designed wherein all the switch are registered, verified (using zero-knowledge proof), and, thereafter, validated in the blockchain using a voting-based consensus mechanism. A deep Boltzmann machine based flow analyzer is deployed at the control plane to identify the anomalous switch requests. The evaluation is performed using a mininet emulator wherein the results obtained depict the superiority of the proposed framework.
Keywords:
Author(s) Name: Maninderpal Singh; Gagangeet Singh Aujla; Amritpal Singh; Neeraj Kumar; Sahil Garg
Journal name: IEEE Transactions on Industrial Informatics
Conferrence name:
Publisher name: IEEE
DOI: 10.1109/TII.2020.2968946
Volume Information: ( Volume: 17, Issue: 1, Jan. 2021) Page(s): 606 - 616
