A novel DAG-blockchain structure for trusted routing in secure MANET-IoT environment

• Published in: Journal of intelligent & fuzzy systems 2024-03, Vol.46 (3), p.5733-5752
• Main Authors: Ilakkiya, N., Rajaram, A.
• Format: Article
• Language: English
• Subjects: Algorithms; Authentication; Blockchain; Clusters; Cryptography; Cybersecurity; Data management; Data transmission; Inspection; Internet of Things; Mobile ad hoc networks; Multiple criterion; Recurrent neural networks; Routing (telecommunications); Topology
• ISSN: 1064-1246; 1875-8967
• DOI: 10.3233/JIFS-232924

Different physical objects can be employed in the modern technological environment to facilitate human activity. In order to connect physical objects with the universe of digital using a variety of networks and communication technologies, an IoT, the cutting edges technological and effective solution, is deployed. Mobile ad hoc networks (MANET) interact with the IoTin smart settings, enhancing its user appeal and boosting its commercial viability. The new system of MANET based IoT and IT-network may be created by integrating wireless sensor and MANET with the Internet of Things. A solution like this increases user mobility while lowering network deployment costs. However, it also raises new, difficult problems in terms of networking considerations. In this, we presented a novel DAG (Directed Acyclic Graph)-Blockchain structure for MANET-IoT security. The network is secured through Multi-Factor PUF (MF-PUF) authentication scheme. With all authorized nodes, the network is segregated into cluster topology. For trusted data transmission, we proposed Jelly Fish Optimization (JFO) algorithm with the consideration of multiple criteria. For deep packet inspection, we proposed a Fully Connected Recurrent Neural Network (FCRNN). Through deep packet inspection, the intrusions are detected and mitigated through blocking system.With help of merged algorithm, the suggested method obtained improved ability in the PDR (Packet Delivery Ratio), production, analysis of time, detection accuracy also security levels. The comparison results clearly indicate that the proposed study outperforms all previous studies in various aspects. Particularly, the suggested methods for cluster creation, data aggregation, routing, encryption, and authentication significantly improve the system of DAG-IDS. Additionally, the planned task exhibits an exceptionally low standard deviation, making the suggested approach highly suitable for a WSN-IoT environment.