Learning Automata-Based Fault-Tolerant System for Dynamic Autonomous Unmanned Vehicular Networks

Abstract

A fault-tolerant routing system is highly needed in autonomous unmanned vehicle (AUxV)-based networks because of the various constraints due to adversarial situations present in the environment of AUxVs. Critical consequences might be resulting even with a minor fault in the system software/hardware in these types of systems due to the involved hazard of their applications such as search and rescue, threat surveillance, and chemical and biohazard sampling. The architecture, capability, application, and power of the internal systems are the parameters that vary among the AUxV network member nodes. Fault-tolerant system design is a key aspect for AUxVs, and heterogeneity is to be considered while designing the fault-tolerant systems' AUxVs. This paper describes an approach to fault-tolerant AUxV networks by designing a routing method. The proposed algorithm for AUxVs is based on the cross-layer design and the learning automata (LA), referred as Unmanned Vehicle Network with LA-based Routing using Cross-Layer Design (ULARC). The optimal path between the source and the destination is obtained using the theory of LA. The effectiveness of the proposed strategy is proved by the proof of convergence and is shown in the Appendix.