The Time-Free Comparison Model for Fault Diagnosis in Wireless Ad Hoc Networks

Abstract

This paper describes a new comparison-based model for fault diagnosis in wireless ad hoc networks. Fault diagnosis is crucial for ensuring the dependability of systems. Wireless ad hoc networks are highly prone to faults as consequence of their dynamical conditions. The comparison approach is a practical diagnosis model that has been used to develop self-diagnosis systems in wired and wireless networks. This approach can detect and diagnose hard and soft faults in systems. The traditional fault diagnostic models were designed for static networks. Thus, they cannot provide complete and correct fault diagnosis in mobile wireless networks. In this paper, we introduce a time-free self-diagnosis model that respects the design requirements of mobile wireless networks. That is, it adapts to the topology’s changes, it imposes no known bounds on time delays, and it requires limited network information. Further, we develop a fault diagnosis protocol that can correctly diagnose faulty nodes undergoing static and dynamic faults in mobile ad-hoc networks (MANETs). Both an analytical model and a simulation study have been used to prove and evaluate the efficiency of our protocol under various scenarios. Furthermore, the performance of our protocol is compared with related protocols. The results show that our proposed protocol is efficient in terms of communication and time complexity.