Localized Fault Tolerant and Connectivity Restoration Algorithms in Mobile Wireless Ad Hoc Network

Due to the dynamic characteristics of mobile wireless ad hoc networks, a node may join or leave/fail at anytime, and links are usually intermittently unavailable. The destruction of the network connectivity would dramatically degrade the performance. To resolve this challenge, topology control is an efficient way to maintain network connectivity, which focuses on the adjustment of node transmit power in the past few decades. However, the adjustment of node movement has not been adequately addressed to restore the connectivity in mobile wireless ad hoc networks. In this paper, we propose three hybrid recovering algorithms, namely, cascade movement based on nearest nodes, cascade movement based on connected dominating set, and cascade movement based on critical nodes. With the adoption of Kalman filter, all these schemes can predict the movement of nodes so as to identify and maintain connectivity of mobile wireless ad hoc networks. Meanwhile, once a node fails or a link breaks down, they can also recover the connectivity immediately. We conduct extensive simulations on the metrics of the average number of participating nodes, the deviation from original direction, and the ratio of cascade movement for the three proposed algorithms, and the algorithm of cascade movement based on critical nodes achieves the best results.