Recovery from multiple simultaneous failures in wireless sensor networks using minimum Steiner tree.docVIP

wireless ad hoc networks, due to the dynamic topology changes, multi hop communications and strict resource limitations, routing becomes the most challenging issue, and broadcasting is a common approach which is used to alleviate the routing problem. Global flooding is a straightforward broadcasting method which is used in almost all existing topology-based routing protocols and suffers from the notorious broadcast storm problem. The connected dominating set (CDS) formation is a promising approach for reducing the broadcast routing overhead in which the messages are forwarded along the virtual backbone induced by the CDS. In this paper, we propose an intelligent backbone formation algorithm based on distributed learning automata (DLA) in which a near optimal solution to the minimum CDS problem is found. Sending along this virtual backbone alleviates the broadcast storm problem as the number of hosts responsible for broadcast routing is reduced to the number of hosts in backbone. The proposed algorithm can be also used in multicast routing protocols, where the only multicast group members need to be dominated by the CDS. In this paper, the worst case running time and message complexity of the proposed backbone formation algorithm to find a 1/(1-ε) optimal size backbone are computed. It is shown that by a proper choice of the learning rate of the proposed algorithm, a trade-off between the running time and message complexity of algorithm with the backbone size can be made. The simulation results show that the proposed algorithm significantly outperforms the existing CDS-based backbone formation algorithms in terms of the network backbone size, and its message overhead is only slightly more than the least cost algorithm. Article Outline 1. Introduction 2. Related work 3. Learning automata, distributed learning automata, and variable action-set learning automata 3.1. Learning automata 3.2. Distributed learning automata 3.3. Variable action-set learning automata 4. DLA-