connectivity restoration in wireless sensor and actor networks using distributed learning automata

In wireless sensor and actor networks, connectivity among actors is very important, especially in critical applications where actors collaborate with each other to provide a report as soon as possible. sometimes a node failure divides the given network into several parts; causing loss of connectivity between actors hence degrades the network performance. several algorithms have been proposed to restore inter-actor connectivity. in these methods, selecting appropriate failure handler to minimize the relocations is crucial. in this paper the issue of nding the best backup is supposed the same as nding shortest path in stochastic graph problem. owing that the solving stochastic shortest path is np-complete, dla is used for failure handler choice. this essay also presents a hybrid algorithm named dla-bus for critical node back up selection based on distributed learning automata. here two methods are proposed; rst one is that each actor node is equipped with a learning automaton so that their cooperation in learning process leads to select the desired backups. the second method states the presentation of dla-mrf to repair stimulant failure of two adjacent actors. in order to show the performance of the proposed algorithms extensive simulations using castalia simulator have been conducted. simulation results demonstrate that the mentioned proposed algorithms outperform existing methods in terms of the number of nodes movement, total distance travels, the percentage of coverage reduction, and energy consumption.