A cellular learning automata-based approach for self-protection and coverage problem in the Internet of Things

Abstract

In the Internet of Things (IoT) sensing layer, targets in the monitoring area are covered by sensor nodes. So the coverage problem is a key issue of IoT application. Self-protection is an effective way to resist external attacks, which can cover the remaining nodes with a set of nodes. This paper considers the coverage problem combined with node self-protection and proposes a cellular learning automata-based method. In this method, nodes are mapped to the cellular structure, and each cell is equipped with learning automata. At the initialization stage, the activation probability is assigned to each node by the residual energy, and updated according to the reward or penalty feedback from neighbors and the random environment. At the adjustment stage, partial self-protection is applied to reduce the number of activated nodes. Consequently critical nodes could acquire protection with high priority than other normal nodes. When the desired coverage is not met, sleeping nodes will be activated according to the coverage increment value. Meanwhile, a local strategy based on the cellular structure is designed to heal the coverage hole caused by failed nodes. In accordance to this strategy, sleeping nodes around the failed nodes could be quickly activated. The simulation results show that the proposed scheme outperforms the existing methods with respect to the number of active nodes, energy efficiency, and coverage reliability.