Learning Automata based Grey Wolf Optimization Algorithm for Adaptive data forwarding and Aggregation in Wireless Sensor Networks

Abstract

Reducing the packet count transmitted within a wireless sensor network is one method for decreasing the consumption of energy. In order to reduce the quantity of transmitted packets, many nodes may have substantially identical information and can be combined in intermediary nodes. Sensor networks typically deploy many redundant nodes to address the issue of node failures. When data packets from all nodes with same information are aggregated together, the aggregate ratio is maximized. Nodes should forward their packets along a path where there are as many nodes as possible that have information similar to that of the sender node. In many real-world scenarios, these path changes frequently and do not remain constant throughout the network lifetime. These changes, which typically cannot be anticipated in advance, may have occurred as a result of alterations to the environment of operating sensor network. In this study, a LA (learning automata) based solution to the aforementioned issue is given. For each node to broadcast its packets towards the sink, the learning automata collectively learn the path and choose the one with the highest aggregation ratio. Initially, the neighbor list of the sensor nodes is estimated using grey wolf optimization (GWO) technique with distance to sink & aggregation cost parameters to identify the paths with fewer count of hops to the sink. This proposed work utilized Adaptive forwarding (AF) factor in LA to determine intermediate nodes for data transmission. The decision about adaptive forwarding and aggregation based on the parameters: load factor & connectivity ratio was taken by LA configured in the sensor nodes makes. Performance assessments and analyses show that the suggested approach decreases energy usage, delays in data delivery, data transfers, and network lifespan.