IMPROVING THE ENERGY EFFICIENCY BY INTEGRATING K – NEAREST NEIGHBOR WITH ARTIFICIAL BEE COLONY OPTIMIZATION ALGORITHM IN WIRELESS SENSOR NETWORKS

Dynamic conditions, which alter quickly over time, are controlled by Wireless Sensor Networks (WSN). Either external variables are responsible for this complex behavior or by device designers themselves. Sensor networks utilize Machine Learning (ML) techniques also to respond to these situations, to reduce the need for excessive redesign. The ML also inspires several realistic solutions that optimize the usage of resources and extend the network's existence. This paper introduces an optimized classification algorithm K-Nearest Neighbors (KNNs) which searches the better WSN sink-node. Sink-node adds data from all sensor nodes and decreases the network energy usage to extend the lifespan of the network. A fitness function has been formulated to choose the best position for the sink-node with the strong residual energy nodes of the neighbor to optimize network existence. The different sensor nodes in the WSN have limited energy resources that seriously impact the network's long-term efficiency. The emphasis of current research is therefore on developing energy-efficient WSN algorithms to enhance network existence. This research introduces the algorithm called Artificial Bee Colony (ABC) in a KNN classification model to enhance network energy ability. The ABC algorithm will boost the internal dynamics of the head and sensor nodes of the clusters in the WSN. The algorithm proposed here will minimize node energy dissipation and balance energy consumption across nodes to optimize the network existence. In contrast to other algorithms, the ABC algorithm includes only fewer control parameters in WSN. It is thus simpler for clustered sensor networks to incorporate. Hence, the ABC method is to cooperate with KNN in a hybrid manner to manage energy consumption for balancing the network. This architecture captures the network status by using acceptable energy measurements and maps them into equivalent costs to determine the shortest route. Thus the performance measures of KNN-ABC were compared with the existing approaches DDAR and OODVRP to prove the energy efficiency.