Hybrid K-Mean PSO Clustering Algorithm for Energy-Efficient of Object Tracking in WSNs

Target tracking can be a time-consuming and resource-hungry process due to the limitations in resource, power, sensing capability, communication, memory, and computational capability, it is a challenge to implement traditional target track solutions in wireless sensor networks (WSNs) applications. As any of the operations of data collection, processing, and dissemination leads to an increase in resource consumption, the target tracking algorithms in WSN should effectively utilize the resource and also reduce the computational complexity. in this paper, a new and effective multiple objects tracking approach was proposed by combining dynamic and static clustering techniques with the predictive tracking technique using the Long Short-Term Memory (LSTM) which is the extension of Recurrent Neural Networks (RNNs). In this work two models from the LSTM were used, where the first model will classify the objects' motion path and then discover the object's new location, and the second model was used to predict the three nodes that will track the objects, and computing the energy consumption for this approach and proposed anew algorithm called Hyper K_Mean PSO Clustering algorithm (HKM_PSO) to improve the energy consumption and increasing the lifetime of the network. The approaches and the algorithm that proposed in this work were very effective. The most important contribution was optimizing the energy consumed in the process of tracking multiple objects and increasing the life of the network by improve the energy through the process of selecting the head cluster and increasing the life of the network.