An Unequal Clustering Algorithm Concerned With Time-Delay for Internet of Things

Abstract

Internet of Things (IoT) enables the devices to exchange data with each other. The wireless sensor network is a key technology for making devices sensible and has been widely concerned. In the clustering routing protocol of wireless sensor networks, the cluster heads have high energy consumption rate since it undertakes data collection, fusion, and forwarding, which causes unbalanced energy consumption. Thus, the network lifetime is limited. In this paper, we present the contribution as follows. First, we propose an improved K-means algorithm to cluster the network and use the weighted evaluation function to optimize the cluster structure. Then, we select to either split or merge the cluster structure according to the evaluation results and in further to obtain a non-uniform clustering structure of the network. Second, in the data transmission phase, the data fusion mechanism is used to improve the energy utilization rate of cluster heads. Given the transmission delay problem caused by the data fusion, we propose a delay-optimized data fusion tree construction-based algorithm. When an active node selects the parent node, the distance and the energy factors are considered. The time slot allocation is optimized through constructing a data fusion tree, and the transmission delay is minimized. Finally, compared with other algorithms in the simulation section, the proposed algorithm can effectively reduce the energy consumption of the network, and the constructing data fusion tree decreases the transmission delay caused by the data fusion process. The service quality of the whole network is therefore improved. The proposed algorithm is suitable for the delay-constraint application of IoT.