Hierarchical clustering algorithms for heterogeneous energy harvesting wireless sensor networks

WSNs are achieving significance importance in the information and communication technologies with the development of Internet-of-Things (IoT). WSNs are connected with the Internet in many industrial applications. However, energy efficiency is one of the challenging problems in wireless sensor networks (WSNs). In this paper, an optimized location of cluster heads (CHs) for energy efficiency in a heterogeneous WSN (that consists of few energy harvesting sensors and mostly non-renewable sensors) is studied. LEACH is an effective method for clustering in a homogeneous WSN, consisting solely of non-renewable sensors with uniform energy profile. Our proposed clustering schemes suggest simple and static clustering strategy for a heterogeneous WSN with good harvesting efficiency. The proposed schemes divide the total network covered area into cells, based on different criterion. In regular grid (RG) approach, a CH is simply placed in the center of each cell. Minimax grid (MG) approach attempts to improve the lifetime of the network by placing the CH at the center of the smallest enclosing circle. K-Medoids (KM) approach first divides the network into clusters and then solves a facility location problem to assign the role of CHs to EH sensors. Simulation results show that RG, MG and KM perform better than LEACH in terms of energy consumption, and consequently, increase the lifetime of the network by upto 200% when the harvested energy is available in healthy amounts. The relative improvement of KM and MG over RG is also discussed in this paper.