Vol.:(0123456789) Wireless Personal Communications https://doi.org/10.1007/s11277-019-06362-6 1 3 Cost and Sub‑Epoch Based Stable Energy‑Efcient Clustering Algorithm for Heterogeneous Wireless Sensor Networks Akshay Verma 1  · Tarique Rashid 1  · Prateek Raj Gautam 1  · Sunil Kumar 1  · Arvind Kumar 1 © Springer Science+Business Media, LLC, part of Springer Nature 2019 Abstract This paper proposes a cost and sub-epoch based stable energy-efcient clustering (CSSEEC) algorithm for heterogeneous wireless sensor networks. In this paper, we pro- vide a cost function for cluster heads selection and a sub-epoch to re-stands the previously selected cluster heads as normal nodes in cluster head selection process for future rounds. Cost function alleviates the energy consumption of sensor nodes by optimum selection of cluster heads and modifed sub-epoch makes the energy balance among the normal nodes. Thereby, the performance parameters like stability period, usable period, throughput and network lifetime are improved remarkably. It is also discerned that the stability period is the paramount parameter than others. By improving this parameter, overall performance of network is improved. Simulation results verifed that proposed CSSEEC protocol is more efcient than existing protocols. Keywords Stability period · Network coverage · Usable period · Wireless sensor networks 1 Introduction A wireless sensor network (WSN) is the collection of low cost and small size sensor nodes that are distributed in the target area for collecting data of interest [1, 2]. Its increasing popularity and growth make it valuable in many research applications [3, 4]. On the basis of network model, WSNs are further classifed into homogeneous and heterogeneous WSNs [5]. In homogeneous WSNs, all the sensor nodes are deployed with same amount of initial energies and phase same environmental conditions in the entire net- work. However, in heterogeneous WSNs, sensor nodes are deployed with diferent initial energies but environmental conditions may or may not be same in the whole network. Since sensor nodes have limited power, memory and bandwidth for processing the data, their energy should be efciently utilized to sense the target phenomena such as temperature, light, pressure, etc [6]. The main focus in network is to save sensor node energy for longer network lifetime and wider netwrok coverage. There are many routing approaches for such * Akshay Verma akshayv.verma49@gmail.com 1 Motilal Nehru National Institute of Technology Allahabad, Prayagraj, India