An energy-efficient artificial bee colony-based clustering in the internet of things☆
Graphical abstract
Introduction
The interconnections of smart devices on the Internet of things (IoT) provides an intelligent landscape for a dynamic world. IoT refers to the interconnection of the networks between heterogeneous wireless sensors and routine physical devices scattered in a monitoring area to sense, gather, share and transmit data throughout the system, and provide a smarter life for human beings [1]. Meanwhile, from the recent advances in technologies and protocols, IoT devices are equipped with sensing, processing, communication, interaction, and corporation capabilities. According to these properties, IoT has penetrated pervasively into the wide-scale aspects of human life, including smart healthcare systems and personal monitoring, smart cities (homes, buildings, traffic monitoring, and urban computing), intelligent commercial, and smart industrial systems [2].
Many IoT devices are powered by batteries with a limited lifetime and deployed in remote areas. Therefore, one of the main concerns while using IoT is energy limitation, as communication and computations on devices might quickly exhaust their battery resources [3]. Thus, providing energy-efficient data processing, aggregation, and transmission mechanisms play a vital role in IoT applications. The amount of energy consumed for wireless communications in the internet-based systems is much more than the processing costs [4]. Therefore, one of the significant challenges on IoT is to achieve an energy-efficient data transmission mechanism from the source to the destination devices. Clustering is the primary operation required for appropriate data transmission on IoT. It is defined as a creative process that groups devices in some clusters and determines a cluster-head for each of them to improve resource consumption [5].
There are different heuristic [6], metaheuristic [7], and fuzzy-based [8] clustering mechanisms in the literature. The heuristic mechanisms try to provide an approximate solution to a specific problem. However, the clustering challenge requires approximation problem-independent solutions for which it can be proved how close the final solution is to the optimal one. Also, the primary purpose of most heuristic mechanisms is to reduce the number of clusters, which increase injected traffic into the cluster-heads and reduce IoT lifetime. The metaheuristic clustering mechanisms focus on the distance and residual energy of devices as the performance criteria, while they ignore the number of one-hop neighbors and data volume. Besides, some metaheuristic mechanisms have suggested the cluster-head selection scheme, while they offer no idea about the device allocation to the clusters. Finally, the solutions from fuzzy-based mechanisms are perceived based on assumptions, while setting exact assumptions, rules, and functions is a complicated process. Furthermore, validation and verification of the fuzzy-based systems need extensive tests. So, it is more efficient to provide a metaheuristic cluster-head selection and clustering mechanism on IoT by considering the data volume generated by each cluster, residual energy of devices, the number of one-hop neighbors, the distance between devices and the sink, and the distance between each device and its one-hop neighbors as the performance criteria.
Since clustering on IoT is an NP-hard problem, metaheuristic algorithms are a good idea to tackle this challenge. Artificial Bee Colony (ABC) is one of the well-known metaheuristic algorithms, which does not need the parameter-setting to address the issues. Also, it yields better performance in comparison to others [9]. So, this paper proposes a novel mechanism for cluster-head selection and device clustering on IoT, which is named ABC-based Device Clustering (ABC-DC). The presented mechanism consists of two phases. The first phase aims to select the near-optimal cluster-heads using the ABC. It considers the residual energy of devices, the number of one-hop neighbors, Euclidean distance between devices and sink, and Euclidean distance between each device and its one-hop neighbors as the performance criteria.
The principal objective of the second phase is to group devices into some clusters. It also exploits the advantages of the ABC to enhance tradeoff between the performance criteria, including energy consumption of the whole system, and data transmission delay. The Euclidean distance between each cluster-head and its members and the data volume generated by each cluster are employed as the performance criteria. Simulation results verify that the proposed mechanism improves IoT lifetime by reducing the energy consumption of devices and data transmission delay in comparison with the recent state-of-the-art approaches.
The fundamental contributions are summarized as follows:
- •
Developing an efficient metaheuristic clustering mechanism on IoT to tradeoff between performance factors, including energy consumption and data transmission delay.
- •
Developing a flexible clustering mechanism to comply with the requirements of the different IoT applications.
- •
Exploiting the advantages of ABC to select the near-optimal cluster-heads considering the residual energy of the devices, the number of one-hop neighbors, Euclidean distance between the devices and sink, and Euclidean distance between each device and its one-hop neighbors as the performance criteria.
- •
Developing an ABC-based mechanism to group IoT devices in some near-optimal clusters considering the Euclidean distance between each cluster-head and its members, and the data volume generated by each cluster as the performance criteria.
The rest of this paper is organized as follows: Section 2 explains the previous clustering mechanisms on IoT. Then, some practical criteria about the IoT modeling, and energy depletion pattern are described in Section 3. After that, the proposed mechanism is discussed in Section 4. Section 5 evaluates the performance of the ABC-DC and its features in comparison with recent state-of-the-art approaches. Finally, Section 6 represents the conclusion part.
Section snippets
Related work
Clustering is one of the primary challenges to satisfy the requirements of energy-aware IoT applications. To address this challenge, kinds of literature have presented a broad range of mechanisms for clustering devices on IoT. The main objective of them is to group IoT devices in certain/uncertain number of clusters, and thus guarantee the energy efficiency, distributed processing, and management hierarchy [5]. We investigate device clustering approaches in three categories, including
System model
In this section, we introduce the system model, including IoT performance criteria, and an energy consumption model. The summary of the considered criteria is mentioned in Table. 2.
Proposed mechanism
In this section, we propose an efficient mechanism for cluster-head selection and device clustering on IoT, which is called Artificial Bee Colony-based Device Clustering (ABC-DC). The overall scheme of the proposed mechanism is shown in Fig. 1.
As shown in Fig. 1, the proposed mechanism consists of two major phases:
- 1.
The main idea of the first phase is to achieve near-optimal cluster-heads using ABC. The residual energy of devices, the number of one-hop neighbors, Euclidean distance between
Performance evaluation
This section explains the simulation results to verify the efficiency of ABC-DC compared with the recent state-of-the-art clustering approaches on IoT. First, we present the simulation parameters, values, and dataset. Then, we analyze the critical performance metrics on IoT, including energy consumption, lifetime, fairness of energy consumption, and data transmission delay.
Conclusion
Data transmission is one of the significant concerns about IoT. Device clustering is the primary operation to satisfy the requirements of an energy-efficient communication in such systems. In this paper, we presented the ABC-DC mechanism that selects appropriate cluster-heads and cluster devices on IoT. The proposed mechanism is composed of two phases. The first phase exploits the ABC algorithm to determine the near-optimal cluster-heads considering the residual energy of devices, the number of
CRediT authorship contribution statement
Shamim Yousefi: Conceptualization, Methodology, Data curation, Formal analysis, Validation, Writing - original draft. Farnaz Derakhshan: Conceptualization, Investigation, Supervision, Writing - original draft. Hadi S. Aghdasi: Methodology, Data curation, Supervision, Writing - review & editing. Hadis Karimipour: Conceptualization, Methodology, Formal analysis, Supervision, Validation, Writing - review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Shamim Yousefi received the B.S. in Information Technology and the M.S. in Computer Engineering from University of Tabriz, Iran in 2013 and 2015, respectively. Currently, she is Ph.D. Student and working as a researcher at Multi-agent Laboratory in University of Tabriz. Her current interests include light-weight methods for clustering and routing on the Internet of Things (IoT).
References (25)
- et al.
Machine learning for wireless communications in the Internet of Things: A comprehensive survey
Ad Hoc Networks
(2019) - et al.
Internet of things applications: a systematic review
Comput Netw
(2019) - et al.
A clustering based routing algorithm in IoT aware Wireless Mesh Networks
Sustain Cities Soc
(2018) A Hybrid Ant Colony and Artificial Bee Colony Optimization Algorithm-based Cluster Head Selection for IoT
Procedia Computer Science
(2018)- et al.
Mobile agents for route planning in internet of things using Markov decision process
- et al.
A review on the applications of multiagent systems in wireless sensor networks
Int J Distrib Sens Netw
(2019) - et al.
A survey of clustering techniques in WSNs and consideration of the challenges of applying such to 5G IoT scenarios
IEEE Internet Things J
(2017) - et al.
A multi-hop clustering mechanism for scalable IoT networks
Sensors
(2018) - et al.
Clustering for load balancing and energy efficiency in IoT applications
- et al.
Enhanced three layer hybrid clustering mechanism for energy efficient routing in IoT
Sensors
(2019)
Reliable data gathering in the Internet of Things using artificial bee colony
Turkish J Electr Eng Comput Sci
A no self-edge stochastic block model and a heuristic algorithm for balanced anti-community detection in networks
Information Sciences
Cited by (0)
Shamim Yousefi received the B.S. in Information Technology and the M.S. in Computer Engineering from University of Tabriz, Iran in 2013 and 2015, respectively. Currently, she is Ph.D. Student and working as a researcher at Multi-agent Laboratory in University of Tabriz. Her current interests include light-weight methods for clustering and routing on the Internet of Things (IoT).
Farnaz Derakhshan is Assistant Professor, and the Director of Multi-agent Laboratory at Faculty of Electrical and Computer Engineering, University of Tabriz, Iran. She received her PhD in Artificial Intelligence from the University of Liverpool, UK. Her main research interests include multi-agent systems and its applications, normative multi-agent systems, multi-agent learning, Internet of Things (IoT) and swarm intelligence.
Hadi S. Aghdasi received his M.S. and Ph.D. in computer engineering from Shahid Beheshti University, Tehran, Iran, in 2008 and 2013, respectively. He has been an assistant professor of Computer Engineering at University of Tabriz, Iran since 2013. His current researches focus on Humanoid Robots, Intelligent Methods in Surveillance Systems and Cognitive Technology, and Wireless Sensor Networks (WSNs).
Hadis Karimipour received the Ph.D. from the Department of Electrical and Computer Engineering in the University of Alberta in Feb. 2016. She is currently an Assistant Professor at the University of Guelph, Guelph, Ontario. Her research interests include application of machine learning on security analysis, cyber-physical modeling, and cyber-security of the smart grids.
- ☆
This paper is for regular issues of CAEE. Reviews processed and recommended for publication to the Editor-in-Chief by Associate Editor Dr. M. H. Rehmani.