Toward a bio-inspired adaptive spatial clustering approach for IoT applications

https://doi.org/10.1016/j.future.2017.05.013Get rights and content

Highlights

  • A new bio-inspired approach applied to clustering in IoT applications.

  • A multi-level clustering approach promoting competition between objects and clusters.

  • A spatial clustering approach for IoT applications.

  • Self-adapting clusters capable of hiring/firing members based on ongoing events.

Abstract

Bio-inspired algorithms have demonstrated effective capabilities to solve Wireless Sensor Network (WSN) challenges. As sensors represent a main component in the emergent domain of Internet of Things (IoT), these algorithms are expected to perform also well in this field while adapting to contextual changes and optimizing the use of the limited resources. In this paper, we propose a new firefly-based clustering approach for IoT applications. Our approach includes a micro clustering phase during which Real-World Things (RWTs) compete and self-organize into clusters. These clusters are then polished during a macro-clustering phase where they compete to integrate small neighboring clusters. We extend our approach to allow the IoT clusters to self-adapt by hiring and/or firing RWTs depending on ongoing events and their expected impact on the network and its current deployment area. Initial simulations are showing promising results where the number of clusters tends to stabilize independently from the density of the network and the various communication ranges of RWTs.

Introduction

Continuous technological advances are shifting communication and computing technologies from large to small and tiny scales. In fact, the last decade has witnessed the emergence of a wide range of small sensing devices, which are capable of self-organizing (into Wireless Sensor Networks (WSNs)) and acquiring and reporting data about different phenomena and events of interest, anytime and everywhere [1]. With the increasing capabilities to insert smartness everywhere, these devices are being used to bridge the physical world where field operations are happening and the cyber world where data processing and decisions are being made. This progress has then led to the emergence of Cyber–Physical Systems [2] and Internet of Things (IoT). The IoT could be regarded as a universal networking infrastructure that deploys data acquisition devices and communication resources to connect physical and virtual objects [3]. In this infrastructure, devices, also called Real-World Things (RWTs), could access and/or exchange their mutual information and services, at any time and any place. However, this mutual support is not always straightforward. For instance, due to communication restrictions and technological incompatibilities, RWTs are not all the time capable of connecting with neighboring devices. Since the IoT is basically an opportunistic network where not all RWTs are IP-enabled, an effective topology management approach is needed to allow the RWTs optimize their communications, save their commonly limited energy resources, and increase their awareness about neighboring peers and services. To meet these goals, the research and development communities have accepted clustering as an efficient approach for its well-known advantages. For instance, clustering allows for data aggregation, topology control, support of Quality of Service (QoS), and energy efficient routing by limiting participating nodes in route establishment [4]. It also allows for improving the network scalability and increasing the network lifetime in large-scale deployments.

Current research and development efforts on IoT clustering (e.g., [5], [6], [7]) are particularly benefiting from the extensive work done in the field of WSNs, particularly since sensors are basic components of the IoT. For instance, the spatially distributed sensors of the IoT applications are now able to adapt to new environmental changes and reason on their own actions. However, these abilities remain limited due to the restrictions on power and processing capabilities as well as the effect of occurring contextual events. To overcome these limitations, sensors, and more generally RWTs, are highly recommended to organize into clusters and collaborate to ultimately achieve goals exceeding their own competencies. In this case, clusters are created based on a variety of criteria, including the distances between cluster-heads (CHs) and base stations, the distributions and sizes of CHs, the residual energies of sensors, and the number of allowed CHs [8]. Clusters can also be created based on the spatial location of CHs, their connectivity degrees, and their semantics.

In order to benefit from the multiple advantages of clustering, many approaches have fundamentally focused on a dynamic selection of Cluster Heads (CHs) as they are commonly expected to bear higher processing and communication loads. Several approaches have been used to elect CHs, including the use of sensors’ IDs (e.g., LCA [9]), sensors’ connectivity, or sensors’ distances from CHs (e.g., Max–Min D-Cluster [10], EEUC [11]). Other works have proposed to perform several iterations for the selection of CHs based on weights assigned to sensors desiring to become CHs. These weights are generally derived from the residual energy of sensors (e.g., EECA [12]) and/or the intra-cluster communication cost (e.g., HEED [13]).

Weights are particularly important as they reflect the obvious differences between sensors, and more generally the RWTs, in terms of their capabilities and locations. They could, therefore, be effectively used to elect the right candidates for CHs, at the right locations, at the right time. In order to reach this goal, biologically inspired algorithms can be adopted, especially since the WSN, and therefore the IoT, could be effectively modeled based on observations of living systems [5], [6], [14]. In this regards, several works have been inspired from Particle Swarm Optimization (PSO) and Artificial Bee Colony. Few other works have proposed firefly-inspired algorithms where sensors carry out intense competitions to be CHs, while trying to attract peers to their clusters. We argue in this paper that this approach is well suited for IoT clustering and we propose a firefly-inspired solution to cluster the IoT based on RWTs’ residual energies, clusters’ sizes, and contextual spatial information. Our solution includes two versions: (1) FiCA (Firefly Clustering Approach) which is consisting in Initialization, Fetching, Intimidation, and Polishing steps and combining a micro clustering phase and a macro clustering phase; and (2) ASFiCA (Adaptive Spatial FiCA) which extends FiCA with spatial capabilities and a Hiring/Firing step where RWTs self-organize to adjust the clusters’ sizes taking into consideration ongoing events and their impact on the IoT components and their deployment areas. In comparison with the current literature, we summarize our main contributions as follows: (1) a new four-steps bio-inspired approach applied to clustering in IoT applications; (2) a multi-levels (macro and micro) bio-inspired approach where individual RWTs compete to become Hs; and (3) an adaptive spatial clustering approach for IoT applications where clusters self-adapt by hiring and/or firing RWTs depending on ongoing events and their expected impact on the network and its current deployment area.

In the remainder of this paper, Section 2 outlines the current literature on bio-inspired IoT clustering. It also focuses on existing firefly-based clustering approaches. Section 3 presents FiCA, our clustering solution, which we extend to SFiCA and ASFiCA in Section 4 by considering the spatial context. In Section 5, we discuss the performance of our solution based on an experimental simulation.

Section snippets

Related work

WSNs have become very attractive candidates to implement the concept of IoT for several reasons, among which we mention two. Firstly, WSNs support the idea of distributed devices that can communicate and collaborate together in order to solve problems of common interest. In this collaboration, sensors could be thought of as any Real-World Thing (RWT) of the IoT. Secondly, WSNs already support clustering, which is an important requirement of IoT, as we mentioned in the introduction. However,

FiCA: Firefly-based clustering approach

Following the findings of Senthilnath et al. [28] (see Section 2), we are proposing an approach inspired by fireflies behaviors, where we assume that the IoT (also the swarm) includes n RWTs (also fireflies). Each RWT r has a set of solutions {xri:i=0,,m} where each xri is a neighboring CH candidate with a fitness value f(xri). Every RWT has a probability to become a CH that matches the brightness of the firefly it represents. The attractiveness βr reflects the strength of the RWT r in

Toward an adaptive spatial clustering approach

In several applications, including environmental monitoring and safety, some spatial locations are given higher attention during related decision making processes. This is the case, for example, of schools which are given high priority during evacuation operations when a sudden disaster (like an earthquake) happens. In such cases, operations within the WSN – and more generally the IoT – could be affected, basically because collecting data on objects and events of interest becomes uneven around

Experimentations

As a proof of concept, we developed a java-based application to test the performance of our approaches FiCA, SFiCA, and ASFiCA. For the sake of illustration, we present in what follows the results of our simulations for FiCA for randomly deployed IoT for different densities of the network. Our experimentations are essentially aiming to study the impact of network density on our clustering process. We report in Fig. 4, Fig. 5, Fig. 6, Fig. 7 the results of the micro FA clustering as well as the

Conclusion

The paper presented a new bio-inspired clustering approach for IoT applications. Based on the way fireflies seduce partners, the algorithm, called FiCA, consists in four steps: Initialization, Fetching, Intimidation, and Polishing. During the first three steps, RWTs continuously compete until the network is clustered. In this competition, Cluster Head Candidates (CHCs) attempt to attract neighboring peers to their clusters based on initial attractiveness values that will fade over time and

Dr. Nafaâ Jabeur is Assistant Professor and Head of the Computer Science Department at the German University of Technology in Oman (GUtech). He received his Master degree and his Ph.D. degree in computer science from Laval University (Quebec, Canada) in 2001 and 2006 respectively. He received his Engineering Degree in computer science from ENSIAS, Morocco, in 1998. Dr. Nafaâ has over than 10 years of experience in the industrial and academic sectors. He is an active researcher in a variety of

References (30)

  • SelvakennedyS. et al.

    A biologically-inspired clustering protocol for wireless sensor networks

    Comput. Commun.

    (2007)
  • SenthilnathJ. et al.

    Clustering using firefly algorithm: performance study

    Swarm Evol. Comput.

    (2011)
  • JabeurN. et al.

    Towards an intelligent web of things

  • N. Jabeur, N. Sahli, S. Zeadally, ABAMA: An agent-based architecture for mapping natural ecosystems onto wireless...
  • CASAGRAS, CASAGRAS Final Report: RFID and the Inclusive Model for the Internet of Things, 2009, pp....
  • YeM. et al.

    An energy efficient clustering scheme in wireless sensor networks

    Ad Hoc & Sensor Wirel. Netw.

    (2006)
  • K.H. Jones, K.N. Lodding, S. Olariu, L. Wilson, X. Chunsheng, Communal, cooperation in sensor networks for situation...
  • AntoniouP. et al.

    Congestion control in autonomous decentralized networks based on the Lotka–Volterra competition model

    J. Artif. Neural Netw.

    (2009)
  • RaniS. et al.

    Novel scheme for an energy efficient Internet of Things based on wireless sensor networks

    Sensors

    (2015)
  • O. Boyinbode, H. Le, A. Mbogho, M. Takizawa, R. Poliah, 2010, A Survey on Clustering Algorithms for Wireless Sensor...
  • BarkerD.J. et al.

    The design and simulation of a mobile radio network with distributed control

    IEEE J. Sel. Areas Commun.

    (1984)
  • A. Amis, R. Prakash, T. Vuong, D. Huynh, Max-Min D-Cluster Formation in Wireless Ad Hoc Networks, in: IEEE INFOCOM,...
  • LiC. et al.

    An energy efficient unequal clustering mechanism for wireless sensor networks

  • W. Heinzelman, A. Chandrakasan, H. Balakrishnan, Energy-efficient routing protocols for wireless microsensor networks,...
  • YounisM. et al.

    HEED: A hybrid energy-efficient distributed clustering approach for ad hoc sensor networks

    IEEE Trans. Mob. Comput.

    (2004)
  • Cited by (15)

    • Distributed resource scheduling in edge computing: Problems, solutions, and opportunities

      2022, Computer Networks
      Citation Excerpt :

      Another work in [183] proposed a solution for decentralized radio access solution in cognitive radio based on social foraging swarming models in birds. A distributed solution has been proposed for clustering in wireless sensor networks using firefly-inspired algorithms [184]. Another related work is for distributed Transmission Power Control in a wireless body area network using the Cucker–Smale model [185].

    • Special issue on future networks & pervasive systems

      2020, Future Generation Computer Systems
    • Proposing a new clustering approach aimed to energy consumption improvement

      2023, Journal of Ambient Intelligence and Humanized Computing
    View all citing articles on Scopus

    Dr. Nafaâ Jabeur is Assistant Professor and Head of the Computer Science Department at the German University of Technology in Oman (GUtech). He received his Master degree and his Ph.D. degree in computer science from Laval University (Quebec, Canada) in 2001 and 2006 respectively. He received his Engineering Degree in computer science from ENSIAS, Morocco, in 1998. Dr. Nafaâ has over than 10 years of experience in the industrial and academic sectors. He is an active researcher in a variety of multidisciplinary fields, including Wireless Sensor Networks, Cyber–Physical Systems, Multiagent Systems, Smart Cities, Geographic Information Systems, Social Networks, Web and Mobile Cartographic Services, Spatial Data Warehouses, and Internet of Things. Dr. Nafaâ has participated in several projects, including Sensor Networks for an Enhanced and Personalized Management and Use of Resources—SNEPMURe, An advanced telemetry system for complex environmental monitoring applications, and An integrated sensor web deployment infrastructure for watershed monitoring. Dr. Nafaâ is the author of a book on Intelligent Web and Mobile Cartographic Map Generation, coauthored a book on Cyber–Physical System Design with Sensor Networking Technologies, has contributed chapters to several books, and authored more than fifty research papers in prestigious conferences and highly ranked journals.

    Dr. Ansar-Ul-Haque Yasar is a professor at the Transportation Research Institute (IMOB) - Hasselt University Belgium since October 2011. At IMOB, he worked on the European FP7 project DATA SIM (2011–2014). He is currently responsible for the European ERA-NET project Smart-PT (2014–2016) with a consortium of several international partners. A part of his current job includes international collaborations, projects and business development at IMOB. He is also responsible for the Intelligent Transport Systems (ITS) course for the masters of transportation sciences students at UHasselt.

    His research interests include ubiquitous computing, context-aware communication, VANETs, intelligent transport systems and mobility management. He received his B.S. degree in Software Engineering in 2001 from Foundation University Islamabad — Pakistan, M.S. degree in Computer Science & Engineering in 2007 from Linkoping University — Sweden and Ph.D. in Engineering in 2011 from Katholieke Universiteit Leuven — Belgium. He has authored more than 40 research articles in renowned international journals, conferences and workshops. As one of his major accomplishments, Dr. Yasar has co-edited a book entitled “Data Science and Simulation in Transportation Research” published by IGI Global in December 2013. Furthermore, he has been involved in organization of many international peer-reviewed conferences, summer schools and other scientific events. Dr. Yasar is also a Technical Expert to evaluate project proposals submitted to the European R&D - EUREKA & COST frameworks.

    Elhadi Shakshuki is a professor and Wheelock Chair in the Jodrey School of Computer Science at Acadia University, Canada. His research interests include Intelligent Agents, Pervasive and Ubiquitous Computing, Distributed Systems, Handled Computing, and Wireless Sensor Networks. He is the founder and head of the Cooperative Intelligent Distributed Systems Group at the School of Computer Science, Acadia University. He received his B.Sc. degree in Computer Engineering in 1984 from Tripoli University, Libya, while his M.Sc. and Ph.D. degrees in Systems Design Engineering respectively in 1994 and 2000 from the University of Waterloo, Canada. Prof. Shakshuki is the Editor-in-Chief of the International Journal of Ubiquitous Systems and Pervasive Networks. He serves on the editorial board of several international journals and contributed in many international conferences and workshops with different roles, as a program/general/steering conference chair and numerous conferences and workshops as a program committee member. He published over 200 research papers in international journals, conferences and workshops. He is the founder of the following international conferences: ANT (2010–2017), EUSPN (2010–2017), FNC (2006–2017), ICTH (2011–2017), MobiSPC (2004–2017), and SEIT (2011–2017). He is also a founder of other international symposia and workshops. In addition, Prof. Shakshuki is the president of the International Association for Sharing Knowledge and Sustainability, Canada, and has guest co-edited over 30 international journal special issues. He is a senior member of IEEE, and a member of ACM, SIGMOD, IAENG and APENS.

    Dr. Hedi Haddad is an Assistant Professor in the Department of Computer Science at Dhofar University, Oman. He received his MBA from Laval University (Quebec, Canada) in 2002 and his Ph.D. degree in computer science from the same institution in 2009. His research interests cover the development of spatial decision support systems in different fields, including Multiagent Geosimulation, Wireless Sensor Networks and Geographic Information Systems. Recently he has been interested in spatially-aware smart applications for IoT. Dr. Haddad has participated in several projects, including Intelligent Monitoring Systems for Cooperating Objects using 6LowPAN, Geosimulation of Communicable Diseases’ Spatial and Temporal Spread Patterns and Evaluation of their Public Health Outcomes, and Adaptive Sensor Networks for Environmental Monitoring. Dr. Haddad has authored/co-authored more than thirty publications in book chapters, international journals, conferences and workshops.

    View full text