Elsevier

Ad Hoc Networks

Volume 101, 15 April 2020, 102094
Ad Hoc Networks

Mobile wireless sensor network lifetime maximization by using evolutionary computing methods

https://doi.org/10.1016/j.adhoc.2020.102094Get rights and content

Abstract

Due to the continuous development and progress of wireless communication technology and sensor network technology, wireless sensor networks (WSNs) have gradually become an attractive technology that facilitates people's lives. Due to the extensive use of WSNs, maximizing the lifetime of WSNs to obtain real-time and effective information has become a critical concern. This paper studies the life of mobile wireless sensor networks (MWSNs). MWSNs are a special type of WSN in that the sensor nodes are movable within a certain area. A system model is developed to prolong the network lifetime of MWSNs. This paper uses five evolutionary computing (EC) algorithms to develop the MWSN lifetime optimization model. Numerical simulations are performed to study the advantages and disadvantages of the five algorithms for solving the model. The comparison and discussion can provide advice for using EC algorithms to solve MWSN lifetime maximization problems.

Introduction

Due to the continuous development and progress of wireless communication technology, network technology, microprocessor technology and sensor network technology, WSNs have gradually become an attractive technology that facilitates people's lives [1], [2], [3], [4], [5]. Moreover, WSNs are a new way to acquire information through real-time monitoring of the environment. Because of their unique way of obtaining information, WSNs are widely used in various fields, such as military defense, biological medicine, smart home technology, industry and agriculture [2,3]. As the capacity of battery of nodes is limited, the operational longevity of nodes is critical. The longevity of a WSN directly affects the overall performance of the network [4].

MWSNs are a special distributed network of many deployed sensor nodes that are movable within a monitoring area. MWSNs form a self-organizing network through wireless communication technology [5]. Unlike in static WSNs, the mobility of sensors or sink nodes in MWSNs causes network topology to change dynamically. Thus, compared to when designing static WSNs, more issues have to be addressed when designing mobile networks [4].

Recently, there have been studies on the lifetime of MWSNs. [6] studied maximizing the lifetime of MWSNs that contained mobile sink nodes. In [7], the exploration and exploitation trade-off was studied, and different methods were compared. According to this study, Thompson sampling and exponential weight algorithm action-section are robust methods for channel selection and transmission power control. In [8], the lifetime of WSNs was maximized by using a method for harvesting solar energy. In [9], a secure and efficient authentication protocol was proposed for multigateway wireless networks. In [10], scheduled transmissions from wireless networks were studied by using learning channel statistics. The proposed greedily constructed schedules show good scalability with the number of network links. In [11], prolonging the lifetime of WSNs was solved by proposed optimization formulations in the case of machine-to-machine communication. In [12], a distributed topology control game method was proposed for WSNs that have many sensor nodes but limited energy resources. In [13], a routing algorithm to reduce energy consumption and delay by MWSNs was proposed.

The contributions of this paper are as follows:

  • (1)

    The objective of the MWSN model is to minimize the residual and consumed energies of all nodes. Minimizing residual energies is useful in preventing nodes from dying quickly. Minimizing consumed energies is useful in prolonging the lifetime of nodes. Both residual and consumed energies are combined to achieve the objective of the MWSN model.

  • (2)

    Unlike current studies, this paper uses EC algorithms to solve the MWSN model. Traditionally, MWSN models have to be relaxed to convex optimization problems [4,[14], [15], [16], [17], [18], [19]. Then, problems can be solved by popular convex optimization solvers. This paper uses EC algorithms, which do not require the MWSN model to have a convex property.

  • (3)

    The properties of five EC algorithms are analyzed and discussed. The results provide useful hints for solving MWSN or WSN models.

This paper presents a lifetime model of an MWSN in Section 2. The paper introduces five EC methods in Section 3. Then, this paper analyzes the lifetime of wireless sensor networks under five EC algorithms in Section 4. According to the simulation experiment of the system model, the optimal EC algorithm is obtained under certain parameters so that the lifetime of the MWSN can be improved. Finally, Section 5 concludes the paper.

Section snippets

Lifetime model of the mobile wireless sensor network

It is assumed that the MWSN system consists of sensor nodes and sink nodes, as shown in Fig. 1. In the figure, the MWSN network consists of 2 sink nodes and 16 sensor nodes. Sink nodes are fixed and have a coverage area with a radius Rc. Sensor nodes can freely move with a certain degree of mobility. In cases where all sensor nodes are covered by sink nodes, the network becomes robust as all nodes are under the control of sink nodes. In cases where sink nodes may not cover all sensor nodes, ad

Evolutionary computing algorithms

EC methods have been applied to many fields [21], [22], [23], [24], [25]. Five evolutionary computing methods are used to solve the lifetime improvement problem: genetic algorithms (GAs) [26], differential evolution (DE) [27], particle swarm optimization (PSO) [28], artificial bee colonies (ABCs) [29] and neighborhood field optimization (NFO) [30,31]. These methods are explained in this section.

Simulation experiment

The simulation experiment consists of three parts. First, the lifetime model at different mobilities is tested and analyzed. Second, the lifetime model with different weights in the objective function is tested and analyzed. Third, the impact of network density is tested and analyzed. Five EC algorithms are used to solve the problems for the three parts.

The settings of the MWSN are as follows: The number of sensor nodes Se is 10. The number of sink nodes Si is 1. The initial energy of the

Conclusion

MWSNs are special distributed networks consisting of many sensor nodes. In MWSNs, unlike in static WSNs, sensors or sink nodes have mobility, thus causing the network topology to change dynamically. In the design of MWSNs, more issues have to be addressed than in the design of static WSNs. Thus, lifetime maximization is an important problem in MWSNs.

This paper presents a new mathematical model for MWSNs. The objective of the MWSN model is to minimize the residual and consumed energies of all

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.

Acknowledgments

This research was supported in part by the National Natural Science Foundation of China (Project no. 61601329, 61603275, 61704122, 61701345, 61801327), the Natural Science Foundation of Tianjin (Project no. 18JCZDJC31900), and the Tianjin Higher Education Creative Team Funds Program.

Xiu Zhang: received the B.Eng. and M.Eng. degrees in biomedical engineering from Hebei University of Technology, Tianjin, China, in 2006 and 2009, respectively, and the Ph.D. degree in electrical engineering from The Hong Kong Polytechnic University in 2012. From 2013 to 2015, she was a Post-Doctoral Fellow with The Hong Kong Polytechnic University. She is currently an associate professor with Tianjin Normal University. Her research interests are mainly focused on numerical methods of

References (32)

  • D.A. Guimaraes et al.

    Increasing the lifetime of mobile wsns via dynamic optimization of sensor node communication activity

    Sensors

    (2016)
  • X. Liu et al.

    Rate and energy efficiency improvements for 5G-based IoT with simultaneous transfer

    IEEE Internet Things J.

    (2019)
  • F. Tashtarian

    On maximizing the lifetime of wireless sensor networks in event-driven applications with mobile sinks

    IEEE Trans. Veh. Technol.

    (2015)
  • H. Sharma et al.

    Maximization of wireless sensor network lifetime using solar energy harvesting for smart agriculture monitoring

    Ad Hoc Netw.

    (2019)
  • H. Guo et al.

    A secure and efficient three-factor multi-gateway authentication protocol for wireless sensor networks

    Ad Hoc Netw.

    (2019)
  • E. Fitzgerald et al.

    Network lifetime maximization in wireless mesh networks for machine-to-machine communication

    Ad Hoc Netw.

    (2019)
  • Cited by (20)

    • A new and reliable intelligent model for deployment of sensor nodes for IoT applications

      2022, Computers and Electrical Engineering
      Citation Excerpt :

      Many researchers have studied the various strategies for sensor node deployment, and there are numerous effective approaches available in the literature [4] gives a comprehensive discussion. As previously stated, current WSN research focuses on meeting certain requirements such as energy reduction [5,6], an ideal number of sensor nodes [7,8], lifetime maximization [9], increased connection and coverage [10–12], and so on. A coverage-aware and efficient planar face topology structure (CAFT) is presented in [13] to reduce sensor node energy consumption.

    • Multi-objective teaching–learning evolutionary algorithm for enhancing sensor network coverage and lifetime

      2022, Engineering Applications of Artificial Intelligence
      Citation Excerpt :

      Heuristics and meta-heuristics were proposed to find the approximate solutions. In Zhang et al. (2020), the authors considered the lifetime of mobile wireless sensor networks. They proposed evolutionary algorithms to minimize residual and consumed energy of all nodes.

    • Dynamic multi-objective cooperative coevolutionary scheduling for mobile underwater wireless sensor networks

      2021, Computers and Industrial Engineering
      Citation Excerpt :

      Hu et al. employed an improved NSGA-III algorithm to solve the problem of the underwater sensor deployment for detecting pollutants (Hu et al., 2018). In addition, Zhang et al. combined the residual and energy consumption into one objective when constructing a lifetime model of sensor networks, and utilized such methods as genetic algorithms, differential evolution, particle swarm optimization, artificial bee colony, and domain search to solve the model (Zhang et al., 2020). In the work proposed by Chowdhury and De, a fuzzy inference system based reverse glowworm swarm optimization algorithm is utillized to move a best sensor to replace one whose battery power is lesser than a threshold value so as to minimize the energy consumption and reduce the redundant coverage (Chowdhury & De, 2020a).

    View all citing articles on Scopus

    Xiu Zhang: received the B.Eng. and M.Eng. degrees in biomedical engineering from Hebei University of Technology, Tianjin, China, in 2006 and 2009, respectively, and the Ph.D. degree in electrical engineering from The Hong Kong Polytechnic University in 2012. From 2013 to 2015, she was a Post-Doctoral Fellow with The Hong Kong Polytechnic University. She is currently an associate professor with Tianjin Normal University. Her research interests are mainly focused on numerical methods of electromagnetic field computation, novel wireless energy transfer systems, and wireless network optimization.

    Xiaohui Lu: received the B.Eng. degree from Weifang University in 2018. Since 2018, she has been a student at Tianjin Normal University and studying for a Master's degree. Her main research interests are wireless networks, wireless power transmission and evolutionary computation.

    Xin Zhang: received the B.Sc. degree from Ludong University in 2006, the M.Sc. degree from the Shandong University of Science and Technology in 2009, and the Ph.D. degree from the City University of Hong Kong in 2013. Since 2015, he has been a lecturer at Tianjin Normal University. He has published more than 50 technical papers, including over 30 papers in international journals. His main research interests are resource allocation, evolutionary computation, and machine intelligence.

    View full text