A hybrid clustering and evolutionary approach for wireless underground sensor network lifetime maximization
Introduction
A wireless underground sensor network includes sensor nodes buried completely at a certain depth [3]. These nodes gather various information from the environment, which is transmitted to one or multiple aboveground data sinks via wireless signals like radio or infrared waves.
WUSNs have many real-life applications. Akyildiz et al. [3] proposed a taxonomy of WUSNs applications, which are divided into four classes. The first class of applications is environmental monitoring. This is often seen in agriculture to control irrigation and monitor underground soil conditions such as water and mineral content [3], [23], [31], [32]. The second class is monitoring infrastructure including pipes, electrical wiring, and liquid storage tanks as well as military applications such as minefield monitoring [3], [16]. The third class is object localization. Many civilian and military applications need to identify objects in a specific area, such as locating people in the event of a building collapse or patients in hospital [3], [9]. The last class is border patrol and security monitoring, where WUSNs can be used to monitor the above ground presence and movement of people or objects [3].
WUSNs have many characteristics that are different from other types of wireless sensor networks. Because sensors are buried underground, the underground wireless channel is influenced by soil properties such as volumetric water content, particle size, density and temperature. Moreover, signals are attenuated when transmitting through multiple environments, primarily caused by the absorption, reflection and refraction at the surface. There are three different communication channels in WUSNs including underground to underground channel (UG-UG), underground to aboveground channel (UG-AG) and aboveground to aboveground channel (AG-AG) [1], [2], [4]. The propagation characteristics of electromagnetic waves in soil are significantly different from those in air. The attenuation in soil depends on several environmental parameters, such as the soil composition in terms of sand, silt, and clay fractions, the bulk density, the particle density, and the volumetric water content. In addition, EM waves encounter much higher attenuation in soil, which severely hampers the communication distance. This is the most important challenge for WUSNs’ lifetime. Network lifetime is the time span from deployment to the instant when the network is considered non-functional. The concept of non-functional is, however, application-specific. In [22], there were three definitions of topological lifetime based on the criticality of a mission:
- •
lifetime: the mission fails when a node in N nodes run out of energy,
- •
lifetime: the mission fails when nodes in N nodes run out of energy,
- •
lifetime: the mission fails when m supporting nodes in N nodes are alive and nodes in N run out of energy.
In this work, we only consider problems in the first definition ( lifetime), which seeks to maximize the lifetime for all network nodes. For example, for applications in earthquake prediction, underground sensor nodes (SNs) are deployed in a large area to track changes in soil properties. If a single sensor shuts down, the collected information is no longer accurate and the whole network becomes non-functional.
In WUSNs, sensor nodes send their data towards base stations or sinks, whose locations are usually fixed. However, sensors are often located far away from their base stations/sinks, and can rapidly exhaust energy. Moreover, it is often impossible to recharge or replace the sensors’ power sources after deployment. Consequently, power-conservation is one of the most prioritized tasks in WUSNs. A viable method is to deploy a layer of relay nodes, positioned close to underground sensors as to reduce the transmission loss from the sensors to the sink. Relay nodes play an important role in the effective and successful deployment of WUSNs. In wireless sensor networks, relay nodes are used for maximizing the network lifetime, energy-efficient data gathering, load-balanced data gathering as well as making the network fault-tolerant [28].
In this paper, we study the problem of relay node placement optimization to prolong the lifetime of wireless sensor networks. The contributions of this paper can be summarized as follow:
- •
Firstly, we present the network model, energy model, and mathematical model of the problem of minimizing the maximum transmission loss between a sensor-relay pair;
- •
Secondly, we prove that the problem is NP-hard as it can be reduced to the Set Cover Problem [8];
- •
Thirdly, we formulate the problem into a mixed integer programming model as a basis to achieve bound solutions. However, the exact Mixed Integer Linear Programming (MILP) solvers have exponential time complexity and do not scale well to large instances;
- •
Finally, we propose two methods to solve the Network Lifetime Optimization (NLO) problem. In the first method, we propose a new clustering-based heuristic for the Load Unrestricted Relay Node Selection (LURNS) problem. We then reformulate the Load Balanced Sensor Node Assignment (LBSNA) problem as a Maximum Flow with Min-max cost (MXF-MC) problem, which can be solved in polynomial time. The output for MXF-MC will satisfy three requirements: uniquely connected sensor-relay pairs, load-balanced relays and maximum network lifetime (for the sensor nodes). Based on the first method, we propose a evolutionary algorithm-based method for further improvement.
The rest of the paper is organized as follows. In Section 2 we review related works. Section 3 is for problem definition and models. Our main contributions will be shown in Section 4 and Section 5. Section 6 presents our evaluations of the proposed methods on benchmark datasets. Finally, we conclude our work and present future works in Section 7.
Section snippets
Related work
There are several works that focus on maximizing the lifetime of wireless sensor networks as well as wireless underground sensor networks. We briefly review them in this section.
Many routing algorithms have used mobile sinks to increase network lifetime [17], [30]. The authors in [30] studied approaches to reduce delay and prolong the lifetime of wireless sensor network, using an efficient routing protocol based on mobile sinks and virtual infrastructure. They proposed an algorithm based on a
Problem formulation
In this paper, we consider the problem of placing and assigning relay nodes in a 2-level wireless underground network, where sensors send data to base stations/sinks via relay nodes. The sensor nodes are assumed to have been deployed, and there is a known finite set of positions where relays can be deployed.
Our goal is to maximize the network’s lifetime by maximizing the shortest sensor lifetime (N-of-N). This in turn becomes the problem of minimizing energy usage. A sensor expends energy for
Min-Max relay placement problem and its NP-hardness
We provide an NP-hard proof for the Min-Max Relay Placement Problem using reduction to the Set Cover Problem. Theorem 1 The Min-Max Relay Placement Problem is NP-hard. Proof We shall present a polynomial time reduction to the Set Cover Problem [8] to Min-Max Relay Placement Problem. Set Cover is a well-known NP-complete problem. Given a “universe” set U with n items, and a collection P1 of m sets whose union equals the universe, the set cover problem
Proposed methods
Yuan et al. [33] proposed a 2-phase approach to the problem. The first phase (Load Unrestricted Relay Node Selection - LURNS) attempts to choose y relay nodes without satisfying the load balance constraint. The second phase (Load Balanced Sensor Node Assignment - LBSNA) then proceeds to reassign relays and sensors to satisfy this constraint. 5 heuristics were also proposed, including LURNS-1 and LURNS-2 for the first phase, and LBSNA-1, LBSNA-2 and LBSNA-3 for the second phase.
By adopting the
Set up
In this section, we consider the following objectives:
- •
Evaluate the performance of different heuristic algorithms, including CluRNS followed by MXF-LB and those proposed in [33];
- •
Evaluate the performance of MXFGA when applied to different experimental scenarios;
- •
Evaluate the effect of CluRNS initialization on MXFGA’s performance;
- •
Evaluate the time/complexity tradeoff of our proposed methods.
Two datasets are created for our experiments. The large dataset is generated according to the configurations
Conclusion
Wireless underground sensor networks have many applications in real world such as in military applications, testing soil properties and water content, pollution control and detection of natural disasters. Despite their potential, the development of WUSNs have faced challenges from the complex underground environment, some of which are related to the communication channel between underground and aboveground nodes. Moreover, the scalability and network lifetime of a WUSN are limited by energy and
Conflict of 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.
Acknowledgment
This research is funded by Vietnam National Foundation for Science and Technology Development (NAFOSTED) under grant number DFG 102.01–2016.03 and Shenzhen Peacock Plan (Grant no. KQTD2016112514355531).
References (35)
- et al.
Wireless underground sensor networks: research challenges
Ad Hoc Netw.
(2006) - et al.
Signal propagation techniques for wireless underground communication networks
Phys. Commun.
(2009) - et al.
An analysis on separability for memetic computing automatic design
Inform. Sci.
(2014) - et al.
Relay node placement in large scale wireless sensor networks
Comp. Comm.
(2006) Reducing delay and prolonging the lifetime of wireless sensor network using efficient routing protocol based on mobile sink and virtual infrastructure
Ad Hoc Netw.
(2019)- et al.
Optimal relay placement for lifetime maximization in wireless underground sensor networks
Inf. Sci.
(2017) - et al.
Terahertz channel modeling of underground sensor networks in oil reservoirs
2012 IEEE Global Communications Conference (GLOBECOM)
(2012) - et al.
Wireless underground sensor networks: channel modeling and operation analysis in the terahertz band
Int. J. Antennas Propag.
(2015) - et al.
Multicriteria adaptive differential evolution for global numerical optimization
Integ. Comp.-Aid. Eng.
(2015) - et al.
Relay sensor placement in wireless sensor networks
Wirel. Netw.
(2008)
A greedy heuristic for the set-covering problem
Math. Oper. Res.
A review of passive wireless sensors for structural health monitoring
Mod. Appl. Sci.
Algorithm for solution of a problem of maximum flow in networks with power estimation
Soviet Math. Doklady
Theoretical improvements in algorithmic efficiency for network flow problems
Advanced border intrusion detection and surveillance using wireless sensor network technology
Int. J. Comm. Netw. Syst. Sci.
Flows in networks
On energy provisioning and relay node placement for wireless sensor networks
IEEE Trans. Wireless Comm.
Cited by (22)
Node depth Representation-based Evolutionary Multitasking Optimization for Maximizing the Network Lifetime of Wireless Sensor Networks
2024, Engineering Applications of Artificial IntelligenceAn energy consumption optimization strategy for Wireless sensor networks via multi-objective algorithm
2024, Journal of King Saud University - Computer and Information SciencesOptimizing wireless sensor network lifetime through K-coverage maximization and memetic search
2023, Sustainable Computing: Informatics and SystemsA Q-Learning-based distributed routing protocol for frequency-switchable magnetic induction-based wireless underground sensor networks
2023, Future Generation Computer SystemsCitation Excerpt :Despite extensive efforts on this issue, as its most critical challenge, the routing decision has not been explored and discussed enough, which places a big limit on frequency-switchable MI-WUSN’s potential development and practicability. In fact, there exists some works trying to explore the routing-decision issue of WUSNs without regard to the frequency-selective property [27–30]. However, in frequency-switchable MI-WUSN, communication link’s feasibility depends on not only the distance between nodes but also its operating frequency selection, which makes most of the existing routing schemes low efficiency and even invalid [31–35].
Multi-objective teaching–learning evolutionary algorithm for enhancing sensor network coverage and lifetime
2022, Engineering Applications of Artificial IntelligenceCitation Excerpt :Yet, these works still did not account for sensors’ energy constraints, which can be daunting to satisfy. The authors in Tam et al. (2019a, b) and Yuan et al. (2017) considered the network lifetime of WUSNs by optimizing relay nodes placement and ignored the coverage problem. Heuristics and meta-heuristics were proposed to find the approximate solutions.
Multifactorial evolutionary optimization to maximize lifetime of wireless sensor network
2021, Information Sciences