Sensor Topology Optimization by Applying Neural Network Configuration

Version	Summary	Created by	Modification	Content Size	Created at	Operation
1		COSTAS CHAIKALIS	--	1685	2023-06-14 11:13:18	\|
2	Reference format revised.	Lindsay Dong	Meta information modification	1685	2023-06-15 04:06:09	\|

This entry is adapted from the peer-reviewed paper 10.3390/s23125422

In dense IoT deployments of wireless sensor networks (WSNs), sensor placement, coverage, connectivity, and energy constraints determine the overall network lifetime. In large-size WSNs, it is difficult to maintain a trade-off among these conflicting constraints and, thus, scaling is difficult. The neural network dynamically proposes and handles sensor placement coordinates in a 2D plane, having the ultimate goal to extend network lifetime. Simulation results show that algorithm improves network lifetime, while maintaining communication and energy constraints, for medium- and large-scale deployments.

sensor placement topology WSNs network lifetime neural networks

1. Introduction

Sensor node deployment is a key design issue related to planning a wireless sensor network (WSN) and is closely related to the domain application requirements and the total energy consumption and robustness of the network. An optimal deployment may reduce communication costs and extend the total network lifetime, under certain constraints imposed by the networking elements and the application. In this way, a WSN can better fulfil its operational tasks, such as sensing and evaluating physical phenomena, data transmission, and inference ^[1].

Sensor deployment methods are closely related to specific applications ^[2] and can be of two types: deterministic and random. In the first type, the nodes’ coordinates are determined a priori by the network design team; this type is preferred in cases where the environment is not harsh. For example, in a precision agriculture application, the sensors’ locations may coincide with trees’ locations ^[3]. On the other hand, in cases where the physical phenomenon is mobile, the environment is harsh, or there are quick alterations in the sensing values, a random deployment is proposed ^[1]^[2]. To achieve the desired coverage ratio, redundant nodes are present, which means that the design is resistant to several node failures. For example, a random deployment is better in cases of measuring chemical gases inside a volcano, when estimating sudden weather changes, or when a cultivation is thick ^[3]^[4]. In the case of random deployments, the coverage constraint requirements may be not so strict. Comparing the two deployment types, one can say that the deterministic case needs careful planning, time, and resources. Moreover, the size of the set holding all possible network topologies is an exponential function of the size of the used sensor nodes. Therefore, deterministic planning is not always cost-effective. On the other hand, in random deployments, there is a high risk of network outage or partition in cases of total ad hoc deployments ^[5].

2. Sensor Topology Optimization by Applying Neural Network Configuration

A WSN designer places sensor nodes inside a field of interest (FoI) according to the application coverage requirements. These nodes, due to their transceivers, have the ability to cover either small or medium areas by sending data to a gateway node. The energy required for data communication is usually greater than the circuitry energy consumption ^[6]. In the majority of WSN and IoT applications, nodes are battery-equipped ^[7]. However, in cases where there is great difficulty in replacing batteries (i.e., harsh terrain, a large volume of nodes, etc.), researchers’ interest focuses on mechanisms to extend the network’s lifetime under communication and coverage constraints. Topology control plays a dominant role in achieving this target. Thus, optimal sensor node placement within the FoI is needed for cost-effective deployment.

The authors in ^[8] propose a sensor placement algorithm, which utilizes a biologically inspired optimization technique to imitate the behavior of territorial predators in marking their territories with their odors, known as the territorial predator scent marking algorithm (TPSMA). The TPSMA technique is based on maximizing a coverage objective optimization function. The problem of determining the location of sensor nodes, such that the terrain target points are covered and the network lifetime is maximized, is called the sensor deployment problem (SDP). Various heuristics and approximation algorithms have been proposed to solve the SDP in WSNs. Two improved versions of the particle swarm optimization (PSO) algorithm are presented in ^[9]. The first one is a cooperative PSO and the second is its improved version, applying fuzzy logic. These approaches do not deal with maximizing the coverage area or prolonging the network lifetime.

A properly designed and applied sensor deployment strategy improves WSN performance and resource management ^[10]. The coverage ratio is directly influenced by the deployment strategy. In principle, there is no positive correlation between energy consumption minimization and coverage maximization. Thus, to maximize the coverage area, sensor nodes should be placed far away from the sink node (SN) or the gateway node, which takes higher transmission power to reach the SN and raises the total energy consumption. This problem is partially solved if an energy-efficient and multihop routing algorithm is applied ^[11] instead of a one-hop communication pattern. However, in several precision agriculture applications ^[3], the terrain’s structure may block multihop communication patterns (i.e., tall trees) and, thus, the one-hop pattern along with an energy- efficient topology design may solve the energy problem.

Following on from this, nature-inspired optimization algorithms are adopted by many researchers in WSN applications ^[12]. While genetics, ants, and particle swarm algorithms are the dominant examples, many others emerge regularly such as the flower pollination algorithm (FPA), which is a novel global optimization algorithm inspired by the pollination process of flowers ^[12]. Based on the multiobjective version of the FPA (i.e., MOFPA) for WSNs, a new approach is proposed in ^[13]. This approach aims to find the optimal topology deployment, taking into consideration conflicting objectives, such as total energy minimization and total coverage maximization, while maintaining connectivity constraints.

The network’s lifetime primarily depends on the total energy consumption, which is mainly related to the nodes’ radio electronics energy consumption. An energy-efficient coverage optimization technique, the Voronoi–glowworm smarm optimization–K-means algorithm, is presented in ^[14]. In this approach, the Voronoi cell structure enhances the area coverage, applying the minimum required number of active nodes. The Voronoi diagram is one of the most famous computational geometrical structures applied in sensor topology design problems to ensure coverage extension ^[15]. Following on from this, a sensor node deployment technique is proposed in ^[16] as a constrained multiobjective optimization (MOO) problem. The proposed algorithm is a multiobjective evolutionary algorithm (MOEA), known as MOEA/D-DE, that uses a decomposition approach and employs differential evolution (DE). The aim is to find a sensor node deployment to maximize the coverage rate, minimize the network energy consumption, maximize the network lifetime, and minimize the number of deployed sensor nodes, while ensuring connectivity between each sensor node and the sink node for proper data transmission. A tree structure between the deployed nodes and the sink node for data transmission is assumed in this approach.

Heuristic approaches are appropriate for near-optimal solutions, especially in NP-complete problems. To this end, a metaheuristic algorithm is proposed in ^[17], which determines the sensors’ positions for area coverage maximization. As a part of the solution, the authors utilized the immune plasma algorithm (IPA), a unique metaheuristic algorithm that is based on the immune plasma therapy concept and the transfer of an antibody-rich fraction of blood from previously recovered patients to others who are considered critically ill. This optimization algorithm is used to determine the optimal solution to the maximum coverage problem, but does not deal with the minimum number of nodes, network lifetime extension, or connectivity assurance.

The authors in ^[18] propose an approach which relies on exploiting the building information modeling (BIM) database to obtain real-time and valid information about a target area. The majority of the proposed schemes present possible solutions without taking into account the terrain structure and potential obstacles during the WSN topology design process. The proposed solution can be integrated as a plugin within BIM tools in order to optimize sensor deployment in real time, taking into account both nodes and obstacles, respectively. In order to optimize WSN deployment, after collecting useful data from sensor nodes and the BIM database, this approach relies on an evolutionary algorithm to solve the multiobjective problem for coverage, cost, and lifetime. The output is each sensor’s optimal location in the smart building application.

One of the major challenges in sensor deployment is to find the trade-off between conflicting network optimization objectives under certain connectivity constraints. As proposed in ^[19], an approach to deal with this is a constrained Pareto-based multiobjective evolutionary approach (CPMEA). It aims to find Pareto-optimal layouts which maximize the coverage and minimize sensors’ energy consumption to prolong the network lifetime, while maintaining full connectivity between each sensor node and the gateway. To cover any type of FoI with a predefined number of sensors, a genetic algorithm is proposed in ^[20] with the purpose of finding the best sensor placement while ensuring maximum network coverage under sensor connectivity constraints. The authors propose the genetic algorithm for area coverage maximization (GAFACM), which covers all shapes of areas for a given number of sensors and finds the best positions to maximize coverage within the FoI, while ensuring connectivity between the sensor nodes.

An additional WSN deployment approach, based on the gradient method and the simulated annealing (SA) heuristic algorithm, is proposed in ^[21], using the minimum number of sensor nodes. However, the work does not deal with maximizing the network lifetime, whereas in ^[22], the SA heuristic algorithm is applied along with an energy-efficient algorithm to arrange the placement of sensors in order to extend the network lifetime. The main function of WSNs is to gather the required information, process it, and send it to remote gateways. A large number of sensor nodes need to be deployed within a field of interest; therefore, finding the best node placement according to several constraints is a hard problem to solve because it escalates. Recent studies in ^[23] focus on solving the deployment problem by applying heuristic and metaheuristic optimization algorithms. In approximately 35% of these studies, the authors apply an improved version of swarm optimization algorithms to solve the sensor deployment problem under constraints. However, network scalability and total energy consumption are not always addressed.

The topology optimization problem in large IoT and WSN deployments is a combinatorial and NP-hard problem to solve in polynomial time. The majority of existing algorithms apply heuristic or nature-inspired rules to reduce the search number of potential problem solutions within a solution space, so as to obtain a suboptimal solution in polynomial time.

References

Jiang, P.; Dong, L.; Pang, X. Deployment Strategy of Wireless Sensor Networks: A Survey. In Proceedings of the 2016 6th International Conference on Machinery, Materials, Environment, Biotechnology and Computer, Tianjin, China, 11–12 June 2016.
Pavithra, L.; Abdullah, M.; Prakash, S.P.; Karthick, S.; Ragavi, B.; Nandhini, V.D. Wireless Sensor Networks: A Review on Sensor Deployment and Routing Protocols for Different Application. In Proceedings of the IOP Conference Series: Materials Science and Engineering, Sanya, China, 12–14 November 2021; Volume 1084.
Oliver, S.T.; González-Pérez, A.; Guijarro, J.H. Adapting Models to Warn Fungal Diseases in Vineyards Using In-Field Internet of Things (IoT) Nodes. Sustainability 2019, 11, 416.
Mao, J.; Jiang, X.; Zhang, X. Analysis of node deployment in wireless sensor networks in warehouse environment monitoring systems. J. Wirel. Commun. Netw. 2019, 2019, 288.
Balister, P.; Kumar, S. Random vs. Deterministic Deployment of Sensors in the Presence of Failures and Placement Errors. In Proceedings of the IEEE INFOCOM 2009, Rio de Janeiro, Brazil, 19–25 April 2009; pp. 2896–2900.
Wang, Q.; Hempstead, M.; Yang, W. A Realistic Power Consumption Model for Wireless Sensor Network Devices. In Proceedings of the 2006 3rd Annual IEEE Communications Society on Sensor and Ad Hoc Communications and Networks, Reston, VA, USA, 25–28 September 2006; pp. 286–295.
Peng, M.; Chenb, H.; Xiaoa, Y.; Ozdemir, S.; Vasilakos, A.V.; Wue, J. Impacts of Sensor Node Distributions on Coverage in Sensor Networks. J. Parallel Distrib. Comput. 2011, 71, 1578–1591.
Husna, Z.A.; Norashidah, D. Sensor Node Placement in Wireless Sensor Network Based on Territorial Predator Scent Marking Algorithm. ISRN Sens. Netw. 2013, 2013, 170809.
Yarinezhad, R.; Hashemi, S.N. A sensor deployment approach for target coverage problem in wireless sensor networks. J. Ambient. Intell. Hum. Comput. 2020, 14, 5941–5956.
Li, R.; Liu, X.; Xie, W.; Huang, N. Deployment-Based Lifetime Optimization Model for Homogeneous Wireless Sensor Network under Retransmission. Sensors 2014, 14, 23697–23723.
Ekhlas, H.; Heba, A. Energy Consumption Analyzing in Single Hop Transmission and Multi-Hop Transmission for using Wireless Sensor Networks. Al-Khwarizmi Eng. J. 2018, 14, 156–163.
Jiao, W.; Tang, R.; Xu, Y. A Coverage Optimization Algorithm for the Wireless Sensor Network with Random Deployment by Using an Improved Flower Pollination Algorithm. Forests 2022, 13, 1690.
Hajjej, F.; Ejbali, R.; Zaied, M. Multi Objective Nodes Placement Approach in WSN Based on Nature Inspired Optimization Algorithms. In Proceedings of the ALLSENSORS 2017: The Second International Conference on Advances in Sensors, Actuators, Metering and Sensing, Nice, France, 19–23 March 2017.
Chowdhury, A.; De, D. Energy-efficient coverage optimization in wireless sensor networks based on Voronoi-Glowworm Swarm Optimization-K-means algorithm. Ad Hoc Netw. 2021, 122, 102660.
Juli, V.V.; Selvi JA, V. Genetic Algorithms and Their Tandem Application in Wireless Sensor Network Deployment. Res. J. Appl. Sci. Eng. Technol. 2014, 8, 2426–2438.
Sengupta, S.; Das, S.; Nasir, M.D.; Panigrahi, B.K. Multi-objective node deployment in WSNs: In search of an optimal trade-off among coverage, lifetime, energy consumption, and connectivity. Eng. Appl. Artif. Intell. 2013, 26, 405–416.
Taşdemir, A.; Demirci, S.; Aslan, S. Performance Investigation of Immune Plasma Algorithm on Solving Wireless Sensor Deployment Problem. In Proceedings of the 9th International Conference on Electrical and Electronics Engineering (ICEEE), Alanya, Turkey, 29–31 March 2022; pp. 296–300.
Zaimen, K.; Brahmia, M.E.A.; Dollinger, J.F.; Moalic, L.; Abouaissa, A.; Idoumghar, L. An Overview on WSN Deployment and a Novel Conceptual BIM-based Approach in Smart Buildings. In Proceedings of the 7th International Conference on Internet of Things: Systems, Management and Security (IOTSMS), Paris, France, 14–16 December 2020; pp. 1–6.
Khalesian, M.; Delavar, M.R. Wireless sensors deployment optimization using a constrained Pareto-based multi-objective evolutionary approach. Eng. Appl. Artif. Intell. 2016, 53, 126–139.
Tossa, F.; Abdou, W.; Ansari, K.; Ezin, E.C.; Gouton, P. Area Coverage Maximization under Connectivity Constraint in Wireless Sensor Networks. Sensors 2022, 22, 1712.
El Khamlichi, Y.; Tahiri, A.; Abtoy, A.; Medina-Bulo, I.; Palomo-Lozano, F. A Hybrid Algorithm for Optimal Wireless Sensor Network Deployment with the Minimum Number of Sensor Nodes. Algorithms 2017, 10, 80.
Xenakis, A.; Foukalas, F.; Stamoulis, G.; Katsavounidis, I. Topology control with coverage and lifetime optimization of wireless sensor networks with unequal energy distribution. Comput. Electr. Eng. 2017, 64, 182–199.
Abdulwahid, H.M.; Mishra, A. Deployment Optimization Algorithms in Wireless Sensor Networks for Smart Cities: A Systematic Mapping Study. Sensors 2022, 22, 5094.

© Text is available under the terms and conditions of the Creative Commons Attribution (CC BY) license; additional terms may apply. By using this site, you agree to the Terms and Conditions and Privacy Policy.

Upload a video for this entry

Information

Subjects: Telecommunications

Contributors MDPI registered users' name will be linked to their SciProfiles pages. To register with us, please refer to https://encyclopedia.pub/register :

George Papastergiou

Apostolos Xenakis

Costas Chaikalis

Dimitrios Kosmanos

Periklis Chatzimisios

Nicholas S. Samaras

View Times: 267

Update Date: 15 Jun 2023

Table of Contents

Video Upload Options

Confirm

1. Introduction

2. Sensor Topology Optimization by Applying Neural Network Configuration

References