Elsevier

Ad Hoc Networks

Volume 71, 15 March 2018, Pages 104-116
Ad Hoc Networks

Connectivity analysis of underground sensors in wireless underground sensor networks

https://doi.org/10.1016/j.adhoc.2018.01.002Get rights and content

Abstract

Wireless underground sensor networks consist of sensors that are buried under the ground and communicate through soil medium. Due to channel characteristics, the connectivity analysis of wireless underground sensor networks is more complicated than that in the traditional over-the-air wireless sensor networks. This paper focuses on analyzing the connectivity of underground sensors in wireless underground sensor networks in terms of the probability of node isolation and path probability which captures the effects of the environment parameters such as soil moisture and soil composition, and system parameters such as sensor node density and propagation techniques. Throughout this paper, both qualitative and quantitative comparisons between electromagnetic wave system and ordinary magnetic induction system for underground communications are provided. More specifically, we derive the exact closed-form mathematical expressions for the probability of node isolation of these two communication systems and validate the correctness of analytical models through simulations. We also provide the simulation-based path connectivity of these two communication systems. The results obtained in this paper provide useful guidelines on the design of reliable wireless underground sensor networks.

Introduction

Wireless underground sensor networks (WUSNs), which is one of the categories in wireless sensor networks (WSNs) [1], consist of wireless sensors buried under the ground, constitute one of the promising areas and enable many important applications such as intelligent agriculture, pineline fault diagnosis, mine disaster rescue, oil exploration, and earthquake disaster prediction [2]. The main difference between WUSNs and the terrestrial WSNs is communication medium, which is no longer air but soil, rock, or water with electrolytes. Consequently, the buried sensor nodes can only communicate wirelessly through soil medium.

There has been a tremendous amount of researches on the channel characteristic anaylysis of WUSNs. In [3], [4], [5], the underground signal propagation for electromagnetic (EM) wave system has been analyzed in detail. The effect of environment parameters such as volumetric water content and system parameters such as operating frequency and burial depth of sensors on the path loss and bit error rate is obtained. The authors of [6] experimentally investigate the link quality characteristics of three communication channels in WUSNs, i.e., the underground-to-underground channel, the underground-to-aboveground channel, and the aboveground-to-underground channel. However, the results of research work gradually show that using EM wave propagation in WUSNs causes some problems [7]. First, EM waves experience high levels of attenuation due to absorption by soil, rocks, and underground water. Second, the environment parameters may affect the performance of communication unpredictably since these factors change with location and vary dramatically over time. Third, operating frequencies in lower ranges are necessary to achieve a practical transmission range. Thus, compared with the communication range, the antenna size will become too large to be deployed underground. Therefore, despite its promising applications, the deployment of WUSNs using EM wave propagation techniques is challenging.

Recent years have witnessed the development of magnetic induction (MI) communication. While techniques using EM wave in underground environment encouter three major problems as mentioned above, MI is an alternative technique which turns out to be an effective communication technology for complex environments such as underground [8] and underwater [9], [10]. In [8], the detailed analysis on the path loss and bandwith of MI system in underground soil medium is provided. In addition, based on the channel model analysis, MI waveguide techniques for communication are developed in order to reduce path losses of traditional EM wave system and original MI system. Similarly, the authors of [11] describe the channel model for both EM wave and MI systems. From that, the multimode model is provided to characterize wireless channels for wireless communication in underground mines and road/subway tunnels.

Because of the complex channel characteristics, the connectivity analysis in WUSNs is much more complicated than that in terrestrial wireless sensor networks. The connectivity in terrestrial homogenous ad hoc networks has been thoroughly analyzed. The author in [12] provides an analytical framework for the calculation of stochastic topological connectivity from different viewpoints: local connectivity and overall network connectivity. Based on this framework, the connectivity of ad hoc networks in shadow fading environment is presented in [13] and the connectivity of cognitive radio ad hoc networks is presented in [14]. The connectivity of wireless sensor network for sandstorm monitoring is analyzed in [15]. Four types of channels that a sensor can utilize during sandstorms are analyzed, which include air-to-air channel, air-to-sand channel, sand-to-air channel, and sand-to-sand channel. Based on these analytical results, the percolation-based connectivity analysis is performed. For underground environment, the authors of [16] and [17] mathematically express the path loss of three channels, i.e. underground–underground (UG–UG) channel, underground–aboveground (UG–AG), and aboveground–underground (AG–UG) channel. Then, the corresponding transmission ranges are calculated from these path loss expressions. Finally, the lower and uppper bounds of the connectivity probability in WUSNs are analytically derived from these channel characteristics, together with the network and mobility models. In [18], the connectivities of WUSNs under varying environment parameters are captured by modeling the distribution of the number of given size clusters and estimating the aboveground communication coverage. An analysis of the path loss and the bandwidth of the MI system in underground soil medium is provided in detail in [8] and [19]. Based on the MI channel characteristics, the MI waveguide technique for communication is developed to reduce the path loss of ordinary MI system. The performance in terms of path loss of the EM wave system, the ordinary MI system, and the MI waveguide system are quantitatively compared. The authors of [20] presents the system architecture and operation framework of MISE-PIPE, a MI-based wireless network for underground pipeline monitoring. This pipeline monitoring technique is believed to provide a low-cost solution for effective real-time leakage detection. In [21], two algorithms are proposed to deploy the MI waveguides to connect the underground sensors in WUSNs. More specifically, the minimum spanning tree (MST) algorithm is used to minimize the number of relay coils. However, the WUSN constructed by the MST algorithm is only 1-connected. Thus, it is not robust against sensor failure. To enhance the network robustness with an acceptable number of relay coils, the triangle centroid (TC) algorithm is used.

In this paper, we focus on the connectivity of underground sensors in wireless underground sensor networks by deriving in detail the exact closed-form mathematical expression for the probability of node isolation and providing the simulation-based path connectivity under the effects of environment parameters and system parameters of both EM wave and MI communication systems. In addition, different from [21] which proposes two algorithms for deploying MI waveguide to improve the connection among underground sensors in WUSNs, we explore another deployment concept. That is, by generating a huge number of random topologies with different sensor locations, we investigate the impact of random initial deployment of underground sensors on their connections and compare the connectivity among underground sensors when they use EM wave and MI for communication. We also notice that for a given sensing area with deployed sensor nodes under the ground, the network may not remain connected over time since both the environment parameters and the properties of soil dynamically change through different seasons in a year. A naive solution is to design and deploy the network for the worst case to guarantee the connectivity. However, this solution is not practical because in the real applications, the environment conditions change dynamically over time and location. Therefore, it is impossible to determine the worst case; and if the costs of deployment and maintenance are considered, the extremely high density of sensor nodes required to keep the network connected in the worst conditions is unacceptable. The results in this paper can help network designers to select an optimal density of underground sensor nodes to maintain stable network connection.

To give a better understanding of our motivation in this paper, we describe two communication scenarios in a WUSN consisting of deployed underground sensors and aboveground stations as illustrated in Fig. 1. In a scenario, due to the change of environment conditions over time, node N1 becomes an isolated node since it has no connection to any neighbor sensor nodes or aboveground stations. In another scenario, node N2 is an underground sensing node (source node) which has no direct connection to any aboveground stations. However, the collected data can be transmitted to one of the aboveground stations through multi-hop connection with the final sensor node, N3 (destination node). From these scenarios, we can intuitively conclude that the connectivity of underground sensors is very essential for the whole system functionalities. Based on this observation, in this paper, we are interested in the question: How do the combined effects of soil properties such as soil moisture and soil composition, and system parameters such as sensor node density and signal propagation techniques affect the connectivity characteristic of underground sensor nodes in WUSNs? To be more specific, considering the initial random deployment of sensor nodes, we focus on how the isolation probability of a randomly chosen node and the possibility of establishing communication path between two arbitrary nodes selected as a source and a destination are impacted by the various environment and system parameters.

The contributions of this paper are summarized as follows.

  • 1.

    Based on [3], [8], [22] and [23], we present thoroughly the expressions of path losses in underground EM wave and MI communication systems. Especially, we provide both qualitative and quantitative comparisons of the path losses of two systems in the same settings of network parameters through various illustrative sample topologies and numerical results to clearly show the effect of environment parameters on the differences of path losses between two systems.

  • 2.

    Contrary to [16] and [17] which provide the lower and upper bounds of the connectivity probability of WUSNs, we focus on the connection among underground sensor nodes and derive the exact closed-form expression of the probability of node isolation from the node degree perspective for both EM wave and MI communication systems. We show that both environment parameter such as volumetric water content and system parameter such as operating frequency significantly affect the probability of node isolation in EM wave communication system. For a given fixed node density λ, the probability of node isolation increases as the operating frequency and volumetric water content increase. Under the assumption that the permeability of the medium is constant, however, the probability of node isolation in the MI communication system is not affected by environment parameters.

  • 3.

    We also give a comparison of path probability between EM wave communication and MI communication systems. We find that the MI communication system gives higher path probability than the EM wave communication system with the same network conditions. Even when soil is very dry, the path probability in the MI communication system is still a bit higher than that in the EM communication system.

The remainder of this paper is organized as follows. In Section 2, the signal properties in underground environment are presented, which consider both EM wave propagation and MI communication. In Section 2.2, the connectivity features of WUSNs are mathematically derived. The numerical results are provided in Section 4. Finally, the paper is concluded in Section 5.

Section snippets

Underground signal propagation

Unlike the over-the-air communication, the performance of wireless sensor communication in soil medium is significantly influenced by environment conditions such as soil composition and soil moisture. Besides that, the burial depth of sensor nodes also affects the channel characteristics. However, in this paper, the sensor nodes are assumed to be buried in soil deep enough so that the influence of the reflection caused by the burying depth is negligible and the single-path model is used. In the

Connectivity analysis

In this section, we formulate the problem of connectivity analysis in WUSNs for both EM wave communication system and MI communication system. We consider both the local network connectivity (probability of node isolation) and the overall network connnectivity (path probability).

Numerical results

We evaluate the probability of node isolation and path probability of WUSNs by running the simulation in Matlab. The simulation is conducted with different settings of system parameters such as propagation technique, operating frequency and sensor density, and environment parameter such as volumetric water content. In each evaluating scenario, sensor nodes are randomly placed in a sensing region. All simulation results are calculated by averaging over 1000 Monte Carlo simulation trials. For

Conclusion

In this paper, we have studied the effects of environment parameters and system parameters on the connectivity of underground sensors in WUSNs. Both EM wave communication system and MI communication system have been considered. We have provided the preliminaries of underground signal propagation, taken into account the path loss, and from that we have derived the exact analytical closed-form expression for probability of node isolation. The accuracy of our mathematical analysis has been

Acknowledgment

This work was supported by the National Research Foundation of Korea (NRF) grant funded bythe Korea government (MSIP) (No. 2017R1A2B4001801).

Hoang Thi Huyen Trang received the Telecommunications and Electrical Engineering degree from Posts and Telecommunications Institute of Technology in 2014 and Master degree in Computer Science from Hongik University in 2017. Her main research area has been connectivity in wireless networks.

References (32)

  • SunZ. et al.

    Magnetic induction communications for wireless underground sensor networks

    IEEE Trans. Antennas Propag.

    (2010)
  • I.F. Akyildiz et al.

    Realizing underwater communication through magnetic induction

    IEEE Commun. Mag.

    (2015)
  • M.C. Domingo

    Magnetic induction for underwater wireless communication networks

    IEEE Trans. Antennas Propag.

    (2012)
  • C. Bettstetter

    On the connectivity of ad hoc networks

    Comput. J.

    (2004)
  • C. Bettstetter et al.

    Connectivity of wireless multihop networks in a shadow fading environment

    Wirel. Netw.

    (2005)
  • LiuJ. et al.

    Connectivity of two nodes in cognitive radio ad hoc networks

    Proceedings of the 2013 IEEE Wireless Communications and Networking Conference, WCNC

    (2013)
  • Cited by (42)

    • A Q-Learning-based distributed routing protocol for frequency-switchable magnetic induction-based wireless underground sensor networks

      2023, Future Generation Computer Systems
      Citation Excerpt :

      One of the core issues in frequency-switchable MI-WUSNs is how to design the data collection schemes to fulfill its innate duty so the network can be optimized for various purposes [17,18]. From existing related researches, it easily follows that this issue includes the directions of transceiver design [19–21], deployment strategy [22,23], cross layer protocol [13,24], and multi-hop networking [25,26]. 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.

    View all citing articles on Scopus

    Hoang Thi Huyen Trang received the Telecommunications and Electrical Engineering degree from Posts and Telecommunications Institute of Technology in 2014 and Master degree in Computer Science from Hongik University in 2017. Her main research area has been connectivity in wireless networks.

    Le The Dung received the B.S. degree in Electronics and Telecommunication Engineering from Ho Chi Minh City University of Technology, Vietnam, in 2008, the M.S. degree and the Ph.D. degree in Electronics and Computer Engineering from Hongik University, Korea, in 2012 and 2016, respectively. From 2007 to 2010, he joined Signet Design Solutions Vietnam as hardware designer. He has been with Chungbuk National University as a postdoctoral research fellow from May 2016. Dr. Dung has 33 papers in referred international journals and conferences. He has served as a Technical Program Committee Member and a reviewer of many international conferences and journals. He is the recipient of the IEEE IS3C2016 Best Paper Award. His major interests are routing protocols, network coding, network stability analysis and optimization in mobile ad-hoc networks, cognitive radio ad-hoc networks, and visible light communication networks. He is a member of the IEEE.

    Seong Oun Hwang received the B.S. degree in mathematics in 1993 from the Seoul National University, the M.S. degree in computer and communications engineering in 1998 from the Pohang University of Science and Technology, and the Ph.D. degree in computer science from the Korea Advanced Institute of Science and Technology. He worked as a software engineer at the LGCNS Systems, Inc. from 1994 to 1996. He worked as a senior researcher at the Electronics and Telecommunications Research Institute (ETRI) from 1998 to 2007. Since 2008, he has been working as an associate professor with the Department of Computer and Information Communication Engineering, Hongik University, Korea. His research interests include cryptography, cyber security and mobile network. He is a member of the IEEE.

    View full text