Elsevier

Computer Networks

Volume 76, 15 January 2015, Pages 1-16
Computer Networks

DynMAC: A resistant MAC protocol to coexistence in wireless sensor networks

https://doi.org/10.1016/j.comnet.2014.10.019Get rights and content

Abstract

The growth of mobile and ubiquitous computing has increased the demand for wireless communications, which in turn raises interference levels and spectrum pollution, causing problems of network coexistence. The coexistence assurance between these devices and wireless sensor networks is a big challenge. This paper proposes a new medium access protocol, DynMAC (Dynamic MAC), which uses mechanisms of dynamic channel reconfiguration, recovery from lost links and reconfiguration of transmission parameters based on the properties of the cognitive radios, to deal with this problem. Simulations and experiments using a real WSN testbed, were performed to validate our protocol. Results show that the proposed mechanisms solve the WSN configuration problems, in noisy and interference environments, and enable the coexistence with different networks and devices operating in the same frequency spectrum, while maintaining application requirements in critical deployment scenarios.

Introduction

The demand for ubiquitous and wireless devices has grown exponentially in recent years as more and more applications were created. However, the extensive use of these devices in the same location causes some problems as most of them use the same available unlicensed radio spectrum known as ISM (Industrial, Scientific and Medical) bands. Furthermore, other devices such as microwave ovens, remote controls, cordless phones, bluetooth communications, Hi-Fi and video systems also use ISM bands. Although most of these devices have small ranges and use signals with low amplitude, their interference in the ISM spectrum is not negligible.

In industrial scenarios, however, the effects induced by common ubiquitous mobile devices are lower due to the strict control imposed on these environments. Nevertheless, the interference problem still arises because of the multiple wireless devices working in the ISM spectrum such as sensor and actuator devices. Furthermore, as industrial applications demand very strict requirements regarding packet delay and packet loss, every interference may lower the expected quality of service (QoS). In the case of Wireless Sensor Networks (WSNs), which are more and more commonly used to replace old cabled monitoring and actuator systems, the coexistence of different networks and devices results in several problems that span from communication failures to inadmissible response times. These problems are especially serious in critical systems.

Wireless interference has been a widely research area. Studies in [1] showed that at some places the 2400 MHz frequency spectrum, which is used by several WSNs, has an occupation of 90%. In addition, [2] predicted that in the near future of 5–10 years the growth of wireless communication networks using the ISM bands would suffer overlapping problems, which may extensively affect WSNs. The coexistence of different networks and devices that operate in the same frequency, or in adjacent frequencies, may lead to harmful interference, which in turn limits the capabilities of the applications and, in some cases, results in the complete shutdown of those networks [3].

In trying to solve the problem of WSN coexistence, a vast amount of research was done and some standard recommendations were produced [4], [5]. However, current devices do not yet support most of these recommendations. The new IEEE 802.19 Wireless Coexistence Working Group was formed to deal with coexistence between unlicensed wireless networks and devices. The objective of this Working Group is to develop standards for guaranteeing the coexistence of future wireless devices (CA – Coexistence Assurance), i.e., to guarantee that wireless devices can operate in the same location without causing significant interference to each other.

Some proposals found in literature, targeting cellular networks and other wireless communication systems, use solutions based in Dynamic Spectrum Allocation (DSA). The mechanisms used in DSA include spectrum sensing, choosing the best channel/frequency available and dynamically reconfigure the device radio. These mechanisms have been used in cognitive or intelligent radios. Akan et al. [6] showed that these same techniques may be applied to WSN, mitigating coexistence problems.

To provide a reliable WSN for industrial environment with performance assurances, several studies and projects have been done. Among them is GINSENG project – Performance control in Wireless Sensor Networks [7], which tried to provide a solution for reliable and timely communication, while achieving energy efficient control, targeting WSN at critical industrial environments. The requirements that were found more relevant in these environments were packet delay and loss rate, which had to comply to the very strict performance boundaries. The GINSENG solution to these requirements was to develop a new reliable medium access protocol called GinMAC [8]. This protocol uses Time Division Multiple Access (TDMA) and contains specific mechanism to improve reliability and assure maximum delays. Also, it uses a specific topology control mechanism and implements message routing.

To validate the GINSENG proposal, a real testbed was installed at the oil refinery at Sines, Portugal. In spite of having good overall results, it was found that some problems existed due to the dynamic noise and interferences from the refinery environment, as stated in [9]. This misbehavior of the WSN existed even in a very controlled area with strict access of personal and wireless devices, within the oil refinery. The cause of the failures was the pollution of the 2400 MHz spectrum and the electromagnetic noise caused by engines and other machinery. In order to solve this type of problem it was necessary to manually monitor all the available channels under 2400 MHz frequency and identify the least interference channel to be used for the sensor network. While in most of the current deployments, the channel used by a WSN is manually set before the deployment, remaining static during the all network operation, it became clear that, in the refinery scenario, the manual solution to choose the best channel did not guarantee the network operation. The reason is that in this environment the noise and interference pattern may change with time, resulting in the unreliability of the network during its normal operation lifetime.

Employing techniques of cognitive radio in WSN is still a big challenge, aggravated by the fact that sensor nodes have resource constraints in terms of communication and processing capabilities. Thus, a cognitive based MAC protocol for WSN must take into account all these hardware limitations and still consider the critical time response essential in the industrial application scenarios. Moreover, the protocol should also be able to sense the wireless spectrum, choose the best channel and share that decision with neighbor nodes, while not increasing the packet loss, the energy consumption, or altering the time constraints of the communication.

This paper proposes a new medium access protocol, DynMAC (Dynamic MAC), which uses dynamic channel reconfiguration mechanisms in order to choose the best communication channel for WSN, while supporting the performance restrictions of critical systems. It has the same properties found in cognitive radios (sensing, decision, sharing and mobility of spectrum) and embeds characteristics of the GinMAC protocol developed under the GINSENG project. Furthermore, to maintain the network resilience, a mechanism was developed to automatically recover the nodes from connection losses.

While, in general, cognitive radios use the primary frequencies (licensed bands) on opportunistic mode to be used by the secondary users, our approach is different. In this paper, the main focus is the coexistence of WSN and devices that only use the ISM bands. In order to accomplish this objective, DynMAC protocol implements methods of classification and reconfiguration of channels. It uses the same techniques of cognitive radios but only uses one channel at a time. Therefore, the process of decision employed in DynMAC can be considered as part of a cognitive process, although this concept is still controversial for some authors. In order to evaluate the viability of DynMAC in critical scenarios, experiments were done using simulation and a real WSN testbed. Results showed that the mechanisms proposed in this paper can dynamically solve reconfiguration problems in WSNs operating in noisy and interference environments, while maintaining the application requirements. Furthermore, it enables the coexistence between the WSNs and wireless devices operating in the same frequency spectrum.

Based on the results it is expected that the DynMAC protocol can be used for any applications that require critical time boundaries and resilience. As WSN are expanding to a broader set of applications and scenarios, where QoS is essential, the importance of guaranteeing that the performance expected initially is maintained over the life of the network is of utmost importance. Although DynMAC was developed mainly to industrial environments there are other examples of scenarios where applications could benefit from the mechanisms studied for DynMAC such as health status monitoring of patients, smart environments, control and monitoring of electrical power, fire system monitoring, gas leaks, industrial plants, and any other general system that require responses in real time.

This paper is organized as follows. Section 2 presents the related works. The GinMAC protocol developed by GINSENG Project is described in Section 3. The characteristics and functionalities of DynMAC are presented in Section 4. Section 5 shows the experimental results using simulation and the real testbed. Finally, Section 6 presents the conclusions and future work.

Section snippets

Related work

WSNs deployed in industrial environments require strict performance control especially in regards to packet delay and loss. Most of the problems arise during packet transmission and relate to the transmission medium (e.g. signal path-loss, noise and interference) and poor hardware. Controlling the transmission power, improving the antennas and carefully placing transmitters and receivers can reduce the effects of noise, path-loss and interference. However, finding the optimal position in which

GINSENG project and GinMAC

The aim of GINSENG project is to propose, develop and deploy a solution for performance controlled WSNs that guarantee reliable and timely data delivery [7]. The targets of this project are critical environments such as the oil refinery in which the time response, packet loss and reliability are bounded by some constraints. In order to fulfill these requirements, the GinMAC protocol was developed to provide a reliable and energy efficient control for wireless sensor networks [8]. GinMAC assures

DynMAC: dynamic channel allocation protocol for WSN

The goal of the DynMAC protocol is to reduce the effects of adjacent and co-channel interference in WSNs installed in environments in which other wireless networks and devices exist. A model of noise that reflects the attenuation and interference imposed by the environment in the signals and packets transmitted between nodes, is shown in [28]. Generally, in these environments the frequency spectrum of WLAN may be very polluted and the level of noise and interference are high. Although the

Experiments and results

In order to evaluate the mechanisms implemented in the DynMAC protocol, the Contiki operating system, together with TelosB sensor nodes was used [31]. As a first step, DynMAC was evaluated using the COOJA simulator provided together with ContikiOS [32]. COOJA simulator was used to make a first evaluation of the settings that would be used later in the real testbed. The structure of the network used for experiments is shown in Fig. 3.

The topology of the WSN is a tree with one sink and 3 levels.

Conclusions and future work

This paper proposes a new MAC protocol, DynMAC, with mechanisms for dealing with the coexistence problem in WSNs. The mechanisms employed in this protocol are based on the cognitive radio functions such as spectrum sensing, analysis, decision, and sharing. The DynMAC was evaluated using both simulation and a real testbed. Experiments were done to test different functionalities of the proposed protocol. Specifically, tests were made to evaluate the ability of the DynMAC protocol for sensing and

Acknowledgment

This work was supported by CAPES in the post-doctoral (BEX 6668/10-0) and had the financial support of agency CNPq (process 306869/2012-8). The work presented also was partially financed by the iCIS Project (Grant CENTRO-07-ST24-FEDER-002003) and by the IST FP7 0384239 GINSENG.

Luiz Henrique Andrade Correia held post-doctoral stage at the University of Coimbra, Portugal (2012). He concluded the PhD in Computer Science for the Federal University of Minas Gerais in 2006, obtained Master’s Degree in Electrical Engineering at the Federal University of Itajubá (1995) and graduated in Electrical Engineering by the Federal University of Juiz de Fora (1990). Since 1997, he works as Associate Professor in the Federal University of Lavras. Currently, he heads a ubiquitous

References (37)

  • O.D. Incel et al.

    MC-LMAC: a multi-channel MAC protocol for wireless sensor networks

    Ad Hoc Netw.

    (2011)
  • M.A. McHenry, P.A. Tenhula, D. McCloskey, D.A. Roberson, C.S. Hood, Chicago spectrum occupancy measurements & analysis...
  • G. Zhou, J.A. Stankovic, S.H. Son, Crowded spectrum in wireless sensor networks, Workshop on Embedded Networked...
  • P. Ferrari et al.

    Coexistence of wireless sensor networks in factory automation scenarios

    Sensors Transd. J.

    (2008)
  • IEEE Std. 802.15.2, Recommended Practice for Telecommunications and Information exchange between systems. Local and...
  • T. Baykas et al.

    Standardization Activities of IEEE 802.19 Task Group 1: Wireless Coexistence in the TV White Space

    (2011)
  • O.B. Akan et al.

    Cognitive radio sensor networks

    Netw. Mag. Global Internetw.

    (2009)
  • FP7, Ginseng project, European Commission FP7, 2010....
  • P. Suriyachai et al.

    Time-critical data delivery in wireless sensor networks

  • T.-D. Tran et al.

    Characteristics of channels of IEEE 802.15.4 compliant sensor networks

    Wireless Personal Commun.

    (2011)
  • T. Rappaport

    Wireless Communications: Principles and Practice

    (2002)
  • IEEE Std. 802.15.4, Wireless Medium Access Control (MAC) and Physical Layer (PHY) Specifications for Low-Rate Wireless...
  • B. Dezfouli, M. Radi, S.A. Razak, T. Hwee-Pink, K.A. Bakar, Modeling low-power wireless communications, J. Netw....
  • W. Yuan, X. Wang, J.-P. Linnartz, A coexistence model of ieee 802.15.4 and ieee 802.11b/g, in: 2007 14th IEEE Symposium...
  • D. Cavalcanti, S. Das, J. Wang, K. Challapali, Cognitive radio based wireless sensor networks, in: Proceedings of 17th...
  • S. Nethi, J. Nieminen, R. Jantti, Exploitation of multi-channel communications in industrial wireless sensor...
  • S. Pollin, I. Tan, B. Hodge, C. Chun, A. Bahai, Harmful coexistence between 802.15.4 and 802.11: a measurement-based...
  • E. Toscano, L. Lo Bello, Cross-channel interference in ieee 802.15.4 networks, in: IEEE International Workshop on...
  • Cited by (23)

    • Adaptive and Beacon-based multi-channel protocol for Industrial Wireless Sensor Networks

      2019, Journal of Network and Computer Applications
      Citation Excerpt :

      Even different channels are used for different clusters, all communication inside the clusters occurs using only one channel, which is a single point of failure. Correia et al. (2015) proposed the DynMAC protocol, which uses dynamic channel reconfiguration, and is based on TDMA. The channel to be used by all nodes in the network (the global best channel) is defined by the sink node, using information collected from the nodes of the network.

    • Spectrum handoff strategy for cognitive radio-based MAC for real-time industrial wireless sensor and actuator networks

      2019, Computer Networks
      Citation Excerpt :

      Consequently, in the case of interference, either the network has to be formed again or a dedicated channel has to be used to recover connectivity. DynMAC [21], based on GinMAC, uses CR to get information about the best channel in its environment. Furthermore, it includes a mechanism to recover links lost due to interference.

    • Design, analysis and implementation of a time-bounded spectrum handoff algorithm for real-time industrial wireless sensor and actuator networks

      2019, Journal of Network and Computer Applications
      Citation Excerpt :

      Consequently, in case of interferences, either the network has to be formed again or a dedicated channel has to be used to recover connectivity. Finally, DynMAC (Correia et al., 2015), based on GinMAC (Suriyachai et al., 2010), uses CR to get information about the best channel in its environment. Furthermore, it includes a mechanism to recover lost links due to interferences.

    • Survey and systematic mapping of industrial Wireless Sensor Networks

      2017, Journal of Network and Computer Applications
      Citation Excerpt :

      Even different channels are used for different clusters, all communication inside the clusters occurs using only one channel, which is a single point of failure. Correia et al. (2015) proposed the DynMAC protocol, which uses dynamic channel reconfiguration. The MAC protocol is based on TDMA, and the channel to be used by all nodes in the network (the global best channel) is defined by the sink node of the network, using information collected from the nodes of the network, through RSSI samples or the Packet Error Rate (PER) calculated by the nodes.

    View all citing articles on Scopus

    Luiz Henrique Andrade Correia held post-doctoral stage at the University of Coimbra, Portugal (2012). He concluded the PhD in Computer Science for the Federal University of Minas Gerais in 2006, obtained Master’s Degree in Electrical Engineering at the Federal University of Itajubá (1995) and graduated in Electrical Engineering by the Federal University of Juiz de Fora (1990). Since 1997, he works as Associate Professor in the Federal University of Lavras. Currently, he heads a ubiquitous computing network group (GRUBi) with several projects in development in the areas: wireless sensor networks, cognitive radio and mobile networks, vehicular networks, security and QoS.

    Thanh-Dien Tran is currently working towards his PhD at the Laboratory of Communication and Telematics of the Department of Informatics engineering, University of Coimbra, Portugal. He received the BS Degree in Computer Science from Can Tho University, Vietnam in 1999, and MS Degree from the University of Glamorgan, Wales, UK in 2004. His research interests are Wireless Sensor Networks, Localization, Web of Thing, and Integrating physical and digital world. He is currently participating in a project at Eneida company in Portugal to find an appropriate solution for locating mobile nodes in sensor networks.

    Vasco Pereira is an Assistant Professor at the Department of Informatics Engineering of the University of Coimbra, and is also a PhD student at the same university. He is a member of the Laboratory of Communication and Telematics of the Centre for Informatics and Systems of the University of Coimbra (CISUC). His main research interests include Wireless Sensor Networks, QoS and evaluation of performance.

    João Carlos Giacomin holds a degree in Electrical Engineering from Federal University of Minas Gerais (1992), MS in Electrical Engineering in the area of Power Electronics, Federal University of Minas Gerais (1998) and PhD in Electrical Engineering in the area of Computer Engineering, Federal University of Minas Gerais (2007). He is currently an adjunct professor at the Federal University of Lavras, Brazil. He has experience in the area of Computer Engineering, with emphasis in Computer Systems, acting on the following subjects: electronic instrumentation, automatic control processes, agricultural instrumentation, microelectronics and wireless sensor networks.

    Jorge Sá Silva received his PhD in Informatics Engineering in 2001 from the University of Coimbra, where is an Assistant Professor at the Department of Informatics Engineering of the University of Coimbra and a Senior Researcher of Laboratory of Communication and Telematics, Portugal. His main research interests are Mobility, Network Protocols and Wireless Sensor Networks. He has been serving as a reviewer and publishing in top conferences and journals in his expertise areas. His publications include 2 book chapters and over 90 papers in refereed national and international conferences and magazines.

    View full text