STETS: A novel energy-efficient time synchronization scheme based on embedded networking devices

https://doi.org/10.1016/j.micpro.2015.07.006Get rights and content

Abstract

Time synchronization is essential in the implementation of large-scale Wireless Sensor Networks (WSNs). However, many approaches of time synchronization suffer from high communication overheads when pursuing high accuracy. Such overheads cause significant shrinkage of the lifetime of WSNs since frequent data communications consume much energy which is extremely limited in each sensor node. The energy consumption increases rapidly with the growth of WSNs density. In this paper, we present a Spanning Tree-based Energy-efficient Time Synchronization (STETS) which effectively incorporates two time synchronization schemes: Sender to Receiver Protocol (SRP) and Receiver to Receiver Protocol (RRP). It reduces the communication overheads while still maintaining high accuracy. In our approach, backbone sensor nodes form a spanning tree and they get synchronized layer by layer through SRP. Other nodes get synchronized through RRP by only listening to the communication between backbone sensor nodes. We evaluated the performances by simulating our approach on NS-2 and implementing it on embedded networking devices STM32W108 with simple MAC protocol stack. The experiment results show that our approach is efficient in both energy consumption and accuracy of time synchronization. Especially, it can get better performances in densely connected WSNs.

Introduction

In recent years, limiting energy consumption and maintaining network lifetime have become the focus of scientific research, development and application for Wireless Sensor Networks (WSNs) [1], [2], [3]. It has been predicted in ‘2020 computing: Everything, everywhere’ [4] and ‘US National Research Council report (Embedded Everywhere)’ [5] that WSNs play an irreplaceable role in future technology due to its satisfactory features as diverse as self-organization, good concealment and high fault tolerance, and its wide range of applications in domains such as military [6], [7], ecosystem [8], [9], preservation of cultural relics [10], and living environment [11], [12], [13], [14], [15]. A majority of WSN applications require that sensor nodes in a certain area have a common time scale so that they can accomplish their tasks such as [16], [17], [18], [19], [20], [21]. Therefore, study on time synchronization has become more and more important. However, WSNs have many weaknesses such as limited energy, limited handling capacity, limited communication capacity, and limited storage capacity. These problems bring difficulty to the researches on time synchronization and thus should be considered when designing time synchronization algorithm. In practical application of WSNs, there exist a large number of closely distributed nodes which result in further expansion of network size. For the large scale WSNs, there are amounts of communication data in processing of time synchronization. Especially, time synchronization becomes increasingly difficult in dense WSNs. At the same time, it causes too much consumption of energy. Therefore, time synchronization approaches based on a large scale of nodes in dense WSNs face to serious challenge.

Time synchronization has become a research focus in recent years. There are two types of Time synchronization schemes which are centralized system and distributed system. It is very important to clarify the differences in methods of obtaining time between centralized system and distributed system. In traditional centralized system, a procedure can obtain time via system call to the kernel. Due to this, records of the order and time in which procedures execute are determined [16]. However, in distributed system, procedure runs in more than one CPUs which cannot be set time to at the same time. This results in different initial time offset [17] among the CPUs. Also, these CPUs are driven by crystal oscillators of which frequency can be different. This slight but non-ignorable difference brings in time drift among CPUs. Such initial time offset and time drift cooperate to cause different local clocks of sensor nodes deployed randomly in WSNs. In many sensor networks such as highway speed system, data detected by different nodes agglomerate into one meaningful result via data fusion. To get accurate data, time synchronization plays an essential role in such sensor networks. In order to achieve high accuracy, many early proposed time synchronization algorithms, such as [17], [18], [22], [23], [24], focus so much on cutting down time error that they also cause serious energy consumption either due to redundant network traffic or over-complicated computing. In the past few years, we have done some related works on time synchronization [25] and modeling for WSNs [26]. In this paper, we present a Spanning Tree-based Energy-efficient Time Synchronization (STETS) that aims at providing a not only energy-efficient but also accurate time synchronization approach for WSNs.

The rest of this paper is organized as follows: we first review related works in Section 2. In Section 3, we present the system model of our algorithm. In Section 4, we give our algorithm design including the layering process and the time synchronization process. In Section 5, we simulate STETS, Time synchronization based on spanning tree for wireless sensor networks (TSBST) and TPSN which are all based on spanning tree for WSNs on NS-2 and present comparisons on time error and energy consumption. We present an implementation of STETS on a hardware platform generally available to users. Finally, we offer our conclusion and describe our plan for future work in Section 6.

Section snippets

Related works

From a practical perspective, there are a considerable number of time synchronization approaches in WSNs literature. FTPS in [27] requires the modification of link layer protocol and this is not desirable for many applications on application layer. A majority of other studies can be separated in two time synchronization types, namely receiver to receiver protocol (RRP) [18], [28], [29] and sender to receiver protocol (SRP) [17], [22], [23], [24]. All of these approaches tend to deal with the

Assumption

We assume that the sensor nodes have unique identifiers. Each node is aware of the set of nodes with which it can directly communicate. These nodes are also termed as the neighboring nodes. Moreover, we also assume that each node knows the distance to each of its neighboring nodes. This can be achieved through various distance estimation algorithms such as TOA (Time of Arrival) [36], TDOA (Time Difference of Arrival) [37], and RSSI (Received Signal Strength Indicator) [38].

Symbols and description

The symbols and

BNs and PNs election

In our approach, the BNs and PNs election process is divided into three phases. The execution of following phases is shown in Fig. 3.

  • Phase 1: at the beginning of the round of time synchronization, the randomly elected b0,0 broadcasts a Sync message to all the UNs in the set B0,0. All the UNs setup a timer after they receive the Sync message. These processes are shown in Fig. 3(a). The initial value of the timer will be discussed in the next subsection.

  • Phase 2: when the timer of one certain UN

Experiments and performance evaluation

There are two experiments: One is simulation on network simulator NS-2 and the other is real experiment on hardware platform STM32W108.

Conclusion

Most of the time synchronization approaches in WSNs suffer from high communication overheads when pursuing high accuracy. For instance in TPSN based on SRP, at least 3 times as many messages as sensor nodes are needed during one single round of time synchronization. Energy consumption should be highly concerned when designing time synchronization approach.

In this paper, we addressed this problem and proposed an approach, namely STETS based on spanning tree structure for WSNs. It combines two

Acknowledgments

This work was supported in part by Natural Science Foundation of P.R. China (Grant Nos. 61202443 and 61402078), and the Fundamental Research Funds for the Central Universities.

Tie Qiu is an Associate Professor and Ph.D. in computer science and technology at Dalian University of Technology, China. His research interests cover embedded system architecture, Internet of things and systems modeling. Tie Qiu received the Ph.D. degree and Master degree at Dalian University of Technology (DUT) in 2006 and 2012, respectively. Now he is an Associate Professor in software school at the Dalian University of Technology of China. His research interests cover Embedded System

References (39)

  • B. Sundararaman et al.

    Clock synchronization for wireless sensor networks: a survey

    Ad Hoc Networks

    (2005)
  • G. Fersi et al.

    Distributed Hash table-based routing and data management in wireless sensor networks: a survey

    Wireless Networks

    (2013)
  • A. Aziz et al.

    A survey on distributed topology control techniques for extending the lifetime of battery powered wireless sensor networks

    Commun. Surveys Tutorials, IEEE

    (2013)
  • M.A. Razzaque et al.

    Compression in wireless sensor networks: a survey and comparative evaluation

    ACM Trans. Sens. Networks (TOSN)

    (2013)
  • D. Butler

    2020 computing: everything, everywhere

    Nature

    (2006)
  • D. Estrin

    Embedded Everywhere: A Research Agenda for Networked Systems of Embedded Computers

    (2001)
  • M.P. Durisic et al.

    A survey of military applications of wireless sensor networks

  • D. Kim, H. Park, S. Yoo, Performance evaluation of routing protocols for wireless sensor networks in military...
  • C.J. Rankine et al.

    Optical wireless sensor networks observe leaf phenology and photosynthetic radiation interception in a Brazilian tropical dry forest

  • K. Smarsly

    Agricultural ecosystem monitoring based on autonomous sensor systems

  • L. Dong et al.

    Wireless sensor networks in relic protection: deployment methodology and cross-layer design

    High Technol. Lett.

    (2009)
  • H. Martín, A.M. Bernardos, L. Bergesio, P. Tarrio, Analysis of key aspects to manage wireless sensor networks in...
  • W. Suntiamorntut et al.

    An elderly assisted living system with wireless sensor networks

  • K. Park et al.

    Human behavioral detection and data cleaning in assisted living environment using wireless sensor networks

    Proceedings of the 2nd International Conference on PErvasive Technologies Related to Assistive Environments

    (2009)
  • J. Heidemann et al.

    Research challenges and applications for underwater sensor networking

    (2006)
  • M. Chen et al.

    A survey of recent developments in home M2M networks

    Commun. Surveys Tutorials, IEEE

    (2014)
  • M.L. Sichitiu et al.

    Simple, accurate time synchronization for wireless sensor networks

    (2003)
  • J. Elson, L. Girod, D. Estrin, Fine-grained network time synchronization using reference broadcasts, in: Proceedings of...
  • J. Wan et al.

    Context-aware vehicular cyber-physical systems with cloud support: architecture, challenges, and solutions

    Commun. Mag., IEEE

    (2014)
  • Cited by (34)

    • Data quality management for service-oriented manufacturing cyber-physical systems

      2017, Computers and Electrical Engineering
      Citation Excerpt :

      Advancements in vertical, horizontal and end-to-end integration covered by Industry 4.0 have led to the widespread use of SOM enterprises in the manufacturing industry [2,3]. Emerging technologies which have ushered in the SOM phenomenon include cyber-physical system (CPS) [4], wireless sensor networks [5,6], cloud computing [7,8], Internet of Things [9,10] and big data [11]. Of these technologies, an efficient CPS with a well-designed structure and intelligent management level is a remarkable feature of competitive SOM enterprises.

    • Concept drift detection for data stream learning based on angle optimized global embedding and principal component analysis in sensor networks

      2017, Computers and Electrical Engineering
      Citation Excerpt :

      As the core components, sensor networks play a significant role in the intelligent monitoring and managements in IIoT. Due to their microminiaturization and intelligence, sensor networks have been widely applied in many practical fields [1], e.g. Internet of Vehicles, mobile phones, health and military etc. In [2], multi-path self-organizing strategy have been comprehensive analyzed and illustrated in IIoT.

    • Cost-efficient traffic-aware data collection protocol in VANET

      2017, Ad Hoc Networks
      Citation Excerpt :

      Over the past decades, we have witnessed the development of automotive electronics [1] and IoTs [2].

    View all citing articles on Scopus

    Tie Qiu is an Associate Professor and Ph.D. in computer science and technology at Dalian University of Technology, China. His research interests cover embedded system architecture, Internet of things and systems modeling. Tie Qiu received the Ph.D. degree and Master degree at Dalian University of Technology (DUT) in 2006 and 2012, respectively. Now he is an Associate Professor in software school at the Dalian University of Technology of China. His research interests cover Embedded System Architecture, Internet of Things, Wireless and Mobile Communications. He published four books and 30 papers. He is a senior member of China Computer Federation (CCF), a Member of IEEE and a member of ACM.

    Lin Chi received B.E. and Ph.D. from Dalian University of Technology, China, in 2008 and 2013, respectively. He has been a lecturer in School of Software, Dalian University of Technology (DUT), China, since 2014. His research interests include pervasive computing, cyber-physical systems (CPS), and wireless sensor networks.

    Weidong Guo is a Master student in School of Software, Dalian University of Technology. His research interests cover embedded system and wireless sensor networks.

    Yushuang Zhang is a Master student in School of Software, Dalian University of Technology. His research interests cover embedded system and internet of things.

    View full text