Elsevier

Computer Networks

Volume 107, Part 2, 9 October 2016, Pages 339-352
Computer Networks

Low-power and lossy networks under mobility: A survey

https://doi.org/10.1016/j.comnet.2016.03.018Get rights and content

Abstract

With the creation of the Routing Over Low power and Lossy networks (ROLL) group, work centered on the Internet of Things (IoT) has been emerging. A routing protocol for Low-power and Lossy Networks (LLNs) named the IPv6 Routing Protocol for Low-Power and Lossy Networks (RPL) has been created recently, though it still has some issues, including its lack of responsiveness to mobility. This article surveys proposed mobility extensions to the RPL and analyzes how the mechanisms introduced affect the requirements for LLNs.

Introduction

In the Internet of Things (IoT), objects that surround us in our daily life will be connected to a network. With the emergence of these connected smart objects, there is a need to build a routing protocol that can run in resource-restrained devices and aims to reduce energy consumption because a considerable amount of nodes in an IoT scenario is battery powered. To this purpose, the Routing Over Low power and Lossy networks (ROLL) working group from Internet Engineering Task Force (IETF) was created, and they introduced the term Lossy Networks (LLNs) to to characterize these network scenarios.

To cope with the resource limitation of LLN networks, the Institute of Electrical and Electronics Engineers (IEEE) created a new Medium Access Control (MAC) protocol, IEEE 802.15.4. The limited capabilities introduced in this standard, such as the low MTU size (127 bytes) and low data rate (250 kbps) [3], put a great barrier in the implementation of IPv6-based routing protocols, as IPv6 has a minimum Maximum Transmission Unit (MTU) size of 1,280 bytes [9]. For this purpose, Low-power Wireless Personal Area Network (6LoWPAN) [24] creates an adaptation layer between the network and a data link to support IPv6 in LLNs.

In the ROLL working group, IETF studied the different application scenarios in which LLNs might be used. The simplest scenario is home automation, which aims to support our in-house daily life activities. Simple applications exist to monitor and control lightening and shutters, appliances or healthcare devices; more complex ones may be used to perform video surveillance, generate security alarms and provide overall energy optimization [5]. These applications use different sensors and actuators that may be battery operated or have quite limited resources. The majority of nodes in home networks are fixed in some part of the house. However, mobility is expected to arise mostly from healthcare devices or remote controls that are carried by people.

Similar to the home automation scenario, though applied to a much wider and more structured space, is the building automation scenario [28]. These areas of application involve mainly the control of air conditioning, elevators, lightning and shutters and the monitoring of fire sensing and security (i.e., motion detectors) devices. Such a variety of applications leads to a more diverse type of node and a higher traffic volume. Although most of the devices are fixed, some applications need to respond well to mobility.

In an industrial scenario, wireless devices are used to monitor and control devices that were not connected in the past [30]. Examples include sensors that report vibration levels on pumps, the state of a fuse or luminary and whether a man is down. Network attributes will vary greatly with the type of industry and the monitored devices; however, most application scenarios require networks of hundreds of devices that might be clustered into smaller networks around different sinks. Industrial applications can involve mobile sensors in containers or vehicles as well as workers carrying Personal Digital Assistants (PDAs). Thus, different types of mobility must be supported. Also, with higher speed values being introduced, the negative impact of mobility on signal quality [33] should be accounted for.

Finally, in an urban scenario, networks are expected to have thousands to millions of nodes that are usually separated through different sinks [10]. Areas of application can include sensors to measure municipal consumption (i.e., gas, water, electricity), meteorological, pollution and ambient data; actuators to control traffic and street lights also constitute an application area. Despite the network scale and diversity, currently, no mobility is foreseen for this particular type of application scenario.

Analysis of the different application scenarios has shown that the routing protocol for LLNs must be able to cope with resource limitation, scalability and Quality of Service (QoS) issues. Apart from the urban scenario, in which the nodes stand still, routing must also support low mobility in all remaining cases.

Several routing protocols have been proposed to handle these issues, and IPv6 Routing Protocol for Low-Power and Lossy Networks (RPL) was considered the de facto standard for routing over LLNs. The diversity of solutions leads to different surveys where the protocols are described and compared, mainly regarding their resource usage. Hence, a survey of the protocols designed for networks with low power motes is presented in [37]. Motes are sensor nodes, which are known for having limited memory and computational power. Also a comparison of RPL with previous energy-aware protocols is presented in [15].

Mobility support is a fundamental issue for the success of LLN. Traditional solutions are often too complex to be adequate for such resource-restrained and large-scale networks, and a routing protocol that can provide such support while considering these aspects will offer better performance. Although different mobility extensions have been proposed for RPL, there is no way to assess them in a unique framework.

This paper aims to compare different routing protocols for LLNs and assess the impact of mobility on the network performance when they are used.

The paper provides a structured study of RPL and the most relevant extensions that have been proposed by different authors to cope with mobility. Using one of the most used mobility models, a simulation study was conducted to assess, under the same framework, the performance of the different protocols under different mobility conditions. The goal was to identify how well they adapt to the requirements of the LLN application scenarios.

The remaining paper is structured as follows: Section 2 provides an overview of low-energy routing protocols and details the RPL protocol; Section 3 presents the mobile RPL extensions, providing an overall analysis at the end; Section 4 assesses their performance under different mobility conditions through a set of simulation studies; and Section 5 concludes the paper.

Section snippets

Routing over low power networks

Power-aware routing solutions and protocols have been presented for a long time, even before the creation of the ROLL group of IETF. Specific metrics and protocols have been proposed, especially for wireless sensor networks, with the aim of supporting nodes with energy consumption limitations. This section presents the most relevant approaches.

RPL mobility extensions

RPL was designed to support LLNs, although there are still some open issues that need to be resolved [32]. Performance studies have shown that the protocol provides a very fast network setup with a relatively high overhead [4] and that mobility support can be greatly improved [25]. Hence, several extensions and modifications have been proposed to solve these problems. Most of them attempt to fix the lack of responsiveness or excessive overhead (due repairs) in mobile scenarios.

The next sections

Simulation

To assess and compare RPL with the mobility extensions, a set of simulation studies were conducted. Extensions to the RPL protocol were developed on top of the ContikiRPL implementation, which runs in the Contiki [1] operating system. The simulations were implemented using COOJA simulator [2], as it is the simulator used to simulate networks with motes running Contiki. The next section details the simulation studies.

Conclusion

This paper surveys routing protocols for LLNs, considering RPL and the most relevant mobility extensions that have been proposed. For each protocol, it analyses the different phases and identifies the different mechanisms and messages that are introduced by the extensions. As is expected, some mechanisms introduce trade-offs mainly in simplicity and scalability against QoS and mobility awareness.

To assess these trade-offs, RPL and some of its extensions were simulated in COOJA under different

Acknowledgments

This work was supported by national funds through Fundaçãopara a Ciência e a Tecnologia (FCT) with reference UID/CEC/50021/2013.

Afonso Oliveira is a PhD student at Instituto Superior Técnico (IST), from University of Lisbon, and is also a Junior Research Fellow in the Communication Networks and Mobility group at INESC-ID, an Associate Laboratory of Portugal. He received his MSc from IST in computer networks with a thesis on design and optimization of optical networks. His current research interests include networks running on low-power such as low-power and lossy networks (LLNs) and body area networks (BANs).

References (40)

  • H. Fotouhi et al.

    MRPL: boosting mobility in the internet of things

    Ad Hoc Netw.

    (2015)
  • O. Gaddour et al.

    {RPL} in a nutshell: a survey

    Comput. Netw.

    (2012)
  • Contiki: the open source operating system for the internet of things, (http://www.contiki-os.org/). [Online; accessed...
  • The cooja network simulator, (http://www.contiki-os.org/). [Online; accessed...
  • Ieee standard for local and metropolitan area networks–part 15.4: low-rate wireless personal area networks (lr-wpans)

    IEEE Std 802.15.4-2011 (Revision of IEEE Std 802.15.4-2006)

    (2011)
  • N. Accettura et al.

    Performance analysis of the RPL routing protocol

    Mechatronics (ICM), 2011 IEEE International Conference on

    (2011)
  • A. Brandt, J. Buron, G. Porcu, Home automation routing requirements in low-power and lossy networks, 2010, (RFC 5826...
  • Z. Cao et al.

    S-AODV: sink routing table over AODV routing protocol for 6lowpan

    Networks Security Wireless Communications and Trusted Computing (NSWCTC), 2010 Second International Conference on

    (2010)
  • I. Chakeres, C. Perkins, Dynamic manet on-demand (dymo) routing, draft-ietf-manet-dymo-14 (work in...
  • C. Cobarzan et al.

    Analysis and performance evaluation of rpl under mobility

    Computers and Communication (ISCC), 2014 IEEE Symposium on

    (2014)
  • S. Deering, R. Hinden, RFC 2460 Internet protocol, version 6 (IPv6) specification, Internet Engineering Task Force,...
  • M. Dohler, T. Watteyne, T. Winter, D. Barthel, Routing requirements for urban low-power and lossy networks, 2009, (RFC...
  • R. Draves et al.

    Comparison of routing metrics for static multi-hop wireless networks

    SIGCOMM Comput. Commun. Rev.

    (2004)
  • H. Fotouhi et al.

    Smart-hop: a reliable handoff mechanism for mobile wireless sensor networks

    Wireless Sensor Networks

    (2012)
  • O. Gaddour et al.

    Co-RPL: RPL routing for mobile low power wireless sensor networks using corona mechanism

    Industrial Embedded Systems (SIES), 2014 9th IEEE International Symposium on

    (2014)
  • C. Gomez et al.

    Impact of LQI-based routing metrics on the performance of a one-to-one routing protocol for ieee 802.15.4 multihop networks

    EURASIP J. Wirel. Commun. Netw.

    (2010)
  • W.R. Heinzelman et al.

    Energy-efficient communication protocol for wireless microsensor networks

    System Sciences, 2000. Proceedings of the 33rd Annual Hawaii International Conference on

    (2000)
  • K.-S. Hong et al.

    Dag-based multipath routing for mobile sensor networks

    ICT Convergence (ICTC), 2011 International Conference on

    (2011)
  • K. Kim et al.

    Dynamic MANET on-demand for 6LoWPAN (dymo-low) routing

    Internet Eng. Task Force, Internet-Draft

    (2007)
  • K. Kim et al.

    6LoWPAN ad hoc on-demand distance vector routing (load)

    Netw. WG Internet Draft

    (2007)
  • Cited by (61)

    • Secured location-aware mobility-enabled RPL

      2023, Journal of Network and Computer Applications
      Citation Excerpt :

      When a node experiences high mobility, it does not wait for the next DIO period to update its parent list table, so it sends DIS messages. The simulation results in Oliveira and Vazao (2016) show that ME-RPL and MMRPL have low responsiveness in terms of topology changes, and ME-RPL produces more control traffic with the increase in the rate of the mobile nodes, which may be due to more use of global repairs. Also, the Packet Delivery Rate (PDR) of MMRPL is less than ME-RPL and RPL.

    • RPL routing protocol over IoT: A comprehensive survey, recent advances, insights, bibliometric analysis, recommendations, and future directions

      2022, Journal of Network and Computer Applications
      Citation Excerpt :

      Furthermore, when compared to existing algorithms, the obtained results provide robustness in mobile networks with minimal control message overhead. The proposal's main flaws are its higher memory consumption and insufficient support for the downward traffic pattern (Oliveira and Vazao, 2016). The authors of (Korbi et al., 2012) introduced ME-RPL.

    View all citing articles on Scopus

    Afonso Oliveira is a PhD student at Instituto Superior Técnico (IST), from University of Lisbon, and is also a Junior Research Fellow in the Communication Networks and Mobility group at INESC-ID, an Associate Laboratory of Portugal. He received his MSc from IST in computer networks with a thesis on design and optimization of optical networks. His current research interests include networks running on low-power such as low-power and lossy networks (LLNs) and body area networks (BANs).

    Teresa Vazão is an associate professor at the Department of Electrical and Computer Engineering, at Instituto Superior Técnico (IST), from University of Lisbon, and is also a senior researcher at INESC-ID, an Associate Laboratory of Portugal. She received her PhD from IST in electrical and computer engineering, with a specialization in computer networks. She has developed research in the field of communications networks and her currently research interests are focused in sensor and vehicular networks, with a special emphasis in routing and performance aspects. She had supervised more than 40 students. Her research work, resulted in international publications in different forums, international journals, conferences and patents. She coordinated INESC-ID participation in several European projects, like MIAVITA – Wireless Sensor Network for Volcanic Monitoring and SENSATION - Advanced Sensor Development for Attention, Stress, Vigilance and Sleep/Wakefulness Monitoring. She also participated, as team member, in several international and national projects.

    View full text