Elsevier

Ad Hoc Networks

Volume 5, Issue 3, April 2007, Pages 324-339
Ad Hoc Networks

A survey on real-world implementations of mobile ad-hoc networks

https://doi.org/10.1016/j.adhoc.2005.12.003Get rights and content

Abstract

Simulation and emulation are valuable techniques for the evaluation of algorithms and protocols used in mobile ad-hoc networks. However, these techniques always require the simplification of real-world properties such as radio characteristics or node mobility. It has been shown that this may lead to results and conclusions which do not reflect the behavior of ad-hoc networks in the real world. Various prototype implementations demonstrate that even simple protocols such as flooding do not behave as it was predicted by earlier simulation. To overcome this problem, real-world experiments are required. In this paper, we present a survey on existing real-world implementations of mobile ad-hoc networks. We report on the technology used for the implementations as well as on key findings from experiments conducted with these implementations.

Introduction

Mobile ad-hoc networks (MANETs) enable mobile users to communicate without the use of a fixed infrastructure. These networks can be used, e.g., to extend the range of access points, to allow communication in disaster areas or to realize inter-vehicle communications. There are a lot of technical challenges in designing MANETs, and for a lot of those challenges, solutions have been presented.

A central problem in this area of research is to prove that a given solution is viable and, possibly, to demonstrate its superiority in relation to other approaches. An established and widely used method for this purpose is network simulation. However, it has become apparent that simulation can only be a first step in the evaluation of algorithms and protocols for MANETs. The key reason for this is threefold:

  • Simulations always require certain assumptions about the real world. These may turn out to be wrong or too coarse to capture all aspects that influence the performance of algorithms and protocols.

  • Some important characteristics of MANETs, like radio propagation or energy consumption, are inherently hard to model accurately in simulators.

  • Simulations do not allow the solutions to be tested in the environment they were designed for.

As a consequence, some of the algorithms and protocols for mobile ad-hoc networks have been implemented and studied in the real wold. Given the effort that is required for real-world implementations it is very surprising that there is a very large amount of duplicated work in this area. It is the key aim of this survey to increase the reuse of prior work by giving a concise summary of existing real-world implementations and pointing out the most important results that have been gained through the experimental evaluation of ad-hoc networks.

In the remainder of this paper we concentrate on the issues connected to the real-world implementation of MANETs. Nevertheless, there are other evaluation techniques such as simulation or emulation and other types of multi-hop radio networks such as mesh and sensor networks. Both will be touched upon as far as this contributes to the better understanding of real-world MANET implementations.

The paper is structured as follows: starting with an overview of the historical development of mobile ad-hoc networks in Section 2 we continue with a classification of techniques for the evaluation of MANETs in Section 3. One of these techniques, emulation, is briefly outlined in Section 4. Experiments conducted with sensor and mesh networks are examined in Section 5. Section 6 investigates real-world experiments with MANETs. The key findings of the experiments are summarized in Section 9. Section 10 highlights the advantages of integrating simulation, emulation and real-world experimentation. An outlook to future directions of research for real-world MANET implementations concludes the paper.

Section snippets

Historical development

Research on multi-hop wireless networks (which were initially called packet radio networks) started in the early 1970’s. The ALOHA [1] project at the University of Hawaii was among the first demonstrations of feasibility for using packet broadcasting in a single-hop system. Based on the knowledge acquired through ALOHA, the DARPA funded PRNET project [39], [36] was started in 1973. PRNET was a multihop Packet Radio NETwork system that reached a size of around 50 nodes and allowed some nodes to

Overview of different evaluation techniques

After an initial theoretical analysis, algorithms and protocols for ad-hoc networks can be evaluated by simulation, emulation and real-world experiments. These methods are necessary to prove or disprove assumptions and to identify interprotocol and interlayer effects which are hard to discover when examining a protocol or an algorithm in a purely analytical way. As a general rule, the number of assumptions influences the accuracy of the results: the fewer assumptions are required by a method

Emulation

An emulator is a combination of soft- and hardware used to mimic the behavior of a network with some of its components being implemented in the real world and others being simulated. There is a lot of published work about emulators for wired and wireless networks, here we focus on emulators for wireless ad-hoc networks. The purpose of those emulators varies, some are built to allow to test protocols on real hardware, others are used to prepare real-world experiments. In the latter case

Sensor networks

Sensor networks consist of small, low-power, low-energy (stationary) nodes used for monitoring parameters such as temperature, humidity, and motion. Algorithms and protocols for these networks often focus on energy conservation and techniques for data aggregation. However, sensor networks are wireless multi-hop networks, therefore they do share some fundamental problems with mobile ad-hoc networks.

In [89] a sensor network consisting of up to 60 nodes was used to measure the packet delivery rate

MANET experiments

Experiments on MANETs have been conducted on static topologies and in scenarios involving node mobility. The following two sections summarize the findings of those experiments.

Testbeds

A testbed is a framework which supports testing, comparing and evaluating algorithms and protocols in the real world.

The only existing testbed for mobile ad-hoc networks used to a larger extent is the ad-hoc protocol evaluation testbed (APE) [52]. APE is a Linux distribution which can be booted directly from CD on regular notebooks. Each experiment participant is instructed to move according to a choreography script. To a certain extent this makes experiments repeatable. Furthermore, the

Software tools

Software tools are intended to ease the task of implementing and evaluating algorithms and protocols for ad-hoc networks. They can be roughly divided into:

  • frameworks,

  • monitoring tools,

  • performance metrics.

Frameworks support the task of implementing MANETs. The PICA API [10] and the “user level framework for ad-hoc routing” [3] shadow the calls to operation system specific functions. With this, a protocol can be developed once and used on different operating systems without porting the

Summary of results

Even though the existing experiences with real-world implementations of mobile ad-hoc networks are quite heterogeneous, there are several observations that can be generalized:

  • A lot of available links in a wireless network are asymmetric. This has been shown for sensor networks [12], [31], mesh networks [17], [18], [19] and MANETs [44].

  • In has been shown for sensor networks that the direction of an asymmetric link can be switched by switching the positions of the two affected nodes [12].

  • Distance

Integrating simulation, emulation and real-world experimentation

Cost8 increases from simulation over emulation to real-world experimentation. If protocols were implemented in a way that allows simulation as well as emulation and real-world experimentation with the same code basis, the advantages can be combined while the disadvantages are avoided: (1) Without the need to reimplement the same code for each step, the work load and also

Conclusions and outlook

The wealth of unanticipated results and information gained through real-world experiments shows that protocols and algorithms for mobile ad-hoc networks must be evaluated in real-world settings. Simulation and emulation are valuable tools but they cannot replace experiments.

At the same time experimentation is not yet a mature methodology in the context of ad-hoc networks: results are often non reproducible and hard to explain. In most cases it is nearly impossible to validate the measurements

Wolfgang Kiess studied computer science at the University of Mannheim from 1997 to 2003. In the academic year 2000/2001 he stayed at the University of Nice Sophia Antipolis, France. He received the M.S. degree in March 2003, the subject of his diploma thesis was “Hierarchical Location Service for mobile ad-hoc networks”. Currently, he is a Ph.d. student at the University of Düsseldorf.

References (89)

  • G. Anastasi et al.

    Understanding the real behavior of mote and 802.11 ad hoc networks: an experimental approach

    Pervasive Mobile Comput.

    (2005)
  • N. Abramson, The ALOHA system—another alternative for computer communications, in: Proc. of the Fall 1970 AFIPS...
  • D. Aguayo et al.

    Link-level measurements from an 802.11b mesh network

    SIGCOMM Comput. Commun. Rev. (CCR)

    (2004)
  • J. Allard, P. Gonin, M. Singh, G.G. Richard, A user level framework for ad hoc routing, in: Proc. of the 27th Annual...
  • How to build, install and run the APE testbed. TeX-file in the APE distribution. Available from:...
  • A. Arora, R.R. et al. ExScal: Elements of an extreme scale wireless sensor network, in: Proc. of the 11th Int. Conf. on...
  • D.A. Beyer, Accomplishments of the DARPA SURAN program, in Proc. of the Military Communications Conference (MILCOM),...
  • J. Bicket, D. Aguayo, S. Biswas, R. Morris, Architecture and evaluation of an unplanned 802.11b mesh network, in: Proc....
  • R. Bruno et al.

    Mesh networks: Commodity multi-hop ad hoc networks

    IEEE Commun. Mag.

    (2005)
  • C.T. Calafate, R.G. Garcia, P. Manzoni, Optimizing the implementation of a MANET routing protocol in a heterogeneous...
  • D. Cavin, Y. Sasson, A. Schiper, FRANC: A lightweight java framework for wireless multihop communication, Technical...
  • A. Cerpa, N. Busek, D. Estrin, SCALE: A tool for Simple Connectivity Assessment in Lossy Environments, Technical Report...
  • W. Chao, J.P. Macker, J.W. Weston, Naval research lab’s mobile network emulator, January 2003. Avaialble from:...
  • K.-W. Chin et al.

    Implementation experience with MANET routing protocols

    SIGCOMM Comput. Commun. Rev. (CCR)

    (2002)
  • J. Crowcroft, M. Conti, A. Passarella, (Eds.), in: Proc. of the IEEE ICPS Workshop on Multi-hop Ad hoc Networks: from...
  • P. De, A. Raniwala, S. Sharma, T. Chiueh, MiNT: A miniaturized network testbed for mobile wireless research, in: Proc....
  • D.S.J. De Couto, D. Aguayo, B.A. Chambers, R. Morris, Performance of multihop wireless networks: Shortest path is not...
  • D.S.J. DeCouto, D. Aguayo, J. Bicket, R. Morris, A high-throughput path metric for multi-hop wireless routing, in:...
  • R. Draves, J. Padhye, B. Zill, Comparison of routing metrics for static multi-hop wireless networks, in: Proc. of the...
  • R. Draves, J. Padhye, B. Zill, Routing in multi-radio, multi-hop wireless mesh networks, in: Proc. of the Tenth Annual...
  • A. Farago, V.R. Syrotiuk, Merit: A unified framework for routing protocol assessment in mobile ad hoc networks, in:...
  • J. Flynn, H. Tewari, D. O’Mahony, JEmu: A real time emulation system for mobile ad hoc networks, in: Proc. of the First...
  • D. Ganesan, B. Krishnamachari, A. Woo, D. Culler, D. Estrin, S. Wicker, Complex behavior at scale: An experimental...
  • J. Garcia-Luna-Aceves, Wireless Internet gateways (WINGs) for the Internet, Technical report, University of California,...
  • L. Girod, J. Elson, A. Cerpa, T. Stathopoulos, N. Ramanathan, D. Estrin, EmStar: a Software Environment for Developing...
  • L. Girod, T. Stathopoulos, N. Ramanathan, J. Elson, D. Estrin, E. Osterweil, T. Schoellhammer, A system for simulation,...
  • The GloMoSim simulation environment. Available from:...
  • J. Gomez et al.

    PARO: supporting dynamic power controlled routing in wireless ad hoc networks

    Wireless Networks

    (2003)
  • R.S. Gray, D. Kotz, C. Newport, N. Dubrovsky, A. Fiske, J. Liu, C. Masone, S. McGrath, Y. Yuan, Outdoor experimental...
  • H. Hartenstein, H. Füßler, M. Mauve, W. Franz, Simulation Results and Proof-of-Concept Implementation of the FleetNet...
  • T. He, S. Krishnamurthy, J.A. Stankovic, T. Abdelzaher, L. Luo, R. Stoleru, T. Yan, L. Gu, J. Hui, B. Krogh,...
  • M. Heissenbüttel, T. Braun, T. Roth, T. Bernoulli, GNU/Linux Implementation of a Position-based Routing Protocol, in:...
  • A. Herms, D. Mahrenholz, GEA: A Method for Implementing and Testing of Event-driven Protocols, in: Proc. of the First...
  • Y.-C. Hu, D.B. Johnson, Design and demonstration of live audio and video over multihop wireless ad hoc networks, in:...
  • S. Jadhav, T. Brown, S. Doshi, D. Henkel, R. Thekkekunnel, Lessons learned constructing a wireless ad hoc network test...
  • J. Jubin, J.D. Turnow, The DARPA packet radio network protocols, in: Proc. of the IEEE 75, January 1987, pp....
  • G. Judd et al.

    Repeatable and realistic wireless experimentation through physical emulation

    SIGCOMM Comput. Commun. Rev. (CCR)

    (2004)
  • J.T. Kaba, D.R. Raichle, Testbed on a desktop: strategies and techniques to support multi-hop MANET routing protocol...
  • R.E. Kahn, S.A. Gronemeyer, J. Burchfiel, R.C. Kunzelman, Advances in packet radio technology, in: Proc. of the IEEE...
  • V. Kawadia, Y. Zhang, B. Gupta, System services for ad-hoc routing: Architecture, implementation and experiences, in:...
  • Q. Ke, D. Maltz, D.B. Johnson, Emulation of multi-hop wireless ad hoc networks, in: Proc. of 7th Int. Workshop on...
  • E. Kohler et al.

    The Click Modular Router

    ACM Trans. Comput. Syst.

    (2000)
  • K. Konolige, C. Ortiz, R. Vincent, A. Agno, Michael, Eriksen, B. Limketkai, M. Lewis, L. Briesemeister, E. Ruspini, D....
  • D. Kotz, C. Newport, R.S. Gray, J. Liu, Y. Yuan, C. Elliott, Experimental evaluation of wireless simulation...
  • Cited by (210)

    • Adaptive Load-Balanced Routing Algorithm Based on Queuing Theory and Multipath Routing

      2023, Proceedings - 2023 6th International Conference on Data Science and Information Technology, DSIT 2023
    View all citing articles on Scopus

    Wolfgang Kiess studied computer science at the University of Mannheim from 1997 to 2003. In the academic year 2000/2001 he stayed at the University of Nice Sophia Antipolis, France. He received the M.S. degree in March 2003, the subject of his diploma thesis was “Hierarchical Location Service for mobile ad-hoc networks”. Currently, he is a Ph.d. student at the University of Düsseldorf.

    Martin Mauve received the M.S. and Ph.D. degrees in computer science from the University of Mannheim, Germany, in 1997 and 2000 respectively.

    From 2000 to 2003, he was an Assistant Professor at the University of Mannheim. In 2003, he joined the Heinrich-Heine-University, Düsseldorf, Germany, as a Full Professor and Head of the research group for computer networks and communication systems. His research interests include distributed multimedia systems, multimedia transport protocols, mobile ad-hoc networks and inter-vehicle communication.

    View full text