COOJA and Testbed Federation/TWIST growing together

Two COOJA plugins and manuals have been published to integrate the TWIST testbed in COOJA and to take checkpoints and perform rollbacks both in TWIST and COOJA.

Call for Papers: CONET/UBICITEC 2013

The 4th International Workshop on Networks of Cooperating Objects for Smart Cities 2013 (CONET/UBICITEC 2013), colocated with CPSWeek 2013, accepts submissions until January 28th, 2013.

Newsletter issue #19

The 19th CONET newsletter has been published. You can read on Virtual Organizations for Multi-Model Based Embedded Systems and on the UvA Bird Tracking System.

Testbed Directory

The following directory provides an overview of the main features of many prominent Cooperating Objects testbeds.


The Ad hoc Protocol Evaluation (APE) testbed [LLN+02] from Uppsala University is a full-scale testbed (no artificial attenuation of the RF signals) for comparative study of different MANET protocols. Node mobility is achieved through choreographed movement of human volunteers carrying the laptops containing the SUT wireless cards. The experimental work is supported by a set of logging and visualization tools. The trace collection is performed in a distributed fashion, relying on timestamps and an off-line aggregation step to recover the global ordering of the events. The APE software framework is publicly available under open-source license.

Casino Lab

The Casino Lab WSN testbed [cas] at the Colorado School of Mines consists of 52 Tmote Sky nodes, hung from the ceiling of a large open industrial space with concrete walls, pipes, ducts and fluorescent lighting. The dimensions of the room is 24.4m×12.30m, and the nodes are deployed in a 4×13 irregular grid. The nodes are connected via USB to 26 Tmote Connect Ethernet bridges providing a wired out-of-band channel for control. The TOSSIM Live extension [Met07] to the TOSSIM simulator, allowing execution of simulations in real-time and their interaction with real testbed nodes was originally developed and validated on the Casino Lab. The testbed is not publicly available.


CitySense [UT] is an urban-scale WSN testbed developed by Harvard University and BBN Technologies. It consists of a set of nodes fixed to rooftops and streetlights (currently 30 outdoor nodes, with a final target of 100 nodes) deployed in Cambridge, MA. The computational unit on the nodes comes in two configurations, one with 533 MHz CPU and 258 MB RAM or with 233 MHz CPU and 128 MB RAM. The nodes run FreeBSD 7 as operating system. The CitySense nodes support an array of sensors relevant for the outdoor urban setting: weather, CO2, noise pollution, etc. Each node has two radios. The communication substrate is formed by mesh routing using OLSR on a Ubiquiti SR9 (900 MHz) radio interface cards. The Wistron CM9 (2.4/5 GHz 802.11a/b/g) radio, is not used by the testing infrastructure and is available for SUT applications. The management of the testbed is performed using a custom software framework called CityMD.


One of the original full-scale MANET testbeds, the Carnegie Mellon University’s testbed for evaluation of the Dynamic Source Routing (DSR) protocol consisted of 5 mobile and 2 stationary nodes The wireless nodes were comprised by an IBM Thinkpad laptop equipped with 900 MHz WaveLAN radio card. Mobility effects were evaluated by driving the mobile units in rented cars in an outdoor area with rough dimensions of 700m×300m. The localization of the units was performed via GPS. One of the fixed nodes was used as gateway, connecting the test network back to the Field Office via a 2.4 GHz point-to-point link.


The DES-MESH is a hybrid wireless mesh and sensor network testbed being developed at the Freie Universität Berlin geared towards long-term studies [GBS08]. It currently consists of 35 hybrid nodes installed in an office setting spanning three floors. The testbed architecture is organized in three tiers: backbone mesh routers, mesh clients and sensor nodes. The mesh routers are equipped with 500 MHz AMD Geode LX800 CPUs with 256 MB of RAM, and have three IEEE 802.11 cards attached via USB hubs. The sensor nodes have 60 MHz ARM 7 cores and Chipcon CC1100 transceivers in the 868 MHz band. The testbed management is realized by a combination of SSH-supported remote command execution and SNMP services. The experiment configuration and control is facilitated by a domain specific language called DES-CRIPT based on XML.


The Diverse Outdoor Mobile Testbed (DOME) by the University of Massachusetts Amherst is one of the longest-running urban-scale MANET testbeds. It is comprised of three categories of nodes: The DieselNet is formed by 40 transit buses with nodes containing a Hacom OpenBrick 1 GHz computing board, a GPS receiver and several different radio cards: a 802.11 a/b/g mini PCI card, 802.11g wireless access point, a 3G modem and 900 MHz RF modem; The “throwboxes” have similar setup but use batteries recharged by solar cells. They can be mounted on bicycles and can act as additional routers improving the connectivity of the DieselNet buses; The third group of nodes is comprised by 26 Cisco 1500-series WiFi access points, mounted on different buildings and light poles, that form a mesh network. The management software is modular and provides services like link management, remote updating, logging and maintenance monitoring.


The Deployment Support Network (DSN) is a testbed framework developed at ETH Zürich [DBK+07], that leverages a secondary multi-hop WSN optimized for connectivity and reliability as a testbed backbone. The DSN-nodes forming this backbone network are in turn connected to the SUT nodes via custom wired interfaces. Currently supported SUT node platforms include the BTnode, TinyNode, Tmote Sky and A80. The testbed backbone is used for SUT image file distribution, for transfer of logging and debug data, and for sending direct commands to the SUT nodes. The operation of the testbed is controlled by a DSN-server that exports the DSN-services via XML-RPC and web based interfaces towards the testbed user. The instance of the DSN framework at ETH Zürich uses the BTnode platform and its Bluetooth radio for the backbone network. The current configuration of the testbed consists of 30 backbone nodes and 30 Tmote Sky and TinyNode SUT nodes.


The Emulated Wireless Ad Hoc Network Testbed (EWANT) [SBBD03], developed at the University of Colorado at Boulder, is a reduced-scale MANET testbed with emulated RF environment using in-line attenuation and RF multiplexing. Mobility is simulated by discrete switching between different antennas connected to the outputs of the 1:4 RF multiplexers attached to the wireless cards.


The Illinois Wireless Wind Tunnel (iWWT) [VBV+05] is a reduced-scale testing environment for wireless networks implemented in an electromagnetic anechoic chamber at the University of Illinois at Urbana-Champaign. The main aim of the testbed is to create a realistic scaled version of the wireless environment maintaining full control over all relevant parameters that affect the performance of the wireless network like obstructions, interferers, etc. Mobility is supported by placing the wireless hosts (laptops, PDAs, sensor nodes) on top of remotely controlled cars. The scaled wireless environment is constructed by combining the effects of several building blocks: Power control module, Multipath module, Doppler module and Scattering Module. Despite these efforts for complete control of the RF environment, repeatability of small-scale experimental results remains elusive due to intrinsic randomness in the evaluated protocols and object positioning errors [ZV08].


The Kansei testbed at The Ohio State University [AER+06] has been initially developed as a testing facility for the middleware services for the final demonstrator of the Extreme Scale Wireless sensor Networking (ExScal) project [ARE+05]. The ExScal demonstration is one of the largest hybrid deployments of sensor and wireless mesh networks ever attempted, with more than 1000 sensor nodes and more than 200 wireless mesh nodes distributed in a 1.3km×300m area. The Kansei testbed is a testbed environment that replicates this heterogeneous architecture at a reduced scale. Its stationary array originally consisted of 210 dual nodes (a combination of one Extreme Scale Stargate (XSS) wireless mesh node and one Extreme Scale Mote (XSM) node) placed on a 15×14 rectangular grid with about 1 m spacing. The XSS nodes have IEEE 802.11b wireless cards, and the XSM nodes, derivatives of the UC Berkeley’s Mica family, operate in the 916 MHz band. Recently, 150 nodes have been upgraded with Tmote Sky boards. In addition to the stationary array, Kansei also has a portable array of 50 Trio sensor nodes and a mobile array of 5 robots from Acorname Inc.. The software architecture of the testbed is organized around the Kansei Director that provides interfaces towards the basic services of the testbed like experiment scheduling, deployment, platform monitoring and management as well as creation and management of testbed arrays and configurations. The Kansei testbed is designed for shared usage, and has open access policy for members of the research community.


The Miniaturized Wireless Network Testbed (MiNT) [DRSC05] at the Stony Brook University consists of eight mobile nodes roaming in an 3.66m×1.83m area. The mobile nodes are built from COTS hardware: Routerboard 230 mini PCs with 1-3 Atheros IEEE 802.11a/b/g cards, placed on top of iRobot’s Roomba as the mobility platform. One of the wireless cards on each node, operated in RF monitoring mode, is dedicated to collecting traces that are transferred to a central node where they can be visualized in real time. The output from the remaining wireless cards is connected to low-gain external antennas via fixed signal attenuators providing about 60 dBm attenuation, limiting the communication range to about 0.6 m. The custom control GUI enables convenient node configuration, editing and execution of traffic generation scripts, mobility scripts and fault injection scripts. The GUI performs merging of the traces collected by the different nodes and extraction of different network statistics. MiNT also supports hybrid execution of unmodified ns-2 simulations over the MAC and PHY layers of the real testbed nodes.


Mirage [CBA+05] is a testbed management system developed by the Intel Research Berkeley (IRB) that applies the concepts of microeconomic resource allocation to the problem of allocating nodes in a sensor network testbed. Users submit bids that are specifying their interest in terms of the nodes and the time they would like to be granted access to, combined with the price they would be willing to pay. The system periodically runs a sealed-bid auction to determine the successful bids based on the demand/availability, while aiming to maximize the aggregate utilization of the testbed. The Mirage framework is used to manage a 100 MicaZ and 50 Mica2Dot node testbed at IRB premises.

Mobile Emulab

The Mobile Emulab [JSF+06] is a robotic wireless and sensor network testbed developed by the University of Utah leveraging their widely used Emulab platform for running network testbeds. The testbed is comprised by four Garcia robots from Acorname, each carrying a Stargate node that is interfaced with a Mica2 node. The mobile nodes are roaming in an area of about 8m×3.5m. The tracking and the identification of the nodes is handled by a vision-based tracking system, using six ceiling-mounted video cameras aimed down towards the floor. A central control daemon is responsible for plotting the movement paths of the robots, so that they can safely reach the user-specified end positions, while maneuvering around any static and dynamic obstacles encountered during the motion. In addition, the testbed is equipped with 25 static Mica2 nodes arranged on the ceiling in a rough 2 m grid and on the walls near the floor. The testbed management software uses the standard Emulab services to provide a batch queued or interactive first-come, first-served usage through a web-based, GUI-driven or programmable XML-RPC user interface.


MoteLab [WASW05] is a very popular testbed solution from Harvard University. In its original design, the testbed was comprised from Mica2 nodes, each connected to Ethernet backbone via dedicated Crossbow interface boards, providing TCP forwarding for the serial ports. The current deployment is one of the largest publicly accessible testbeds with 190 Tmote Sky nodes deployed over 3 floors of Harvard’s Engineering building. The SUT nodes are connected to the testbed backbone via Tmote Connect units. The testbed server provides a web interface that lets users monitor the status of the testbed and register jobs. The system uses a quota system to limit the time that each user has at its disposal for the outstanding test jobs. At one given time, only a single user has control over the complete resources of the testbed. The jobs are started by a job scheduler that configures the SUT according to the job description (installing and configuring the SUT images, etc.) and starts logging of the SUT output to a local database. The users can also get raw data access by connecting to the TCP forwarded serial ports of the SUT nodes during the assigned job slot.


The Motescope testbed [mot] at the University of California, Berkeley is an update of the sMote testbed installed in the Soda Hall. The original 78 Mica2Dot nodes in sMote have been replaced with MicaZ nodes in Motescope. The testbed provides convenient web interface for configuration and control of the experiments. The testbed has open access policy for the members of the academic research community.


The Open Access Research Testbed from Next-Generation Wireless Networks (ORBIT) [OSSS05] at Rutgers University is massive indoor grid of 400 wireless nodes each equipped with two IEEE 802.11a/b/g mini-PCI cards, deployed in a 20 by 20 grid with 1 m spacing. The testbed provides full remote control over the nodes during the assigned time slots, including installation of custom system images. The topology control is performed by a topology generator that leverages a wired remote control infrastructure to switch some of the nodes on or off. Similarly, mobility is simulated by transferring the state of a “mobile” node from one grid node to another. ORBIT provides extensive libraries for collecting, analyzing and accessing measurement data.


The Omega testbed [ome] is another testbed at the University of California, Berkeley. It consists of 28 TelosB nodes, connected via daisy-chained USB hubs to the central control server. This wired back-channel is used for powering, programming and debugging of the SUT nodes. The testbed has open access policy for the members of the academic research community.


The Re-Mote framework [Dat] developed by the Datalogisk Institut (DIKU) at the University of Copenhagen is comprehensive testbed management suite with four main software components: mote control infrastructure, database scripts, web services and GUI client. The database component uses the MySQL relation database and tracks the static and dynamic state of the testbed through the mote, mote host site and user session models. The core of the architecture is formed by the mote control infrastructure that is implemented in C++, separated in a low-level Mote Control Host (MCH) and Mote Control Server (MCS) parts. The web services component provides a loose interface between the testbed services and the clients. The Java-based client interacts with the Java web services component and the MCS allowing users to authenticate to the testbed and to program and monitor the SUT nodes. The Re-Mote framework has been tested in a deployment at DIKU with 36 Freescale DIG528-2 development boards as SUT nodes connected to 5 host PCs.


The RoofNet testbed [BABM05] at the Massachusetts Institute of Technology is an experimental multihop IEEE 802.11b Internet access network comprised of nodes positioned on the roofs of various buildings in Cambridge, Massachusetts, covering an area of about 4 square kilometers. In the original design, each node was built out of a mini PC and a 802.11b card connected to a roof-mounted omnidirectional antenna. The cards are operated in simplified IBSS ad-hoc mode which does not use beacons. The RoofNet routing protocol Srcr, implemented in Click [MKJK99], uses a combination of link-state and on-demand mechanisms and selects routes that minimize the estimated transmission time. The RoofNet deployment demonstrated that urban mesh-based Internet access networks can provide sustained DSL-level performance. The original software framework for RoofNet was open sourced and serves as basis for many RoofNet replicas like the Berlin Roof Net [Hum] and the NetEquality Portland deployment [Net]. The RoofNet technology was commercialized and is being further developed by Meraki [mer].


Sensei [RHF+08] is a nomadic testbed that can be used in a lab, for in-situ experiments and as prototype deployments. Physical sensors are normally attached to a Linux host consisting of a stationary computer, a laptop, a broadband router or a Personal Data Assistant (PDA) but there is also a cellphone implementation that enables using a cellphone as sensor proxy or as a sensor itself. The testbed supports usage mobile sensor nodes, carried around on random or predefined paths by either humans or robots. Control and management of the testbed can be handled remotely via a GUI that is highly customizable for different purposes. The nomadism supports testing the same experiment in different locations or at different testbed installations. The support for repeatable mobility enables repeating similar tests including mobile nodes multiple times, and the highly flexible GUI makes it possible to create experiment-specific visualizations of testbed activities.


Tutornet [tut] in the Ronald Tutor Hall at the University of Southern California is a tiered testbed with 13 node clusters, formed by a Crossbow stargate cluster-head, connected to several mote-class nodes via USB cables. The stargate tier uses the EmStar development platform [GSR+04] and operates as a multihop wireless network, using AODV-based routing over IEEE 802.11b links. The mote tier is currently populated with 81 Tmote Sky and 13 MicaZ nodes. The testbed has open access policy for the members of the academic research community.


The WASAL testbed [dL] at the École Polytechnique Fédérale de Lausanne consists of 25 TinyNode nodes connected to a wired testbed back-channel via custom serial-to-ethernet devices that act as passive communication bridges and range extenders. The WASAdmin management tool uses an XML-based configuration language and concurrently executes a separate shell instance and script parser for each target node in the SUT.


The Wireless Industrial Sensor Network Testbed for Radio-Harsh Environments (WINTeR) [SMP+08] is a testbed facility for the Canadian Petroleum Applications of Wireless Systems (PAWS) project. It aims to replicate the harsh RF conditions of an offshore oil platform, where the dense piping and other large metallic structures create complex multipath effects, combined with strong noise and interference and challenging environmental conditions. To this end, a platform mockup can be crated out of six modules each with dimensions of 2.44m×1.83m×2.74m. The mockup includes typical industrial metal structures like beams, pipes and tanks. The testbed uses a customized node platform consisting of a mote, DC power source, embedded CPU, micro power meter and programmable attenuator placed in an industrial enclosure. The emulation of the RF environment is further supported by an VSG-based EMI generator. The software architecture of the testbed is based on a modified version of Harvard’s MoteLab suite.


[AER+06] A. Arora, E. Ertin, R. Ramnath, M. Nesterenko, and W. Leal. Kansei: a high-fidelity sensing testbed. Internet Computing, IEEE, 10(2):35–47, March-April 2006. doi:10.1109/MIC.2006.37.


[ARE+05] A. Arora, R. Ramnath, E. Ertin, P. Sinha, S. Bapat, V. Naik, V. Kulathumani, Hongwei Zhang, Hui Cao, M. Sridharan, S. Kumar, N. Seddon, C. Anderson, T. Herman, N. Trivedi, M. Nesterenko, R. Shah, S. Kulkami, M. Aramugam, Limin Wang, M. Gouda, Young ri Choi, D. Culler, P. Dutta, C. Sharp, G. Tolle, M. Grimmer, B. Ferriera, and K. Parker. Exscal: Elements of an extreme scale wireless sensor network. In Proceedings of the 11th IEEE International Conference on Embedded and Real-Time Computing Systems and Applications (RTCSA ’05), pages 102–108, Aug. 2005. doi:10.1109/RTCSA.2005.47.


[BABM05] John Bicket, Daniel Aguayo, Sanjit Biswas, and Robert Morris. Architecture and evaluation of an unplanned 802.11b mesh network. In MobiCom ’05: Proceedings of the 11th annual international conference on Mobile computing and networking, pages 31–42, New York, NY, USA, 2005. ACM. doi:10.1145/1080829.1080833.


[BLFM05] T. Berners-Lee, R. Fielding, and L. Masinter. Uniform resource identifiers (uri): Generic syntax, 2005.


[cas] Casino Lab, Colorado School of Mines. [Online; accessed: 14.05.2009].


[CBA+05] B.N. Chun, P. Buonadonna, A. AuYoung, Chaki Ng, D.C. Parkes, J. Shneidman, A.C. Snoeren, and A. Vahdat. Mirage: A microeconomic resource allocation system for sensornet testbeds. Embedded Networked Sensors, 2005. EmNetS-II. The Second IEEE Workshop on, pages 19–28, May 2005.


[Cro06] D. Crockford. The application/json media type for javascript object notation (json), 2006.


[Dat] Datalogisk Institut (DIKU), University of Copenhagen. Re-Mote Testbed Framework. [Online; accessed: 14.05.2009].


[DBK+07] Matthias Dyer, Jan Beutel, Thomas Kalt, Patrice Oehen, Lothar Thiele, Kevin Martin, and Philipp Blum. Deployment Support Network - A Toolkit for the Development of WSNs. In Proceedings of the 4th European Conference on Wireless Sensor Networks (EWSN), pages 195–211, Delft, The Netherlands, 2007.


[dL] École Polytechnique Fédérale de Lausanne. Wasal testbed. [Online; accessed: 14.05.2009].


[DRSC05] P. De, A. Raniwala, S. Sharma, and T. Chiueh. Mint: a miniaturized network testbed for mobile wireless research. INFOCOM 2005. 24th Annual Joint Conference of the IEEE Computer and Communications Societies. Proceedings IEEE, 4:2731–2742 vol. 4, March 2005. doi:10.1109/INFCOM.2005.1498556.


[DVGIF02] L. Daigle, D.W. Van Gulik, R. Iannella, and P. Faltstrom. Uniform resource names (urn) namespace definition mechanisms, 2002.


[FGM+99] R. Fielding, J. Gettys, J. Mogul, H. Frystyk, L. Masinter, P. Leach, and T. Berners-Lee. Hypertext transfer protocol – http/1.1, 1999.


[Fie00] R.T. Fielding. Architectural Styles and the Design of Network-based Software Architectures. PhD thesis, University of California, Irvine, 2000.


[FN96] N. Freed and Borenstein N. Multipurpose internet mail extensions (mime) part two: Media types, 1996.


[GBS08] M. Gunes, B. Blywis, and J. Schiller. A hybrid testbed for long-term WSN studies. In Int. Workshop on Sensor Network Engineering (IWSNE), 2008.


[GFJ+09] Omprakash Gnawali, Rodrigo Fonseca, Kyle Jamieson, David Moss, and Philip Levis. Collection tree protocol. Technical Report SING-09-01, Stanford University, 2009.


[Goo] Federated Login for Google Account Users. [Online; accessed: 02.10.2009].


[GSR+04] Lewis Girod, Thanos Stathopoulos, Nithya Ramanathan, Jeremy Elson, Deborah Estrin, Eric Osterweil, and Tom Schoellhammer. A system for simulation, emulation, and deployment of heterogeneous sensor networks. In SenSys ’04: Proceedings of the 2nd international conference on Embedded networked sensor systems, pages 201–213, New York, NY, USA, 2004. ACM Press. doi:10.1145/1031495.1031519.


[HHP+08] I. Haratcherev, G. Halkes, T. Parker, O. Visser, and K. Langendoen. PowerBench: A Scalable Testbed Infrastructure for Benchmarking Power Consumption. In Int. Workshop on Sensor Network Engineering (IWSNE), 2008.


[HKWW06] Vlado Handziski, Andreas Köpke, Andreas Willig, and Adam Wolisz. Twist: a scalable and reconfigurable testbed for wireless indoor experiments with sensor networks. InREALMAN ’06: Proceedings of the 2nd international workshop on Multi-hop ad hoc networks: from theory to reality, pages 63–70, New York, NY, USA, 2006. ACM. doi:10.1145/1132983.1132995.


[Hum] Humboldt Universität zu Berlin. Berlin Roof Net. [Online; accessed: 14.05.2009].


[IAN] IANA. [Online; accessed: 02.10.2009].


[JSF+06] D. Johnson, T. Stack, R. Fish, D. M. Flickinger, L. Stoller, R. Ricci, and J. Lepreau. Mobile emulab: A robotic wireless and sensor network testbed. INFOCOM 2006. 25th IEEE International Conference on Computer Communications. Proceedings, pages 1–12, April 2006. doi:10.1109/INFOCOM.2006.182.


[Ken] Sun Cloud API. [Online; accessed: 02.10.2009].


[Ker] Kerberos Authentication Protocol. [Online; accessed: 02.10.2009].


[LLN+02] H. Lundgren, D. Lundberg, J. Nielsen, E. Nordstrom, and C. Tschudin. A large-scale testbed for reproducible ad hoc protocol evaluations. Wireless Communications and Networking Conference, 2002. WCNC2002. 2002 IEEE, 1:412–418 vol.1, Mar 2002. doi:10.1109/WCNC.2002.993531.


[mer] Meraki Website. [Online: 14.05.2009].


[Met07] Chad Metcalf. TOSSIM live: Towards a testbed in a thread. Master’s thesis, Colorado School of Mines, 2007.


[MKJK99] Robert Morris, Eddie Kohler, John Jannotti, and M. Frans Kaashoek. The click modular router. SIGOPS Oper. Syst. Rev., 33(5):217–231, 1999. doi:10.1145/319344.319166.


[mot] Motescope Testbed, University of California, Berkeley. [Online; accessed: 14.05.2009].


[Net] NetEquality Portland Deployment. [Online; accessed: 14.05.2009].


[OAu] OAuth. [Online; accessed: 02.10.2009].


[ome] Omega Testbed, University of California, Berkeley. [Online; accessed: 14.05.2009].


[Ope] OpenID. [Online; accessed: 02.10.2009].


[OSSS05] M. Ott, I. Seskar, R. Siraccusa, and M. Singh. Orbit testbed software architecture: supporting experiments as a service. Testbeds and Research Infrastructures for the Development of Networks and Communities, 2005. Tridentcom 2005. First International Conference on, pages 136–145, Feb. 2005.


[Pro] ProtoGENI. [Online; accessed: 02.10.2009].


[RHF+08] Olof Rensfelt, Frederik Hermans, Christofer Ferm, Per Gunningberg, and Lars-Åke Larzon. An interactive testbed for heterogeneous wireless sensor networks. In Demo Proceedings of the 4th International Conference on Distributed Computing in Sensor Systems (DCOSS 2008), 2008.


[SBBD03] S. Sanghani, T.X. Brown, S. Bhandare, and S. Doshi. Ewant: the emulated wireless ad hoc network testbed. Wireless Communications and Networking, 2003. WCNC 2003. 2003 IEEE, 3:1844–1849 vol.3, March 2003. doi:10.1109/WCNC.2003.1200667.


[Shi] Shibboleth (R). [Online; accessed: 02.10.2009].


[SMP+08] Jeff Slipp, Changning Ma, Nagesh Polu, James Nicholson, Martin Murillo, and Sajid Hussain. Winter: Architecture and applications of a wireless industrial sensor network testbed for radio-harsh environments. Communication Networks and Services Research, Annual Conference on, 0:422–431, 2008. doi:10.1109/CNSR.2008.42.


[tut] Tutornet Testbed, University of Southern California. [Online; accessed: 14.05.2009].


[UT] Harvard University and BBN Technologies. Citysense testbed. [Online; accessed: 07.10.2009].


[VBV+05] Nitin H. Vaidya, Jennifer Bernhard, V. V. Veeravalli, P. R. Kumar, and R. K. Iyer. Illinois wireless wind tunnel: a testbed for experimental evaluation of wireless networks. In E-WIND ’05: Proceedings of the 2005 ACM SIGCOMM workshop on Experimental approaches to wireless network design and analysis, pages 64–69, New York, NY, USA, 2005. ACM. doi:10.1145/1080148.1080162.


[WASW05] Geoffrey Werner-Allen, Pat Swieskowski, and Matt Welsh. Motelab: A wireless sensor networks testbed. In Proceedings of the Fourth International Conference on Information Processing in Sensor Networks (IPSN’05), Special Track on Platform Tools and Design Methods for Network Embedded Sensors (SPOTS), 2005.


[Wis] Wisebed. [Online; accessed: 02.10.2009].


[ZV08] Yuhao Zheng and Nitin H. Vaidya. Repeatability of illinois wireless wind tunnel. Technical report, University of Illinois at Urbana-Champaign, 2008.