A Demonstration of the MeshTest Wireless Testbed

Brenton Walker Ian D. Vo Matthew Beecher Charles Clancy

brenton @ltsnet.net idv2101 @columbia.edu beechema@iastate.edu clancy @ltsnet.net

Laboratory for Telecommunications Sciences

US Department of Defense

Abstract-MeshTest is a laboratory-based multi-hop wireless

testbed that can subject real wireless nodes to reproducible
mobile scenarios. It uses shielded enclosures and an RF ma-
trix switch to dynamically control the attenuation experienced
between pairs of nodes. The testbed is an ideal platform for
experimenting with MANET and DTN implementations, offering
convenient experimental control and data management.

I. PURPOSE OF THE DEMO

Our goal is to demonstrate that the MeshTest is an ideal
platform for mobile wireless experimentation, development
and debugging. We have installed a variety of mobile network
software on the testbed nodes and can demonstrate a variety
of experiments. It is also possible to place other wireless
devices, including most laptops, in the shielded enclosures to
test different types of hardware and software.
II. TESTBED DESCRIPTION
MeshTest consists of a rack of computers in shielded
enclosures, an RF matrix switch, and a server that provides Fig. 2. Several shielded enclosures, three RF switches, and a testbed mobility
experiment control, as depicted in Figure 2. The RF from server.
each computer's WiFi card is cabled through the enclosures
and into the matrix switch of programmable attenuators. The
matrix switch allows us to control the attenuation experienced place and move nodes around on a virtual map. Additionally,
between pairs of devices. By varying the switch settings in each node can be given a trajectory, designated by a series
real time we are able simulate the effects of arbitrary physical of target destinations. Nodes can travel at different speeds
scenarios and mobility. (O-IOOkm/h) between different pairs of destinations, rest at
MeshTest uses ORBIT's testbed management software to a destination for a specified time, and toggle their RF status
control the nodes. We also developed a Java Swing application (active/inactive).
to provide a portable, graphical interface in which users can The user-end aUI sends XML-encoded physical scenarios
to the testbed server at one second intervals during an exper-
iment. Upon receiving a physical scenario the testbed server
must compute the appropriate path loss matrix for the given
File
Map Information
Magnification: 253.15131879405598%

node arrangement, compute the appropriate attenuator settings

Map State: Break in simulation. changes are NOT ALLOWED
and apply the settings to the switch. We use a simulated
Node Information
speed
annealing algorithm to find the attenuator settings that best
approximate the desired physical scenario. The testbed server
Post Delay: 0 sec
also collects information on the states of the nodes. This
Trave!ingPower:
information can be queried by the user-end application and
displayed in the aUI during the experiment.
More background on the theory and performance of Mesh-
Stop Simulation

Test can be found in [I], [2].

1127.0.0.1

REFERENCES
[1] B. Walker and T. C. Clancy, "A quantitative evaluation of the meshtest
wireless testbed," in TridentCom 2008, March 2008.
Fig. 1. The user-end testbed GUI. Each square in the grid represents a square
[2] T. C. Clancy and B. D. Walker, "Meshtest: Laboratory-based wireless
kilometer. The orange lines are motion paths of nodes.
testbed for large topologies," in IEEE TridentCom 2007, pp. 1-6, 2007.

Authorized licensed use limited to: ULB Darmstadt. Downloaded on December 09,2022 at 18:47:54 UTC from IEEE Xplore. Restrictions apply.