Hardware/Computing Resources in NSL

The Network Systems Lab is well-equipped to conduct networking and multimedia research. The figure below depicts some of the computing resources available in the Network Systems Lab.

File:Nsl.jpg

Wireless Sensor Testbed

The testbed is used to implement and test our coverage and connectivity protocols for wireless sensor networks. It contains:

• 10 sensors
• software

== Access to PlanetLab WAN Testbed == 

PlanetLab is composed of several hundred machines distributed all over the Internet. We are a member of this testbed. We use this testbed to test the systems that we develop in realistic environments. For example, our pCDN system was tested with thousands of clients distributed over a few hundreds PlanetLab nodes.

Servers and Cluster

There are several decent servers in our Lab, as well as a cluster of machines for conducting large-scale experiments and simulations.

• nsl: Lab web and file server. Has more than 2 TB of (RAID) storage. It has 8 cores (4 cores per processor), 8 GB memory.

• nsl-cpu: Fast compute server to run simulations and large-scale experiments. It has 8 cores It has 8 cores (4 cores per processor), 8 GB memory.

• nsl-win: Windows Terminal Server for remote access. It has most of the needed Microsoft software.

• nsl-cluster: A number of machines (currently 12) interconnected through a gigbit Ethernet switch. They can be configured in different topologies to test our code. They could form an isolated network for experimentation. They also can be used for general processing such as running multiple replicas of a simulation code.

• Workstations: For use by students.

Miscellaneous Items

• Wireless Interface Cards (3): Model xxx , they have open-source Linux driver. We use them to build a WLAN testbed to test power-aware video streaming in wireless networks.

• Routers and NAT boxes (5): Different types/models: xxx, xxx We use them to test our NAT traversal schemes being developed for the pCDN system.

• GPUs xxx

• HDTV cards xxx

These computing resources are partially funded by an NSERC RTI Grant.