High Performance Wireless Research and Education Network January 30, 2002

HPWREN Team Experiments with Small, Low-Power Network Performance Measurement Machines for User Sites

While network measurement machines typically consist of relatively large desktop or rack-mountable computers, the HPWREN team is currently experimenting with a 3"x3"x3" computer that examines the quality of links at user sites. An early deployment includes the Pala Native American Learning Center, which is connected to HPWREN. This machine allows for reachability as well as throughput tests and is available for additional applications, such as weather stations.

"This Internet-ready, low-power machine gives us the capability to efficiently measure the performance of 802.11b HPWREN's access links," says Hans-Werner Braun, HPWREN principal investigator. "If the machine performs as well as anticipated at Pala, we will deploy additional performance machines of this type at other access points."

Todd Hansen shows off HPWREN's new "supercomputer": the Intrinsyc CerfCube, which measures 3"x3"x3".
The CerfCube, which runs a Linux operating system, consists of a high-performance Intel StrongARM 1110 processor with 32 MB RAM and 16 MB Flash.

An additional application for this type of miniature computer is being explored by the ROADNet team of researchers. "We are particularly interested in using the CerfCube as a gateway for sensors into our real-time data distribution system," says Todd Hansen, ROADNet systems coordinator. "Although we will have to add some flash memory into the system, we feel that our VORB system can greatly benefit from such a device and look forward to experimentation."

Additional photographs regarding the HPWREN measurement machine at Pala are available at https://cdn.hpwren.ucsd.edu/images/20020112.

The HPWREN team continues to research small, low-power network performance measurement machines for user sites. If you have insight regarding this area, please contact us.


 back to top

 back to HPWREN news

field research
network analysis
 ~ university of california, san diego ~ © 2000 ~