Undergraduate Ben Eckart receives national NSF award for poster

A Tennessee Tech University undergraduate recently received national recognition for a new protocol for transferring large amounts of computer data to a distant location more quickly and with more stability than is now possible.

Ben Eckart, a senior from Clarksville, Tenn., received an award from the National Science Foundation for "Best Poster" created through its Research Experiences for Undergraduates program, a project led by his adviser, Xubin He, TTU electrical and computer engineering professor. Eckart is working to improve how extremely large quantities of data are ferried to remote locations.

The list of applications, which is quickly growing, includes large-scale remote operations, such as those produced by large department stores or commercial businesses, and remote archival of data sets that run on supercomputers.

"Basically, any high-performance, data-intensive, long-distance application would benefit from my work," said Eckart. "The trouble is that the current standard protocols, the agreed rules, for communicating data over long distances do not produce good data transfer speeds and tend not to be stable. It is not a design flaw; they were designed for markedly different applications than what we are doing now."

Eckart's research works to mend this disconnect by constructing a protocol that stably and dynamically maximizes network speeds. So far his design has seen very good results, beating all other current protocols in its class.

"Ben is one of the best students I have every advised, and he is conducting top-notch research although he is an undergraduate student," said He, who will continue to advise Eckart as a graduate student.

Online data storage doubles almost every nine months in this country due to an ever-growing demand for networked information services, says He, who directs Eckart's work through TTU's Storage Technology and Architecture Research Laboratory, or STAR Lab.

A paper based on Eckart's work has been accepted by the 22 nd IEEE International Parallel and Distributed Processing Symposium, one of the top conferences in the field.