Research agency wants new approach for an ad hoc network with thousands of nodes.
The Defense Advanced Research Projects Agency wants “far out” ideas to improve mobile wireless networks, and those ideas should not rely on Internet protocols to relay data between nodes.
DARPA, in a notice posted on the Federal Business Opportunities website Tuesday, said it and other government research organizations want to develop Mobile Ad-hoc Networks software that can scale to between 1,000 and 5,000 nodes. This is likely to be a challenge, as DARPA said “it is difficult to field a MANET with 50 nodes.”
The Army has tested small scale mobile ad-hoc networks at its ongoing Network Integration Evaluation exercises at White Sands Missile Range, N.M., for the past three years. But DARPA said previous work focused on attempts to extend the Internet into the ad hoc networks. It is asking researchers to take a “completely different tack” based on “a postulate … that the MANET environment is fundamentally incompatible with underlying concepts of the Internet, such as routing, link reliability, end-to-end connectivity, high stability, opaque packets, source addressing, fixed infrastructure and client-server distribution patterns.”
Research on a large-scale ad hoc network should start “unencumbered by existing protocols,” DARPA said. “Start with a clean slate, informed by experience and previous research but free from any constraints or restrictions.”
The research agency said it is interested in protocols that take advantage of features of the mobile ad-hoc network environment, including peering between radios, duplicity of roles and many-to-many distribution patterns.
DARPA will hold a symposium in Arlington, Va., on Aug. 7-8, on what it formally calls its Novel Methods for Information Sharing in Large-Scale Mobile Ad-hoc Networks project. It asked researchers to provide it with one-page abstracts of their ideas before the start of the conference.
The symposium is “not about redesigning or rearchitecting the Internet; there are other ongoing efforts focused here,” the agency emphasized. “It is not about developing protocols for use in commercial applications or in areas with well supported, ubiquitous infrastructure…. New concepts and aggressive ideas are encouraged.”