DARPA sets sights on improving analog-to-digital conversion

ARLINGTON, Va., 22 July 2005. Scientists from the U.S. Defense Advanced Research Projects Agency in Arlington, Va., are asking industry to consider revolutionary new ways of digitizing analog signals.

Jul 22nd, 2005

ARLINGTON, Va., 22 July 2005. Scientists from the U.S. Defense Advanced Research Projects Agency in Arlington, Va., are asking industry to consider revolutionary new ways of digitizing analog signals.

Toward this goal, DARPA is soliciting industry proposals for the Analog-To-Information (A-to-I) program to investigate alternatives to standard analog-to-digital (A-D) converter technology.

Specifically, DARPA scientists are trying to find better ways to digitize analog information than today's approach that implements the usual quantized Shannon representation -- or uniform discretized samples at the Nyquist rate or better.

The fundamental question of the A-to-I program is whether additional prior knowledge, hypotheses, and/or ancillary measurements can enable practical data conversion approaches, DARPA officials say.

DARPA experts want to find ways that more effectively apply system resources to find the useful information content embedded in a complex radio-frequency environment and directly measure it in a concentrated form.

Many military applications must digitize broadband analog signals with extreme fidelity, which severely stresses current A-D converter technology. Shannon's representation exploits only minimal prior knowledge about such RF environments, namely the bandwidth, DARPA officials say. Consequently, standard digitization often measures much more data than is actually useful.

From industry, DARPA is looking for proposals on data-conversion approaches that not only directly measure concentrated information in compressed format out of broadband analog environments, but also that reduce the digitization sample rate or sample density compared to classical Shannon representation -- yet with no loss of fidelity.

Proposals should describe one-year studies that seek to determine the minimum sampling necessary to characterize and reproduce a particular signal of interest with no information loss, and then determine how to achieve this in practice.

Additional information is in the Broad Agency Announcement 05-35 (BAA 05-35) proposer information pamphlet available online at http://www.fbo.gov/spg/ODA/DARPA/CMO/BAA05-35/Attachments.html.

DARPA officials say they expect awards for one-year study efforts toward the end of this year, and are encouraging collaborative efforts and industry teaming. More information on teaming for this program is online at www.davincinetbook.com/teams/Login.asp

The technical point of contact for the program is Dr. Dennis M. Healy. He can be reached by fax at 703-696-2206, or by e-mail at dhealy@darpa.mil.

Companies should submit full proposals by Aug. 30 to be considered during the initial round of selections. DARPA will, however, evaluate proposals through July 2006. Submit proposals to DARPA/MTO, 3701 North Fairfax Drive, Arlington, VA 22203-1714 (Attn.: BAA 05-35).

The administrative addresses for this BAA are: Fax: (703) 351-8616 (Addressed to: DARPA/MTO, BAA 05-35), Electronic Mail: BAA 05-35@darpa.mil Mail: DARPA/MTO, ATTN: BAA 05-35 3701 North Fairfax Drive Arlington, VA 22203-1714.

More in RF/Analog