08.17.09

DARPA RFI on CR and Machine Learning

Posted in artificial intelligence, military, public safety at 9:56 am by jody

Regular blogging may resume this week as I’m almost caught up.
In the mean time, here’s a quick solicitation of likely interest.
(link) DARPA-SN-09-60 Machine Learning for Behavioral Control of Cognitive Radios (ML BCCR)
Blurbing from the pdf:
DARPA is seeking information regarding government, academic, and industrial base capabilities for autonomous and man-guided adaptive systems that allow cognitive radios to cooperate with and, if necessary, learn to control the behavior of other cognitive radios while in the field. These systems could potentially be used for Humanitarian and Military uses. Research issues of potential interest to DARPA include:
  • Radically new system architectures for behavioral control of radios and networks
  • Low size, weight and power adaptive broadband receivers and transmitters
  • Automated methods for locating, direction finding and classification of emitters
  • Computationally efficient learning, synthesis and assessment algorithms
  • Quick response approaches to control known radios
  • Techniques for incrementally improving system libraries for new and unknown signals, internal policies and behavioral control waveforms.
  • Approaches that consider and implement Force-on-Force, Force-on-Net and Net-on-Net strategies.

The potential impacts of these systems and technologies include the reduction of development cycles from months to minutes, control of new continually evolving cognitive technologies that are driven by international standards and, ultimately, saving lives. This RFI is exploring a possible paradigm shift by determining methods for taking the laboratory to the field instead of new radios to the laboratory for analysis.

 There’s an industry day in September with invitations based on white papers submitted by August 28.