Thursday, July 29, 2010

DARPA wants electronic warfare systems that learn from their mistakes to keep pace with the latest communications technology

Posted by John Keller

Jamming RF signals in electronic warfare operations is getting a lot harder these days. It seems the ability of many military wireless communications devices like cell phones, battlefield radios, and command-and-control networks are developing the ability to adapt automatically to their environments to maintain the highest-quality signals possible.

This phenomenon is called adaptive communications -- or the notion of communications devices able to change quickly in response to conditions in the environment to make sure their signals get through. That presents a problem to the electronic warfare guys -- those whose job it is to jam the bad guy's signals to prevent them from getting through.

Unfortunately, today's adaptive communications technology seems to be outpacing electronic warfare. That means -- for the U.S., at least -- that the bad guys can change their communications faster to keep information flowing than the electronic warfare systems experts can tweak their electronic jammers to keep pace. Now DARPA is trying to change all that with the Behavioral Learning for Adaptive Electronic Warfare program.

This program -- BLADE, for short -- seeks to develop machine learning technology to enable future electronic warfare systems to adapt their jamming techniques just as quickly as the adversary adapts his communications.

Imagine that: an electronic warfare system that sniffs around for changes in the communications patterns of the enemy, learns his system, and sends out the appropriate jamming signals in response. In other words, fool me once, shame on you, but fool me twice, shame on me. Sounds like DARPA is on the right track.

It worries me, though, that commercial off-the-shelf communications technology seems to be able to turn inside developments in electronic warfare technology, but that's fodder for a future blog.

No comments:

Post a Comment