Wednesday, October 8, 2008

Col. Wayne Parks - Cognitive Network Theory

October 8, 2008
A month and a half ago I made reference to a “self healing network” in the DoD Blogger’s Roundtable. This blog post explains what I actually meant. While visiting the University of Kansas (KU) and Kansas State University (KSU), I ran across projects associated with cognitive networks or telecommunications. If you had asked the Army what cognitive network theory was, you would likely have been pointed towards a graduate level psychology professor. “Cognitive” drums up visions of hours of lectures on human psychology and words like “dissidence,” “reasoning,” “abnormal” and “processing.” As it turns out “reasoning” and “processing” relate, only instead of human psychology we’re talking about technology; scarily advanced computers and telecommunications, but computers and telecommunications none the less.

Cognitive Network (CN) Theory is the newest addition to the United States Army Computer Network Operations and Electronic Warfare (USACEWP) Information Library. The idea behind it is deceptively simple. You build a technological system using a matrix that is composed of electronic parts, through adaption and reasoning, actively adept to varying computer network conditions and stimuli in order to produce a set of performances and actions. Cognitive networks provide the possibility of pervasive computing, seamless mobility, ad-hoc networks, and dynamic spectrum allocation, among others. Cognitive network design can be applied to any type of fixed or wireless sensor networks. Cognitive computer and telecommunications networks could provide better protection against security attacks and network intruders. Such networks could benefit the service operator as well as the consumer by offering self-governed networks.

The not so simple part is the similarity these networks bear to a truly human system. The scientist building these networks have to take into account a) how each component will pull power and information for its own exclusive use; b) how each will respond to the other components pull for the same power and information; c) how the degree of variables and lack of information will influence each component, and d) finally, how much control each component can exert over the network as a whole at any given time. As I mentioned: scarily advanced computers and telecommunications.

The thing to keep in mind is these are created by human minds. The groups working at KU and KSU are our own Great Plains example, the boys and girls next door proving that homegrown American’s provide an amazing amount of intelligence and initiative for the future. As scary as this technology can be if extrapolated to the degree of say, The Terminator movies, these are still tools. Tools which purpose is to; while not truly think for itself, adapt enough to be trusted in situations where humans would be in danger. In a machine or device run by cognitive networks this means the ability to spill more hydraulic fluid or bits than blood. That’s the truly incredible part. That’s the truly important part.


Anonymous said...

Hi. I'm a faculty member in the psychology department at Kansas State University. I'd be interested in learning who at KSU and KU are working on this.


- Jack said...

I'll see what I can find out.