Wednesday
I picked up the following two items from a Janes newsletter and thought they might be of interest:
US military airborne laser passes first in-flight engagement The US military's airborne laser (ABL) successfully completed its first in-flight test against an instrumented target missile on 10 August, the prime contractor Boeing said in a statement on 13 August. The US Missile Defense Agency (MDA) is testing the viability of using the high-powered laser to destroy enemy missiles in the boost phase.
Standard Missile 3 Block IB cleared to begin flight tests The Standard Missile 3 (SM-3) Block IB programme to develop an improved missile for the US Missile Defense Agency's sea-based Aegis Ballistic Missile Defense System has completed its critical design review, Raytheon announced on 13 July. The new missile is expected to begin flight tests in 2010. SM-3 Block IB offers significant improvements over the SM-3 Block IA version currently deployed on US Navy Aegis cruisers and destroyers and Japanese Maritime Self-Defense Force destroyers to defend against short- to intermediate-range ballistic missile threats in the ascent and midcourse phases of flight.
So things are still plodding along on all fronts and all becomes simpler as technology improves. I still believe the key number for missile interception is the figure of merit I wrote of a long time ago: instructions per meter. That is the number of machine code instructions that a CPU can process in the time it takes for the relative positions of the target and the interceptor (or laser station) changes by one meter. When this number gets large, the targeting system has more time to ponder what is going on and more time to analyze fused sensor data. Another way of looking at it is that time effectively runs more slowly for the targeting software as the number gets larger.
This is yet another side effect of Moore's Law. Our processing capabilities are growing to the point where either very sophisticated predictive programs may be used... or very unsophisticated and unoptimized programs will become 'good enough'.
Interesting that instrutions per meter concept.
Meanwhile http://home.kyodo.co.jp/modules/fstStory/index.php?storyid=456615
Russia deploys ballistic missile interceptors to counter N. Korea threat
MOSCOW, Aug. 26 KYODO
Russia has deployed S-400 air defense missile complexes, capable of intercepting ballistic missiles, in the Russian Far East to counter the threat posed by North Korea's missile tests, the country's top general said Wednesday.
Which once agains shows how the Western Media is in the other side.
Posted by lucklucky at August 27, 2009 08:38 AM
True enough. Look at Java code. Dog-slow, inefficient. 1980s assembler would run a lot faster.
But computers have plenty of horsepower now so who cares, Java works.
Posted by EUBanana at August 27, 2009 12:32 PM
Just to be wholly accurate, the measure I speak of is the change in seperation, so it is actually the number of instructions/sec divided by the change in the magnitude of the vector distance between A and B per second. Hopefully I got that right in words... It's almost easier to write the equation than to spell it out, but this blog interface doesn't do math :-)
Posted by Dale Amon at August 27, 2009 01:31 PM
These defences may get a noncomputer "test" very soon.
Although, sadly, even the best missile test can not defend against such things as a nuke in a truck (or a ship) in or near a major city.
Posted by Paul Marks at August 27, 2009 01:41 PM
One problem with measuring instructions per meter is that it assumes you're using a conventional general purpose CPU. I'd imagine it would make sense to build targeting systems for laser missile defense using ASICs to do specialised jobs very quickly, in much the same way as we use specialised ASICs to do the graphics for computer games.
But I get your point. Moore's law means we can build more complex and faster ASICs, too.
Posted by Rob Fisher at August 27, 2009 02:16 PM
Heh. I was involved in the testing of an early version of a BMD kill vehicle. I'd go to work, process successful test data, then come home and see some boob from the Union of Really Concerned Physicists and Dental Hygienists on the television explaining how such a thing could never be made to work. So we should cut the funding. It was science, he said.
On second thought, I suppose he was right. If you cut the funding it can never be made to work.
Posted by Eric at August 27, 2009 06:53 PM
