On May 10, 12:21 pm, "Paul J. Adam"
wrote:
Ken S. Tucker wrote:
On May 10, 4:13 am, "Keith Willshaw"
wrote:
Its not simple stuff, a MRBM is doing anything up to 4,000 m/sec
on rentry. The plasma around the reentry vehicle is going to make
most sensors useless while also making radical manoeveurs next
to impossible.

It's a sub-orbital ballistic missile that breaks to subsonic
at high altitude, then it has a lot of time (by electronic
standards) to search, select, aim and fire.

Making itself a marvellous TBMD target for a SM-3... and suddenly much
of the attraction of an anti-ship ballistic missile is gone.

((what's TBMD?)), anyway, I'll play this game a bit more.
The inbound is changing velocity rapidly and unpredictably,
reducing interception probability. It's subsonic at 80k feet,
strips, and fires at 75k, (consider 1960's ASROC).
Suppose they fire 10 $1million missiles at an asset (CVN)
with a value of $10Billion, then successive vollies.
We need to understand the problem before we can solve
it, and *rose-colored* glasses won't work.

Note that while Pershing II used a synthetic aperture radar system
for terminal guidance this was an ancillary to the INS and compared
radar maps of the terrain with the on board maps. Its inclusion
was simply to reduce the CEP from the 400m of the Pershing I to
30m. This system did not have the capability to search for, locate and
guide the warhead to a moving target that may be 30 miles from the aim
point.
Keith

Things haved changed. A missile can shoot down a satellite
going 15,000 mph, yet you Keith steadfastly hold to the idea
that hitting a huge CVN doing 30 mph is very difficult.

The satellite's location is known and its ability to change speed and
direction very limited. A carrier can cover thirty miles in an hour, in
any direction it chooses: this gets you not only the physics problem of
manoevering to hit it, but the target identification issue.

So the enemy peppers the region.

Keith, a young fella like yourself has probably never seen a
Telex machine.

Keith's older than I am and we had a Telex in Registry until relatively
recently.

Well I always enjoy youthful optimistic exuberance.

Classified military electronics is likely 10-15 years ahead of
what is publically known.

Having worked on the stuff, fielded military electronics is a few years
behind civilian. Back in the 1970s, the military took something like 25%
of all integrated-circuit production and could set standards and lead
technology: now it's probably not even one per cent and the innovation
is pushed from the civilian sector. Hence the demise of MILSPEC
components... manufacturers weren't interested in getting the
certification for the size of orders available.

When you want a few thousand ruggedised CPUs for your guided weapon
(total production run over several years) you get in the queue behind
the motor manufacturers who are buying that many every *week*. You
design to the planned "next best thing" and keep options open, because
when you start the design process Intel are talking about possibly
taking the 486 CPU to fifty megahertz and memory costs forty pounds a
megabyte.

By the time you've got a frozen design it's getting hard to source a
ruggedised 486 and nobody sells SDRAMs smaller than eight megabytes.

By the time the production contract gets placed the 486 is a distant
memory and the question now is "dual or quad core, and how many
gigabytes of RAM would Sir like with that today"?

And that's to get stuff off the drawing board and into service. Once
it's fielded and frozen, you'll find logos of long-lost companies on
mission critical kit (the Ferranti logos scattered around the Radar 911
tracker office, for example).

What you wrote is correct, (in my experience),
but there is much more to it than the CPU!
Consider imagers and transducers that feed CPU.
Ken