On Tue, 25 Oct 2005 04:21:01 GMT, "Steven P. McNicoll"
wrote:


wrote in message
...

That's what I was saying. Although it is pretty sad that the nationwide
system cannot keep everything together in a unified way. It's not that
it's that much data... just that the system is so old that it cannot
handle it.
If a couple megabytes of memory in a GPS can handle every APT, VOR,
ADF, INT for all of north america, regular computers shouldn't have
trouble.


Right. Let's put all the flight data in one system, give every controller
access to all of it, and make it easy to shut down the entire system.

The simpler you make a system from the user's point of view the more
complicated the programming, the more likely you'll end up with side
effects, and the more easily compromised the system and or data.

In a sensitive system the data needs to be compartmentalized and
redundant. It also needs to be isolated from the rest of the world,
but if stations/centers are to talk to each other then the data is
open to compromise.

It's true the data is *relatively* simple and takes very little memory
for each way point. A hand held can hold all the data for the low
level enroute charts and approaches. It might be pressed to have all
of the low level and high level but lets assume it can. This is
static data in that it doesn't change, or has little change.

Then there are the many thousands of flights in the system daily. This
data is dynamic and takes many times the computing power to handle
compared to the static data. The varying data must be coordinated
through centers and with airports. As airplanes join and leave, climb
and descent they affect others in there area or possibly approaching
their area. If their speed changes, arrival times at reporting points
change this has a snow ball effect on planes that may not even be in
the air yet.

Add to this random failures in the system. By that I mean both
aircraft, RADAR, and the computer system itself. No system is without
some faults and they have a way of surfacing at the most inopportune
times.

Probably the best system I can think of would be a distributed
computing system that works much like a PtP system with systems
sharing information, but with checks and balances to provide data
integrity and security. Each system would *almost* work
independently, but share information and operate based on feed back as
well.

The worst and most dangerous approach from a system being compromised
and loss of data integrity would be a single central system feeding
various points/centers and airports. It lacks the ability to do the
checks and balances of a true and integrated distributed computing
system.

However any time a system has the requirement that parts in different
locations must communicate, that system is open to outside compromise
so additional measures such as encryptions and various methods of
verification must be used in addition to firewalls and other
protective software.

It's not nearly as simple as it sounds.

Roger Halstead (K8RI & ARRL life member)
(N833R, S# CD-2 Worlds oldest Debonair)
www.rogerhalstead.com