[mod.ai] AI in Network Protocols.

ramarao@umn-cs.UUCP.UUCP (02/28/87)

topic : EXPERT SYSTEMS OR AI IN NETWORKS AND PROTOCOLS

	I am trying to find out if there has been any attempt at 
applying AI techniques, AI languages to the field of network protocols.

Can anyone give me some references. I would like to know why it is
not feasible to implement network protocols, etc. in a non-procedure
based approach.

I am trying to find out if it is feasible to design network protocols
in LISP or Prolog  or any AI languages.

-Bindu Rama Rao    (ramarao@umn-cs.arpa)
(612)-625-9637

**** Keep smilin  (-:            

roberts%cognos@math.waterloo.edu (Robert Stanley) (03/06/87)

In article <8702280810.AA09316@cs-gw.D.UMN.EDU> ramarao@umn-cs.UUCP.UUCP writes:
>
>topic : EXPERT SYSTEMS OR AI IN NETWORKS AND PROTOCOLS
>
>	I am trying to find out if there has been any attempt at 
>applying AI techniques, AI languages to the field of network protocols.
>
There seems to be more than one question being asked here, and some fairly
fundamental confusion about so-called AI languages.  It is perfectly
possible to create communications protocols of almost any kind using LISP
or PROLOG (or less well-known object-oriented languages such as NEON) provided
that there is a clean interface to the underlying system.  The InterLisp-D
world on the Xerox 1100 series machines is a good example.  However, this
has nothing to do with AI.

Is it possible to apply AI techniques in creating protocols?  Yes, of course,
but most of the work that has come to my attention appears to have been
tackling the problems of network configuration, diagnostics, and routing.  Why
would anyone want to to "use AI" to write a protocol?  This is a hard-science
engineering issue, and one that is pretty well nailed shut for existing
protocols.  The key to protocols has tended to be universal standardization, so
that many people can use them.  One area that it would be interesting to see
AI applied is the smart "automatic" conversion between protocols.

I suppose the last interptretation of the question is: could a new protocol be
created which is based on AI techniques?  Again, one supposes so, but such
things have a habit of being created only when there is a clearly apparent
need.  I have not currently got any communications problems (telecommunications
problems, anyway) that can't be solved by applying existing protocols, and I
find that 10megabit LAN and 1.27megabit GAN are sufficiently wide bandwidths
for all my current needs.

Perhaps the original posting could be rephrased or expanded slightly, to open a
possibly interesting topic for discussion.

Robert Stanley


-- 
Robert Stanley             decvax!utzoo!dciem!nrcaer!cognos!roberts
                    Voice: (613) 738-1440 (on EST) Tuesdays only
                                        don't ask-----'
Cognos Inc., 3755 Riverside Drive, Ottawa, Ontario, K1G 3N3  CANADA