Newsgroups: comp.ai.alife
Path: cantaloupe.srv.cs.cmu.edu!bb3.andrew.cmu.edu!newsfeed.pitt.edu!gatech!newsfeed.internetmci.com!uwm.edu!math.ohio-state.edu!usc!sdd.hp.com!hplabs!hplextra!hplb!jco
From: jco@hplb.hpl.hp.com (Joao Oliveira)
Subject: ALife and Flow Control
Sender: news@hplb.hpl.hp.com (Usenet News Administrator)
Message-ID: <DM1J9y.180@hplb.hpl.hp.com>
Date: Wed, 31 Jan 1996 10:17:58 GMT
Nntp-Posting-Host: incl05.hpl.hp.com
Organization: Hewlett-Packard Laboratories, Bristol, England
Lines: 25

	As anyone used ALife/Agents/Genetic Algs on Network Flow 
Control? 

	Imagine a network where messages are sent by each node randomly (it
can be bursty once in a while). I'm not looking for routing solutions but
for flow control - i.e. control of rates at which messages are sent, so that
nodes don't get overflown by fast senders. No assumptions about network con-
nectivity level should be made. It might be a ring network. I'm looking for
adaptive solutions as new nodes might be inserted and other removed at any 
instance. Direct communication channels exist between any two nodes that are
exchanging messages - i.e. messages are not forwarded ... they are always
consumed. 

	Solutions that depend on dedicated infra-structures are not suitable
as this is supposed to be fairly portable to any environment, programming
language, etc.   

	Tips on papers, known algorithms, etc are tremendously welcome ;)

joao
--
+ Joao Barreto Fernandes - Intelligent Network Platform Dep. (research stdt) +
+ HP Labs (INCL), Filton Road, Bristol, BS12 6QZ England    +44 0117 9228248 +
+ jco@hplb.hpl.hp.com  http://www-incl/~jco  http://www.di.fc.ul.pt/~barreto +
+      'All points of view are my own and not necessarily HP's as well'      +
