From: Janet Davis (jlnd_at_cs.washington.edu)
Date: Sun Feb 08 2004 - 14:10:39 PST
I've gotten a couple of requests for clarification of question 1 on
homework 2. The scheme is just like learning bridges; here's an example.
Suppose we have the following topology:
0----1
\ \
\ \
2--3--4--5
First, suppose 0 floods a broadcast ping throughout the network. Each
node notes who it receives the first copy of the packet from. In the
diagram below, the * is the direction the node received the packet from:
0---*1
\ \
* *
2-*3--4-*5
Now, suppose 5 sends a ping reply back to 0. Each node will use this
learned information to send the packet towards node 0, rather than
broadcasting it to all neighbors. Here's the path the ping reply will
follow:
0<---1
^
\
4<-5
The goal is to communicate using less packets than flooding. It's
distinct from routing in that it's done purely through soft state and
learning, rather than through an explicit exchange of information about
the network topology.
Cheers,
Janet
-- Janet Davis jlnd_at_cs.washington.edu http://www.cs.washington.edu/homes/jlnd/ _______________________________________________ Cse461 mailing list Cse461_at_cs.washington.edu http://mailman.cs.washington.edu/mailman/listinfo/cse461
This archive was generated by hypermail 2.1.6 : Sun Feb 08 2004 - 14:10:43 PST