Deprecated: The each() function is deprecated. This message will be suppressed on further calls in /home/zhenxiangba/zhenxiangba.com/public_html/phproxy-improved-master/index.php on line 456 SWAP
JITS - Just In Time Switching of high speed networks
Cooperation Report
JITS - Just In Time Switching of high speed networks
Cooperation Report to SWAP up to and Including November 1995
Author:
Hans Hansson, DoCS, UU
Date:
November 20, 1995
We
have participated in the project meetings, Per Gunningberg has made
a visit to Lund, and that we participated in the joint AASIC, SWAP and
SDT workshop. Also, Per Gunningberg has been responsible for
the Uppsala ATM pilot project and he has initiated a series
of meetings with graduate students and MSc. thesis workers
working in ATM related projects at Uppsala univ.
Hans Hansson has been visiting Univ. of York (Alan Burns group).
There are several reasons why our cooperation with
other SWAP partners have been limited, including
JITS started 6 months later (Jan 1st 1995) than other SWAP
projects.
The main activity in JITS started from scratch as far as ATM,
and ATM technology is concerned.
The focus in JITS is slightly different than that of many other
SWAP projects (more computer science and less hardware).
JITS has had a lower than planned activity level. Ken Tindell
left and Hans Hansson has been occupied with an SFSR initiative.
(JITS has consequently consumed less resources than planned.)
Also, the full time engaged PhD student (Mikael Sjödin) is newly
recruited (January 1st, 1995) and has consequently spent a large part
of his time on graduate course-work.
Nevertheless, some interesting results have been developed in JITS.
In particular, the theory for fixed priority scheduling has been
adopted to an (increasingly more) faithful ATM-switch model.
Also, a simulator has been developed in Erlang.
JITS is now in a phase where it is interesting
to interact with other SWAP projects, in particular
JITS would benefit from
feedback on the proposed switch model (Are our assumptions realistic?;
Is the assumed buffer handling realistic?;
How many priority levels can conveniently be handled?).
We have recently had some interaction with Johan M Karlsson's group in Lund,
and we
are planning to have discussions about switch design and switch modelling
with Tenhunen group at KTH.
JITS can to other SWAP projects offer
analysis, modelling and simulation of the
(hard) real-time aspects of their switch designs.