SIP in higher education

[(Dennis Baron)]

SIP.edu Conference Call April 13, 2006

*Attendees*

Dennis Baron, MIT
Candace Holman, Harvard 
Jerry Keith, UC Riverside
Jeff Kuure, Internet 2
Don McLaughlin, UCSD
Steve Roeper, Independent Consultant
John Todd, Tello
Dave Zimmerman, Berkeley

*Discussion*

Today's call begins with Dennis talking about the upcoming Internet2
Spring Member Meeting. Currently there is a workgroup meeting
scheduled for lunch on Tuesday, and on Wednesday Dennis will be trying
to drum up interest from schools to join the SIP.edu initiative.

Dennis has noticed a renewed interest in VoIP deployment on
campuses. Some of this seems to be focused on PBX and desk phone
replacement at the expense of SIP.edu, which gives him mixed feelings
as it might distract from the email addressing and connectivity
between campuses that is the primary focus of SIP.edu. Additionally,
the Real-Time Communication advisory group has began work on a
strategy guide for campuses, but has moved on towards addressing
organization issues within this area in Internet2. There is proposal
under discussion which would probably disband the SIP.edu working
group and makes it part of a larger organization with a broader focus.

Candace mentions that the SIP.edu working group began as a spin-off of
the VoIP working group and now has much more active participation
than the original group. But this does not help people who are
interested in protocols other than SIP. Dennis mentions that the VoIP
group is looking into a new charter, with one recommendation being
that the VoIP group focus on PBX and hard phone replacement.

Jerry Keith says that the management at UC Riverside sees VoIP and SIP
as a commercial issue, ultimately threatening established telecom
monopolies as well as campus telecommunication departments. This is
inevitable, but will disrupt things along the way. This leads to a
discussion about cost recovery for campus telephone and data
service. Jerry says that they realize that the same things may happen
with VoIP that happened with their long distance model following the
increase in cell phone usage by students.

Dennis asks if people would be interested in discussion about staffing
and funding changes as traditional voice traffic migrates to
VoIP. The consensus seems to be that this would be desirable.

John Todd asks about other real-time communications such as Jabber or
XMPP, and if these could be part of SIP.edu, despite not being SIP or
VoIP. Dennis says that there was an Internet2 IM group that declared
themselves dormant, so most of the discussion about XMPP and Jabber
has been in the Presence and Integrated Communication group. Dennis
feels that there shouldn't be separate SIP group and Jabber groups
that are charting different courses, as their integration should be
investigated. Candace concurs that the shouldn't be a working group
for every protocol, such as XMPP.edu or H323.edu. John mentions that
they do share the same addressing scheme, and XMPP will be more widely
supporting a voice transport layer with the Jingle protocol so the
differences are becoming less distinct.

Dennis asks about the general idea of getting more campuses
connected. Do participants see value in this? Candace believes that
there is value for the campuses who are looking to get
connected. Dennis feels that they have made it somewhat easy for
people to receive SIP calls, but need to make it easier for campuses
to make calls to other institutions, which ISN may help with. Steve
Roeper asks about ISN dialing, and how it differs from ENUM. Dennis
gives a quick overview of the issues of control and number assignment
with ENUM and how ISN is more open and does not require an E.164
number. John Todd mentions the ISN cookbook and discussion that is
available at www.freenum.org.

John feels that more actual examples for SER, Asterisk, and other
commercial offerings would be beneficial for promotion of
SIP.edu. Guides help organizations go from a concept to a test lab
easily, and, particularly with open-source software, provide the
how-to as well as the necessary software. Dennis asks if SIP.edu
should be encouraged experimentally, or as part of the supported
infrastructure on a campus. John does not see the difference, as this
is a decision that the organization will have to make. Providing
production quality examples and experiences from other universities
will make it easier for people to pitch SIP.edu to management.

Dennis is still looking for a compelling case as to why campuses
should investigate SIP.edu. There are a lot of valid local reasons,
including experimentation, staff development, or the replacement of a
Centrex system. Dennis is wondering what would inspire a CIO to return
from the member meeting with a desire to join SIP.edu. Candace sees
the promotion of SIP.edu as sort of a difficult thing, as it can lead
to diminished campus telecom revenue and corresponding staffing
cuts. However, she sees trunking between campus and reduced long
distance cost as a compelling reason for SIP.edu investigation.

This leads to a discussion of trunking, and if this would involve ENUM
and E.164 addressing. General consensus is that there needs to be some
way of knowing if a number that you dialed is reachable over a
specific trunk, which requires some sort of central SIP proxy with a
redirection databases that handles the calls. Alternately, there would
have to be a central repository of numbers that could be queried by
other universities to get the correct SIP URI. Candace considers SIP
trunking to be direct gateway connections between organizations, which
does not necessarily have anything to do with number dialing. John
says that there would still need to be a way to route calls and know
which trunk to use. Candace says that this is the problem that needs
to be solved, and that there is an IETF group looking at this. Dennis
and John think that this still sounds like ENUM.

Jerry thinks that trunking campuses together will eventually be a good
thing that will lead to a co-existence between institutions and will
be evolutionary from a technology standpoint. Currently within the UC
system, however, this idea has been shot down as the traffic analysis
has indicated that the capital investment will not pay for itself due
to low toll costs.

Dennis thinks that once an organization invests in a SIP gateway and
associated training, a lot of options appear. MIT and Harvard, for
example, have tielines between the two campuses. They're renting
facilities from Verizon that connect the two campuses despite having
dark fiber between them. It would take some effort, but they should
eventually convert use the fiber directly. They could just as easily
have a tieline to Stanford, for example. Other compelling cases for
CIOs could involve disaster recovery, which would be more reliable
for voice traffic between campuses in the event of an emergency.

Finally, Dennis will be at the Internet2 member meeting but should be
back in time for the next call.

The next call will be on April 27th.