Transcript Eliminate the Coax-Or Else!

Eliminate the Coax-Or Else!
Lessons Learned from Implementing
IPTV in a Large Campus Network
Environment
David Allred, Systems Design Engineer
Paul Hardin, Senior Project Manager
Craig Malquist, Network Design Engineer
PCH
BYU Demographics
• 33,000 Undergrads and Grads
BYU Demographics
• 33,000 Undergrads and Grads
• Employees
– 4,000 F/T
– 1,300 P/T
– 14,000 P/T Student Employees
BYU Demographics
• 33,000 Undergrads and Grads
• Employees
– 4,000 F/T
– 1,300 P/T
– 14,000 P/T Student Employees
• Church Sponsored – The Church of Jesus
Christ of Latter-day Saints
BYU Physical Plant
• 295 Buildings on 560 Acres
– 87 Academic
– 150 Housing
– 58 “Other”
• On-Campus Housing
– 900 Married Student Apartments
– 4,800 Beds for Single Students
• Salt Lake Center
Brief Technology History at BYU
• <1970 Campus TV Cable Installed
• Mid-1980’s TV Cable first used for data
• Early 1990’s First fiber optic cable installed for
data network
• 2001 last major data network upgrade (for IP
Phones)
• 2011/2012 Major data network upgrade
planned for data/voice/video convergence
DA
WHY the Mandate to Abandon Coax
• Main Driver – TV industry direction to move
from analog to digital television over the air—
wakeup call for change
WHY the Mandate to Abandon Coax
• 40 year old campus TV cable plant; losing
amplifiers
WHY the Mandate to Abandon Coax
• Where best to invest technology $$?
– Follow lead of IP phones and converge video &
data networks as much as possible
WHY the Mandate to Abandon Coax
• Where best to invest technology $$?
– Follow lead of IP phones and converge video &
data networks as much as possible
– Minimize cable plant
WHY the Mandate to Abandon Coax
• Where best to invest technology $$?
– Follow lead of IP phones and converge video &
data networks as much as possible
– Minimize cable plant
– Upgrade only data side of cable plant
• Needed to convert for digital TV anyway
• Needed to upgrade data network anyway
Possible Directions
• Build a digital coax system on campus
• Contract out services to CATV cable provider
• Go with IPTV and do it on our own—we
THOUGHT it would be cheaper (we chose this
one)
Decisions, Decisions, Decisions!
PCH
HOW We’ve Implemented IPTV To-Date
• Began upgrade of network infrastructure in
2001 to accommodate VOIP
• By 2005 we recognized the need to upgrade
additional network equipment (primarily edge
switches) to deal with multicast
• Prepared multicast for implementation
• Implemented 24 SD channels in 2006
• Built our own portal/web page using VLC
HOW We’ve Implemented IPTV To-Date
• “Finished” upgrades to network
infrastructure; ongoing throughout process
• Had to change cable TV content provider;
former provider didn’t want to be on the
network
• 2 years after initial 24, we add another 24 SD
channels in 2008 and added portable
encoders
HOW We’ve Implemented IPTV To-Date
• Later, we bought a portal, but because of web
browser/OS incompatibility problems, we
rebuilt our own over 2-3 months in 2010
WHAT the Challenges Have Been
DA
WHAT the Challenges Have Been
• NAC (Network
Admission Control)
– We didn’t want to
burden the NAC
servers with streaming
traffic load
– Had to develop
separate multicast path
into Campus Housing
WHAT the Challenges Have Been
• Portable encoder (to originate anywhere on campus)–
Inconvenient because of network architecture and
policies
– Public jacks on separate segment from campus
• Portable encoder can’t broadcast to campus from public
jacks
• Want to broadcast from multiple locations on campus
– Need to request ops to reconfigure jack
temporarily (different VLAN) to put it on a different
segment of the network that allows the multicast
WHAT the Challenges Have Been
• Watching at computer
– Done in conjunction with developing the web
portal (2010)
WHAT the Challenges Have Been
• Developing IPTV Administration Application
– Done in conjunction with developing the web
portal (2010); Source listing for web portal, STBs,
and dynamic play list for VLC
WHAT the Challenges Have Been
• Redeveloped interface for
computer IPTV portal
• When new channels were
introduced, we found old
interface hard coded
• Wanted more flexibility
with a dynamic channel
lineup and interface that
accommodated that
flexibility
• We can control which
channels are displayed in
different parts of campus
(based on IP address)
WHAT the Challenges Have Been
Changed STB Interface—unintended additional
work - Pushed new code to 600+ STBs
WHAT the Challenges Have Been
TEC Podiums needed modified
» Coordinated redesign of
controller GUI
» Had to reprogram
controller
• 500+ TEC Rooms had
same code
• Custom rooms done
separately
WHAT the Challenges Have Been
• Transcoding Problems
– Transmit with different resolutions throughout the
day; changing resolutions as broadcasters first
started digital broadcasting
– E.g. commercial would be in different resolution
from program content
– Resolved through firmware upgrades
WHAT the Challenges Have Been
• Change in direction on Off-Air channels
– Partway through IPTV project we were directed to
not receive over the air channels from antenna,
» BECAUSE some broadcasters wanted to be paid for oncampus distribution of content
» BECAUSE we (OIT) thought we were in license violation,
so were going to remove content
» BUT University lawyers now feel we are within the law
» SO we are now investigating getting it from an antenna
again
WHAT the Challenges Have Been
• Getting people to use IPTV (instead of campus
cable)
– Cable turned off to only two existing buildings; new
buildings don’t get coax
– Put more content on IPTV to get people to use it;
Housing especially complained about lack of
content
• Some of Housing has CAT3 cable - outdated
• Edge Switches were outdated - have been upgraded
• We update network as buildings are updated
WHAT User Reaction Has Been
• Users dislike Standard Definition
– Looks pretty bad when blown up BIG
– Flat panels especially exacerbate the digital artifact
problem when scaling up image to fill display
• Users compare picture quality with viewing
experience at home
PCH
Pixelated image in SD on
Large LCD display
(simulated)
Same image in HD on
Large LCD display
(simulated)
Some SERIOUS
pixelization…
Even High Def won’t
help this!
WHAT User Reaction Has Been
• High Res vs. Standard Res
– HR consumes more bandwidth
» SD: 2 MB/Channel x 48 channels ~~ 100 MB
» HD: 4-6 MB/Channel x 48 channels ~~200-300 MB
350
300
250
200
Bandwidth
150
MB/CH
100
50
0
SD
HD
WHAT User Reaction Has Been
• High Res vs. Standard
Res
– Bandwidth restricted
by policy; no room to
send HD signal into
residence halls;
trying to prevent
bandwidth hogging
WHAT User Reaction Has Been
• High Res vs. Standard Res
– HD Encoders costlier
– HD Content cost higher
WHAT User Reaction Has Been
• Sports
– High motion content results in lots of digital
artifacting—blah!
WHAT User Reaction Has Been
• Don’t use/need STBs in a lot of environments
– Housing: one dorm prepared for 50% distribution of
STBs (over 100); ended up renting 5
– On-line viewing much more than TV—online content
providers, some of which are free; sports activities
– OIT doesn’t have a way to track TV usage (cable or
IPTV)
– Customers want to know about sub-channel content
(e.g. 11.1, 11.2); we haven’t dealt with that yet
WHAT User Reaction Has Been
Able to watch content on
computer at desk or in
room—users like
anyplace access
WHERE We Plan to Go From Here
• Removing coax is not a
high priority—it’s still
functioning—Standard
definition analog
– As coax & amplifiers get
less reliable, we will pull
it out; death by attrition
over next few years
– We will eventually
schedule shutdown to
campus buildings
DA
WHERE We Plan to Go From Here
• Not expanding cable plant in new construction or
remodels
WHERE We Plan to Go From Here
• IPTV is reasonably good enough for SD to replace
analog “in some peoples’ opinion”
– BUT Big screens produce big pixels
– BUT we are considering adding one HD channel for
sports in Fall 2011 ($8-10K)
– BUT we are adding local, isolated, single-building HD
» If customer use requires it
» If building network architecture can handle it
WHERE We Plan to Go From Here
• Bigger issue than HD vs SD is, is traditional TV dead?
– The traditional model is breaking down more quickly than
we want to admit
– Replaced by video on demand; JIT taking its place: Hulu,
NetFlix, YouTube, network sites, e.g. Korean dramas, etc.;
– However, JIT doesn’t satisfy need for real-time sports, live
events, news (e.g. Japan earthquake, Libya)
WHERE We Plan to Go From Here
• Latest addition has been 4 HD internal channels
in new BYU Broadcasting Building
WHERE We Plan to Go From Here
• Implementing internal IPTV channels in campus
buildings to replace in-building modulators
– We have six foreign language channels
– Library has 15 internal channels
• Already have several portable IPTV encoders for
live campus events
• Need to add HD Channels when politically or
financially viable
WHERE We Plan to Go From Here
• NextGenNet Project
– General network upgrade looking to next 10 years
– Upgrade infrastructure for HD
– Track IPTV channel usage
Conclusion
Eliminating coax costs $$--What are
the benefits?
• Long term recovery of costs—
maintain single infrastructure
• Keep pace with user interests