Connecting the Dots: Switching and VOD
by Leslie Ellis // March 27 2006
Maybe it’s happened to you lately. You’re in a meeting. Engineers are present. They’re talking about switched digital video. (Or maybe they call it “switched broadcast,” or “SDV,” or “SDB.” They’re interchangeable.)
You drift off to your checklist of basics: Switching is a bandwidth conservation technique driven by the simple fact that most people watch mostly the same stuff, most of the time. That leaves dozens of channels gumming up the pipe, with no (or few) eyeballs on them. Save room. Only send the channels tuned by the viewers.
From what you’ve heard, the math seems promising. In a digital video broadcast lineup of 160 channels, half aren’t watched. A video switch makes 160 channels look like 80 channels — and that’s a conservative estimate.
You tune back in to the conversation happening around you. You could swear you only mentally skipped out for a few sentences — yet your technical colleagues seem to have careened into a deep-dive on the architectures used for video on demand (VOD) delivery.
This week’s translation seeks to connect the dots between the seemingly unrelated technology precincts of switched digital video, and video on demand.
It turns out that the switching of video has a lot in common with the storing of video, especially when the storage happens inside the network.
The first commonality between switched video and VOD is perhaps obvious: They aren’t broadcast. When you’re in front of the TV, and you pick something from the on-demand roster offered by your cable company, you’re setting up a “session.” The session links your set-top with a remotely-located storage server.
Only you get what you selected. Not everyone on the system. (Tech people call this “unicast.” Translation: You-cast.)
Switched broadcast techniques also use “sessions,” to fetch the channel you pick, when thumb-surfing the remote.
That’s why technology people talk about “SRM,” for “session resource management,” when they’re talking about either technique. They mean everything thats required, behind the scenes, to set-up that box-to-server link, maintain it, and tear it down. Amongst your tech-side peers, SRM is a chewy topic.
The second commonality between switched video and VOD is the technology used to deliver those sessions to and from a set-top box. This is where the conversation usually veers into modulation talk, and especially “QAM,” for Quadrature Amplitude Modulation.
Sorry. I know how easy it is to tune out, when QAM pops into a conversation. Heres a hopefully easy way to put it in perspective: QAMs are what moves digital channels down the pipe. One digital channel equals one QAM, equals one chunk of bandwidth capable of hauling 40 Megabits per second down the pipe.
QAMs enter conversations about switched video and VOD through the door marked “inventory management.” Say a cable operator sets aside four digital channels for its on-demand offerings. Engineers will say this as “I’ve got four QAMs for VOD.”
Right now, any operator considering a video switch is working out how many channels in their digital shelf space — how many QAMs — they should allocate to the switch.
That raises the question: Should that switch also manage the four channels allocated to on-demand, and the QAMs that support them? And those other digital channels, allocated to IP (Internet Protocol) services, like voice-over-IP and broadband. They also use QAMs to move bits to homes. Should the switch manage those, too? And if so, are the switch people tuned in to the DOCSIS 3.0 chapter, which carries similar implications?
Pull up, the engineers say. Right now, the installed base (a million or so) of digital modulators (QAMs) is service-specific. VOD QAMs carry VOD. Broadcast QAMs carry the “regular” digital video channels. Data QAMs carry the bits of voice-over-IP, web surfing, and everything moving over the broadband connection.
In short, all bits stay in their respective “lanes.”
Since switching video is a nascent technique, the time is now to decide whether those “single purpose” QAMs should be re-configured, to carry whatever traffic comes at them — data, video, or voice.
Ultimately, operators seek a video switch that can speak in multiple languages. They want the QAMs, located near the edges of the network, to perhaps be able to handle tons of VOD streams on Friday night, and tons of broadband sessions at 4p on weekdays, when kids get home from school and log on.
Because switched video is just emerging as a technical and strategic priority, the vendor community is predictably overlapped. VOD suppliers make session resource managers. So do the makers of video switches. And thats not even factoring in the mass of voice switch vendors with video intentions. Translation: Partnerships are imminent.
VOD and switching hang out together, conversationally, because they have overlapping features. More on that, and the differences between VOD and switching, next time.
This column originally appeared in Multichannel News.
Getting to Know “Over the Top” Video
by Leslie Ellis // March 13 2006
Not too long ago, the term “over the top,” to describe those companies sending voice or video services over a broadband network owned by someone else, was pretty much contained to the technical community.
Now, and especially as the “network neutrality” debate starts to look like something Federal regulators will consider, as they mull whether AT&T should buy Bell South, “over the top” is stepping into the mainstream.
So far, when “over the top” enters a conversation, the spotlight tends to beam on companies unaffiliated with broadband network owners, who provide fat downstream services, like video.
In practice, it goes like this: Consumer Jane pays for a 6 Megabit per second downstream connection. Maybe she uses half of it to watch a video stream offered by someone else — some “over the top” video provider.
This is all fine, now, when downstream speeds tend to double every 18 months or so, as a function of the cable v. telco speed wars. So far, there’s room.
Over the Upstream
But what about “over the top” services on the upstream, home-to-headend direction? In a way, it’s almost “over the bottom.” Technologists used to fret that this would start in a measurable way with “user generated content” — sending the digitized home video of Claire’s first big snowstorm to grandma and Aunt Connie.
They didn’t imagine something like SlingBox, developed by Sling Media to give consumers a way to “take their TV with them.” SlingBox, like TiVo, is one of the growing list of innovators that start at the “analog hole” — the spigot, on many consumer devices (including cable set-tops) that allows an incoming digital signal to move to a connected device, like a TV, in analog form.
In Sling’s case, they grab the analog signal, encode it (digitize), compress it — and then sling it up the network, using the broadband pipe. From there, it goes toward the person, somewhere else, who wants to watch it — usually via laptop.
Putting aside the obvious questions of legality about Sling, and “location shifting” — at the Consumer Electronics Show this year, a gaggle of programming side executives were overheard joking about who should sue Slingbox first — it also carries big network implications.
Here’s why: That upstream path, connecting cable homes to cable networks, is skinny and hostile. Relative to the total bandwidth available on a contemporary cable system (860 MHz), the upstream path, sometimes called the “reverse path,” is about five percent.
Operators, of course, are fully aware of their upstream shortcomings, and are working up cures. One is to go to higher orders of modulation in the upstream — that’s tech talk for finding better ways to imprint more data onto the carriers that move that data around. Adelphia, among others, has already gone as far as 64 QAM (Quadrature Amplitude Modulation) in the upstream, which translates to about 27 Mbps.
Widen the Upstream
And then there’s that Feb. 17, 2009 date, approved by President Bush, as the date the analog stuff goes away. Poof! Gone. Maybe it’s a stretch, but, there are those who believe that once the cable channel lineup doesn’t have to start at channel 2, maybe that upper boundary of the upstream spectrum can finally move up higher — and give operators, and their customers, some breathing room.
Either way, it’s probably time for two things: A better name for “over the top,” and a better way to deal with “over the bottom” (upstream path) providers.
This column originally appeared in Multichannel News.