WebRTC Expert Feature

July 01, 2013

Why WebRTC Needs its Sessions Controlled

I enjoyed a lot of the conversations I had in Atlanta last week at the WebRTC Conference and Expo, starting with the editorial briefings, where I heard from companies that are implementing WebRTC with compatibility strategies for Flash and H.264 (with and without SVC).

Back when session controllers crossed into VoIP, staff members from various carriers came up to me and – having not yet heard of these devices -- asked if they needed them. As a recovering Bell Head, my answer was a simple one: If you think of the network as “your network” then “you” need one, however if you are an over-the-top player, you likely do not. This time around, though, the strategy is quite different, mainly because even the over-the-top folks are going to be confused.

Here are the key factors as I see them:

  1. Détente on the H.264 vs. VP8 issue is unlikely in the near future. The good news for Android is that it will be in the chipset, however I bet it won’t be in Nokia, Apple and certain phones from other vendors as specified by carriers. In addition, we may see the chipset folks add a price difference on their VP8, the result being that even browser-to-browser will require transcoding.
  2. It’d be nice to be able to say that browser-to-browser will be end-to-end, but the reality is that session controllers will likely be seen even in this area, particularly in enterprise VPN implementations and as Flash 2 (aka, Flex brings life back to other strategies). Note, though, that although VP8 is a great codec, some of the animation features done in WebGL may warrant other codec support, and thus Enterprise Application Servers will probably need them (or even embed them).
  3. The “fear factor” cannot be discounted, especially if Nokia presses on intellectual property issues. Indemnification may be a problem for some of the smaller players and working with hosted services that include session controllers may become popular, thus many media servers are offering transcoding with session controller functions that may be a way to ensure that no patent violations occurred.

From all of this, we can infer that the ownership of the network will not be as relevant as the content being transmitted and the types of devices used, and thus the answer will be a lot foggier as to when a session will need outside control.

Seamus Hourihan of Acme Packet (now Oracle) spoke about the future of session control beyond the telco when we last met at 4GWorld, and the vision he shared then now seems to be coming true in spades. Session controllers at their root manage edges, and we are now seeing that in this case each session is getting edgier. While I believe in the Internet model of end-to-end, the reality is that end points will be in the eye of the controller, and it is entirely possible that three or more session controllers will be involved in WebRTC at various levels. 

Given the edge is going to be split into several sections, a piece of advice I would offer – and you can take this to the bank – is not to short session controller stocks at this time.

Edited by Rachel Ramsey
Get stories like this delivered straight to your inbox. [Free eNews Subscription]


Free WebRTC eNewsletter

Sign up now to recieve your free WebRTC eNewsletter for all up to date news and conference details. Its free! what are you waiting for.