|

Peer-To-Peer, Changing The Face Of Broadcasting

Peer-To-Peer, Changing The Face Of Broadcasting

Will Morton The advent of peer-to-peer streaming technology may mean that the internet is finally ready as a medium for radio and television. So argues Will Morton, managing director of Robonobo Ltd, an online company made up of a group of individuals who undertake projects designed to promote collaboration and community through digital technology…

The adoption of the internet as a broadcasting medium has been promised since it became mainstream. Real (formerly Progressive) Networks broadcast the first live baseball game over the internet in 1995; it seemed that a massive explosion of online radio and TV was just around the corner. However, reality has disappointed, as broadcasters have found that although it is easy to find the demand for streaming content, creating a working business model is harder.

The reasons for this are twofold. Firstly, consumers have demonstrated that they are not willing to pay for receiving streaming content. This may be down to historically free radio, or a more general lack of willingness to pay for online content in any form. Secondly, notwithstanding the extraordinary progress in network technology, the costs of broadcasting audio and video content to a large audience are very high.

These high costs stem from some basic facts about networks. Running high-bandwidth links over a short distance is easy and cheap – your office, and possibly your home, has 100Mb (Megabits) connectivity throughout. Running high-bandwidth links over long distances, however, is very costly; a 100Mb line a mile long costs thousands of pounds per month.

Thus the internet, from the lowest levels, has a hub-and-spoke structure. There is a small number of so-called ‘Tier One’ (also known as ‘backbone’) network operators, who run a few very high bandwidth links between major centres (known as ‘Points of Presence’).

These Tier One providers then link their networks to other (‘Tier Two’) network operators, who pay handsomely for the privilege, and then provide a larger number of lower-bandwidth connections from these PoPs to local businesses and consumers. This pattern is repeated again for ‘Tier Three’ network operators. (For pretty pictures of this in action, see the Internet Mapping Project at http://research.lumeta.com/ches/map/gallery/index.html.)

As a broadcaster, you have a streaming server on the internet, which serves up a stream to each listener. In computer science parlance, this scheme ‘scales linearly’, i.e. your server needs ten times as much bandwidth to serve 1,000 listeners as it does for 100 listeners.

Due to the architectural limitations mentioned above, this gets expensive very quickly. The maximum size of a streaming network is currently in the tens-of-thousands range, and this is limited to large organisations willing to spend millions of pounds on bandwidth, such as the BBC with their World Cup coverage.

Regardless of future developments in bandwidth technology, this system is never going to scale to the tens of millions of listeners/viewers currently receiving traditionally-broadcast media, and so a system called ‘IP Multicast’ was developed, which changes the way that data is transmitted by internet routers. Unfortunately this requires changes to the way network operators configure their routers, and it has not been implemented outside closed, private networks such as cable TV.

Ironically for broadcasters, the answer to their prayers is coming from their greatest enemies, online pirates, who have had no choice but to solve the problem of cheap large file distribution.

This solution is the now-infamous ‘peer-to-peer’ or P2P paradigm. Instead of receiving a file from a single centralised source, under P2P the file is sent from multiple sources simultaneously, and everyone who downloads the file also uploads it to others, thus spreading the load.

This works better than a centralised system because it adopts the underlying structure of the internet rather than fighting against it. A receiver can select sources that are close to it on the network, thus saving long-distance bandwidth (which is expensive) at the cost of more local bandwidth (which is cheaper).

The P2P paradigm has been so successful that its sworn enemies are adopting it, the record labels and movie studios, for who cost savings are just as important. However, for technical reasons it has thus far been limited to the distribution of files rather than streams.

A new wave of software companies (one of which, www.robonobo.net, I work for) is seeking to extend the P2P paradigm to streaming, bringing massive cost savings to broadcasters. Our goal is to bring ‘Broadcast Economics’ to online streaming, i.e. the broadcaster’s costs are fixed, regardless of how many listeners they have.

By making online broadcasting affordable for the many, rather than only the largest and richest, we hope to make the internet the broadcast medium that has been promised for so long.

Media Jobs