In the 2010 video-on-demand trial carried out at the Dutch Public Broadcasters we assessed the efficiency of peer-to-peer streaming and the effects on the overall bandwidth usage patterns. The general outcome of this pilot is that an overall bandwidth reduction of 40% is feasible.

Download the report

During this test videos of the popular catchup-TV service called "Uitzending Gemist" were available through a separate portal and player. It appears that concurrency, the number of viewers watching the same video at the same time, is a good predictor for the expected peer-to-peer efficiency. We show that whenever concurrency increases, p2p efficiency increases too. In our report, we describe the experimental results, the models and the expected bandwidth reduction for populare site such as "Uitzending Gemist".  

Management Summary

This report describes the results of the pilot project to study peer-to-peer streaming of video, and that has been carried out by the Dutch Public Broadcasters (Nederlandse Publieke Omroep / NPO) together with the Dutch company Bitnomica. This pilot ran from June 29, 2010 until November 15, 2010. This pilot was positioned as a test-environment for the popular catch- up-TV service “Uitzending Gemist” In this pilot, we proved that by using P2P streaming technology, we can substantially lower the bandwidth usage for on-demand video applications. At some moment these savings were above 50%. Furthermore, we proved that whenever more people watch, and more people watch the same video simultaneously, this efficiency skyrockets. In general, we can expect efficiency (bandwidth savings) of more that 50% for the most popular content. For less popular (long-tail) content these savings will be lower or absent. However, since popular content will constitute the majority of the traffic, the savings yielded by the popular content largely defines the average savings. In general we expect that by using P2P , a bandwidth (in Mbit) reduction of at least 35% is possible. Furthermore, we expect that a reduction of the peak bandwidth capacity (in Mbit/s) of least 41% is possible. This is important especially for designing the streaming infrastructure. Furthermore, since bandwidth deals typically use the peak capacity as a metric to calculate the total monthly bill, P2P could reduce the monthly bandwidth bills by 40% or more. If we include bandwidth that has been saved when we did not need to retransmit a video, if the viewer wanted to see the video again, online or offline, the savings where about 60%. Of course this heavily depends on the reviewing habits of the users, or the possibilities of the application. This pilot has also shown that there is a demand amongst catch-up-TV users to watch videos whilst being offline. Many users used this functionality.

Bitnomica has started development of a new social television app. The exact idea we keep to ourselves for a while, but come back soon, since we may leak a few details when we are a few steps on track. But what is a Social Television anyway. Isn't Television already social, you watch it with other, you discuss it with others. You twitter about it. So what's more? Social Television is all about integrating your television experience with social networks, in order to enrich your television experience, to make good recommendations, and to present you relevant information. This sounds vague? What we try to do to make an App that does this, and that will change the way you watch television.

Bitnomica neemt voor het tweede achtereenvolgende jaar deel aan het YesDelft Network Event.

Het YesDelft network Event staat dit jaar in het teken van de opening van het nieuwe gebouw. Sinds Augustus 2010 zit Bitnomica ook in het nieuwe Yesdelft gebouw. Tijdens dit event presenteren alle YesDelft bedrijven zich aan de bezoekers. Bitnomica zal de laatste ontwikkelingen tonen van de producten die zij maakt.

Zie hier voor meer informatie.