special report the connected world supplement
Changing the infrastructures
The operator now has to deal with fundamentally different types of video consuming devices. On one hand there is the traditional IPTV infrastructure - highly managed content, typically dictated by the network owner, but providing a sophisticated range of additional services such as fast channel change, network PVR, restart TV, pause live TV and many others. This is an optimal viewing experience that is hard to beat. On the other hand, consumers want access to other types of content including sports, additional programming, older movies, and TV programmes that are not available through the operator system, or they want to view the operator content on a different device. The biggest challenge the operator now faces is finding a solution that will allow them to take back control of their network.
Taking back control: scaling bandwidth and deeply distributed delivery
The first step towards taking back the control over how subscribers access content is to embrace the opportunity rather than be cowed by it. Subscribers want to be able to access content of their choice through any means available to them. In order to stay competitive, it is up to the operator to answer this demand. The second step is to provide an easy access to the content through the operator’s own infrastructure with the highest level quality of experience possible - meaning the highest quality appropriate for the device and access methodology.
Third, operators need to recognise that current highly centralised or inefficiently distributed content libraries will fall under their own weight and deploy a solution that provides high quality of service. Centralised architectures designed to stream and cache content from one centralised location cause major overload on the bandwidth. In addition, if there are 10 device types in a poorly distributed network, with five different bit rates supported that is potentially 50 versions of the same content being stored around the network. Now imagine that there are 100 or even 1000 different types of device and bandwidth combinations. While transcoding, distributed
containerisation and segmentation take care of some of this, many of those solutions are by no means deployable yet.
The only viable solution is effective demand-based video distribution which provides video serving capability as close to the consumer or subscriber as possible. It becomes almost impossible to completely centralise or ‘distribute by force’, ie distribute the whole content library around the network because it is just too huge and inefficient.
Operators need to realise that traffic patterns will inevitably change with the types of content being made available. This especially applies to potentially challenging issues of micro-charged sports or other key live events.
A good example of this is a European operator that wanted to provide access to a range of devices for viewing live sports in a country where that sport is essentially a national religion. However, to ensure the highest QoE for the viewers and correct levels of available bandwidth, the operator was forced to cap the number of ‘tickets’ that it sold to access the game. Thus, subscribers that wanted to view the match on their device would potentially face a ‘sold out’ message and be forced to walk away, causing the operator to lose customers and potential profit. Successful multiscreen streaming requires a delivery model that allows the operator to distribute virtually unlimited tickets and burst up to ten times the typical baseline bandwidth for key live events. This can be deployed using highly scalable servers, such as Edgeware, that allow reliable bursting up to the full potential of the network infrastructure. This means allowing the use of the full available bandwidth of the operator network without the need for the massive capital expenditure associated with building and constantly redeploying traditional servers to support the massive
Centralised vs distributed delivery - network bandwidth = money.
number of connections required only sporadically. Once the event is over, the bandwidth can be scaled back down to lower levels.
In addition to scalability, another key element of managing the required levels of bandwidth is caching content closer to the subscriber. This also ensures high quality of service (QoS). Centralised architectures have a lot of issues with available bandwidth especially during key live events. Video servers that can be positioned at any point in the network and with the support of video aware file systems allow for caching to happen as and where needed. Content that is delivered through servers distributed throughout the network allows for far more predictable traffic and thus significantly improves not only the scalability, but the overall QoE for the viewers.
A peak bandwidth scaling example.
Deeply distributed and scalable hardware enables operators to build a network of highly accelerated video servers which can be distributed deep into the operators’ networks and adapt to different models of caching. This allows the operator to not only deliver high quality IPTV services, but using the same infrastructure, stream high quality videos to Internet connected devices without putting a strain on the core network or investing into building new infrastructures that would support different types of video delivery.
www.ibeweb.com l the connected world supplement march/april 2011 l ibe l 43
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68