paradoxically it may come from a large number of HD HDR channels which more organisations will be able to justify launching .” DEPENDENCY . “ In terms of IP , there is perhaps more of a dependency on what form of IP is being used . Existing IPTV systems use UDP ( User Datagram Protocol ) which is inherently designed for streamed media . However , TCP ( Transmission Control Protocol ) based schemes are becoming more widely used ( notably for HTTP Adaptive Streaming ). While for UDP , bandwidth is pretty much the only consideration , for TCP many other network characteristics and system configurations can play a role ,” he advises .
According to Richard Brandon , CMO at Edgware , there is no doubt that 4K , and potentially 8K , will become as popular as the move from SD to HD television . “ But this is just the start of the problem in terms of network capacity . This new multiple in bandwidth consumption is hugely compounded by the trend from ‘ multicast ’ to ‘ unicast ’ traffic - as viewers want to watch programmes when they want - pausing them , time-shifting them , and ‘ rewinding ’ them back to watch highlights . Then there ’ s the desire to show personalised ads as well , or a choice of camera angles . All of this means a move from a single stream of packets per show ( albeit a larger one with 4k TV ) to millions of streams of packets - one per viewer , each personalised in terms of timing , content , ads and even device type .”
“ So , the above all sounds like pretty bad news for networks unless fibre everywhere is deployed , and plenty of it . But thankfully this isn ’ t necessary . The first thing to remember is that network bottlenecks are not all created equally . The last mile into the home isn ’ t shared by anyone else - so while it needs to be fast enough to cope with a few simultaneous channels , it can still be supported over copperbased technologies ,” he says . DEEPER . According to Brandon , the real problem manifests itself deeper in the network , when these multiple individual IP streams are aggregated together . “ Just throwing extra bandwidth at that problem can cause all sorts of problems . However , all is not lost there , because every TV stream from the origin location to the metro or last mile distribution point doesn ’ t necessarily have to be delivered . TV caching technologies , like Edgeware ’ s , are far lower cost than building out new network capacity - and delivering TV without delays or packet loss when the core ,
“ The industry must look towards reducing bitrate requirements much further .” – Fabio Murra , V-Nova
metro or aggregation networks are congested .”
“ This is achieved by simply keeping a copy of the most popular content - live or recorded – close to the viewer . The content can then be streamed out as needed , enabling pausing or ‘ rewinding ’ as required . The content can then be repackaged for the correct device format and personalised ads inserted if needed . Netflix has already realised this with its own TV delivery architecture , deploying local delivery servers that keep down its network or CDN costs and ensures TV is delivered with higher quality . So in conclusion , do network operators and TV content providers need to address this problem ? Absolutely . Does it require a massive new investment in fibre ? Thankfully not ,” he states .
“ Making broadband scale for video delivery is a major challenge for ISPs ,” admits Trow . “ Multicast offers the potential to scale IP delivery , but it is a fundamentally different network technology to the current , ubiquitous , unicast-based infrastructure . Upgrading broadband provision for multicast is expensive and only really a viable option for the largest operators , who can offset the significant CAPEX associated with such an upgrade . Another vital tool is caching content to reduce the burden on congested networks . However , caching is very much a containment exercise in a scenario where 1G performance is needed , particularly for live video streaming applications .” BREAKTHROUGH . “ Big sporting events are definite drivers for advanced television formats , but the major breakthrough of 4K is yet to come on a global scale ,” says Narjus . “ The advanced image resolutions are already putting more pressure on data transmission capacities and the development is guaranteed to continue strongly . If we think about
millions of consumers watching football world championships , and simultaneously streaming the content on 4K over IP , it is easy to forecast that the lack of sufficient bandwidth causes IP based networks to collapse . However , the current DVB-C can carry the load easily . In future , the logical phase after the DVB-C delivery will be all-IP television , but issues such as network capacity , service availability and even the availability of attractive consumer devices will ensure that linear television stays important far to the future ,” he predicts .
“ There is a clear shift towards IP-delivered services in television , and with the advent of higher-resolution formats and the proliferation of video-capable devices in and outside of the home , IP networks are going to be under pressure ,” warns V-Nova ’ s Murra . “ According to multiple reports , video traffic growth is well into double-digits and cannot therefore be tackled simply by the 50 per cent typical incremental performance benefits we get every 10 years ( roughly ) from legacy compression approaches . It ’ s going to be a mixture of new technologies , from costly fibre upgrades , 5G deployments and – importantly – new approaches to video processing .”
“ It is clear that HDR and 8K are firmly on the TV industry agenda , with 4K already embedded in consumer expectations , they are currently purchasing high-quality video consumption devices . Demand for highquality video is growing faster than the speed at which new infrastructure is being deployed , creating a divide between those that can receive services and those that cannot . Netflix , for example , still recommends a 25 Mbps minimum connection for its 4Kp24 services .” STRAIN . “ The good news is that nextgeneration codecs will help both alleviate the networks ’ strain and target the devices already in the field , making high-quality IP delivered video a mass market reality , whether 4K , full HD over DSL or SD video on 2G mobile networks . For example , in trials with EE , we have already demonstrated full 4Kp50 video over 4G / LTE at 6 Mbps . Let ’ s not forget that the problem of delivering the highest-quality 4K services starts with the capture and contribution of premium content , such as sports . Fibre connectivity can still be a bottleneck in certain venues , due to both
20 EUROMEDIA