What’s ahead for switched digital video
Next generation will bring personalization, information about viewership
The holy grail of the television industry is an ability to deliver infinite video choices and increasingly personalized content in a scalable and economical manner.
Among the technologies needed to achieve this vision, switched digital video (SDV) stands out as one with great potential, and the speed with which the technology has been embraced by cable operators attests to this perception being widely embraced. SDV has moved very quickly from proof-of-concept trials to commercial deployments. The five largest U.S. cable operators are implementing SDV and commercial deployments of the technology now pass an estimated seven million households.
Cable operators need no convincing of the ability of switched digital video to reclaim spectrum and allow them to expand their programming line-ups in standard and high definition. Operators have already used SDV to offer dozens of new channels in nine different languages, add more than 100 high definition and standard definition channels to the switched programming tier and save close to 75 percent of the bandwidth in a single service group.
SDV, however, is about much more than mere bandwidth reclamation: the segmentation and individualization of narrowcast programming distribution opens possibilities for further value enhancements.
Figure 1: The bandwidth consumption of switched unicast trends toward that of switched multicast as service group size decreases.
Virtually infinite programming
One of the most impactful attributes of SDV is that it can remove the shackles that chain the amount of programming that a cable operator can provide to its network capacity. Instead, in an SDV environment, bandwidth demands scale with the number of active viewers in each service group. This demand-side paradigm means that, as the amount of high-quality, long-tail content increases exponentially, there is virtually no limit to the programming that can be made available to subscribers. Other approaches to bandwidth expansion – including compression techniques such as H.264, plant upgrades or migration to only-digital delivery – are generally more expensive, and in some cases, don’t generate as much bandwidth savings as switched digital video.
Increasingly personalized content
Addressable advertising is a frequently cited example of content personalization and one that can boost revenues for cable operators. An ability to customize ads more closely to subscriber interests can incentivize marketers to pay more to the cable operators that support this capability, while interactivity and telescoping can further enrich these experiences. Revenue models for addressable advertising are still in the formative stages but it is conceivable that additional revenues generated by advanced advertising could be shared among content and service providers.
Switched unicast is a form of switched digital video in which each active subscriber receives a unique stream. The technology employs a similar tuning process to switched multicast, except that a subscriber is not sharing the bandwidth with additional subscribers in his/her service group tuned to the same program.
Switched unicast can enable cable operators to deliver different ads and embed customized promotional messages that map to individual viewers’ interests. Switched unicast can lead to higher advertising revenues for operators since advertisers will often pay more for a targeted ad. Web-based advertising can consistently generate a considerably higher cost-per-impression than today’s television advertising, and early trials of advanced television advertising suggest that an uplift of at least 25 percent can be attained by targeted television ads, compared to non-targeted counterparts.
Advanced advertising, however, is far from being the only opportunity for personalization. Other examples include customized mosaics, news bulletins and informational searches.
While switched unicast may initially seem to require many more resources than a switched multicast mechanism, it can be shown that as the number of programs offered increases, the overall bandwidth required per service group is increasingly related to the actual number of active viewers. As a result, the amount of additional resources required to implement switched unicast trends toward that required by switched multicast.
Refer to the whitepaper entitled, “Evolving Switched Broadcast Beyond First Generation Deployments,” presented at the SCTE Cable-Tec Expo conference in 2006, for a broader examination of switched unicast (the paper is available on the BigBand Networks’ Web site: http://www.bigbandnet.com/index.php/
By abstracting services from the underlying networks, cable operators can react more quickly to market dynamics and accelerate the delivery of new services. Effective management of edgeQAM resources enables cable operators to produce superior experiences for their customers by instructing their networks to deliver exactly the right mix of resources for the services being delivered. The proper scheme also provides a means to implement business policy decisions, making it possible to maximize the monetary return of the HFC network investment.
Universal edgeQAMs demonstrate this potential perhaps better than any other technology as they allow switched digital video programming to be allocated to QAMs that may also carry on-demand, data and voice traffic during other times of day. Policy-based management of resources, such as edgeQAMs, can allow a more deterministic and automated approach to delivery of triple-play services across shared network assets.
The introduction of DOCSIS 3.0 channel bonding allows cable operators to utilize their DOCSIS resources not just for data, but for video services also. Carrying video over DOCSIS creates a path for delivery of video to a wealth of end-point devices including DOCSIS-based set-top boxes, hybrid MPEG/DOCSIS transport-capable set-top boxes, and most importantly, other IP-based equipment in the household. While allowing an important paradigm shift in the end-point equipment, subscriber consumption experience and flexibility, the back-office video elements for linear and on-demand video can remain unchanged, allowing operators to protect their current investments.
SDV can be extended to support streams that are carried over DOCSIS. Upon receiving a channel change request, the SDV Session Manager, in conjunction with an Edge Resource Manager (ERM), can make an intelligent decision whether to deliver the stream over DOCSIS or over MPEG transport, based on the tuner capabilities, the available resources, the currently available streams in the service group, and any operator-defined business rules. QAM assets can be re-purposed between MPEG and DOCSIS transport, allowing for better utilization of cable operators’ resources.
This direction also calls for alignment between, and possibly consolidation of DOCSIS policy servers and ERMs, in order to allow for an operator to easily define business rules that are applicable to both MPEG and DOCSIS transports and to seamlessly re-purpose QAM resources between the two transport mechanisms.
Enhanced video quality
Early SDV implementations adopted certain aspects of on-demand, particularly the use of constant bit rate (CBR) program streams. Improvements in bandwidth efficiencies and picture quality, however, can be attained by moving to a variable bit rate (VBR) model, where multiple single-program-transport-streams (SPTS) with variable bit-rates are multiplexed at the edge. A VBR implementation can allow 14 to 15 SD streams to be switched in a QAM while attaining the same overall video quality achieved today when multiplexing only 10 CBR streams. While VBR is a well-known concept for digital video, introducing VBR into switched digital video provides operators with the flexibility to deliver high-quality video while optimizing QAM capacity. In essence, this allows higher bit-rates and quality when network resources are under-utilized and higher stream density in cases where viewership statistics require it.
Any scalable and cost-effective implementation of variable bit rate SDV should avoid the need for additional devices at the edge of the network. This is due to the direct correlation between the number of tuners in subscribers’ homes and the number of edge devices required to support them. Instead, the processing and analytical functionality required to implement VBR switched digital video are best located in the headend where cable operators will benefit from scalability and cost advantages.
Switched digital video is a technology of choice in more ways than one, endorsed by a growing number of operators, and already expanding programming options for subscribers. SDV’s bandwidth-saving capabilities are well established, and progress toward next-generation advances, such as personalization, is underway. The network infrastructures that support these services are taking strides toward becoming more versatile, while architectures are increasingly characterized by openness and multi-vendor collaboration. SDV is already changing the way television is delivered, but future implementations, with the promise of virtually infinite, addressable programming, will cement its legacy. It’s certainly a future worth watching.