Managing Cloud-Based Cable TV DVR Migration Prepares to for IP Prime Part Time 1 An Executive Summary of Deployment Techniques and Technologies Series Introduction: Market Drivers and Tech Challenges John Ulm, Fellow of the Technical Staff
Introduction While DVR services are still relatively young, today s consumers are already seeking new capabilities when it comes to time-shifted television. Whether it s recording more shows at once, storing additional hours of programming, viewing saved programs in any room, or even taking their content with them on the road, today s savvy subscribers are beginning to clamor for a better DVR experience. But these advanced new features can push the limits of traditional set-top boxes, leaving service providers to investigate new architectural options for delivering nextgeneration DVR services. After multiple fits and starts, cloud-based DVR services are beginning to once again gain popularity as a means of advancing DVR capabilities for subscribers. However, several challenges to this network-based DVR implementation remain, ranging from substantial infrastructure investments to enhanced content management capabilities to expensive business agreements with programmers. This paper focuses on the techniques and technologies that may prove useful as service providers seek to overcome these issues and bring cloud-based DVR deployments into prime time. Overcoming the Challenges of Cloud-Based DVR Deployment By moving DVR storage into the cloud, service providers can sidestep many of the limitations of today s home-based DVR implementations and unlock several new benefits as well. For consumers, cloud-based DVR can mean recording more programming simultaneously, eliminating storage limits, and making saved content readily available throughout the home and on a host of IP connected devices. For service providers, it can dramatically reduce the overall storage budget, lower failure rates for set-top boxes, and help offset the costs of networkbased video processing resources for VoD and multiscreen services. Even with all of the benefits, several challenges must be resolved before cloudbased DVR deployments go mainstream. While centralized storage is more cost effective than distributed set-top storage overall, it still requires a significant upfront investment. And once content is stored in the cloud, it must then be transcoded and streamed to subscribers, which requires new spending on ingest, network, server and CDN resources. But the most significant challenge for service providers may be in the new agreements that need to be forged with programmers that enable a single copy of content to be shared among multiple subscribers. Overcoming these challenges requires that service providers use every tool at their disposal. This means maximizing the utilization of centralized pools of storage and processing power, leveraging new capabilities in transcoding and packaging, and 2 ARRIS 2013. All rights reserved.
deploying both network DVR (ndvr) and remote storage DVR (RS-DVR) architectures. Maximizing Centralized Storage and Processing Moving the storage of DVR content from the home to the cloud carries with it substantial capital and operational expenses. In order to deliver this comprehensive service, infrastructure is required for storage, ingest and networking and managing these resources requires software and personnel. However, all of these investments have one important thing in common- they are centralized and can be shared across multiple services. The storage pool that will be used to hold DVR content, along with the ingest capabilities such as transcoding and packaging, can all be leveraged to enable video on demand and multiscreen content as well. While these resources may be partitioned to ensure service fidelity, they can also exist as centralized storage pools and content processing engines that can be managed and scaled as one to achieve significant capital and operational efficiencies. Similarly, the network bandwidth required for cloud-based DVR can be expanded to benefit broadband performance, channel capacity, video quality and the delivery of multiscreen video delivery. Leveraging New Transcoding Techniques and Capabilities In addition to the savings associated with resource sharing, service providers can take advantage of emerging transcoding and packaging techniques to enable a more efficient approach to cloud-based DVR video delivery. This begins by 3 ARRIS 2013. All rights reserved.
creating a single FIGURE 1: Cost Comparison Between Content Storage of all formats (left) vs. Single Mezzanine Copy with Transcoding On Demand (right) mezzanine video file that can be used to create all of the file versions needed for streaming to set-top boxes and IP-connected devices. The mezzanine file enables the correct ABR version to be transcoded on-the-fly, rather than transcoding and storing multiple versions for playback on multiple devices. Using this technique, service providers can save significant storage costs for both cloud-based DVR and multiscreen video delivery, as shown in Figure 1. As multiscreen service growth continues, so too do the innovations in transcoding. These innovations, including two emerging technologies known as cooperative transcoding and dynamic profile selection, may also prove useful in cloud-based DVR delivery. By eliminating redundant video processing functions, cooperative transcoding can significantly reduce the burden on today s transcoders, freeing up these resources to process additional multiscreen or DVR content. Dynamic profile selection takes into account the uniqueness of each piece of content before selecting the correct profile and creating video files for streaming. This avoids the creation of ABR files that consume more storage and more network bandwidth than needed, without degrading video quality. With the transcode requirements for multiscreen, VoD and cloud-based DVR services increasing annually, these two emerging innovations may be crucial to keeping costs low for service providers. Deploying a Hybrid Common Copy and RS-DVR Architecture One key obstacle standing in the way of cloud-based DVR deployment is the current contractual agreement between programmers and providers that require copies of recorded content to be stored separately for each subscriber. With homebased DVR, this happens naturally because all recorded content is saved on the local hard drive within the set-top box. And moving DVR storage into the cloud is no different, as long as a unique version of each piece of content is saved each time a user records it. This remote storage DVR (RS-DVR) implementation uses the same amount of storage as home-based DVRs, but that storage is centrally located and managed, which achieves capital efficiencies and reduces operational expenses by eliminating truck rolls associated with hard drive failures in the home. But cloud-based DVR implementations can be made far more efficient when service providers move to a common copy approach, where a single version of each piece of content is stored centrally and shared among multiple subscribers. The common copy approach is the preferred storage method for service providers 4 ARRIS 2013. All rights reserved.
seeking to drastically reduce their storage infrastructure requirements and associated storage management costs. But moving to the common copy approach requires new unique copy content agreements, which will carry a premium over today s multiple copy deals. To help determine the best architecture for cloud-based DVR, ARRIS recently analyzed consumer viewership patterns for 100, 200, and 400 channel lineups. In this study, ARRIS determined that even across the largest channel lineups, the top 20 programs on TV account for a very large percentage of the viewers. All of the remaining long tail programming is then watched by a much smaller proportion of the viewership. Summarized in Figure 2, these usage patterns indicate the potential for a hybrid approach to cloud-based DVR that mixes both common copy and RS-DVR. FIGURE 2: Viewing Patterns for 100, 200 and 400 Channel Lineups By securing unique copy agreements for only the top 10-20 programs, service providers can reduce their network storage costs by 50% to 70%. Although 90% of content still requires RS-DVR, the storage impact of this content is minimized because it is recorded and viewed far more infrequently than popular content. By utilizing viewing patterns and deploying a hybrid common copy/rs-dvr architecture, service providers use the most cost-efficient method to deliver their entire content lineup. Conclusion 5 ARRIS 2013. All rights reserved.
While the excitement around cloud-based DVR has been palpable for years, several challenges have obstructed the road to widespread deployment. However, the video delivery industry has begun overcoming many of these obstacles. Infrastructure costs and their associated operational expenses can now be shared with other services such as VoD and multiscreen video delivery to help offset costs. Innovations in transcoding are making it more efficient to process large amounts of video quickly. And better information on users and the content they watch are yielding new architectural choices that can reduce the expenses of both content itself and infrastructure. As the challenges of cloud-based DVR deployment begin to subside, service providers have begun to contemplate the many business opportunities associated with this exciting new service. They see the efficiencies that come with unifying their infrastructures for multiscreen video delivery, VoD and cloud-based DVR into a more efficient virtualized architecture. They calculate the possibilities of improving retention by giving subscribers the advanced DVR capabilities they crave. And they imagine new revenue opportunities associated with targeted advertising, expandable networked storage, and a host of must have cloud-based apps that provide new opportunities for differentiation. The key to bringing cloud-based DVR into prime time is in developing a path to deployment that leverages current investments and establishes opportunities for both consolidation and efficiency. Related Reading The Economic Drivers for Cloud-Based DVR Services this paper models key cost drivers and demonstrates the economic impact of offering cloud based DVR services, addressing the issues of capital costs and the trade-offs between storage, transport and transcoding. Dynamic Profile Selection & Cooperative Transcoding: An Efficient Future for the Multiscreen World this paper presents a co-operative transcoding technique that can substantially reduce the processing requirement of generating multiple representations as well as produce a uniform quality of user experience across the multiple representations. Motorola Mobility Media Engagement Barometer this independent global study details video consumption habits among 9,500 consumers in 17 countries. The study looks closely at new and emerging content trends, such as multi-screen habits and recording behaviors, which are dramatically shifting the way audiences are consuming video. 6 ARRIS 2013. All rights reserved.
ARRIS Enterprises, Inc. 2013 All rights reserved. No part of this publication may be reproduced in any form or by any means or used to make any derivative work (such as translation, transformation, or adaptation) without written permission from ARRIS Enterprises, Inc. ( ARRIS ). ARRIS reserves the right to revise this publication and to make changes in content from time to time without obligation on the part of ARRIS to provide notification of such revision or change. ARRIS and the ARRIS logo are all trademarks of ARRIS Enterprises, Inc. Other trademarks and trade names may be used in this document to refer to either the entities claiming the marks and the names of their products. ARRIS disclaims proprietary interest in the marks and names of others. The capabilities, system requirements and/or compatibility with third-party products described herein are subject to change without notice. 7 ARRIS 2013. All rights reserved.