We are entering a new period in the Internet-video era, in which proliferation will reach new heights and video traffic will predominate on Internet-protocol networks.
Technologies have emerged to address many of the challenges posed by scaling video delivery over IP networks, but obstacles must be overcome before the vision of “one subscriber, any device, any network” can be realized. In particular, new adaptive technologies are required to extend video services over wireless devices and provide consistent cross-device user experiences.
A recent Cisco Systems forecast projects that video will account for 90% of all consumer IP traffic and 64% of all mobile traffic by 2013. Increases in online viewership, HD-quality video consumption, video on demand and other MSO-content catalogs, and free Internet video services such as YouTube all contribute to this rising tide. Triple-play TV, telephone and Internet service packages are extending to quad-play packages that include mobile and go deeper than mere billing integration. Both cable operators and carriers have a strategic interest in this new market.
IPTV technologies are addressing the most immediate infrastructure problems posed by the coming high-volume phase of the Internet video era. Unlike their Web cousins, video CDNs (vCDNs) are designed to meet the stringent latency requirements of VOD at Internet scale and provide a cost-effective way of distributing large volumes of on-demand video content without incurring heavy storage or transport costs. Also, as the amount of available video content increases, supporting a national MSO footprint will require ever more intelligent video caching and storage that does not hamper delivery of network-based DVR services.
vCDNs are designed to perform two distinct functions. First, the core network is designed to transport on-demand video that can cut through edge and multiprotocol label switching routers and intelligently cache content on demand. Second, the vCDN edge — what we might call the vCDN “last mile” — is responsible for adaptation and delivery of content to various end points. Thus, the edge is where “one subscriber on any device, over any network” is made possible. However, the devil is in the details here. Delivery of content to mobile end points at the last mile is daunting and is probably the least-understood area in the industry.
The notion of “one subscriber, any device, any network” implies that the subscriber retains his/her personal condition (or state) over multiple sessions across multiple content-consumption endpoints, including set-top boxes, desktop computers, mobile phones and netbooks. But additional intelligence is required at the network edge to provide an integrated user experience, sometimes referred to as nomadic state management, across platforms. In addition, presenting a unified user experience across all platforms requires extending current desktop-Web standards to dynamically adapt presentation and delivery to mobile screens at the last mile.
Delivering mobile video is a significant challenge. Today’s video-streaming protocols are optimized for desktop delivery, not for bandwidth-constrained mobile networks. Unless mobile streaming can achieve high-quality video while adapting to uneven network bandwidth and cellular-network congestion, user experiences will continue to be less than desirable. Indeed, bandwidth availability for cellular networks is too sporadic to guarantee a consistently acceptable user experience without considerable intelligence inserted at the point where the browser or application client shakes hands with the storage and delivery network — at the edge.
Content protection, including user-rights management and digital rights management, is also a key challenge to managing a common, multiplatform subscriber ecosystem. Device-level DRM, while common for the desktop, is still in its infancy for mobile endpoints, largely due to the constrained device operating-system environment and scarce CPU and system resources.
Thus, vCDN edge technology should be advanced toward a common set of services for managing nomadic state, DRM and interactivity for all three screens. Meanwhile, delivery to each screen must be adaptive to account for the characteristics, constraints and network conditions of any given end point in real time. A mobile-specific adaptive streaming technology for volatile, lower-bit-rates traffic is a fundamental requirement.