The transition to IPv6 will be necessary sooner rather than later.
One by one all communications services and applications have migrated or are in the process of migrating to Internet Protocol (IP)-based transport. IP enables the convergence of all data, voice and video services onto a common infrastructure that delivers both increased returns on investment and reduced operational costs.
The open-standards nature of IP eases integration of new capabilities and features, increases the agility and scalability of the infrastructure and facilitates the creation of new or enhanced services through mash-ups of existing ones. Broadcast services infrastructures are on track to inevitably converge to IP. While the production and contribution components of the infrastructure might lag behind distribution in adopting IP, the characteristics of current and future content sources and content consumers demand a converged, end-to-end IP network. (See Figure 1.)
The business and technical case for IP-based broadcast networks is not based only on the envisioned opportunities offered by the converged infrastructure, but also on the capabilities of today's IP and multiprotocol label switching (MPLS) networks. According to the Broadcast Engineering article “Video over IP” from February 2008, high-bandwidth, Ethernet-based transport media features enable network convergence within 0.5 and 0.05 seconds, low jitter and error correction capabilities, along with protocols such as multicast and quality of service (QoS), which enable optimal use of the infrastructure and make IP networks fully capable of supporting video services.
The next generation of IP
With the case for a migration to IP infrastructures made, there is only one step left to justify the very existence of this article: explain its focus on IPv6, the next generation of IP. The main characteristics of IPv6 and the case for its adoption were covered in a December 2008 article in Broadcast Engineering titled “IPv6 is coming.” It emphasized the need for a large IP address space to sustain the rapid and continued adoption of the Internet to support the increasing number and types of devices connected via IP and to enable services that rely on peer-to-peer and mobility-focused communication. While IPv6 is not dramatically different from IPv4, the version of IP currently in use, it provides the resources needed to scale up IP-based infrastructures and services. The adoption of IPv6 is picking up steam, so it makes sense to keep it in mind when planning the deployment of broadcast services over IP.
The adoption of IPv6 in certain areas of broadcast services infrastructure is easily justified. With end systems such as mobile handsets and home networking devices adopting IPv6, it makes practical and strategic sense to consider IPv6 for distribution. There are already broadband access providers that deliver video content over IPv6.
For other areas, however, the case for IPv6 adoption might be less evident, which could lead to a mix of IPv4 and IPv6 infrastructures, each supporting specific aspects of the broadcast service. It is important to remember that the two protocols are incompatible, and the insertion of protocol translation gateways that lead to loss of functionality and performance should be avoided. Instead, the boundaries between the IPv4 and IPv6 domains should be aligned with the boundaries for the various components of the broadcast service. For example, content sources and production centers could be connected over an existing IPv4 network, while distribution is done over IPv6.
Although the coverage of today's IPv6 networks might not be sufficient to support the entire broadcast services chain, planning should be done for end-to-end IPv6 transport.
The architecture, design and products needed to deploy broadcasting services over IPv4 are well understood and are already being used. Because, for the most part, IPv4 and IPv6 are similar, the same architecture and design principles apply to both protocols:
The forwarding features and performance for IPv4 and IPv6 are similar.
Routing concepts are the same, so design is IP version agnostic but topology and environment dependent.
Similar network convergence times can be achieved with either protocol.
IPv4 multicast capabilities and architectures are available with IPv6.
The same features and capabilities available for implementing QoS for IPv4 are available with IPv6.
For the most part, the same security policies apply to both IPv4 and IPv6. While some threats are eliminated by IPv6, new ones are introduced, which require explicit mitigation.
There are finer differences, however, between the two protocols, which provide alternative design options and service implementation or enhancement:
The new IP header structure provides the resources to enhance content delivery services and to extend the protocol as needed to support new services.
IPv6 offers a new rendezvous point (RP) discovery with the introduction of embedded RP, while the overall deployment of multicast is made easier in IPv6 through the new address architecture.
Mobile IP provides more scalable deployment options with IPv6.
For infrastructures leveraging MPLS, there is no real differentiation between the way IPv4 and IPv6 is being tunneled. The control plane of the MPLS infrastructure might be supported and implemented only in IPv4; however, this has limited impact on transporting video content over IPv6.
In reality, the main concern of the broadcaster who is implementing broadcast services over an IPv6 infrastructure most likely is the full support of the protocol by all the elements involved in supporting the service. Everything from encoder/decoders and headends to video gateways and networking equipment needs to be evaluated for IPv6 readiness.
The adoption of the protocol is ongoing, and so is its maturity in terms of product support. Any equipment that leverages hardware assistance needs to be designed with IPv6 in mind. Major manufacturers have been supporting IPv6 for some time, and feature or performance gaps (in comparison with IPv4) are fewer and less constraining. Nevertheless, as part of the planning process, it is important to identify gaps and future needs while working closely with vendors to mitigate the gaps.
The convergence of broadcasting services onto an IP infrastructure is inevitable. If your service already leverages IPv4 transport, you need to start planning for the integration of IPv6 and a migration and expansion of existing services to IPv6. Most technical and operational aspects of operating an IP network for broadcasting services will apply to the new infrastructure, which at first will operate in parallel over the same hardware as the IPv4 one.
If you are looking to implement IP transport, then you should plan it in a version-agnostic (IPv4/IPv6) way and focus on IPv6 wherever possible; plan to leapfrog. (See Figure 2.) IPv6 transport or capabilities might not be available in all infrastructures you need; however, with expected exhaustion of the IPv4 address space by 2011, the transition to IPv6 will become necessary sooner rather than later. The important thing is to be prepared, to understand where IPv6 will become critical first (distribution, for example) and to address the needs of existing customers and those of new ones. (At the consumer level, IPv6 adoption is faster in Asia and Europe.)
It is never too early to start inventorying IPv6 capabilities and the needs of the IP infrastructure supporting your current or future broadcast services. Early planning ensures timely readiness and lower implementation and operating costs.
Ciprian Popoviciu, PhD, CCIE, leads the Systems and Infrastructure Data Center architecture team at Cisco Systems. He is co-author of two IPv6 books, “Deploying IPv6 Networks” and “Global IPv6 Strategies,” and is a senior member of the IEEE.