Flexible Networks for Media Services

To help understand what is evolving in the transport of these services, we will explore some of the networks, transports and switching mechanisms employed today--with an eye on the future of media distribution.
Author:
Publish date:

The deployment of networking technologies is reaching deep into the content space, making it possible for the user to enjoy new experiences using a flexible set of alternatives. To help understand what is evolving in the transport of these services, we will explore some of the networks, transports and switching mechanisms employed today--with an eye on the future of media distribution.

(click thumbnail)
Ultra-high band digital transport for media has been under development for many years. One of the transport mediums is the use of the hybrid-fiber/coax (HFC) cable system, a combination of high-speed fiber and last-mile copper.

The introduction of this technology has enabled networks originally designed for video services to provide reliable bandwidth for interactive video, data and voice services.

Those large telecom companies that have overlaid video on fiber-in-the-loop (FITL) architectures efficiently carry analog video on synchronous optical network (SONET) backbones. This structure solves the power challenge imposed by dense wave division multiplexing (DWDM) on optical amplifiers in long-haul networks.

In the United States, with its 300 million analog television sets--most of which are cable-ready--the analog structure must remain intact for years. This reality has driven fiber and hybrid technologies. Today, much of the current bandwidth is still used for analog video, which highlights a fundamental difference between HFC and copper-based networks.

Copper's bandwidth limitations, coupled with significant advances in digital subscriber line (DSL) technology, have forced a switched approach, resulting in untenable interface costs. For mass deployment, and to address the growing volume of interactive traffic, it was necessary to expand the service set offered over HFC. Changes were necessary in both in access and transport.

TRANSPORT EVOLUTION

Modern media services are deployed from servers that may be located throughout the network, necessitating that transport be flexible and connections efficient. Since services rendered from these servers typically start out at the headend, significant strain is placed on the network at this point. The goal in providing a suitable video and interactive service is to maximize the reusable bandwidth per user, while providing the flexibility to connect servers anywhere on the network.

A critical element in building the high-speed multimedia backbone for digital transmission is the deployment of video-optimized SONET multiplexers. Technologies integrated into this network are enabled through wave division multiplexing (WDM), which not only increases bandwidth but permits optical routing, and results in lower access costs. Furthermore, as the network grows, so does the amount of carriage on fiber. Accordingly, passive optical technologies become essential from both the cost and performance perspective.

SONET is a physical network designed to provide a universal transmission and multiplexing scheme through the standards of the other OSI layers. SONET offers cost-effective transport, both in the access and core areas of the network. SONET transport is widely relied upon in telephone and data switches for interconnection. Because the services (i.e., video, voice and data) require a resilient network, SONET was adapted to address resiliency--the ability of the network to cope with the loss of a link or node and still provide alternative routes for the transport of traffic.

SONET, for cable operators, would be more costly than current HFC architectures. However, once operators want to add high-speed data and telephony, they must add switching. An Asynchronous Transfer Mode (ATM) or IP network is better suited to this kind of traffic.

ATM, a connection-oriented high-speed general-purpose transport mechanism, can be implemented on several physical networks, including SONET. ATM, being a cell-based technology designed for switching and multiplexing, is excellent for carrying multimedia formats such as video, audio, text and graphic imaging. The extremely high speed of ATM is accomplished by fixed-length cells switched through hardware-based processing and routing systems.

Transporting ATM traffic over SONET is accomplished by continuous mapping of the ATM cells into the SONET frame. Additionally, IP packets can also be formatted into ATM cells, enabling content generated within an IP network to be transported over the ATM data transport.

However, IP transported over ATM is not terribly efficient because of the overhead requirements of the ATM cell structure, and the inherent overlap of IP datagrams that map well outside the ATM cell size. Internet service providers are considering running IP content directly over a SONET to improve efficiency.

From and advantage/disadvantage perspective, IP over ATM does offer bandwidth management, which IP over SONET does not. ATM further provides for end-to-end quality of service, whereby SONET is strictly point-to-point.

Addressing and routing for IP over SONET is possible, but requires extensive provisioning. ATM already has simple, fast provisioning built in. ATM by design offers flow control, whereas SONET has none. Both architectures for IP provide fault tolerance, albeit through desperate means (SONET uses dual-ring; ATM, dynamic routing protocol).

All of these transport issues seem to be relatively in the background from an end-user perspective, but as the deployment of high-speed multimedia services steps to its next level, so will next-generation networks. The impact will affect content delivery providers, including broadcasters, networks and cable services. Those new players in the marketplace will certainly have something to crow about once a successful deployment is in place. Until then, service providers continue to deploy new services over existing transport systems.