06.20.04

Fabrice Beer-Gabel, BigBand Networks

All network operators including cable, telecommunications, wireless, and satellite, are facing steep financial pressures as they are expected to leverage the investments they've made in their plants by increasing subscribership and expanding existing revenue streams.

Technology advances are allowing for greater competition and many operators are expanding into markets previously not available to them. In the Telco market, Independent Operating Companies (IOCs) and Regional Bell Operating Companies (RBOCs) are competing with cable and wireless operators to sustain their voice and high speed data revenue, while expanding their competitive sphere by entering the video business, leveraging their high bandwidth xDSL or FTTx network deployments.

Telco operators, as opposed to other service providers, have a unique opportunity in that they are investing in building a video operation from the ground up. They are able to make the most economically and technologically sound decisions without having to support existing deployments with legacy technologies. Their investment decisions must be made to build a scalable infrastructure that supports more services and more subscribers over time. The goal is not only to increase revenue in a highly competitive environment but also to quickly achieve sustainable and growing profitability.

This paper offers headend and network design recommendations for both large and small Telco operators to leverage their core competency and build a profitable video operation. These recommendations are centered around building a flexible switched video architecture designed to exceed customer expectations of services provided, and maximize profitability through advantages such as local advertising revenue and reaching more subscribers optimizing use of both RF and IP access networks in mixed environments. At the core of this architecture resides a programmable switch digital video (SDV) platform that enables non-disruptive service evolution and standard migrations, and leverages low-cost and robust Gigabit Ethernet transport to provide high-quality video services at life-line reliability.

Benefits of a switched video architecture

Very much like telephony, the video world has become highly competitive, with subscribers able to easily change providers. Service downtime can be disastrous in increasing not only customer churn but also operational expenses required to maintain large customer call centers and service dispatch capabilities. Telco operators enjoy the benefits of a switched network architecture, a perceived high service reliability by their customers due to their long history of life-line service delivery, and  strong ties with local communities. A programmable switched digital video platform takes advantage of and builds on these strengths. Such a platform provides the following benefits:

• Greater service reliability: a carrier-class platform in modular chassis form with built-in redundancy minimizes video channel interruptions by reducing the number of separate network pizza boxes or video processing appliances that each introduce more failure points, cabling problems, and hardware or software compatibility issues among various discrete units and suppliers. The platform's rich media processing can be configured to de-jitter programming distributed over low cost GigE networks and its active intelligence can detect video quality degradation across the network to trigger dedicated service-level redundancy functions between interconnected video headend facilities.
• Flexible service evolution: programmable hardware processing modules such as Field Programmable Gate Arrays (FPGA) and Digital Signal Processors (DSP) provide maximum flexibility to evolve the network with new services (e.g. advertising, HDTV and network based PVR), new compression standards (e.g. MPEG-4 H.264 or Windows Media VC9), video quality compression and transcoding advancements and software upgrades. For example, the operator can start with a basic digital turnaround module that controls channel line-ups including dynamic bit rate adaptation of MPEG-2 programming by rate-shaping techniques that act according to the access network bandwidth requirements and upgrade the platform over time to support advertising, cost efficient transcoding of HDTV programming to new compression standards and low cost GigE transport to reach remote subscriber communities over either IP or RF access networks. This provides a cost-effective means for the operator to control expansion and scaling of services, with experience gained along the way used to choose among flexible options that the modular platform approach provides.
• Efficient management of services: All services flowing through the platform are controled from singular management that provides a global view of bandwidth availability across access networks and the ability to monitor any program from a single location. The programmable platform enables remotely loading of software upgrades and new applications, while the switching core enables to make programming changes without re-cabling and via a simple drag-and-drop operation in the management system.

Building a profitable advertising revenue stream

Cable operators generated $3.65B in revenue in 2001 from local advertising . This figure is expected to grow past $6B by 2010, representing 40% upside on top of digital subscriber revenue . Not only does advertising represent a significant revenue opportunity for Telco operators deploying video services, it helps them mitigate business risks by offering diversification of revenue while offering a valuable and differentiating community service to local businesses. As operators plan the deployment of advertising services, they must consider the multiple implementation options offered to them and understand their respective impact on profitability. An all-digital implementation in the Switched Digital Video platform is recommended to minimize the investment and offer the most flexible and scalable solution.

The majority of the advertising infrastructure deployed today relies on analog insertion systems. These systems use a combination of analog switches and video decoders to insert commercials in selected broadcast channels. Their use is primarily explained by legacy infrastructure requirements where programming distribution to the operator and its delivery to subscriber were in analog format. These systems weren't designed to support the video industry's migration to all digital distribution and as a result lack the flexibility offered by digital systems. Moreover, Telco operators deploying IP video services may elect to avoid carrying an analog tier and would incur un-needed costs to digitize and distribute this content on their network.

Figure 2: Analog advertising systems in IP networks offer limited flexibility and un-wanted infrastructure costs

The massive migration to digital services has brought with it digital commercial insertion standards and technologies that introduce efficiency and flexibility in the ad-insertion process. In a digital insertion system, a television channel is received from a programmer as a compressed digital stream based on the ISO MPEG-2 standards. The insertion process is handled via standard-based communication and processing between an ad server and a splicer. Within the stream, specially placed cue markers provide the exact timing location of advertising opportunities or "avails" for the program. The cue messages are specified by the Society of Cable and Telecommunications Engineers (SCTE) 35 2001 standard and are forwarded to an ad server using another standardized protocol, SCTE 30 2001. The ad server then streams an advertisement (also in MPEG-2 digital format) and the SDV platform splices it into the existing program stream. The media processing capabilities present in the splicer allow this splicing to occur in both a precisely timed and visually seamless fashion. The splicing process is handled entirely in the compressed digital domain, eliminating the expense of decompressing and re-compressing the program.

Figure 3: All-Digital advertising systems migrate towards targeted services

Telco operators deploying IP services can greatly benefit from implementing an all-digital insertion architecture to reduce their implementation costs. With most broadcast content being available in digital form, an all-digital advertising architecture eliminates the need for costly encoders for all programming sourced in digital format. Moreover, the integration of the splicing function in the SDV platform increases the system's flexibility and future proofs the network. In addition to providing a single point of management for all browdcast services, the use of the switch enables operators to allocate the insertion module to any channel across the entire lineup by a simple drag-and-drop operation in the management software, eliminating the need for complex re-cabling. Finally, this implementation enables a smooth migration towards targeted advertising services. This is made possible by using the SDV platform to duplicate broadcast channels while the splicing modules insert different commercials in each copy of the channel targeted to either different geographic zones or demographic segments. Operators deploying video services over an RF network will want to consider an all-digital advertising architecture to leverage low cost GigE transport technology to efficiently target advertising across a large subscriber base without the need to maintain parallel networks. These transport capabilities are further described in the next section.

Designing for IP and RF multi-access environments

Whether they deploy the triple-play over fiber or copper physical networks, Telco operators often face the need to support a mix of IP and RF access networks. Several drivers get them into RF. First, a Telco operator may acquire an existing cable plant to accelerate entry in the video business and understand the dynamics of program acquisition. An operator can also acquire CATV assets as a result of expansion and consolidation to increase the subscriber base and build economies of scale. Secondly, many customers may wish to avoid having to use expensive and intrusive Set-top boxes on 2nd, 3rd of 4th TV sets.  This not only keeps customers happy, but also saves the operator millions of dollars in set-top box investments by leveraging a cable-ready TV-set RF tier.  Finally, an operator may select an RF transport architecture that best fits its current network architecture and enables the use of mature technology to launch a wide portfolio of services such as HDTV until new compression techniques allow delivery of these services over more bandwidth-limited IP access networks. Whatever the driving scenario is, it becomes essential to build upon a headend that can grow in terms of services and subscribers over both IP and RF access networks while minimizing capital and operational expenses in order to accelerate profitability. Several architecture choices can be made, but a digital-centric switching  core is recommended.

An obvious path to implementing a mixed IP/RF system could be to split signals at the program acquisition stage and to send them to either the SDV sub-system for IP distribution, or to analog and Quadrature Amplitude Modulation (QAM)  modulators for RF distribution. Several challenges  exist in this implementation:

• The resulting complex headend cabling prevents easy modifications and additions to the channel lineup.
• Supporting advertising across these networks requires the addition of multiple analog inserters and digital splicers, adding points of failure to the network and increasing the complexity of maintaining the service.
• No single management system is provided to provision, monitor and maintain services across networks significantly burdening the cost of operating such a network.
  

Figure 4: Multi-access headend implementation with limited flexibility and high maintenance costs

A digital-centric approach can alleviate these problems by bringing all programming into the digital realm for processing, and delivering access network specific streams as a last step in the distribution to the subscriber. This approach leverages the multicast capability of the SDV platform across both network types to packetize, multiplex and transport content according to network specific requirements. This is achieved under a single management platform with greater flexibility in handling programming changes and increased reliability by eliminating points of failures. Program sources are fed into the video switch where they are duplicated internally and processed as required by each network type. One set of program sources is converted to single program transport streams (SPTS) encapsulated over UDP/IP for IP distribution. The other set is turned into QAM modulated and up-converted multi-program transport streams (MPTS) for RF distribution. The SDV sub-system hands off respectively an IP feed as well as an RF feed for the IP and RF access networks. Channels destined for analog RF distribution are sent from the SDV system to a low cost multi-channel decoder after having gone through digital commercial insertion. This step into digital enables next generation advertising techniques and uses a single low cost transport network to distribute analog content to remote hubs or consortium members. A single management system makes it easy to provision, monitor and maintain services on both networks while maintaining independent service offerings on each network. For example, the IP delivery network could provide more broadcast channels while the RF network supports HDTV channels to test market demand for such service in the community. Digital advertising insertion is implemented in the SDV platform and delivers commercials across the entire subscriber community or to each targeted network. This architecture eliminates the need for splitting cables and provides the ability to easily launch,  rearrange and maintain services on the network.

Figure 5: Digital Centric implementation of a multi-access headend provides maximum management efficiency, flexibility and reliability

GigE network expansion with QoS and localized services

The advances in optical transport technology bring a new economic paradigm in digital transport that enables building low cost GigE networks. As operators expand their headend reach to remote hubs and neighboring operators, GigE networks provide a cost efficient and flexible way to transport services over new and existing networks. With GigE however, the real-time jitter sensitive nature of video necessitates the use of redundancy and quality of service (QoS) mechanisms to guarantee a sufficient level of video quality. An optimized and intelligent SDV platform enables Telco's to build and operate complex video networks that cope with these GigE network limitations. It also provides a highly reliable and scalable network that leverages the headend investment over large subscriber communities without limiting the ability to provide local services.

Modular Optical Transceivers enable transport of 1Gbps worth of video traffic over 80 miles for less than two thousand dollars. This new economic paradigm in digital transport is accomplished with the use of Small Form-factor Pluggable (SFP) modules that exist in various flavors from multi-mode to single-mode 1550nm, passive CWDM and DWDM options. These commodity devices are competitively offered by a variety of manufactures enabling operators to build complex networks with virtually unlimited bandwidth scalability at a very low price. The integration of these modules in a network intelligent SDV platform enables operators to affordably build elaborate ring architectures using Optical Add Drop Multiplexing (OADM) to reach subscribers in geographic areas that may have otherwise proven economically unfeasible to serve, therefore increasing subscriber revenue. It is also cost-effective to share the headend costs among a consortium of operators or geographically diverse locations to help accelerate profitability of the video business.

The recommended architecture utilizes a route diverse fiber ring to interconnect multiple sites using GigE Layer 2 or Layer 3 transport. The headend performs all the programming acquisition, media processing, ad splicing and transcoding into IP video over GigE or RF networks to local communities. The platform's multicast capability duplicates outputs for dual route diverse fiber feeds on either separate fibers or separate lambdas and enabling remote hubs or consortium member operators to receive the programming for local distribution.

The SDV platform at the receiving site is optimized to insure QoS and localization of services. It monitors the programming received from both fiber feeds and enables instant switching to the redundant route in case of a fiber cut.  It removes jitter introduced by the network, integrates the local broadcast and PEG channels, inserts local commercials that may be delivered from a centralized ad server, performs rate-shaping to adjust the program bit rates to meet the access network's bandwidth requirements, and delivers either IP streams or QAM modulated RF feeds or both to the local access network.

As the headend expands its subscriber reach, service reliability becomes increasingly essential to maintain customer satisfaction. A few minutes of service downtime can represent lost advertising revenue but more importantly cause very unsatisfied customers who all have easy access to alternate service providers. A second redundant master headend can be connected to the fiber ring and exchange program level redundancy with the primary headend, which helps to mitigate service interruptions. The SDV platform at the hub sites connected to the fiber ring are able to instantly switch to any program of the redundant headend should a service degradation or a service interruption occur in one or all of the programs in the primary headend. This service-level redundancy scheme protects not only from fiber cuts in the network but also from possible failures of headend equipment (e.g. receivers, satellite dishes, etc.) or video quality degradation that may occur due to damaged cables, environmental factors or other equipment problems.

Figure 6: Leveraging GigE networks to expand subscriber reach

Triple play operators should consider the use of recent advances in optical transport technologies to expand their headends' subscriber reach and accelerate their time to profitability through increased subscriber revenue or shared headend infrastructure cost. This can be achieved with adequate quality of service and without compromising localization of services for each community served by the network

Conclusion

This paper offers several headend and network design recommendations for Telco operators who plan their entry in the video business as part of their triple play plans. The video business has several incumbents including DBS and MSOs that enjoy years of experience and although Telco operators have certain inherent advantages, they need to build the infrastructure with the most scalability and reliability in mind to enable establishing a profitable and competitive operation as quickly as possible. Building a switched video architecture around an all-digital and IP centric core network with a modular platform that provides the flexibility to deploy advertising services, serves multiple types of access networks and offers economically efficient expansion capabilities over GigE fiber backbones is the best path to achieve these goals. Even if they start small, Telco operators who design their network with these capabilities in mind will yield maximum return on the headend and network investments as the operation grows over time.