Full-Service Networking
A major issue is whether the Access Network provider will offer full-service networking. Full-service networks (FSN) are Access Networks that are intended to provide everything, including broadcast TV, Internet access, video on demand (VoD), and even voice telephony. FTTx and cable are considered to be the likely alternatives for FSNs. However, xDSL networks could also participate. An important alternative carrier offering full service is RCN (Nasdaq: RCNC, www.rcn.com ). RCN is building on its relationships with Boston Electric Power and Potomac Electric Power, two large utility companies, to offer full service over fiber built along utility rights of way.
Proponents argue that FSNs offer the lowest-cost alternative for subscribers to receive all services—certainly lower than subscribing independently to multiple Access Networks. They also argue that it would be easy for the subscriber to use, because she would have a single interface to a variety of services. Finally, for the carrier, offering an FSN would tend to reduce customer churn because a consumer would be more reluctant to part with a single service if that service came in a bundle. The control of customer churn is an important marketing requirement of any monthly service.
Case Study: Time Warner Pegasus
A leading example of a system offering video services and interactivity is the Time Warner Pegasus system. This is not strictly an FSN, because it does attempt to offer telephone service. Furthermore, high-speed data services are offered by another Time Warner company called Roadrunner, which shares a common ATM backbone but uses the Access Network differently. Nonetheless, Pegasus is an interesting case because of its plans to carry broadcast video, video on demand, and IP connectivity on a single network. A closer look at this particular network reveals some of the general questions and problems facing network implementers who are contemplating this most complete and integrated option of RBB methods.
After some experience with a preliminary FSN trial in Orlando, Florida, Time Warner will implement a new end-to-end, full-service network called Pegasus (see www.timewarner.com/rfp ). Whereas the original Orlando trial used ATM end to end, Pegasus uses ATM in the Core Network, and MPEG and IP in the distribution network. At the set top, video is separated from data using frequency-division multiplexing or MPEG demultiplexing for in-band data.
The Orlando FSN proved the technical capability of HFC as a full-service network. Its costs were high because it was an innovative experiment. Pegasus is billed as a deployable form of the Orlando FSN.
The Pegasus system is mainly designed to provide services for analog TV, digital TV, VoD, and two-way data service for control purposes, but it also provides services for minor data applications through a single set-top box.
These services are provided through three types of channels. The Forward Application Transport (FAT) channel is a 6 MHz channel encoded using QAM 64 or QAM 256. VoD and broadcast TV are distributed on the FAT channels. The Forward Data Channel (FDC) is a 1 MHz channel encoded using QPSK to yield a T1 data channel in the forward direction. The Reverse Data Channel (RDC) is a QPSK-modulated 1 MHz T1 data channel in the reverse direction and is compatible with the DAVIC signaling specification. The FDC/RDC provides the following:
Two-way real-time signaling for VoD
Delivery of set-top authorization messages
Network management of the set-top box
IP connectivity for set-top applications, such as Web browsing (in conjunction with the FAT channel for high-bandwidth services)
Figure 8-4 illustrates the major components of the Pegasus network. Analog broadcast is received at the head end and is retransmitted on a FAT channel. Digital broadcast is received at the head end in MPEG packets with QPSK modulation. Satellite delivery generally is encoded in QPSK. Cable delivery is encoded using QAM 64. In addition, the encryption techniques used by satellite differ from those used by cable. The Broadcast Cable Gateway (BCG) remodulates and re-encrypts the satellite digital video into a form standardized on cable networks.
Figure 8-4. Pegasus Full-Service Network
The Interactive Cable Gateway (ICG) is used in the early phase of Pegasus to enable the integration of video with data into the FAT channels. This is required, for example, to embed Web content into a program or to enable electronic commerce if it needs video. Video and data are received from the Core Network over ATM at links of 155 Mbps. ICG is required to match the speeds of 155 Mbps to the FAT channel, which runs at 27 Mbps. Newer versions of Pegasus will require that media servers perform the integration of video and data onto a FAT channel by using specifications standardized by the DVB.
A Data Channel Gateway (DCG) provides a two-way real-time data communication path between the Pegasus set top and both the application servers (which perform such tasks as program selection) and the Digital Network Control System (DNCS), which performs set-top authorization, set-top management, and purchase information collection. The DCG is essentially an IP router with QPSK interfaces.
A schematic of the Pegasus set top is shown in Figure 8-5. Beginning on the left side of the schematic, cable input is received from the drop cable through the QPSK receiver (FDC channel) and the RF tuner (FAT channel). An integrated MPEG and IP transport processor performs PID selection and forwards MPEG packets to the MPEG decoder and graphics processor; IP packets are forwarded to the CPU bus.
Figure 8-5. Pegasus Set Top Courtesy of Time Warner Cable
For MPEG processing, a digital TV channel is demodulated in the QAM demodulator and is forwarded to the MPEG and IP transport processor. Audio PIDs are directed to the AC-3 decoder and then to an audio out socket. Video PIDs are decoded in the MPEG decoder and the graphics processor, then are fed to an NTSC encoder, and finally are sent to video out (either S-Video, baseband, or RF).
Analog video is received on a FAT channel and forwarded to the Video ADC, the analog-to-digital converter. The digital feed goes to the MPEG decoder, where it is processed as digital video.
IP packets are forwarded to the CPU bus and then to a set-top CPU. Outputs are available for Ethernet and Universal Serial Bus (USB). USB is a new peripheral attachment standard that provides for 12 Mbps. A USB socket on Pegasus will enable standard peripherals, such as keyboards, joysticks, and printers, to be connected to the Pegasus set top. (Further information on USB is available at www.usb.org .) Infrared input (IR input) is received by the set-top box for television controls. Controls are sent in IP packets upstream through the QPSK transmitter (RDC channel). Finally, an Ethernet connector is used to connect Pegasus to a PC or a game console.
Note that in this architecture, no requirement exists for routing a Residential Gateway (RG). A network with this amount of bandwidth might not require downstream demultiplexing because each piece of terminal equipment (TE) can be granted a forward data stream from the carrier premises. Consider the case of 500 residences sharing 2.7 GB, each residence with 2.7 active users watching TV or browsing the Web. In this case, each active user in the neighborhood can be granted exclusive use of 2.0 Mb with the downstream demultiplexing performed at the head end rather than in the residence. This moves intelligence out of the home and into the carrier premises, thereby reducing total system cost.
With the introduction of digital broadcast, more channels will originate in digital form, and the amount of analog bandwidth will drop over time. In this environment, cable networks can achieve their FSN promise.
Benefits of Full-Service Networking
FSNs tend to reduce complexity to the consumer (because there is only one network to deal with) and reduce overall costs to consumers who want a wide variety of services. A single NIU, RG, and STU is amortized over a variety of services. There is also a single billing system and customer service infrastructure. All-in-one service motivates long-distance carriers to offer combined television, video on demand, voice, and Internet services. Instead of dealing with a number of providers, the consumer pays one bill and possibly achieves certain discounts accordingly.
For the carrier, FSNs tend to create better account control and reduce customer churn. A customer cannot simply drop one service; he must drop a bundle of services if he wants to change carriers. FSNs also provide a better marketing information database. By offering more services to the subscriber, more is known about the consumer, so it is possible to target the market more aggressively.
Challenges of Full-Service Networking
For many consumer products and services, there is always competition between a full-service model and an appliance model. For example, many consumers prefer stereo sound equipment in components rather than in a single, integrated boom box. The appliance model permits piecemeal optimization of each component as technology advances. This is also a lower-cost entry for consumers who want to try only one or two services. It has been argued that in the consumer space, very few products have been integrated successfully, except for the integrated VCR/TV and perhaps the boom box. Full-service networking faces similar technical and commercial challenges.
FSN Versus Cell Phones
A large part of the value proposition of FSN is voice service. For cable, CLEC, and wireless operators, it is better to have a small percentage of the voice business than a large percentage of the Internet and TV businesses, both of which are dwarfed by the telephone business. Cable, for example, has revenues on the order of $40 billion, whereas telephone companies have revenues in excess of $200 billion. However, widespread voice on an FSN must overcome the success of cellular phones. Cell phones may very well limit the appeal of voice on FSNs, especially among higher-income households.
Place Your Bets Up Front
Many carriers fear that moving to a full-service network is an all-or-nothing proposition—place your billion-dollar bets up front. It will take two to three years to see what competition emerges, and five to ten years to get an accurate picture of financial results.
It would be financially less risky if the carrier could add services and bandwidth capacity incrementally. But FSNs must develop capabilities for multiple services simultaneously rather than developing a single capability, such as voice only or data only. In addition, those combined services depend on having all the required bandwidth at the outset. For both services and bandwidth, a significant up-front investment is required.
Incomplete information on customer preferences, pricing sensitivity, competitive landscape, scaling properties of the technology, and the regulatory environment make FSN seem like a long-odds gamble to many companies. As an indicator of what other potential carriers can expect, the Time Warner FSN in Orlando was not an entirely reassuring experiment. Only familiar services such as VoD proved popular. Unfamiliar services such as interactive shopping and news-on-demand did not fare well. It is unclear whether the market drivers mentioned in Chapter 1, "Market Drivers," will in fact be popular enough in the early years of RBB rollout to keep the carriers' interest.
Biting Off Too Much
Telephony, video, and data delivery systems individually are large systems that have been subject to optimization for years by lots of talented people. When optimizing large systems, it is not necessarily true that one obtains a globally optimal result by optimizing everything at once. Sometimes better results are achieved by segmenting the problem, optimizing each subproblem independently, and then adding the results together. This is especially true when each subsystem is substantially different. Instead of trying to optimize voice, video, and data simultaneously in a single system, it might be more cost-effective to optimize each system independently and simply provide three networks.
Billing Complexity
Finally, this chapter ends on a light note. The diversity of user services available on a single full-service network invoice can lead to complicated billing problems. A customer could end up with an invoice such as the one shown in Table 8-2.
There is a degree of user-friendliness as well as an unbundled costing that will take some thought to present to the consumer.