Cable Plant Tactics
Over the last 10 years, cable operators have spent considerable time and revenue upgrading their systems to 750 MHz or 870 MHz. At those elevated upper bandwidth limits, the plan was to collapse the analog tier and create more space for digital channels to provide cable operators with all the bandwidth they needed. Several variables have interceded, and today cable operators are faced with a bandwidth crunch. While digital video penetration has grown impressively, the revenue contribution from analog tier customers is still at a level that makes it difficult to simply turn off the analog spigot. And the rise in popularity for high-definition (HD) and on-demand programming, high-speed data and voice over Internet protocol (VoIP) services is placing new strains on already crowded bandwidth. Add the increased competition from satellite service providers and the looming launch of IPTV video service from telcos, and cable operators are faced with a future that needs careful, well-thought-out plans for bandwidth availability to enable the delivery of the types and quantities of services that consumers will expect. Optimization, expansion As cable operators completed upgrades to 750 MHz or 870 MHz for their systems, their focus turned toward looking at ways to use their newly available bandwidth to generate new income. As a result, they have loaded their systems with larger quantities of digital programming, expanded on-demand libraries, seen HD’s popularity grow significantly, and launched lucrative data and voice traffic that has started to crowd the forward and reverse paths. Bandwidth optimization solutions to meet this bandwidth crunch range from moving more channels to the digital tier and shifting from 64-QAM (quadrature amplitude modulation) to 256-QAM, to employing dual pass encoding/closed loop encoding and transrating, and deploying switched digital video technology. Additionally, a new phase of bandwidth expansion is an option. Because the costs for the move from 550 or 750 MHz to 870 MHz are virtually identical to a move up to 1 GHz, some operators are choosing to move to 1 GHz to create an optimal bandwidth environment for immediate and future needs. By creating this high-bandwidth, robust backbone, cable operators can prepare themselves to meet the growing challenges from other service providers. New plant In addition to addressing the day-to-day bandwidth challenges, cable operators are also looking at a playing field that is expanding all the time. Depending on how fast housing starts are growing in a service area, new plant miles needing service are growing between three and 12 percent per year, per system. As a result, operators are forced to decide how these new greenfield service areas “fit” with the spectrum already installed in adjoining areas. Is 750 MHz enough? Expand existing bandwidth to 870 MHz and deploy that in the greenfield? Pull fiber to serve the new areas? Jump to 1 GHz? The encouraging news is that whatever decision is made, it most likely results in an opportunity for a good return on investment (ROI) because greenfield development yields new customers. In contrast, rebuilds to boost bandwidth don’t offer as rapid an ROI, but they will present new revenue opportunities from the additional capacity that can support delivery of more services. Greenfield costs Looking at a greenfield build for 25,000 homes, cable engineers have identified and detailed the capital expenditure (capex) costs per homes passed for an HFC network. Based on the number of amplifiers in a cascade and the number of homes served by a node, there are definite economies at certain points and some obvious points where costs will begin to rise steeply. The analysis conducted was based on a 25,000-home greenfield deployment with 70 percent aerial and 30 percent underground construction. From 700 down to 200 homes passed per node, the capex per home is about the same. At 50 homes passed per node, it rises by about 40 percent and then escalates even higher as the number of homes passed decreases. (See Figure 1.) The critical takeaway here is that segmenting the network into smaller areas and reducing the number of amplifiers needed is about the same down to about 100-200 homes passed, providing the operator with increased operating flexibility without drastically impacting capex. One potential area of product improvement is the creation of optical nodes with higher output levels. Advances in gallium arsenide field effect transistor (GaAsFET) technology may make this possible in the foreseeable future. The result of these new products would be a leveling of HFC deployment costs from N+6 to N+0 designs where N+0 is marginally more expensive than a traditional N+6 architecture. Another upside to the smaller service areas and fewer network components is increased system reliability and service for the consumer, along with lower operating expenditure (opex) maintenance costs. A node plus six, four or one amp cascade can deliver the same video quality to the consumer, again making the flexibility available a positive option for the operator. Once the decision on node sizes is made for the greenfield, the operator’s attention needs to be directed at looking at existing network infrastructure and deciding how to upgrade the existing cable plant to have similar serving group sizes. The completed greenfield deployments provide valuable opportunities for cable operators to look at alternative technologies, do things a little bit differently, and decide how to combat some of the competitive threats that cable operators are starting to see right now. Developing a greenfield plan and putting it into action presents excellent opportunities to explore new avenues for delivering services. Because you are not dependent on past architectures in these new areas, they are ripe for deploying new technologies designed to support the new services customers want today and to establish a platform for future service and revenue enhancements. To better illustrate the greenfield options, here are two greenfield scenarios: HFC passive cable network. This architecture (see Figure 2) drives a traditional HFC network to an optical node-only design, eliminating all actives between the node and the customer. The optical node size is driven to a level where it supports approximately 100 homes passed. Although this does not create an HFC fiber to the home (FTTH) solution, it does significantly eliminate actives and improve plant reliability. Benefits of the passive HFC architecture include: Lowest-cost outside plant via shared single coax
Easily modified or enlarged service area w/ RF amplifiers
Least amount of field electronics of any network
No additional provisioning for subscriber activation Challenges associated with the passive HFC architecture include: Powering and maintenance for field electronics
No simple plug-and-play operations for amplifier cascades greater than Node + 1 HFC mini-node. This deployment, where the nodes get even smaller, looks similar to the previous configuration except that the products and architecture change as you get to the remote terminal and beyond. This is fundamentally an HFC fiber to the curb (FTTC) solution, with each node existing at an eight-homes-passed level in the system architecture. With the deployment of DOCSIS 3.0 and a “Burst-Mode transceiver,” this becomes very similar to a HFC passive optical network (PON) FTTC solution, with minimal coax existing between the node and the customer premises. It would be possible to deploy this technology deeper by reducing the gain in the FTTC node and pulling fiber to each home passed. Of course, it also increases the cost per home passed vs. other solutions. (See Figure 3.) The benefits of the mini RF mode architecture include: • Equivalent optical input power, plug-and-play
• Reduced leakage detection requirements from low RF
• No additional provisioning for subscriber activation
• Minimized office home-run fibers, minimized restoration time Challenges associated with the mini RF mode architecture include: • Powering and maintenance for field electronics
• Four to 10 times the number of field nodes as compared to HFC-PCN
• Partial drop installation on street frontage, capex impact Limit downtime One standardized approach in the entire greenfield and existing plant upgrade scenario is to look at it from the standpoint of keeping the “destruction and disruption” to a minimum because of the volume of very sensitive services already deployed. As a result, the goal is to keep downtime to a minimum. As the greenfield or upgrade plan is implemented, there inevitably comes a point where costs will go up significantly because of infringement into areas of the cable plant’s backbone. Cable operators need to have a plan that balances construction costs against node size, and careful planning will help you avoid driving deployment to a level that’s so low that it doesn’t make operational or financial sense. For example, node plus one in an upgrade is 30 homes passed, whereas in the greenfield it may be 150 or 200 homes passed. 750 MHz to 1 GHz In recent initiatives with operators taking a 750 MHz system to 1 GHz, there have been three main areas of focus based on the technology available. First, deploying an electronics-only upgrade can minimize plant disruption and customer downtime. Here, cable technology engineers develop a plan to help maintain spacing between actives. Second, maintaining all tap values helps ensure that the customer experience doesn’t change. Finally, ensuring that existing 750 MHz levels are maintained means that current plant capabilities or services are not compromised, but rather enhanced by adding more spectrum. The goal is to extend the tilt out past 750/870 MHz and make sure that when operators extend to 1 GHz bandwidth, they’re not going into compression or issues with amplifier performance. If this is accomplished, when the tilt is extended out, plant performance should be acceptable out to 1 GHz. This approach has worked well in 1 GHz deployments that operators have completed already. This performance can be delivered by using standard HFC design maps, specifications and performance requirements, drop-in and maintenance of spacing, and the creation of a plant that functions to 1 GHz with minimal plant disruption and construction costs. Because of extensive construction costs, upgrades to 750 and 870 MHz in the past cost as much as $17,000-$18,000 per mile. In comparison, electronic upgrade costs somewhere in the neighborhood of $4,000-$5,000 per mile because of the affordability of the new components and the much lower labor costs they require for installation. As a result, by moving to 1 GHz, the operator gets more bandwidth than moving to 870 MHz at approximately the same cost and at a significantly reduced cost target than the original HFC upgrades. With all the variables in play today, operators need to identify and set a course that is both cost and bandwidth efficient. The right moves will expand bandwidth for existing plant and greenfields and maintain a high level of customer satisfaction at a minimal cost. The endgame is a cable plant that works hard and meets the competition. Bob Tynan is director, Business Strategy, Transmission Networks Systems, and Bob Loveless is director, Advanced Planning and Technology, Transmission Network Systems, both for Scientific Atlanta, a Cisco Company. Reach them at [email protected] and [email protected].
