Extending the Life of Copper Based Broadband Whitepaper
In a perfect world, all broadband connections to the customer would be fiber-to-the-premise (FTTP). The benefits of FTTP are tremendous including the impressive bandwidth capabilities, the lower maintenance costs, and it’s potentially “future proof ” nature. But we don’t live in a perfect world and copper based connections are still the dominant method for broadband access across the globe. Deploying FTTP is still an expensive proposition and can’t be justified everywhere, at least not yet. In the U.S., some estimates put copper facilities at 80% of all route miles, even with the impressive FTTP momentum of the past few years.
Copper based facilities will not disappear anytime soon and broadband service providers who utilize them need to evaluate how to best maximize their capabilities. The demand for bandwidth will continue to accelerate and copper based broadband does have its limitations. There is a considerable amount of research and development taking place to improve the capabilities of copper based broadband. There are literally tens of millions of DSL connections across the globe, representing billions of dollars of network investment. Broadband service providers want to find ways to exploit that stranded investment and extend the life of copper for as long as possible.
WHY BOTHER WITH COPPER PLANT?
The simple answer is because there is a lot of it. There has been significant momentum with FTTP construction in the U.S. and abroad. But as Figure 1 below illustrates, copper based broadband still dominates and copper based broadband providers need to continue to leverage their investment for years to come.
Beyond the obvious, copper based plant is still quite functional and in good condition as well.
Most telephone companies, particularly smaller independent carriers, have done an excellent job of maintaining their copper plant. It’s not FTTP capable in terms of bandwidth throughput, but can be adequate, and even more than adequate in some cases. Considering the average broadband speed in the U.S. is 8.7 Mbps, copper facilities are quite capable. There are techniques that allow DSL to deliver tens of megabits per second to the home and even 100 Mbps or more on very short copper loop lengths.
Copper based plant is also relatively simple to upgrade. Broadband carriers have already invested in capacity through broadband loop carriers and fiber fed DSLAMs that push fiber deeper into the network. Leveraging that ongoing investment for additional benefit does not require tremendous effort and cost, at least in comparison to replacing it with FTTP. Broadband access manufacturers continue to innovate their access equipment as well, with smaller scale broadband digital loop carriers as an example. Replacing copper facilities with FTTP is simply not feasible everywhere, at least in the short to mid-term timeframe.
HOW MUCH BANDWIDTH IS ENOUGH?
There is significant discussion of late regarding ultra- fast broadband, with several high profile gigabit (or 1,000 Mbps) FTTP deployments in the news. Google, AT&T, CenturyLink, and many smaller regional and municipal broadband carriers now tout gigabit FTTP projects. But the aforementioned average broadband speed in the U.S. is roughly 8.7 Mbps. The FCC’s current broadband plan calls for rural markets to receive at least 4 Mbps downstream (to qualify for Connect America Fund funding), and calls for 50 Mbps downstream to be within reach of 100 million homes by 2015. By 2020, the FCC’s plan calls for 100 Mbps to 100 million homes.
There is no denying that consumers want and use more bandwidth and will certainly continue to do so. In most markets today, a broadband connection of 30 – 50 Mbps is considered very fast and can adequately meet the demand of most subscribers. But faster speeds are being introduced and there is growing demand for it. Verizon reports that 27% of their FiOS Internet customers subscribe to FiOS Quantum, which offers speeds of 50 Mbps to 300 Mbps.
Higher bandwidth demand is driven primarily by growing consumption of video over the network. A growing number of households now possess multiple HDTVs. Recent research from the Leichtman Research Group reveals that 75% of U.S. households now have at least one HDTV, with 51% having more than one . The growing penetration of tablets and smartphones in the home also extends video consumption beyond the TV. In today’s multiscreen world, multiple streams, often multiple HD streams, can now be in demand and delivered via the Internet.
This impact of HD video on bandwidth is only going to multiply. The next generation of HDTV, or 4K TV, is now beginning to ship. This latest version of HDTV has a resolution of 2160 x 3840 pixels, compared with today’s HD which generally offers 1920 × 1080 pixels, or roughly 4x the resolution and 4x the amount of bandwidth necessary to transmit it. Netflix is now testing 4K video streams and intends to offer the service for OTT streaming . Beyond 4K, there is 8K HDTV on the long term horizon, which doubles the pixels and bandwidth requirements for 4K HDTV.
So while the current average broadband speed is 8.7 Mbps, there is ample evidence to suggest
that it will increase, and significantly so. Akamai tracks global broadband trends, and according to some of their recent research, the number of U.S. households receiving 10 Mbps or more downstream broadband (currently 19%) jumped 90% between fourth quarter 2011 and fourth quarter 2012 . Broadband service providers who will continue to utilize copper networks will need to keep pace and ensure their broadband offer continues to meet the demand of their customers with a quality broadband experience.
ADVANCED DSL TECHNOLOGY
Fortunately for broadband service providers who rely on DSL, there is significant ongoing research and development in improving DSL performance. Unfortunately for rural carriers with lower density market characteristics, much of this research focuses on short copper loop lengths. There are a number of technologies worth noting.
Perhaps the most widely used DSL performance enhancement today is pair bonding. With pair bonding, a service provider “bonds” two or more copper pairs from between the closest electronics in the network and the customer premise to increase broadband performance. While the increases in speed are not exactly proportional to the number of pairs that are bonded (two pairs that individually could produce 5 Mbps won’t give you 10 Mbps when bonded), the increase is still significant. For example, two bonded pairs on good quality copper and under good other environmental conditions could increase a single 10 Mbps service to 18 Mbps at 8,000 feet. For VDSL2+ bonding could potentially increase a 60 Mbps service at 3,500ft to 80 Mbps.
Many pair bonding solutions exist and they require two DSLAM ports to be utilized, as well as a bonding supported DSL modem at the customer premise. Other applications for bonding exist and some carriers have bonded as many as 8 pairs to achieve 75 Mbps over 10,000 feet to feed cabinets, small node sites, and even for wireless backhaul.
The underlying technology used for DSL vectoring is relatively old, but is being applied in new ways for DSL. In simple terms, DSL Vectoring can detect interference in frequency ranges caused by noise and/or cross talk that is present in binder/feeder cables, which are common in the network. With vectoring, these affected frequencies can be isolated and effectively “removed,” concentrating power and throughput to the remaining frequencies and improving performance and throughput.
The longer the loop length, the less effective vectoring becomes. DSL modems at the customer premise need to be compatible with vectoring to take advantage of the advances. VDSL2 vectoring solutions can improve DSL performance by 50% or more, depending on loop length and line conditions.
xDSL repeaters have been active for ADSL2+ for some years and are now becoming more active for VDSL2. Repeaters are often used instead of pair bonding, where “spare” pairs may not be available. They offer similar performance bumps to pair bonding. Repeaters are typically powered from the central office and some can work on 48 volt DC power. They are most effective when placed as close to the middle of the span as possible. Typical costs for repeaters are in the $250 range, which is more expensive than pair bonding.
Phantom technology is also an older technology, originally used in analog systems, that now has applications for DSL. Phantom technology works when the voltage between two bonded pairs is measured correctly to determine the differential between the two. That identified differential can be exploited for a third “phantom” pair. Data can be transmitted over the phantom pair, creating three pairs in effect.
Phantom pairs are best suited for shorter loops, perhaps 1,500 feet and shorter. In theory, Phantom technology could increase the bandwidth throughput of a 130 Mbps VDSL2 with vectoring connection to close to 300 Mbps. The original 130 Mbps, plus 100 Mbps from the second bonded pair, and an additional 70 Mbps from the phantom pair. Phantom technology also leverages the aforementioned vectoring approach to potentially increase the overall performance by an additional 50%. Using Phantom technology for this approach is relatively new with trials now underway.
G.fast is the newest copper based broadband technology improvement and it aims to dramatically improve broadband performance on very short loops of 500 feet or less. G.fast aims to deliver 500 Mbps to 1 Gbps over these copper loops and may be positioned as an alternative to FTTP. For urban FTTP applications, the drop to the customer premise may well be the most expensive component of a FTTP build. G.fast hopes to eliminate the need for fiber all the way to the premise, by using copper drops instead for similar FTTP performance.
G.fast dramatically expands the frequency ranges over today’s DSL that can be used across a copper medium. G.fast is initially looking to expand the the available frequency range over 100 MHz with future plans to go above 200 MHz, whereas ADSL2+ maxes out at 2.2 MHz or 30 MHz for VDSL2. G.fast is currently under development and is on the fast track at the ITU for standards definition. Trials are expected to begin in 2014 with deployments perhaps as early as 2015/2016.
The above technologies can help improve DSL performance. Here are some practical best practices for doing the same:
• Make sure that vectored pairs remain in the same binder group throughout the entire path
• Stay above the manufacturer’s minimum signal-to- noise-ratio (SNR), especially with video
o ITU recommends minimum of 6 dB which may not be enough for video
o Minimize retrains to reduce the chance of tiling and freeze frames in video
o Reduces intermittent problems with video
• Additional xDSL lines in a cable/binder will impact all circuits so take an active bandwidth management plan when adding more xDSL pairs to a cable
• Where possible, reduce upstream rates, especially if offering video
• Look to place video on separate pairs and modems
Copper based broadband is and will remain for the foreseeable future, the dominant broadband access technology across the globe. Broadband service providers who rely on copper loops for broadband access have options to improve broadband performance and extend its life. Choices between DSL technologies, fiber technologies, or a hybrid of the two should be rooted in a network planning exercise, first and foremost.
Indeed, service providers should not begin with deciding which technology architecture to deploy, rather they should determine the services and applications that are necessary, the desired coverage area, and the implementation timelines and milestones. Once determined, the proper technology can then be selected, based on these criteria. Without these planning steps, service providers risk deploying inadequate or excessive network capacity, both of which can be very expensive mistakes.
By setting the proper goals and conducting the proper planning, service providers can deploy the right network, which will involve both fiber and copper components, and will be prepared to meet the evolving demands of customers. Finley Engineering can provide support, guidance, and counsel to clients for this planning process. We’re prepared to help clients realize their network and company goals.
1.Telecompetitor – “Akamai: U.S. Broadband Connectivity Plateau?,” http://www.telecompetitor.com/akamai-u-s- broadband-connectivity-plateau/
2. Fierce Telecom – “Verizon FiOS additions drove up consumer revenue 4.3% to $3.6 billion,” http://www. fiercetelecom.com/story/verizon-fios-additions-drove- consumer-revenue-43-36-billion/
3. LRG Research Notes, 1Q 2013, http://www.leichtmanresearch.com/research/notes04_2013.pdf
4.PC Mag – “Netflix Testing ‘Ultra HD’ 4K Video Streams,” – http://www.pcmag.com/article2/0,2817,2426728,00.asp
5. The State of the Internet 4Q12, http://www.akamai. com/stateoftheinternet/?WT.ac=soti_banner
6. Finley Engineering – “Latest Copper Broadband Standard for 1 Gbps Broadband, G.fast on Fast Track at ITU,” http://www.fecinc.com/resources/blog/entry/latest- copper-broadband-standard-for-1-gbps-broadband-g- fast-on-fast-track-at-itu