– Cablecos Winning the Broadband War with 83% of New US Subs in Q1
– Next Step for Telcos’ Copper Wires Is G.fast
Here’s a statistic that should frighten the world’s telcos: US cablecos took 83% of the new broadband subscribers in the first quarter of 2014 and telcos took only17%. Even though AT&T and Verizon added 732,000 broadband subscribers via U-verse and FiOS, the two together lost 638,000 DSL broadband subscribers, no doubt many to themselves but also many to the cablecos. The situation is not as bad in Europe and Asia but only because cablecos’ footprints are not nearly as widespread. Even there, telcos are frightened by the growth in the cablecos’ broadband subscribers.
Vectoring is the telcos’ next best copper wire hope because it can provide speeds in the 80-100 Mbps range, more than enough for most homes.
Telcos’ xDSL technologies, including vectoring, have one major advantage over the cablecos’ DOCSIS: It is not shared with other cableco subscribers in the neighborhood. If an xDSL subscriber has 80 Mbps, he has 80 Mbps all the time. When his xDSL-using neighbors start streaming videos, his speed does not decrease; it’s always 80 Mbps or 40 Mbps or whatever he has been promised. The net speed of the cablecos’ broadband depends on what the other cableco subscribers on the same neighborhood node are doing. A cableco subscriber who is working at home might have the promised speeds during the day but they drop dramatically when the neighbors get home and start online gaming and streaming videos from Netflix. For some reason, telcos have not communicated that advantage to consumers.
Alcatel-Lucent, a leading supplier of both copper wire and all-fiber gear to the telcos, says that sales of both are picking up. However, all-fiber networks cost more and take longer to build so it is expected that in the near term, sales of vectoring gear will continue to increase. Its experience is that first a telco buys quantities of vectoring-capable line cards. Then, after upgrading to vectoring its copper wire network to the home, telcos order the system level processing gear that’s needed to turn on vectoring so they can begin marketing the increased vectoring speeds. Stefaan Vanhastel, marketing director for AlcaLu’s wireline division, told The Online Reporter that the company is seeing that trend being followed by telcos.
AlcaLu said last week that it has now shipped more than 5 million pieces of vectoring gear and is now shipping more vectoring than non-vectoring VDSL2 gear.
Asked about the increased interest in all-fiber networks in the States, Vanhastel said Google may have lit a spark with its fiber rollouts and announcements that it’s looking at more.
There are three major differences between vectoring and all-fiber, he said:
– Cost: With vectoring, the fiber only needs to be deployed to cabinets in the neighborhoods but all-fiber requires it to be deployed all the way into the residence.
– Time-to-market: Cablecos, government regula-tors and consumers are pressuring telcos for faster speeds. All-fiber takes much longer to deploy.
– Installers inside the residence: Deployment of all-fiber networks requires the telco to send a trained installer inside the home to complete the installation. That often requires installing new wires, something that most residents prefer to avoid. Vectoring does not require any work within the residence. Once vectoring has been deployed, the telco need only ship a vectoring-capable modem for the resident to install (the beloved self-install) or remotely upgrade the subscriber’s existing modem if it can be upgraded to vectoring.
It would take telcos years, he said, to build nationwide all-fiber networks, which are capable of up to 1 Gbps. VDSL2 vectoring allows telcos to quickly upgrade existing DSL subscribers to 100 Mbps, earn new revenues from these services, and then use these to fund the longer-term FTTH infrastructure.
A broadband technology called G.fast that’s being developed may shorten the time to 1 Gbps and as a bonus it eliminates the need to send an installer into the home. To deploy it, telcos will have to install fiber closer to the residence — within 100 meters or so. That would especially ease the installation process at Europe and Asia’s many MDUs, especially the older ones, because the telco could use the building’s existing copper wires and still provide fiber-like speeds. At standalone residences, the telco would run fiber near to the home or even to the side of the home. There, the installer would connect fiber to the residences’ copper wires.
Vanhastel said the development schedule for G.fast is proceeding and a final, approved specification (“the standard”) is expected by the end of 2014. Because chip and equipment makers are involved in the development process, they can speed their time to market. Vanhastel said G.fast chips will appear in 2015 and equipment by the end of 2015. That means a few deployments could begin in 2016.
He cautioned that phase 1 of G.fast only provides speeds of up to between 500 Mbps to 700 Mbps. Stage 2, which will come later, will provide 1 Gbps. G.fast’s stated speeds are its aggregate speeds, including both up and down.
There are 3 factors that can impact real-world G.fast performance:
– It might not be able to use the full frequency. Telcos might need to skip the VDSL2 spectrum to avoid interference, or not be allowed to use spectrum reserved for FM radio or military use.
– The quality of the telephone wiring being used may impact performance.
– Some wiring practices (like bridge taps) also have a negative effect on performance
Asked whether G.fast will be as successful as vectoring, Vanhastel said that the business case for vectoring is more obvious — and its increased speeds are needed now. G.fast is more expensive since you have to get really close with fiber, and you can only serve a handful of subs from each node, so at first sight, the business case for G.fast looks difficult – the cost is very close to that of FTTH. But the real attraction is that it allows the telco to stay out of the home, and that can save time and money.
Vanhastel said the real interest for G.fast is that “it’s almost fiber to the home.”
Asked why consumers need more than about 40-50 Mbps, Vanhastel said…
For the complete article and latest edition, please write email@example.com or click here to register for a four week free trial