Broadband Communities

OCT 2014

BROADBAND COMMUNITIES is the leading source of information on digital and broadband technologies for buildings and communities. Our editorial aims to accelerate the deployment of Fiber-To-The-Home and Fiber-To-The-Premises.

Issue link: http://bbcmag.epubxp.com/i/416055

Contents of this Issue

Navigation

Page 12 of 94

6 | BROADBAND COMMUNITIES | www.broadbandcommunities.com | OCTOBER 2014 BANDWIDTH HAWK P rodded by competition from Google and other gigabit fber-to-the-home builds, Comcast and other cable operators have moved up their planned deployments for DOCSIS 3.1, originally slated for 2016 – despite the fact that DOCSIS 3.1 electronics are unlikely to be produced in quantity until next year. In addition, they are testing the option that allows use of EPON from DOCSIS nodes into households, which makes possible connection speeds up to 10 Gbps downstream and 1 Gbps upstream, along with the point-to-point Ethernet network diagnostic tools that help ensure reliable, seamless service. DOCSIS 3.1 is technology-agnostic at the customer end, and some vendors are experimenting with GPON rather than EPON. Te overall technical efort has been pushed by U.S. cable companies and the other large cable deployer, China. As China favors EPON, that standard has moved faster, and high equipment volume should promote lower costs. Equipment is already becoming available, despite the fact that the draft standard is just on the verge of approval and the fnal standard is not due until next year. In a typical brownfeld MDU deployment, the use of DOCSIS 3.1 would translate to fber to the basement (if the building is big enough to have its own DOCSIS node) or fber to the curb for smaller multifamily dwellings or even single- family neighborhoods. Te fnal jump to users would usually be over existing coax, using the pending EPoC (EPON over Coax) standard that is part of IEEE 802.3. However, in greenfeld deployments and in existing buildings that have unsuitable coax but a good business case for the bandwidth, the signal path would be fber all the way – true FTTH from cable companies. Te DOCSIS node becomes a local point of presence for active Ethernet (and maybe, eventually, GPON). New carrier-quality wireless gateways can reduce the need for broadband providers to rewire inside customer premises and therefore signifcantly reduce the cost of connecting customers to gigabit services. Te gateway (essentially an ONT with a fancy Wi-Fi antenna) is all that most residential and small-business customers should need for on-premises connectivity. Comcast alone is working on at least a dozen brownfeld and greenfeld MDU and planned-unit development projects and seems set to announce them this fall. It has also confrmed ongoing tests of DOCSIS 3.1 to single-family homes in Chicago and Denver. Will there be enough capacity in downstream fber from the DOCSIS nodes to cable aggregation points? And what about reliability? Carrier Ethernet gains reliability by connecting to multiple access points and by using a ring architecture. If a metro Ethernet ring sufers a failure, the signal fow's direction can be reversed to bypass the failure point. However, core cable networks were originally designed for sending massive amounts of video in smoothly fowing streams toward all customers at once, not for redundancy. DOCSIS 3.1 attacks the reliability issue in multiple ways. First, even partial conversion to EPON brings better management tools and allows the use of existing and ever- cheaper components to wring more bandwidth out of a strand of fber – most notably via wave-division multiplexing and Ethernet burst mode. DOCSIS 3.1 also uses 4,096-bit QAM, which requires between 800 and 900 MHz (by bonding of about 125 channels) to deliver 1 Gbps. Tat would be in one direction, typically downstream toward users, which suggests that even more video channel space will be devoted to delivering raw bandwidth and that large cable companies may be accepting the idea of à la carte programming (over the bonded channels or the fewer available "video only" channels). Comcast has been talking quietly about converting entirely to IPTV over the next fve years or so. Tis change would make its network more reliable, reduce power consumption and allow longer distances from aggregation points to DOCSIS nodes. All-digital EPON or GPON signals can travel 12 to 60 miles; analog video signals more typically travel less than 10 miles and require more power on the network edge. IPTV also allows cable companies to abandon set-top boxes if they wish. v Contact the Bandwidth Hawk at steve@bbcmag.com. Cable Companies Join the Gigabit Party Deployment of true, reliable gigabit services by cable companies has begun even before issuance of a fnal standard for DOCSIS 3.1. By Steven S. Ross / Broadband Communities

Articles in this issue

Links on this page

Archives of this issue

view archives of Broadband Communities - OCT 2014