FTTxtra » HFC

RFoG Overview

John Bartell — Tue, 27 Oct 2009 13:38:52 +0000

RF over Glass (RFoG) is a simple Hybrid Fiber/Coax (HFC) technology that replaces the coax access network with a passive optical network (PON). It is one form of what is known as DOCSIS PON or DPON. With RFoG, the HFC head end access fiber multiplexer is replaced by a WDM device that divides the upstream and downstream wavelengths and delivers them on separate fibers to other devices in the headend. Fiber splitters, instead of HFC nodes, are deployed to serve neighborhoods. A fiber NID is installed at each house.

The image below (click to enlarge) provides an architectural overview of a traditional HFC network.

The RFoG architecture in the outside plant (OSP) is identical to that of a PON, and in fact RFoG can be deployed in conjunction with a PON. The image below provides an overview of a combined GPON and RFoG implementation. RFoG’s biggest advantage is its ability to leverage legacy systems and electronics in delivering Cable TV service. Set Top Boxes (STBs) within the home remain unchanged in a shift from traditional HFC to RFoG. The RFoG ONU on the side of the house generates/receives RF coax signals identical to those used with HFC. The operational systems that tend the headend and the CPE with an HFC access network work just fine when RFoG is used instead.

With RFoG, downstream traffic is typically transmitted at 1550nm, and upstream traffic (with each STB transmitting in turn) is typically carried at 1310nm, 1590nm, or 1610nm. If 1310nm is used for the upstream, no upgrade path to GPON or most other PON technologies is possible since they use 1310nm for upstream transmission.

Several vendors provides RFoG products today. Standards work is ongoing for RFoG in the Society of Cable Telecommunications Engineers (SCTE) subcommittee IP SP 91.

A Cornucopia of Broadband

John Bartell — Thu, 15 Oct 2009 14:03:05 +0000

Carriers have many choices for delivering broadband servies to residential and business subscribers they choose to serve. The majority of broadband services are delivered with ADSL2+ or VDSL2 over twisted pair copper or DOCSIS cable modems over hybrid fiber coax (HFC) networks. However, wireless carriers will compete with 4G technologies to deliver tens or hundreds of megabits per second to fixed subscribers. Note that all types of carriers are attempting to deliver the same thing: high bandwidth IP packets with low latency.

A compelling factor in choosing a broadband technology is what local loop facilities are already installed. The majority of the cost for a new local loop is just installing the medium to reach the customer, and much of this effort in no way follows Moore’s law.

Four common mediums for delivering broadband services are twisted pair copper, hybrid fiber coax (HFC), fiber optics, and wireless. These are discussed below.

Twisted Pair Copper

The telcos have a tremendous amount of already installed copper local loops, and they will do everything possible to use this copper for delivering broadband. Avoiding installing new local loop (for instance fiber) to a subscriber can eliminate $1000 or more of capital, so telcos are keen to use their copper for DSL. However, competition is sometimes forcing telcos to overbuild fiber just to effectively compete (Verizon FiOS is a good example). The diagram below shows the architecture of a typical DSL access network serving residential subscribers in single family and multi-dwelling units.

The major players in DSL these days are ADSL2+ and VDSL2. VDSL2 is the relative newcomer, and its chipsets typically support a fallback mode to ADSL2+ because of VDSL2’s much shorter reach than ADSL2+. Together, these two technologies just about wring out all the broadband that a copper local loop can deliver. Further improvements will have to rely on multi-pair techniques like Dynamic Spectrum Management (DSM).

Both ADSL2+ and VDSL2 are technologies used for Fiber to the Node (FTTN) and Fiber to the Curb (FTTC). FTTN and FTTC are hybrid technologies combining fiber and DSL for delivering broadband rates in the tens of megabits per second. Their attraction is avoiding that final fiber link to a subscriber’s premises.

VDSL2

Though rates typically delivered are much lower, VDSL2 can deliver broadband at rates of up to 100 Mbps over local loop copper. Its greatest limitation is reach. It really does not provide any advantage in bandwidth over ADSL2+ beyond about 3500 feet, and it effectively does not function beyond about 6000 feet. Many VDSL2 chipsets support a fallback mode to ADSL2+ to extend their reach.

ADSL2+

ADSL2+ is an evolution of ADSL, which has been a broadband workhorse for the telcos for over a decade. ADSL2+ reaches up to 18000 feet with useable bandwidth, and its top rate is about 24 Mbps over short distances. The advantage ADSL2+ has over VDSL2 is cost, though the difference is very small when considering the entire cost of delivering broadband to a subscriber. Another advantage is that, since its complexity is lower, ADSL2+ heat dissipation is substantially lower than VDSL2’s. This lower heat dissipation is a significant advantage in OutSide Plant (OSP) cabinets, which tend to be heat constrained.

Fiber Optics

Essentially, there are three fundamental architectures for delivering fiber directly to the home: point-to-point, switched, and Passive Optical Network (PON). All three Fiber to the Home (FTTH) architectures require an aggregation device in the CO (the Optical Line Terminal or OLT), and all three require an optical to electrical converter (Optical Network Terminal or ONT) in or on the home. These three architectures differ mainly in what type of device (if any) is installed between the CO and the home.

Active or Point-to-Point Fiber

Active or Point-to-Point (P2P) Fiber is the simplest of all three fiber broadband architectures. With an Active Fiber network, a fiber (typically only a single fiber) runs directly from a subscriber’s house into the Central Office (CO) serving that subscriber. This architecture is simple, but it does require dealing with a large number of fiber optic cables in the CO. The CO contains a high port count aggregation device (the Optical Line Terminal or OLT) which has one port per subscriber.

Each subscriber house has an Optical Network Terminal (ONT) installed. It is placed either on the side of the house, which is typical in the US, or inside the house, which is not typical in the US, but more common in Europe and Asia. Active Fiber has an advantage in that all access ports are dedicated and unshared, and troubleshooting access network problems is simple.

With Active Fiber, problems can be easily identified. And, this simple architecture can deliver the most bandwidth of any fiber access technology. Each fiber added increases the aggregate bandwidth of the access network.

Active Ethernet

Tremendously decreasing the number of fibers that must be terminated in the CO, an Active Ethernet architecture has many of the advantages of active fiber in terms of scaleability and bandwidth. However, Active Ethernet requires active electronics to be installed in the OutSide Plant (OSP), which adds capital and operational costs. This equipment is necessary to aggregate fibers delivered directly to subscribers.

PON

Similar in architecture to the Active Ethernet architecture, a Passive Optical Network (PON) architecture, however, requires no OSP electronics. A completely passive optical splitter replaces the OSP Ethernet switch. The optical splitter combines the light from the ONTs and divides the light from the OLT. Optical splitters instead of active electronics reduces the cost of aggregation since the optical splitter is quite inexpensive, need no local power, and requires almost no maintenance of any kind. Up to 32 subscribers are typically served per fiber, though higher ratios are possible. See diagram below for the architecture of a typical PON access network.

Common types of PON deployed today are BPON, EPON, and GPON. On the horizon are 10G EPON and 10G GPON. WDM PON is another PON technology, though it operates differently. It uses an Arrayed WaveGuide (AWG) to divide wavelengths for individual delivery to subscribers instead of an optical splitter.

Hybrid Fiber Coax (HFC)

The cable companies have spent billions upgrading their access networks to Hybrid Fiber Coax (HFC). They will do everything possible to use these networks for delivering broadband services.

DOCSIS

DOCSIS is an upgrade to Hybrid Fiber Coax (HFC) networks to deliver high bandwidth broadband Internet service. Cable TV companies use DOCSIS to compete with telcos for broadband access subscribers. The fundamental attraction to DOCSIS is that, since it uses the existing HFC network to deliver hundreds of megabits of bandwidth, it avoids expensive network upgrades.

There are three major versions of DOCSIS: versions 1, 2, and 3. Both DOCSIS 1 and DOCSIS 2 are limited to a single analog channel for downstream bandwidth, but DOCSIS 3 allows bonding of several channels. Four bonded analog channels allows for downstream rates of over 100 Mbps. Note that an analog channel can be used for DOCSIS, or for an analog TV channel, but not both (of course), so the more channels dedicated to DOCSIS, the fewer analog TV channels available.

DOCSIS 3.0 requires upgrades to the Cable Modem Termination System (CMTS) and the cable modems, but other than this, the HFC network requires no changes. The only change is in equipment that is installed in the headend and on the subscriber premises.

RFoG

RF over Glass (RFoG) is a more drastic upgrade to a cable company’s network, and it implements a passive optical network (PON) to replace a Hybrid Fiber Coax (HFC) network. It is one technology known as DOCSIS PON or DPON. With RFoG, the headend multiplexer is replaced by a Wave Division Multiplex (WDM) device that delivers the RFoG wavelengths on separate fibers to other devices in the headend.

A fiber splitter serves a neighborhood or portion of a neighborhood. A fiber NID is installed at each house. See below for a diagram of a typical RFoG installation.

The attraction of RFoG is that it is compatible with HFC headend gear (except for the HFC fiber mux), and it is compatible with HFC subscriber equipment, including all those STBs. The RFoG NID delivers RF coax signals identical to those used for HFC. Legacy management systems can still be used when RFoG replaces an HFC network.

4G Wireless

4G wireless technologies deliver broadband Internet rates over wireless connections. These should be popular in rural areas as they can serve a large geographic area for relatively little capital compared to the wired alternatives. 4G deployments are just beginning. WiMAX is already deployed, and ClearWire has an extensive nationwide WiMAX network. Long Term Evolution (LTE) is quite popular today, with LTE-Advanced following and providing up to 1 Gbps of throughput.

There are two forms of 4G wireless broadband: mobile and fixed. Mobile is, like it sounds, for users on the go. Fixed wireless broadband provides higher bandwidth and is competition for DSL, DOCSIS, and Fiber access technologies to serve residential and business locations.

WiMAX

WiMAX (Worldwide Interoperability for Microwave Access) is a broadband wireless technology standardized by the IEEE and currently capable of rates of up to 10-15 Mbps. A new version of WiMAX known as 802.16m or WiMAX-m provides rates of up to 1 Gbps. WiMAX uses OFDMA for both the uplink and downlink. Clearwire is deploying WiMAX and will continue to do so for the next couple of years owing to agreements it has in place with its backers.

LTE

LTE (Long Term Evolution) is a competitor to WiMAX and is being standardized by the 3GPP (3rd Generation Partnership Project), which operates under authority of the ITU. LTE provides low latency and high data rates compared to 3G technologies and is similar to WiMAX. LTE-Advanced is is a new version of LTE providing data rates up to 1 Gbps. The 3GPP is still working on the LTE-Advanced specification, and services based on this new technology should be available in perhaps a few years.

Like WiMAX, LTE uses OFDMA for its downlink. However, a big difference between LTE and WiMAX is that LTE uses SC-FDMA for its uplink instead of OFDMA. A big advantage of SC-FDMA is its lower power consumption compared to OFDMA, so LTE should allow greater battery life in handsets/terminals.

A Cornucopia, but Some Still Go Hungry

Carriers have a plethora of choices when it comes to delivering broadband. What they deploy often has more to do with where they have been, rather than where they would like to go. Many subscribers, especially rural ones, would be happy to get anything at all.

Whither VDSL2, DOCSIS, and FTTH?

John Bartell — Thu, 24 Sep 2009 19:40:38 +0000

Choosing a new broadband network technology is no small task for a carrier. When moving to a better broadband technology, a carrier must consider what access infrastructure is available and already deployed. There are many factors to be considered, but by far, what a carrier has already deployed tends to be the biggest consideration. Using existing network infrastructure is imperative to keep installation costs down. It eliminates the most expensive part of installing a new service, which is trenching the ground to install new cables.

The cost to trench new cable for a single subscriber can be as high as $500 to $1500 per subscriber or more, and it can be especially high in rural areas. This cost can often be justified in new installations, but where other access facilities are available, such as coax or twisted pair copper, a carrier rationally must leverage these already installed access facilities.

Telcos have many billions of dollars of copper local loop installed, so they will gravitate to VDSL2 and ADSL2+, at least in the short term. Cablecos have invested tremendous capital in Hybrid Fiber Coax (HFC) networks, so they will continue with their DOCSIS upgrades (1.1 to 2.0 to 3.0) for quite a while. And those carriers with no infrastructure, like many municipalities, will make the leap to Fiber to the Home (FTTH), like Lafayette, LA (my home town), famously did. There are very few arguments for deploying any technology but FTTH in brand new wireline access networks.

VDSL2

Telcos have local loop copper, billions and billions of dollars worth, so they are keen to deploy Digital Subscriber Line (DSL) technologies like ADSL2+ and VDSL2. Many people complain that these are deficient when compared to FTTH, and they are right in many ways. Practically, VDSL2 delivers perhaps 50 Mbps at reasonable distances, and ADSL2+ substantially less.

The cost to deploy DSL where copper is already available is probably under 20% of the cost to deploy new fiber facilities along with splitters and the FTTH Optical Network Terminal (ONT) and Optical Line Terminal (OLT) electronics. This substantially lower cost for DSL makes the likelihood of a subscriber getting any broadband at all much more likely when this option is available. Telcos will still deploy FTTH in greenfield buildouts where they have no existing copper local loops. Over time, with competition from cablecos and munipalities, telcos will deploy FTTH networks in brownfield areas as well.

DOCSIS

Over the last decade or so, Cable TV companies (also know as MSOs) have installed billions of dollars of Hybrid Fiber Coax (HFC) access networks. They do not need a new broadband technology to deliver quite good broadband rates. DOCSIS 3.0 allows bonding of multiple HFC analog channels to provide rates of over 100 Mbps, though this is shared among a number of subscribers. However, most subscribers use only a very small fraction of their available bandwidth, so this is not too much of a constraint.

For an MSO to move to DOCSIS 3.0, it is usually just an upgrade to the Cable Modem Termination System (CMTS) in the headend and new DOCSIS 3.0 cable modems in subscriber residences. One requirement for DOCSIS 3.0 upgrades is that several analog TV channels must be sacrificed for broadband, but this is a much easier sacrifice than investing in an entirely new fiber access network.

Broadband is delivered to my house over an HFC network. The picture below (click to enlarge) shows the gray coax NID installed on the outside of my house. This NID delivers voice, analog TV, and cable modem service to my house.

The Cable TV companies will probably be the slowest to migrate to FTTH. Their HFC networks can easily deliver broadband services at rates in the hundreds of megabits per second. One issue they have is that their HFC networks do not reach many business customers they would like to serve, so their initial access fiber deployments will serve their business customers.

FTTH

Fiber to the Home (FTTH) (a subset of Fiber to the Premises or FTTP) is seen by many as the ultimate broadband delivery platform for residential subscribers, and they are correct. Nothing compares with the bandwidth capabilities of FTTH. However, in brownfield areas, nothing is as expensive as deploying a new FTTH network. Interestingly, all wireline carriers, including munipalities, cablecos, and telcos, are moving to FTTH architectures, each at their own pace. FTTH is a broadband access architecture common to all these carriers. Once one carrier in an area deploys FTTH, other carriers will have to respond, and often this will be with another FTTH network.

FTTH, although capable of tremendous broadband rates, is not a perfect solution. Unlike coax and twisted pair copper, fiber is unable to deliver electrical power of any useful magnitude (there is technology for delivering photovoltaic power over fiber, but it only delivers a very small amount of power, and very inefficiently at that, over the single mode fiber used in FTTH). Most carriers do not leave or install any copper local loop along with the fiber, so subscribers must deal with a battery to ensure lifeline access for FTTH services.

Almost everyone agrees that FTTH is the future for all wireline access carriers. Each carrier will proceed to this ultimate broadband destination at their own pace, a pace determined largely by the alternative networks already in place and the level of local broadband competition.

VDSL2 Versus DOCSIS Smackdown

John Bartell — Fri, 04 Sep 2009 12:23:51 +0000

Light Reading is reporting that Cox is offering 50 Mbps DOCSIS cable modem service in Arizona on Cisco gear to counter Qwest’s 40 Mbps VDSL2 service offering. Prices are $89.99 per month for Cox’s DOCSIS-based offering and $109.99 per month for Qwest’s VDSL2-based offering. The informal poll on FTTxtra indicated that the majority of participants were willing to spend as much as $100 per month on broadband access service, so the prices Cox and Qwest are charging make sense as initial conditions.

The bulk of HFC electronics are deployed in OutSide Plant (OSP) cabinets serving a neighborhood or portions of a large neighborhood, and with DOCSIS, all of this substantial investment is retained by the MSO. These node cabinets are fed with fiber and serve many homes with telephone, DOCSIS cable modem, and television services over coax. A typical HFC node cabinet (this is the actual one serving my neighborhood) is shown below.

HFC Cabinet

The picture below shows the power, telephone, and cable connections into my house. The grey box is the HFC coax NID, which is span powered from the cabinet above. The HFC NID has a coax connection to a three-way splitter installed above the NID and a two-wire telephone connection to the brown box on the top right.

Telcos often deploy VDSL/VDSL2 services in FTTN (at the node) or FTTC (at the curb) architectures. The length of some copper local loops served from a FTTN node (typically up to 5000 feet or so, depends on the telco) tend to be a bit long for VDSL2, but sometimes FTTN is a telco’s only feasible option for installing VDSL2.

FTTN VDSL2 nodes are installed in locations with access to the copper local loop, usually through what is called a Feeder Distribution Interface or FDI, shown below. The DSL electronics accessing the loops at this FDI are housed in the beige boxes located behind the FDI. The FDI houses no electronics, just cables and cross-connects. A typical concentration level at an FDI is one upstream copper pair (feeder pair) for two downstream (distribution) copper pairs, so on average, before any electronics are installed at the node, only perhaps half of the local loops delivered to subscribers are already in use.

DSL Feeder Distribution Interface (FDI)

If you want to see what an HFC network can do with broadband, do not miss the video at the end of the article on DOCSIS 3.0 showing a speed test on a Comcast DOCSIS 3.0 network.

RFoG (DPON) Tutorial

John Bartell — Mon, 31 Aug 2009 11:01:32 +0000

RF over Glass (RFoG) replaces a Hybrid Fiber/Coax (HFC) access network with a passive optical network (PON). It is one form of what is known as DOCSIS PON or DPON (another technology also known as DPON is DOCSIS over EPON, which is being marketed by Salira). With RFoG, the HFC head end access fiber multiplexer is replaced by a WDM device that divides the upstream and downstream wavelengths and delivers them on separate fibers to other devices in the headend. With RFoG, fiber splitters, instead of HFC nodes, are deployed to serve neighborhoods. A fiber NID is installed at each house. So the RFoG architecture in the outside plant (OSP) is identical to that of a PON. The image below (click to enlarge) provides an architectural overview of a traditional HFC network.

The image below provides an overview of an RFoG implementation. Note that Set Top Boxes (STBs) within the home can remain unchanged in a shift from traditional HFC to RFoG.

The RFoG form of DPON allows MSOs to use their existing operational infrastructure and much of their existing equipment with the new access technology. RFoG is compatible with most of the MSO headend gear (except for the HFC fiber mux), and it is compatible with the in-home gear (including all those STBs) currently in use with HFC networks. The RFoG ONU on the side of the house generates/receives RF coax signals identical to those used with HFC. The operational systems that tend the headend and the CPE with an HFC access network work just fine when RFoG is used instead.

Standards work is ongoing for RFoG in the Society of Cable Telecommunications Engineers (SCTE) subcommittee IP SP 91. The document they are developing is entitled RFoG System Overview.

So, what are the major advantages of RFoG?

RFoG provides better reach and requires much less equipment in the access network. From the HFC node, coax requires repeaters every 1000 feet or so. RFoG does not require an active piece of gear be installed in the access network, only passive splitters, and can reach as far as 20km.
RFoG is green. It does not require an active node or repeaters, so it avoids the power consumption of these devices.
With RFoG, the ultimate signal delivered to the subscriber is better because optical transmission is impervious to electromagnetic interference, especially important in the return path.
The RF bandwidth provided in an RFoG network is greater than what is provided by HFC. This allows a provider to deliver services above 1 GHz, even another complete channel lineup.
Since it requires no access network electronics, RFoG imposes much less of a hardware maintenance requirement than does HFC. Additionally, fiber is much better than coax at handling changes in temperature, humidity, and other environmental conditions further reducing the operational burden.
RFoG, if deployed with compatible wavelengths, can easily be upgraded to work with GPON or other PON technologies to provide tremendous data bandwidth.

RFoG is a smart way for carriers to deploy a simple, proven RF video solution along with high bandwidth PON services. The diagram below (click to enlarge for detail) shows what a combined PON and RFoG implementation looks like.

One of the largest advantages of RFoG is that it provides better service at a lower operational cost than comparable HFC networks.

For further reading, Broadband Properties has this article describing what Cable companies are doing with fiber. Cisco offers this whitepaper describing their RFoG+ solution. Tom Anderson of Alloptic wrote this article on Why RFoG?. Motorola offers this whitepaper describing their RFoG solution.

PON, VDSL, and DOCSIS 3.0 Demand Up

John Bartell — Wed, 26 Aug 2009 15:44:17 +0000

From a Dell’Oro Group press release issued today…

“Two important factors impacting demand for Access equipment are broadband subscriber additions and network upgrades,” said Tam Dell’Oro, President of Dell’Oro Group. “Total broadband subscriber additions have been lower since the peak year of 2006, and are having a negative impact on equipment demand, especially for a slower-speed technology such as ADSL. On the positive side are the upgrade projects that are being driven by competition, increasing internet traffic, government incentives, and the desire by operators to enable new revenue-generating services such as TV over broadband. These upgrade projects increasingly drive demand for higher-speed PON, VDSL, and Cable DOCSIS 3.0 equipment,” Dell’Oro added.

DOCSIS 3.0 Tutorial

John Bartell — Thu, 13 Aug 2009 13:47:09 +0000

HFC DOCSIS Cable Modem and Router

DOCSIS 3.0 (sometimes DOCSIS 3) is a standard developed by CableLabs to upgrade Hybrid Fiber Coax (HFC) networks to deliver high bandwidth broadband Internet service at rates of up to several hundred megabits per second. It is used by MSOs (cable TV companies) to compete against Telcos using FTTH and FTTN/FTTC with VDSL2.

Cost

DOCSIS 3 is estimated to cost approximately $70 per subscriber, mostly in the headend, in addition to the cost of a new DOCSIS 3 compliant cable modem required on the subscriber premises. This is in line with the cost of installing DSL in an existing OutSide Plant (OSP) cabinet. What DOCSIS avoids is drastic upgrades to an HFC network for this extra bandwidth, and this is its fundamental appeal. Cable companies can deliver very high Internet access bandwidth for very little additional investment with the various DOCSIS standards, especially DOCSIS 3.

Bandwidth

The most important thing about DOCSIS 3.0 is bonding cable TV channels for more bandwidth. Previous versions of DOCSIS only used a single channel and were limited in their throughput. With 4 channels, bandwidths of about 160/120Mbps are possible. With 8 channels, rates of 320/120Mbps are possible. Of course, the cable companies with have to give up these analog video channels to devote them to DOCSIS 3.0, but this can be done incrementally in the headend, and it does not have to be done throughout the network all at once.

Equipment

The equipment required to deliver DOCSIS 3.0 services is the Cable Modem Termination System (CMTS) and the DOCSIS 3.0 cable modems on each subscriber premises. Both must be upgraded to support DOCSIS 3.0 over an existing HFC network, shown below before an upgrade. Note that the HFC network is unchanged with the upgrade to DOCSIS 3.0. The only change is in equipment that is installed in the headend and on the subscriber premises.

HFC Network

Below is a Motorola DOCSIS 3.0 CMTS, which would be placed at the headend shown above.

Motorola DOCSIS 3.0 CMTS

And here is a Motorola DOCSIS 3.0 cable modem, one of which is required in each home to be served with DOCSIS 3.0. Watch the video at the end of this article to see this cable modem in action. Other vendors of DOCSIS cable modems include Cisco and Arris.

Motorola HFC DOCSIS 3.0 cable modem

DOCSIS 1, 2, and 3

This figure below provides a good overview of the various DOCSIS versions and how 3.0 compares. Also note that Europe’s EuroDOCSIS standard is different than the US DOCSIS standards (1.x, 2.0, and 3.0) as the frequency allocations are different.

DOCSIS and EuroDOCSIS 1, 2, and 3

Channel Bonding

And this one, from Broadcom, provides a graphical representation of what DOCSIS 3.0 is doing with channel bonding. Bonding dispersion distributes the traffic across several channels, and bonding combining technology combines the multiple streams into a single one. All this is done with support for channel bonding in the headend CMTS and in the DOCSIS 3.0 cable modems.

DOCSIS 3.0 Channel Bonding

More Info

For more information, see this overview (from Cisco), follow this link is to a great tutorial with further technical details, or watch the overview video below.

[youtube 43qqOpJmbWg]

And finally, to see how much bandwidth DOCSIS 3.0 really provides, you MUST watch this video of a speed test on a Comcast network.

[youtube MLAvj6S0zCY]

A good overview presentation from Cisco.

DOCSIS 3.0, a DSL nightmare

John Bartell — Mon, 10 Aug 2009 21:33:57 +0000

As much as the DSL folks would like to disbelieve, DOCSIS 3.0 is here today, works quite well, and will be ubiquitous in just a few years. Below are a few quotes indicating progress and expected penetration for DOCSIS 3.0.

From Broadcom’s 2Q09 earnings call (he means DOCSIS 3.0 when he refers to DOCSIS),

Scott McGregor, Broadcom’s CEO, said “Absolutely, I do believe we were seeing continued move over to DOCSIS. That’s more next year phenomena, but HD continues to be a phenomenon. People will buy a new TV set and they really want to get the better picture quality.”

Pike and Fisher have released a report covering DOCSIS 3.0, and they predict universal coverage in a few years. From their report summary…

We conclude that the top cable operators will have DOCSIS 3.0 covering 100% of homes passed by the end of 2013, and that MSOs can significantly shorten the time to achieve a return on their DOCSIS 3.0 investments by aggressively targeting business customers.

Comcast is going full speed ahead with DOCSIS 3.0. From DSLReports…

“We have raised our goal to roll out DOCSIS 3.0 to nearly 80% of our national footprint before the end of this year,” Comcast spokesman Charlie Douglas tells us. “That would mean passing about 40 million homes and businesses with the new wideband offerings Extreme 50 and Ultra 22, in addition to doubling speeds for existing Premier customers for no additional cost,” he notes.

And finally, the following is a video of an internet speed test over a Motorola DOCSIS 3.0 modem.

[youtube MLAvj6S0zCY&]

Competition Drives FTTP

John Bartell — Wed, 05 Aug 2009 19:42:57 +0000

What is the big driver for FTTP deployments? In a word–competition, if this was not already obvious. FTTP is not necessary with captive customers. DSL is just fine for the Telcos thank you. And the MSOs without competition would romp with 3 Mbps cable modem service.

What is interesting is that both the telcos and MSOs are driving to the exact same architecture with FTTP. The names telco and MSO will soon lose their distinct meanings. Both will become broadband access providers. The resulting differences will be due to their legacy networks and systems, but these differences will not be consequential. Both will be delivering the exact same services across virtually the exact same access architecture. Since they will be selling largely undifferentiated services, they will be forced to compete on price and bunding.

With both having to run just to keep up, the telcos and MSOs are in what is known as a Red Queen scenario. In Lewis Carroll’s Through the Looking Glass, the Red Queen said, “It takes all the running you can do, to keep in the same place.” And this is what will be so good for our fair industry. Lots of fiber will have to be installed by the telcos to keep up with the MSOs, and the MSOs will not be able to rest on their laurels with DOCSIS 3.0. Who benefits from this race? Why, it is the consumer of course.