Why "Net Neutraliity" is a Bad Idea

If no Internet service provider can give preference to any streams or protocols, it is not possible to create a service a user actually does want: the ability to give preference to voice streams when on a Skype call or conference, when watching a webcast or a movie.

It is understandable that some advocates worry about ISPs giving preference to their own streams over others. News: they already do. It's called "cable TV." The issue is whether an end user can prioritize certain protocols at certain times, to optimize their own experience.

Strict net neutrality, which allows no prioritization, or minimal traffic shaping to maintain overall user experience when a very small number of users are straining a network, also will prohibit any other "user friendly" forms of prioritization.

http://venturebeat.com/2009/03/23/netflix-were-not-slowing-our-streamsyet/

Seattle Weighing Own FTTH Build?

The city of Seattle seems t be pondering building and perhaps even operating its own fiber to the home network. Planners think such a network, offering speeds of perhaps 25 Mbps, might ultimately get 24 percent market share.

That certainly is within the realm of possibility, though lots of executives with experience in the overbuilder business, where a network competes against both a telco and a cable company, might point out that the business case is tough without the ability to offer both voice and video services.

A full FTTH network with on-going revenues from access alone might be very difficult.

http://seattletimes.nwsource.com/html/businesstechnology/2004293823_brier20.html

Mobile Access Less Strategic, Vodafone, Telefonica Signal

Telefónica and Vodafone have confirmed wide-ranging plans to share mobile network assets--starting with tower sites--across Europe.

As part of the collaboration, Telefónica and Vodafone also are actively exploring opportunities to cooperate in related areas such as the provision of transmission services, which would extend sharing to radio capacity as well. The companies will share mobile network assets in four European markets where both have operations - Germany, Spain, Ireland and the UK - while discussions are ongoing to extend the agreement to a fifth market, the Czech Republic.

The joint building of new sites and or consolidation of existing 2G and 3G tower sites, with one site housing the equipment of both companies where previously two would have been used, is expected to lead to a significant reduction in the total number of towers in operation.

The move also suggests a perception of where value lies. Towers are deemed not to be strategic, though required, and radio resources, though likewise required, might not offer much end-user value either. That tends not to be as true for more-expensive wired infrastructure, where most executives with the ability to do so tend to prefer exclusive use of their own facilities, essentially forcing would-be competitors to undertake their own expensive construction projects.

The proof point here is what has happened in markets where robust wired network wholesale access has been enforced. In such markets, incumbents tend to lose huge chunks of market share while end user prices fall dramatically. As good as that is for consumers, it is problematic for service providers.

For that reason, regulators are being much more cautious about extending current rules for copper access to new optical access, taking care to ensure that investment opportunities are attractive enough to fund such networks.

http://www.vodafone.com/start/media_relations/news/group_press_releases/2009/telefonica_and_vodafone.html

How Much More Can be Wrung from DSL?

Ericsson engineers have demonstrated in the lab a way to bond together six unshielded copper wire pairs to push 500 Mbps about 500 meters. The technique is promising for the same reason all such technology developments are: they suggest it is possible to wring yet more performance out of a widely-deployed infrastructure that has some real-world performance limitations.

That doesn't mean the lab performance easily can be ported to a real-world environment. For starters, lab tests are not encumbered by all the messy real-world complications. Real-world access plant faces hazards such as squirrels, moisture ingress, corrosion and electromagnetic interference from other services that might be running inside the same cable bundle.

The other practical limitation is that most consumer locations might not have six available spare pairs to use. On the other hand, as fewer people use landlines, there theoretically are more unused pairs to tap. It's a messy process, though, and likely would require installing new drop cables and resplicing the drops to the distribution wiring. That sort of heavy craft work isn't so interesting for any service provider that might contemplate trying to introduce such a service.

On the other hand, in some urban markets in countries with dense housing, copper drops might be short enough that something could be practical, perhaps not delivering 500 Mbps, but still providing enough bandwidth to be quite interesting.

As often is the case, spot deployment seems more promising. Mobile tower backhaul, for example, could benefit, as the upside from deploying a relatively short length of new cable, when a suitable fiber node is available, could be quite helpful. Some enterprise locations also could benefit, much as these locations now do for "Ethernet-over-copper" technologies.

In the consumer market, at least in the United States, where loop lengths generally are longer than in Europe, bonding two pairs is much more practical. As always is the case when trying to market a service to the consumer market, varigated real-world physical constraints are important. Two pairs is typical. In some cases four pairs might be fairly standard. But four pairs might not generally be available everywhere in a region or city.

No question, any developments that increase DSL bandwidth are helpful. How helpful really "depends" on a variety of messy details.

Big Gains Seen for SIP Trunking, Ethernet

Nemertes Research says 53 percent of respondents to a recent survey are evaluating, deploying or planning to deploy SIP trunking to reduce access costs and take advantage of new services for in-bound and out-bound call routing.

In addition, 62 percent are using, or planning to use, Ethernet as a WAN connectivity technology to reduce bandwidth costs, with 85 percent planning to increase Ethernet usage.

Cox Aims SIP Trunking at 20-99 Employee Segment

Cox Business will be rolling out SIP trunking later this year, aiming at the 20 to 99 employee organization, after introducing its hosted IP PBX service.

http://voip.biz-news.com/news/en_US/2009/03/19/0001/cox-business-to-roll-out-sip-trunking-later-this-year

Broadband Stimulus: Too Much Optimism About What Can Be Done

Hype is running far ahead of what can be done even when all of the broadband stimulus funds are awarded and projects deployed. Broadband access is a hugely capital-intensive process, especially for the last couple of percent of locations that are so remote, in areas so thinly settled, that nearly all the cost must be recovered from such a small number of potential customers.

Physics and demography, not lack of will, are the key problems. In an urban area, some forms of broadband can be deployed for less than $2,000 a location, and often for as little as $1,000 a location.

In remote areas, serving one location can cost $10,000 to $50,000. You can build your own spreadsheet to figure out how long it would take to break even on that sort of investment, when the customer is expected to pay $40 to $50 a month. Don't forget the cost of interest on borrowed money, operating costs, maintenance and repairs, as well as the need, at some point, to replace the entire infrastructure because of age.

There likely will be some incremental benefits. But the "problem" of access in rural areas, or the quite-different problem of "under-used" broadband, will not be solved. Not by a long shot.

http://ow.ly/1dPd