Tag: "symmetry"

Posted November 30, 2010 by christopher

So Comcast and Level 3 are in a peering dispute following the Netflix partnership with Level 3 to distribute their streaming movie service. Studies suggest Netflix movie streaming has become a significant chunk of Internet traffic, particularly at peak times.

A quick primer on peering: the Internet is comprised of a bunch of networks that exchange traffic. Sometimes one has to pay another network for transit and sometimes (commonly with big carriers like Comcast and Level 3) networks have an agreement to exchange traffic without charging (one reason: the costs of monitoring the amount of traffic can be greater than the prices that would be charged). (Update: Read the Ars Technica story for a longer explanation of peering and this conflict.)

Comcast claims that Level 3 is sending Comcast 5x as much traffic as Comcast sends to Level 3 and therefore wants to charge Level 3 for access to Comcast customers. Of course, as Comcast only offers radically asymmetrical services to subscribers, one wonders how Level 3 could be 1:1 with Comcast…

At Public Knowledge, Harold Feld ties the dispute to network neutrality:

On its face, this is the sort of toll booth between residential subscribers and the content of their choice that a Net Neutrality rule is supposed to prohibit.  In addition, this is exactly the sort of anticompetitive harm that opponents of Comcast’s merger with NBC-Universal have warned would happen — that Comcast would leverage its network to harm distribution of competitive video services, while raising prices on its own customers.

Susan Crawford

Susan Crawford wrote a lengthier piece about Comcast, Netflix, network neutrality, set-top boxes and NBC that is well worth reading (as is just about anything she writes).

However, for the purposes of this post, we will assume the 5x traffic imbalance is true (and unique and...

Read more
Posted November 21, 2010 by christopher

One of the key differences between community owned networks and those driven by profit is customer service. Community-driven providers spend more and create more jobs in the community to ensure subscribers' needs are met. The massive private companies instead choose to outsource the jobs to call centers (sometimes in the U.S., sometimes outside) in order to cut costs (and jobs - see the report from the Media and Democracy Coalition).

We've seen a few examples of the big carrier approach in this arena - as when Cablevision billed apartment residents $500 after a fire for the DVR that was consumed in the blaze... stay classy, Cablevision.

Another difference between community networks and the big carriers is that big carriers see little reason to upgrade their anemic networks to ensure communities remain competitive in the digital age. As Free Press has long documented [pdf] big companies like AT&T have been investing less in recent years as the U.S. has continued falling in international broadband rankings.

Up here in Minnesota, Qwest has invested in FTTN - what they call fiber-to-the-node. We call it Fiber-to-the-Nowhere. For those who happen to live very close to the node, they get slightly faster DSL speeds that are still vastly asymmetrical. Meanwhile, Qwest has branded this modest improvement for some as "fiber-optic fast" and "heavy duty (HD)" Internet, misleading customers into thinking they are actually going to get faster speeds than Comcast's DOCSIS 3.

Much as I hate to praise the middling DOCSIS 3 upgrade, it certainly offers a better experience than any real results we have seen with Qwest. But as we carefully documented in this report, community networks offer more for less.

Two friends recently moved to Qwest. One, J, was convinced by a Qwest salesperson that Qwest would be much faster so he signed up for a 20Mbps down package. Fortunately, he didn't cancel the cable immediately because he was back on it quickly - he says Qwest dropped out 4 times in...

Read more
Posted November 9, 2009 by christopher

The Media and Democracy Coalition put together an impressive report examining a number of policy options to put communities first in telecommunications infrastructure. The report discusses the fundamental importance of broadband - noting that it enables the right to communicate. Having establishing its importance, the report notes that good policy must be well informed and goes on to make multiple recommendations.

Policy should promote competition, innovation, localism, and opportunity. Locally-owned and -operated networks support these core goals of Federal broadband policy, and therefore should receive priority in terms of Federal support. Structural separation of ownership of broadband infrastructure from the delivery of service over that infrastructure will further promote these goals.

The report also touches on other key issues - including Universal Access, a non-discriminatory Internet (network neutrality), symmetrical connections, and privacy. But the most important focus from our perspective is that of localism:

For decades, American communities — both rural and urban — have been neglected and underserved by absentee-owned networks, whose business models clearly do not work in smaller or economically challenged communities. By contrast, in the communities in which they are based, locally-owned networks are more likely than absentee-owned networks to provide rapid response to emergencies, enhanced services, and value-added, social capital benefits such as job-training, youth-mentoring, and small business incubation. In addition, local networks are less likely to outsource jobs, thereby strengthening local and regional economies, while creating more opportunities for community-based innovation and problem-solving. Federal broadband policy that prioritizes support for local networks will produce more competitive markets, consumer choice, and opportunities for innovation.

The first two recommendations in this section calls for federal policy that discourages absentee ownership:

  1. To fulfill the goal of extending broadband service to un- and underserved areas, federal broadband policy should prioritize support for locally-owned and -operated networks, including those owned by local governments, nonprofits and cooperatives, and public-private partnerships. Local, for-profit networks receiving federal broadband subsidies should certify that any...
Read more
Posted October 30, 2009 by christopher

Many publicly owned community fiber networks offer symmetrical connections - allowing subscribers to both upload and download content at the same speeds. This approach treats the subscriber as both a producer and consumer of content (one of the reasons I generally avoid calling a subscriber a "customer" or "user").

Nearly all private network offerings are asymmetrical - DSL and cable are more less subject to constraints that encourage asymmetry, but in the case of fiber, one might assume that private companies are generally more interested in selling content to subscribers rather than encouraging them to create their own.

These companies have generally argued that symmetrical connections are just not necessary because most people are inherently more interested in downloading content than uploading - and note that on existing networks, people tend to download more than they upload.

However, the aggregate data of some 7,000 users on a fast, symmetrical network in Europe suggests that when subscribers have the opportunity, their upload usage balances the downstream usage.

We should continue pushing for increased upstream capacity from providers - especially providers that have to listen to their community. As for absentee-owned companies only interested in profits, well, good luck.

Which brings me to the flu. One would rationally expect that when a profit-maximizing company builds a telecommunications network, it will make different trade-offs when it comes to redundancy and spare capacity. Planning for high-impact, low probability events is not as high on the priority list of a company looking out first for shareholder interests. On the other hand, communities are more likely to be concerned.

Suburban community Lakeville in Minnesota, has been significantly motivated in its attempts to improve fast broadband access by a recognition that an epidemic or pandemic would leave the community paralyzed and its networks unable to cope with a many telecommuters. DSL and cable networks cannot handle a sudden surge in usage.

To some, this appears to be a surprise though the recent GAO Report rightly notes that full fiber networks are less susceptible to falling apart when they are...

Read more

Pages

Subscribe to symmetry