Thursday, June 19, 2014

FCC report checks if your internet speed lives up to the ads, and why that's not fast enough

Source: http://www.engadget.com/2014/06/19/fcc-broadband-america-report/

Operations During US Internet Fiber Optic Cable Installations

Since 2011 the FCC has collected data on the wired (there's a separate report for wireless) broadband speeds US residents are actually receiving to release in its Measuring Broadband America report, and now the most recent one is here. First, the good news -- based on its data (collected from "Whitebox" devices sent to around 10,000 participants that performed automated tests during September 2013), most ISPs were meeting or exceeding their advertised speeds even during peak hours. Four that couldn't deliver 90 percent or more of their advertised rate during peak hours included Verizon, Frontier, Qwest and Windstream -- all of which can expect a letter from the FCC asking why not, for whatever good that will do. So if ISPs are delivering 101 percent of advertised speeds, why are users still seeing buffering notifications and experiencing slowdowns?

[Image credit: Bloomberg via Getty Images]

There are a few reasons (not including the shameful state of U.S. broadband former commissioner Mike Copps called out in recent comments), and one that it looked closely at is consistency. The way the FCC's report measures that, is by showing how many of the service's customers get a certain percentage of the advertised speed, a certain percentage of the time. With the bar placed at 80 percent of the customers, 80 percent of the time, the only services pushing higher-than-advertised download speeds were Verizon FiOS and Cablevision. Another issue is that DSL lagged behind cable, fiber and even satellite in being able to meet its advertised speeds -- averaging 91 percent of the promised download speeds during peak periods, as compared to 102 percent for cable.

Of course, as everyone has recently become familiar with, another issue is apparent slowdowns at "certain interconnection points" during the test period. The FCC didn't include that info in this report, but says it will have new tests ready to measure the impact of network congestion by the testing period planned for later this year. It's working with services like Netflix, YouTube and Hulu to analyze the data, and previously announced it was looking into the interconnection deals between Netflix and Comcast/Verizon. In a separate blog post, the FCC specifically called out drops in broadband performance during times when internet middleman Cogent was having its disputes, and says it will release the raw data it collected.

Also, data on interconnection congestion will be made public, & @FCC will work w. Netflix, Hulu & others 2 further analyze.

- Gigi Sohn (@GigiBSohnFCC) June 18, 2014

The million dollar question(s) however, are how this report will play into the commission's actions on net neutrality, peering agreements and proposed mergers by communications giants. If DSL doesn't measure up as a competitor to cable, does this mean the Time Warner Cable / Comcast deal shouldn't go through, but maybe AT&T / DirecTV should? And who exactly is telling the truth between Netflix and the ISPs? The FCC is receiving more scrutiny than ever about its handling of all of the above -- if you need some supporting info for your comments, just hit this link.

Filed under: , , , ,

Comments

Source: Measuring Broadband America 2014, Internet Traffic Exchange, FCC (1), (2) (PDF)