x
Monitoring & Assurance

SlideshowBy the Numbers: Network Latency in Texas

Wireless vs. Wireline Latencies in Texas
Each dot represents a different company operating a broadband network in Texas; the y axis is latency in milliseconds. The grouping on the left includes four wireless carriers; Sprint and AT&T are close together at about the group average; T-Mobile is doing notably better by this measure than its rivals; Verizon is doing notably worse. The grouping on the right includes five wireline companies. Verizon's and Comcast's dots completely overlap at 31ms of latency. Charter Communications, Cox Communications, and AT&T all have higher network latency, their numbers are similar and so their three dots are clustered together.
Each dot represents a different company operating a broadband network in Texas; the y axis is latency in milliseconds. The grouping on the left includes four wireless carriers; Sprint and AT&T are close together at about the group average; T-Mobile is doing notably better by this measure than its rivals; Verizon is doing notably worse. The grouping on the right includes five wireline companies. Verizon's and Comcast's dots completely overlap at 31ms of latency. Charter Communications, Cox Communications, and AT&T all have higher network latency, their numbers are similar and so their three dots are clustered together.

kq4ym 6/19/2016 | 10:19:52 AM
Re: proof is in the details! Yes, it will be interesting to see how the measurements might vary depending on the methodology and then again there's the "traffic load that can vary monthly, weekly, daily, and sometimes even hourly," that's going to play into just how, when, and where the latency measurements are made I would think.
Pete Mastin 6/8/2016 | 4:34:34 PM
Re: proof is in the details! Hi Jayakd0, 

I am Pete Mastin from Cedexis and I agree with your comments that more granularity is better! With regard to the methodology that was used I can help there. Basically Cedexis used Real User Measureents to measure public infrastrcutrue such as ISPs, Clouds and Content Delivery networks. Because our JS tag is so broadly distributted we get Billions of measurements a day of all this public infrastructure. We render the results of these measurements in near realtime to our Portal. You can read more about our methodology here. You can also get much more granular measurements over time and every percentile. 

http://www.cedexis.com/radar/methodology.html

 

Thanks
jayakd0 6/8/2016 | 8:10:34 AM
proof is in the details! It will be great to know how these latencies were measured. Also better to state the variance also (than just mean). Monthly average figures, whether it will be that useful? unless we have better visibility on day vs night averages, weekday vs weekend averages
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE