T-Mobile and Verizon recently came out ahead of AT&T and Sprint in terms of network latency, which is important for consumers and the app community
Mobile operators are constantly being graded on network and operational performance, with a seemingly never-ending parade of new reports showing how domestic carriers rank across various metrics.
One aspect that is only just recently garnering attention is in terms of network latency, which is the time it takes for a consumer request to be completed via a mobile data connection. The lower the latency, the quicker the response for consumers.
This measurement is expected to become increasingly important as mobile operators move towards “5G” network technologies and greater support for the “internet of things,” which are likely to require significantly lower latency in order to support planned services. Many analysts have noted the current threshold of around 100 milliseconds seen from top performing “4G” LTE networks will likely need to improve to under 10 milliseconds to fully support IoT.
On this week’s Carrier Wrap, we spoke with Andrew Levy, co-founder and chief strategy officer at Apteligent, about the firm’s recent survey that found T-Mobile US and Verizon Wireless posting significantly better network latency results compared with AT&T Mobility and Sprint, and what that means for consumers and the application community.
Apteligent said its data showed T-Mobile US’ network posting average latency of 299 milliseconds, edging out Verizon Wireless for top spot among domestic carriers. T-Mobile US was also applauded for having the “most predictable network performance,” with Apteligent noting only 16% of requests on T-Mobile US’ network required more than 500 milliseconds to complete.
At the other end of the scale, regional carrier U.S. Cellular was dinged for having the “slowest” network, with tests showing average latency of 337 milliseconds and 22% of requests taking more than 500 milliseconds to complete. U.S. Cellular is still in the midst of rolling out LTE services, and has a much shallower spectrum position than its larger rivals, though it also supports considerably fewer customers.
Beyond the network conditions, Apteligent cited several reasons for lagging application request performance, including network architecture.
“There are several strategies to deal with slow network performance,” Apteligent wrote. “If you have a content heavy app, consider deploying a content delivery network in regions with high latency. An app can also be designed to ‘hide’ slow downloads by bundling content with the app download and intelligently downloading content ahead of when the app needs it. Another common mistake is to misuse or abuse cloud service APIs. A good practice to reduce data transfer is to add parameters to your API calls that restrict the results based on time or location.”
Make sure to check us out again next week when we are scheduled to speak with Berge Ayvazian from Wireless 20/20 to get an update on the ongoing FCC 600 MHz incentive auction proceedings.
Bored? Why not follow me on Twitter