Ericsson’s Mobility Report: why network performance matters, and how to measure it

Ericsson's launch of its Mobility Report 2013 contained some interesting nuggets for mobile network watchers.

In amongst a lot of detail lot of detail the company highlighted its Apps Coverage concept. Essentially this looks at network performance not in terms of geographic coverage but in terms of how many sessions achieved performance that was sufficient to support certain apps.

Presenting the June 2013 data the company presented network performance in terms of the applications that you can carry out with that level of performance: so with 100kps at cell edge you can send texts and make voice calls, at 1Mbps in the mid-cell you are in browsing, IM and audio steaming in terms of your app coverage, and at 10Mbps close to the site your App Coverage is full video streaming.

Ericsson’s contention, by aggregating together a huge number of speed tests from Ookla, is that at the moment 78% of Ookla’s Speedtest measurements in the USA and Western Europe showed downlink speeds in the mid-tier of App Coverage. European users were seeing 17% high level app coverage, and US users 31%. That means there is still quite poor coverage in terms of support for apps requiring higher bandwidths.

Patrick Cerwall, Head of Strategic Marketing and Intelligence, Networks, Ericsson, said that thinking about network performance in terms of app coverage could enable operators to plan networks in a way that takes account of all the factors that could account for performance, including backhaul, content networks and core capabilities.

In fact, if you take the cell edge issue that Ericsson used to demonstrate its conceptual view of app coverage, this is one of the issues being addressed within LTE-Advanced developments. Solutions such as relay nodes and the greater co-ordination of small cells within feeder macro cells are intended to smooth out the peaks and troughs of performance within a macro cell. That could mean that operators could achieve a greater “Apps Coverage” without necessarily visibily extending their geographic coverage.

Another aspect worth considering is Ericssons Minutes of Network Use measure. Although Ericsson was not shy in bringing to readers’ attention the dominance of video in traffic volume terms, it also highlighted the fact that certain apps consume more of a network’s signalling resources by being very chatty in signalling.

Mobile network performance is the chief driver of customer loyalty

The industry has been switched on for quite a while to the potential impact of chatty apps that consume relatively little bandwidth but impact heavily on the network, typically those with some aspect of presence and/or social network apps is not new. Now Ericsson has provided some insight into the apps, but also the devices and form factors, drive the most signaling chatter. Larger screens, for instance, drive greater usage of video but have little impact on usage of social networking apps. And you don’t get the temporal spikes with chatty apps that you do with video – background apps stay on 24 hours a day.

Finally, for those with design authority within the mobile network, Ericsson had a handy back-up stat. An assessment of net promoter scores across a sample of 9,040 users spread across the world showed mobile network performance is the chief driver of customer loyalty It clearly outscored aspects such as price, customer support, value for money and device range.