Industry Blog

The Surprising Effect of the Weather on Your Network

Time to read: 3 minutes

The weather is an important topic of conversation in most people’s daily lives as we plan our activities and events around its prediction. It’s important enough that a number of smartphone providers provide it as a pre-loaded application on the device, and a lot of us by default have it on our devices’ home screen. But comfort of knowing the weather comes at a surprising cost for mobile data and is representative of a growing trend in background applications.

We used our Mobile Data Analytics solution to explore the effects of weather apps on phones, which is considered a “silent app” compared to more obvious activities such as social networking. The research yielded some surprising observations.

Silent apps, such as the weather app, affect resources because when the apps connect via the operator’s radio network the application establishes a radio link, and then a data link. Each radio link uses up to four times more device power when idle.  Each time the weather is accessed, the link is kept alive for approximately eight seconds; this can be configured from network to network but is a good indicator of the length of time it is in use.

Surprisingly, the background weather app accessed its information on average once every five minutes, consuming eight seconds of network access per retrieval – or roughly 38 minutes in a 24 hour period.

Although there is not a large amount of data being transferred each time, the frequency and time consumed per day should cause users and data service providers to take note. Weather apps are updating every five minutes or so and although the data transferred is small, the effect on the device and network is significant. As radio access looks to maximize utilization by increasing the number of subscribers connected, data service providers have to consider the impact of allowing simple applications to be enabled for background access.

A network of a large number of devices behaving the same way will have a significant effect in the long term capacity planning for networks. Additionally, the benefits of manipulating transmission links for maximum utilization of radio resources will be undone by background access checking for updates. This problem will only grow more acute with the increase in machine-to-machine (M2M) traffic and IPV6. In enterprise environments we have been involved with, this can be greater than 50 percent of the devices registered.

The group effect of multiple applications in the background acting individually to get their update creates problems for the device battery, user consumption and network costs. In their ‘Coalesced Updates’, iOS7 has a feature of multiple applications sharing connectivity in an attempt to preserve battery life. This feature may encourage more cooperative behavior among applications resident on the device to make optimal use of scarce resources, which is in the interest of the user and data service provider to encourage.

Analyzing data from the perspective of applications being used is a key asset for network management, and will become increasingly important from a capacity perspective (not just in data volume but in signaling, battery life and user experience).

ackground applications and the exchange of information is very surprising – a recent study by leading U.K. news organization, Channel 4,] found that a smartphone, with average settings and popular applications, will contact as many as 315 servers per day and even when it was perceived by its user to be in ‘idle’, contacted 76 servers per day. In other words, the smartphone is always active. Here is a map of the locations the phone contacted when in its idle state.

The study raises public awareness of key issues, such as privacy and access to information. At a fundamental level it has the practical implications on battery life, the network (as resources are consumed on the radio side) as well as an impact to the user’s wallet as these costs are passed on.

The cumulative effect on the battery is a source of concern. In the isolated case of weather, it may not seem to be such a big deal. If we assume that when a device is online that 1 percent of the battery time equates to 5 minutes of online time, then we can see that 1 day of weather will consume 7 percent of the battery life. The accumulated effect of multiple background applications independently accessing the network is important to consider; the average number of applications users have downloaded is over 30. If some of these have background activity, then the accumulated effect of background activity (advertisement refresh, location, weather update etc.) becomes much more significant.

The article was published by Mobile Market Portal on Sept 25. Click here to read more!