This blog is by Gerald Reddig from Nokia Networks.
Subscribers love the convenience of listening to music on their mobile devices. Yet as more customers adopt “music streaming” services, operators are increasingly asking: What is the impact of streaming music on network performance and on device battery life?
The team at Nokia Smart Labs explored this question by testing two popular music streaming applications, Spotify and Pandora, on the two most popular mobile operating system platforms. All tests were done in a controlled, end-to-end LTE and WCDMA lab environment. The team used the standby and busy profiles defined by Smart Labs to emulate real-life usage and behavior for the applications.
Test results show application differences
As you might expect, the tests found that different music streaming applications have different characteristics. What may be surprising, however, is that the same application can have different network impacts based on the device platform.
Data traffic, data connections
• Spotify consumes more than double the amount of data (measured in megabytes) compared to Pandora when the user is actively listening to music. For users with a 1 GB per month data plan, this difference means they could listen to approximately 28 hours of music with Spotify and 54 hours with Pandora.
• Pandora used up to 6 times more concurrent data connections compared to Spotify in one of the operating system platforms tested.
• In standby mode, Spotify used a frequent keep-alive mechanism in one of the device platforms, which resulted in a very high signaling load of up to 7 times more than Pandora.
• In busy mode while playing songs, both applications performed similarly. However, because the data was delivered in bursts to the user, the music stream caused a relatively high signaling load.
• In one of the device platforms tested, Spotify had more than double the power consumption compared to Pandora when in standby mode. This higher level is due to the frequent keep-alive mechanism used by Spotify.
• During active listening, both applications consumed similar levels of battery when running on the same device. But in one device platform, power consumption was significantly higher.
What these results mean for mobile networks
The impact of streaming music on signaling load is the most significant concern for mobile operators because of the danger of overwhelming the radio access network (RAN). Unlike data consumption plans, there is no metering of the signal load consumed by subscribers, so they have no incentive to choose applications with lower signaling demands.
To address this challenge, operators should optimize their network parameters to accommodate the signaling loads of streaming music so that no problems arise in other applications and so that user experience isn’t impacted.
Additionally, application developers can help reduce battery power consumption by limiting the frequency of keep-alive messages from streaming music applications.
How to put Smart Labs to work for you
Are you keen to learn how certain applications and smartphone behavior might impact your network? Smart Labs consults with mobile network operators and application developers to help guard against network outages caused by overloads. Our labs are fully equipped to assist your testing with end-to-end LTE and 3G technologies and are located in America, Europe and Asia. Check out our services on the Smart Labs web page or contact us.
To read more on this topic, please see Gerald’s blogs listed below under “Related posts”: “Smart Labs looks at smartphone energy consumption for video streaming” and “New Smart Labs Online Service: performance benchmarking at your fingertips”.
To share your thoughts on the topic, join the discussion with @NSNtweets using #NetworksPerform #LTE #SmartLabs #Signaling.