liu.seSearch for publications in DiVA
Change search
ReferencesLink to record
Permanent link

Direct link
Optimizing Communication Energy Efficiency for a Multimedia Application
Linköping University, Department of Computer and Information Science, Software and Systems.
2016 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
Abstract [en]

Mobile devices have evolved rapidly in recent years and increased usage and performance are pushing contemporary battery technology to its limits. The constrained battery resources mean that the importance of energy-efficient application design is growing and in this regard wireless network accesses are a major contributor to a mobile device's overall energy consumption. Additionally, the energy consumption characteristics of modern cellular technologies make small volumes of poorly scheduled traffic account for a substantial share of a device's total energy consumption. However, quantifying the communication energy footprint is cumbersome, making it difficult for developers to profile applications from an energy consumption perspective and optimize traffic patterns.

This thesis examines the traffic patterns of the Android client of the popular multimedia streaming service Spotify with the intention to reduce its energy footprint, in terms of 3G energy consumption. The application's automated test environment is extended to capture network traffic, which is used to estimate energy consumption. Automated system tests are designed and executed on a physical Android device connected to a 3G network, shedding light on the traffic patterns of different application features.

All traffic between the Spotify client application and the backend servers is encrypted. To extract information about the traffic, the application code is instrumented to output supplementary information to the Android system log. The system log is then used as a source of information to attribute data traffic to different application modules and specific lines of code.

Two simple traffic shaping techniques, traffic aggregation and piggybacking, are implemented in the application to provide more energy-efficient traffic patterns. As a result, 3G energy consumption during normal music playback is reduced by 22-54%, and a more contrived scenario achieves a 60% reduction. The reductions are attained by rescheduling a small class of messages, most notably data tracking application usage. These messages were found to account for a small fraction of total traffic volume, but a large portion of the application's overall 3G energy consumption.

Place, publisher, year, edition, pages
2016. , 60 p.
Keyword [en]
3G, energy consumption, energy efficiency, traffic patterns, software testing, test automation, Spotify
National Category
Computer Science
URN: urn:nbn:se:liu:diva-125789ISRN: LIU-IDA/LITH-EX-A--16/007--SEOAI: diva2:910711
External cooperation
Spotify AB
Subject / course
Computer Engineering
2016-03-23, Alan Turing, Linköpings universitet, Linköping, 13:15 (English)
Available from: 2016-03-11 Created: 2016-03-04 Last updated: 2016-03-15Bibliographically approved

Open Access in DiVA

Optimizing Communication Energy Efficiency for a Multimedia Application(921 kB)109 downloads
File information
File name FULLTEXT01.pdfFile size 921 kBChecksum SHA-512
Type fulltextMimetype application/pdf

Search in DiVA

By author/editor
Green Olander, Jens
By organisation
Software and Systems
Computer Science

Search outside of DiVA

GoogleGoogle Scholar
Total: 109 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Total: 2092 hits
ReferencesLink to record
Permanent link

Direct link