Framework for automated energy measurement of video streaming devices

Video streaming devices are globally omnipresent in our daily lives. From small mobile devices to large TV sets, streaming devices surrounding us offer an indispensable access to visual content but at the same time they account for a large proportion of energy consumption. On our journey to analyze and understand the impact of each individual part in the video streaming supply chain we initially focus on the question how to perform energy measurements on end devices efficiently and repeatable. We define a comprehensive set of attributes describing a holistic environmental setting of common streaming setups and present a framework enabling the automated execution of energy measurements of video streams. Our framework utilizes server and network assisted DASH (SAND), as used in video streaming to assess the quality of streamed videos based on acquired metrics, like the switching between different resolutions. We extend the concept by adding new metrics to track energy measurement related data enabling coherent data sets for further analysis and model training.