Scope
The paper focuses on the energy consumption of video streaming devices and proposes a framework for automating energy measurements within video streaming setups. It highlights the omnipresence of streaming devices in modern life and their significant energy usage, necessitating a closer examination of energy consumption patterns. The framework categorises attributes influencing energy usage, such as content, device, and network characteristics. It outlines a controlled, synchronised, and automated testing environment for efficient energy measurements. Key components include Player Worker and Measurement Worker modules, along with integration with an energy monitoring device. Automation and synchronisation ensure consistent and reliable data collection across various scenarios. The goal is to optimise energy efficiency in video streaming, addressing environmental concerns while maintaining streaming quality.
Summary
The paper’s main aim is to understand the energy consumption patterns of video streaming devices, particularly focusing on end-user devices. It considers factors such as content, device, and network attributes that influence energy usage.
Measurement Framework: A framework for automated energy measurements in video streaming setups is proposed. This framework includes components for measuring energy-related metrics such as voltage, current, and energy consumption, as well as integrating with an energy monitoring device.
Automation and Synchronisation: Automation is a key aspect of the proposed framework, enabling efficient and repeatable energy measurements across different devices and scenarios. Synchronisation ensures an accurate correlation between real-time streaming data and measured energy values.
Optimisation Strategies: By understanding energy consumption patterns, the paper aims to identify strategies for optimising energy efficiency in video streaming setups. This includes potentially adjusting streaming parameters such as bit rate, resolution, and frame rate to minimise energy usage.
Environmental Impact: The goal is to reduce the environmental footprint associated with video streaming by optimising energy efficiency.
Overall, the paper encompasses understanding, measuring, and optimising the energy consumption of video streaming devices, with the ultimate goal of reducing environmental impact while maintaining the quality of streaming experiences.
Framework Overview: The measuring system is presented as part of a broader framework for automated energy measurements in video streaming setups. This framework is designed to facilitate efficient and repeatable measurement of energy consumption across different devices and scenarios.
The measuring system comprises several components:
Player Worker: Responsible for executing streaming sessions and reporting metrics related to video playback, such as quality adjustments.
Measurement Worker: Performs sequential actions similar to the Player Worker but focuses on measuring energy-related metrics, such as voltage, current, and energy consumption.
Network-connected Computing Unit: Paired with an energy monitoring device, this unit measures the device under test’s energy consumption during streaming sessions.
Both the Player Worker and Measurement Worker connect to components in the service plane of the framework through HTTP REST interfaces. This integration allows for coordinated execution of streaming sessions and energy measurements. The measuring system operates within a controlled testing environment, using well-defined streams and devices in known network and environmental conditions. This ensures that measurements are conducted under consistent and reproducible conditions. Automation is a key aspect of the measuring system, enabling parallel measurements on different devices and repeatable tests on the same device. This automation improves efficiency and ensures that measurements are conducted systematically.


Relevance for EXIGENCE
EXIGENCE Relevance for energy measurements, resource interconnectivity and orchestration focusing on the network side, a worker that monitors the energy behaviour could be of interest, but in this paper the focus is the end user and the metrics regarding video playback and power consumption.