FAMIUM Green: An Automated Hardware-in-the-Loop Framework for Energy Characterization of Video Streaming Devices

The rapid growth of video streaming, which now accounts for over 70% of global internet traffic, has intensified scrutiny of the energy footprint of end-user streaming devices. This paper presents FAMIUM Green, a fully automated hardware-in-the-loop measurement framework designed for reproducible energy characterization of heterogeneous streaming devices, including smart TVs, streaming boxes, and mobile devices. The framework synchronizes hardware-level power acquisition with streaming session metadata via MPEG-SAND protocol events, allowing power changes to be linked to specific content and streaming parameters. We evaluated 14 smart TVs spanning three panel technologies (OLED, QLED-LCD, and LED-LCD) across 746 measurement sessions. Purpose-built reference content sequences, including grayscale luma ramps, TimeScapes luma ramps, and random pixels, were used to isolate the effects of content luminance, decoding complexity, and scene dynamics on device energy consumption. Reproducibility validation across N=5 independent sessions yielded a mean inter-session coefficient of variation (CV) below 0.3% for all tested panel types. Empirical measurements across OLED, QLED, and LED display technologies demonstrate that content luminance is the dominant driver of power consumption on emissive OLED panels (monotonically increasing from ∼55 W to ∼180 W across the luma ramp), while LED-backlit panels show near-constant power draw independent of content brightness. Comparative analysis confirms that the display subsystem constitutes over 98% of total system power (∼180 W for the display versus ∼3 W for the streaming box). Pearson correlation analysis reveals that streaming parameters, including bitrate, resolution, and frame rate, have a statistically negligible influence on total device power (|r|<0.20), whereas eco mode is the most effective controllable factor for reducing energy consumption on LED-LCD and QLED-LCD panels (r=−0.636 to −0.783). The effect is weaker on OLED panels (r=−0.339), where energy savings are bounded by the content luminance. These findings suggest that adjusting display power settings can save much more energy than tuning streaming bitrates or codec selection.

Authors

Institutions

Publication Details

Journal
Computers
Published
2026-09-25
DOI
https://doi.org/10.3390/computers15100653
Primary Topic
Green IT and Sustainability
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
article

FAMIUM Green: An Automated Hardware-in-the-Loop Framework for Energy Characterization of Video Streaming Devices

Robert Seeliger, Görkem Güçlü, Stefan Arbanowski, Stephan Steglich et al.
Computers
Green IT and Sustainability
article

FAMIUM Green: An Automated Hardware-in-the-Loop Framework for Energy Characterization of Video Streaming Devices

Robert Seeliger, Görkem Güçlü, Stefan Arbanowski, Stephan Steglich, Minh Nguyen, Martin Lasak, Moustafa Ghaddar
article en

Abstract

The rapid growth of video streaming, which now accounts for over 70% of global internet traffic, has intensified scrutiny of the energy footprint of end-user streaming devices. This paper presents FAMIUM Green, a fully automated hardware-in-the-loop measurement framework designed for reproducible energy characterization of heterogeneous streaming devices, including smart TVs, streaming boxes, and mobile devices. The framework synchronizes hardware-level power acquisition with streaming session metadata via MPEG-SAND protocol events, allowing power changes to be linked to specific content and streaming parameters. We evaluated 14 smart TVs spanning three panel technologies (OLED, QLED-LCD, and LED-LCD) across 746 measurement sessions. Purpose-built reference content sequences, including grayscale luma ramps, TimeScapes luma ramps, and random pixels, were used to isolate the effects of content luminance, decoding complexity, and scene dynamics on device energy consumption. Reproducibility validation across N=5 independent sessions yielded a mean inter-session coefficient of variation (CV) below 0.3% for all tested panel types. Empirical measurements across OLED, QLED, and LED display technologies demonstrate that content luminance is the dominant driver of power consumption on emissive OLED panels (monotonically increasing from ∼55 W to ∼180 W across the luma ramp), while LED-backlit panels show near-constant power draw independent of content brightness. Comparative analysis confirms that the display subsystem constitutes over 98% of total system power (∼180 W for the display versus ∼3 W for the streaming box). Pearson correlation analysis reveals that streaming parameters, including bitrate, resolution, and frame rate, have a statistically negligible influence on total device power (|r|<0.20), whereas eco mode is the most effective controllable factor for reducing energy consumption on LED-LCD and QLED-LCD panels (r=−0.636 to −0.783). The effect is weaker on OLED panels (r=−0.339), where energy savings are bounded by the content luminance. These findings suggest that adjusting display power settings can save much more energy than tuning streaming bitrates or codec selection.

ComputersVol. 15(10)
Fraunhofer Institute for Open Communication Systems (DE)
Affordable and clean energy
Openalex Percentile: Top 21%
Green IT and Sustainability
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.