The considerable surge in energy consumption within data centers can be attributed to the exponential rise in demand for complex computing workflows and storage resources. Video streaming applications are both compute and storage-intensive and account for the majority of today's internet services. In this work, we designed a video encoding application consisting of codec, bitrate, and resolution set for encoding a video segment. Then, we propose VE-Match, a matching-based method to schedule video encoding applications on both Cloud and Edge resources to optimize costs and energy consumption. Evaluation results on a real computing testbed federated between Amazon Web Services (AWS) EC2 Cloud instances and the Alpen-Adria University (AAU) Edge server reveal that VE-Match achieves lower costs by 17%-78% in the cost-optimized scenarios compared to the energy-optimized and tradeoff between cost and energy. Moreover, VE-Match improves the video encoding energy consumption by 38%-45% and gCO₂ emission by up to 80% in the energy-optimized scenarios compared to the cost-optimized and tradeoff between cost and energy.

This paper highlights the current trends and the necessary research directions for achieving carbon-effective streaming services. Collaboration between the end-to-end delivery pipeline vendors is encouraged to design standards and technologies.

Video streaming is a widely used and energy-demanding online service, which contributes to CO2 emissions and environmental issues. In this paper, we investigate the technological feasibility and benefits of green streaming technologies, which aim to optimize the energy efficiency and carbon footprint of streaming content across the whole supply chain. Our work focuses on three key technologies: context-aware encoding, green media players, and energy-aware content steering. We present experiments and simulations to evaluate the performance and impact of these technologies. We also explore their economic viability and potential for creating new business opportunities in the streaming industry. Our work provides a holistic evaluation of green streaming technologies, which can support the global efforts for climate action and environmental protection.

Current OTT linear deployments are rather static infrastructures. The setup can be altered occasionally by an operator manipulation, based on anticipated audience or events. In this article, several real time optimizations are presented to decrease the traffic and hence to achieve energy and costs savings. Finally, the potential of central orchestration between encoding and CDN is outlined.

For adaptive streaming applications, low-complexity and accurate video complexity features are necessary to analyze the video content in real time, which ensures fast and compression-efficient video streaming without disruptions. State-of-the-art video complexity features are Spatial Information (SI) and Temporal Information (TI) features which do not correlate well with the encoding parameters in adaptive streaming applications. To this light, Video Complexity Analyzer (VCA) was introduced, determining the features based on Discrete Cosine Transform (DCT)-energy. This paper presents optimizations on VCA for faster and energy-efficient video complexity analysis. Experimental results show that VCA v2.0, using eight CPU threads, Single Instruction Multiple Data (SIMD), and low-pass DCT optimization, determines seven complexity features of Ultra High Definition 8-bit videos with better accuracy at a speed of up to 292.68 fps and an energy consumption of 97.06% lower than the reference SITI implementation.

In this paper, we discuss the evolution in energy efficiency for software-based encoding on general-purpose processors, along with recent trends that enable increasingly power-efficient video encoding. We provide an overview of different contributing factors that have led to a massive increase in software-based performance, along with innovations that are leading to further efficiency improvements.

In this paper, we study decoding energy reduction opportunities using temporal-domain filtering and subsampling methods. In particular, we study spatiotemporal filtering using a contrast sensitivity function and temporal downscaling, i.e., frame rate reduction. We apply these concepts as a pre-filtering to the video before compression and evaluate the bitrate, the decoding energy, and the visual quality with a dedicated metric targeting temporally down-scaled sequences. We find that decoding energy savings yield 35% when halving the frame rate and that spatiotemporal filtering can lead to up to 5% of additional savings, depending on the content.

In live video streaming, the size of Dynamic Adaptive Streaming over HTTP (DASH) manifest grows as the number of periods increases and/or the overall time duration of DASH manifest increases. The bigger the manifest size in bytes is, the more computation for manifest generation, manifest storage and parsing and network traffic there will be on service side, the more data to be downloaded, manifest refresh latency, manifest parsing and storage in memory there will be on device side. In this paper, we analyze the techniques and algorithms available to optimize and reduce DASH manifest size with their pros and cons, and limitations of each technique in different scenarios, and propose further optimizations and the adoption of technique(s) in each video streaming scenario according to product requirements, operational cost, system complexity and the requirement of quality of video playback experiences.