Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

NeuSaver: Neural Adaptive Power Consumption Optimization for Mobile Video Streaming

Kyoungjun Park, Myungchul Kim, Laihyuk Park

TL;DR

Experiments showed that NeuSaver effectively reduces the power consumption of mobile devices when streaming video by an average of 16.14% and up to 23.12% while maintaining high QoE.

Abstract

Video streaming services strive to support high-quality videos at higher resolutions and frame rates to improve the quality of experience (QoE). However, high-quality videos consume considerable amounts of energy on mobile devices. This paper proposes NeuSaver, which reduces the power consumption of mobile devices when streaming videos by applying an adaptive frame rate to each video chunk without compromising user experience. NeuSaver generates an optimal policy that determines the appropriate frame rate for each video chunk using reinforcement learning (RL). The RL model automatically learns the policy that maximizes the QoE goals based on previous observations. NeuSaver also uses an asynchronous advantage actor-critic algorithm to reinforce the RL model quickly and robustly. Streaming servers that support NeuSaver preprocesses videos into segments with various frame rates, which is similar to the process of creating videos with multiple bit rates in dynamic adaptive streaming over HTTP. NeuSaver utilizes the commonly used H.264 video codec. We evaluated NeuSaver in various experiments and a user study through four video categories along with the state-of-the-art model. Our experiments showed that NeuSaver effectively reduces the power consumption of mobile devices when streaming video by an average of 16.14% and up to 23.12% while achieving high QoE.

NeuSaver: Neural Adaptive Power Consumption Optimization for Mobile Video Streaming

TL;DR

Experiments showed that NeuSaver effectively reduces the power consumption of mobile devices when streaming video by an average of 16.14% and up to 23.12% while maintaining high QoE.

Abstract

Video streaming services strive to support high-quality videos at higher resolutions and frame rates to improve the quality of experience (QoE). However, high-quality videos consume considerable amounts of energy on mobile devices. This paper proposes NeuSaver, which reduces the power consumption of mobile devices when streaming videos by applying an adaptive frame rate to each video chunk without compromising user experience. NeuSaver generates an optimal policy that determines the appropriate frame rate for each video chunk using reinforcement learning (RL). The RL model automatically learns the policy that maximizes the QoE goals based on previous observations. NeuSaver also uses an asynchronous advantage actor-critic algorithm to reinforce the RL model quickly and robustly. Streaming servers that support NeuSaver preprocesses videos into segments with various frame rates, which is similar to the process of creating videos with multiple bit rates in dynamic adaptive streaming over HTTP. NeuSaver utilizes the commonly used H.264 video codec. We evaluated NeuSaver in various experiments and a user study through four video categories along with the state-of-the-art model. Our experiments showed that NeuSaver effectively reduces the power consumption of mobile devices when streaming video by an average of 16.14% and up to 23.12% while achieving high QoE.

Paper Structure

This paper contains 25 sections, 7 equations, 8 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Motivation and Background
  3. Similarity Trend Analysis in Videos
  4. Perceptual Similarity Method
  5. Overview
  6. RL Model
  7. Training Methodology
  8. A3C Training Algorithm Basics
  9. Reward Function Design
  10. Support for Various Situations
  11. Implementation
  12. Evaluation
  13. Methodology
  14. Perceived Quality vs. Frame Rate
  15. Energy Saving
  16. ...and 10 more sections

Figures (8)

  • Figure 1: Variation of average SSIM between video chunks for three video types. The length of each video chunk is 2 seconds.
  • Figure 2: Correlation between Y-Diff and SSIM of three different types of videos (Pearson correlation coefficient: -0.8162).
  • Figure 3: Overview of adaptive video streaming with NeuSaver.
  • Figure 4: Overall architecture of learning agent that generates the AFR policy based on observations.
  • Figure 5: NeuSaver's parallel training design with 1 central agent and 16 learning agents.
  • ...and 3 more figures