Modeling the Energy Consumption of the HEVC Decoding Process

TL;DR

This paper introduces an improved energy model for HEVC decoding that explicitly includes inloop filters and demonstrates superior accuracy across various software and hardware systems, with less than 7% and 15% error respectively.

Contribution

It extends existing models by explicitly modeling inloop filters and provides a comprehensive evaluation framework for energy estimation accuracy.

Findings

Model achieves less than 7% mean error on software systems.

Model achieves less than 15% mean error on hardware systems.

Outperforms seven existing energy models from literature.

Abstract

In this paper, we present a bit stream feature based energy model that accurately estimates the energy required to decode a given HEVC-coded bit stream. Therefore, we take a model from literature and extend it by explicitly modeling the inloop filters, which was not done before. Furthermore, to prove its superior estimation performance, it is compared to seven different energy models from literature. By using a unified evaluation framework we show how accurately the required decoding energy for different decoding systems can be approximated. We give thorough explanations on the model parameters and explain how the model variables are derived. To show the modeling capabilities in general, we test the estimation performance for different decoding software and hardware solutions, where we find that the proposed model outperforms the models from literature by reaching frame-wise mean estimation errors of less than 7% for software and less than 15% for hardware based systems.