Energy Optimization for Latency- and Qaulity-Constrained Video Applications

This paper proposes an energy optimization technique for latency and quality constrained video applications. It consists of two key techniques: frame-skipping technique and buffering technique. While buffering increases the slack time utilization at the OS level, frame skipping increases the slack time itself at the application level, and both enhance the effectiveness of the DVS technique. We use an H.263 encoder application as a test vehicle to which the proposed technique is applied. Experiments demonstrate that the proposed technique achieves noticeable energy reduction satisfying the given latency and video quality constraints.National Research Laboratory Program No. M1- 0104-00-0015 and the Brain Korea 21 Project supported this work. The Institute of Computer Technology at Seoul National University provided research facilities