H.264 Video Encoding-based Edge-assisted Mobile AR Systems: Network and Energy Issues
Anik Mallik, Jiang Linda Xie · 2022
Edge-assisted mobile augmented reality (Edge-MAR) systems have emerged as effective ways to support computation-intensive and latency-sensitive applications for mobile devices due to the offloading capability of heavy computational burdens. However, the network- and energy-resource utilization of such systems is high. Video encoding schemes like H.264 can help Edge-MAR systems reduce latency and bandwidth utilization but at the cost of increased energy consumption. In this paper, we present a comprehensive study of Edge-MAR using H.264 video encoding with a focus on network condition, resource utilization, detection accuracy, and energy consumption of various mobile devices. We collect latency, energy, transmitted data size, and accuracy data for each segment of an object detection pipeline measured through experiments with testbeds, and analyze the non-linear behaviors of Edge-MAR. Following this, we demonstrate the challenges associated with the experiments conducted to test the system as well as the ways to overcome them. Finally, we propose regression-based models to analytically compute different Edge-MAR parameters to achieve desired outcomes. This extensive study provides essential guidelines to network- and energy-aware H.264 video encoding-based Edge-MAR system design.