Division of Computer Science and Engineering, Sun Moon University, Asan 31460, Korea.
Division of Computer and Information Engineering, Hoseo University, Asan 31499, Korea.
Sensors (Basel). 2018 Sep 6;18(9):2982. doi: 10.3390/s18092982.
Recently, various technologies for utilizing unmanned aerial vehicles have been studied. Drones are a kind of unmanned aerial vehicle. Drone-based mobile surveillance systems can be applied for various purposes such as object recognition or object tracking. In this paper, we propose a mobility-aware dynamic computation offloading scheme, which can be used for tracking and recognizing a moving object on the drone. The purpose of the proposed scheme is to reduce the time required for recognizing and tracking a moving target object. Reducing recognition and tracking time is a very important issue because it is a very time critical job. Our dynamic computation offloading scheme considers both the dwell time of the moving target object and the network failure rate to estimate the response time accurately. Based on the simulation results, our dynamic computation offloading scheme can reduce the response time required for tracking the moving target object efficiently.
最近，人们研究了各种利用无人机的技术。无人机就是一种无人飞行器。基于无人机的移动监控系统可应用于各种目的，例如目标识别或目标跟踪。在本文中，我们提出了一种感知移动性的动态计算卸载方案，可用于在无人机上跟踪和识别移动目标。该方案的目的是减少识别和跟踪移动目标所需的时间。减少识别和跟踪时间是一个非常重要的问题，因为这是一项非常紧迫的工作。我们的动态计算卸载方案同时考虑移动目标对象的驻留时间和网络故障率，以准确估计响应时间。基于仿真结果，我们的动态计算卸载方案可以有效地减少跟踪移动目标对象所需的响应时间。
