Energy-aware Path Planning for Obtaining Fresh Updates in UAV-IoT MEC systems
Hao Chen, Xiaoqi Qin, Yixuan Li, Nan Ma
Abstract
The ubiquitous computing resource at UAVs and IoT devices can be exploited in conjunction to form a UAV-IoT edge computing system for low-cost and responsive environmental monitoring. Under stochastic computational task arrival at IoT devices, one major challenge is how to realize path control for multiple UAVs and energy efficient computation offloading in real time. Moreover, the freshness of obtained updates is of critical importance to the system performance under such time-critical scenarios. In this paper, we employ the concept of age of information (AoI) to quantify the timeliness of updates at IoT devices, and formulate an energy minimization problem by jointly considering UAV path planning, energy consumption of computation offloading and age evolution of updates. To solve the formulated problem, we propose a deep reinforcement learning based solution to achieve fast decision making. Simulation results show that the performance of proposed solution is competitive in terms of obtaining fresh updates at low energy cost.