Data collection based on Q-learning method in terrestrial networks assisted by UAV

Abstract

Nowadays, Unmanned Aerial Vehicle (UAVs) are playing an integral and sustainable function in many verticals touching exclusive components of our lives; such as civil, public and military applications. The objective is to appoint a self-trained UAV as a flying cell unit gathering data from ground sensor nodes fairly distributed in a given geographical area through a predefined period of time. In this approach, Q-learning (QL) algorithm is employed to train the UAV to learn the environment and provide appropriate scheduling to accomplish its data collection mission while minimizing the data collection time. As a matter of fact, collecting information from sensors may face some noticeable challenges. However, due to use the flexibility of UAVs, collecting information would be much easier. In this thesis, we tried to figure out about UAVs that support data collection over WSN (wireless sensor network), while maximizing the amount of collected data and minimizing the flight duration of time.