Abstract:
Unmanned Aerial Vehicle (UAV) applications and services have gained a huge deployment and adoption in different fields, such as the military domain (Defense or reconnaiss...Show MoreMetadata
Abstract:
Unmanned Aerial Vehicle (UAV) applications and services have gained a huge deployment and adoption in different fields, such as the military domain (Defense or reconnaissance) and the civilian domain (Healthcare, surveillance, and transport). UAV operations are generally critical and require, during operations, a control link with the drones, which should be reliable with very low latency. To ensure low-latency, 5G architecture intends to deploy Mobile Edge Computing (MEC) servers, which provide cloud computing capabilities close to the end-users. Consequently, it is envisioned that the AutoPilot application will be deployed at the MEC in order to ensure a low latency connection to the drones. However, the high mobility of drones makes the migration of the AutoPilot applications among MEC servers unavoidable; in order to maintain a low latency connection with the flying drones. This may lead to frequent downtime of the service, which may impact the AutoPilot performances, and hence service migrations should be limited as much as possible. Accordingly, this paper aims to reduce the number of service migrations of drones by introducing novel algorithms that act at the mission planning phase, where the path of the drones is defined.
Date of Conference: 07-11 December 2020
Date Added to IEEE Xplore: 25 January 2021
ISBN Information: