A Measurement Study on Edge Computing for Autonomous UAVs

Callegaro, Davide; Baidya, Sabur; Levorato, Marco

doi:10.1145/3341568.3342109

Cited by 17 publications

(7 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thanks to our 'ready-to-fly' framework design, the trained DRL agents can be easily instantiated on real hardware, a procedure that is straightforward and only involves enabling the hardware drivers in the DRL DPPS. We will carefully analyze the performance of future experimental fieldings to identify discrepancies with the simulation environment [55] and employ techniques of transfer learning to further reduce the emulationexperimental performance gap. On-line learning: We will also focus on how to optimize the performance of the distributed DRL agents in the case of compromised nodes and outages.…”

Section: Discussionmentioning

confidence: 99%

Learning to Fly: A Distributed Deep Reinforcement Learning Framework for Software-Defined UAV Network Control

Cheng

Bertizzolo

Buczek

et al. 2021

IEEE Open J. Commun. Soc.

View full text Add to dashboard Cite

Control and performance optimization of wireless networks of Unmanned Aerial Vehicles (UAVs) require scalable approaches that go beyond architectures based on centralized network controllers. At the same time, the performance of model-based optimization approaches is often limited by the accuracy of the approximations and relaxations necessary to solve the UAV network control problem through convex optimization or similar techniques, and by the accuracy of the channel network models used. To address these challenges, this article introduces a new architectural framework to control and optimize UAV networks based on Deep Reinforcement Learning (DRL). Furthermore, it proposes a virtualized, 'ready-to-fly' emulation environment to generate the extensive wireless data traces necessary to train DRL algorithms, which are notoriously hard to generate and collect on battery-powered UAV networks.The training environment integrates previously developed wireless protocol stacks for UAVs into the CORE/EMANE emulation tool. Our 'ready-to-fly' virtual environment guarantees scalable collection of high-fidelity wireless traces that can be used to train DRL agents. The proposed DRL architecture enables distributed data-driven optimization (with up to 3.7x throughput improvement and 0.2x latency reduction in reported experiments), facilitates network reconfiguration, and provides a scalable solution for large UAV networks.

show abstract

Section: Discussionmentioning

confidence: 99%

Learning to Fly: A Distributed Deep Reinforcement Learning Framework for Software-Defined UAV Network Control

Cheng

Bertizzolo

Buczek

et al. 2021

IEEE Open J. Commun. Soc.

View full text Add to dashboard Cite

show abstract

“…In order to solve it, researchers and industrial propose computation offloading from UAVs to MEC servers. The offloading of computation has become very popular, [21][22][23][24][25][26] because it allows drones to consume less energy. 23,27 Callegaro et al, 24 survey the MEC performance when UAVs offload compute-intense tasks.…”

Section: Existing Efforts On Uavmentioning

confidence: 99%

“…The offloading of computation has become very popular, [21][22][23][24][25][26] because it allows drones to consume less energy. 23,27 Callegaro et al, 24 survey the MEC performance when UAVs offload compute-intense tasks. Messous et al, 25 propose an optimal UAV offloading strategy using a game theory approach.…”

Section: Existing Efforts On Uavmentioning

confidence: 99%

6G‐edge support of Internet of Autonomous Vehicles: A survey

Ibn‐Khedher,

Laroui,

Alfaqawi

et al. 2023

Trans Emerging Tel Tech

View full text Add to dashboard Cite

With the commercial deployment of 5G mobile communication systems, 6G is proposed to meet the needs of new use cases including connected and autonomous vehicles (CAVs). In this article, we surveyed different 6G embracing technologies such as mobile edge computing, unmanned aerial vehicles, network function virtualization, software defined networking, and artificial intelligence that can enhance the Internet of Autonomous Vehicles (IoAV) network performance comparing to existing 5G enabled solutions. Then, we highlight the convergence path of 6G and IoAV to satisfy the next generation of CAV applications. Further, we propose 6G cellular communication to support the quality of next IoAV applications. Furthermore, the article attempts to study the serious challenges in the 6G‐IoAV research field and propose potential future directions.

show abstract

“…Regarding experimental work, different authors have analysed the performance of latency‐aware applications 6,25‐27 . Motlagh et al evaluate the processing time and energy consumption of a facial recognition service based on a video stream captured by a drone, by comparing onboard computing with computational offloading to the network edge 6 .…”

Section: Related Workmentioning

confidence: 99%

Is the edge really necessary for drone computing offloading? An experimental assessment in carrier‐grade 5G operator networks

et al. 2022

View full text Add to dashboard Cite

In this article, we evaluate the first experience of computation offloading from drones to real fifth‐generation (5G) operator systems, including commercial and private carrier‐grade 5G networks. A follow‐me drone service was implemented as a representative testbed of remote video analytics. In this application, an image of a person from a drone camera is processed at the edge, and image tracking displacements are translated into positioning commands that are sent back to the drone, so that the drone keeps the camera focused on the person at all times. The application is characterised to identify the processing and communication contributions to service delay. Then, we evaluate the latency of the application in a real non standalone 5G operator network, a standalone carrier‐grade 5G private network, and, to compare these results with previous research, a Wi‐Fi wireless local area network. We considered both multi‐access edge computing (MEC) and cloud offloading scenarios. Onboard computing was also evaluated to assess the trade‐offs with task offloading. The results determine the network configurations that are feasible for the follow‐me application use case depending on the mobility of the end user, and to what extent MEC is advantageous over a state‐of‐the‐art cloud service.

show abstract

A Measurement Study on Edge Computing for Autonomous UAVs

Cited by 17 publications

References 9 publications

Learning to Fly: A Distributed Deep Reinforcement Learning Framework for Software-Defined UAV Network Control

Learning to Fly: A Distributed Deep Reinforcement Learning Framework for Software-Defined UAV Network Control

6G‐edge support of Internet of Autonomous Vehicles: A survey

Is the edge really necessary for drone computing offloading? An experimental assessment in carrier‐grade 5G operator networks

Contact Info

Product

Resources

About