UAVs as an Intelligent Service: Boosting Edge Intelligence for Air-Ground Integrated Networks

Chao Dong,Yun Shen,Yuben Qu,Kun Wang,Jianchao Zheng,Qihui Wu,Fan Wu

DOI: https://doi.org/10.1109/mnet.011.2000651

IF: 10.294

2021-07-01

IEEE Network

Abstract:The air-ground integrated network is a key component of future sixth generation (6G) networks to support seamless and near-instant super-connectivity. There is a pressing need to intelligently provision various services in 6G networks, which is challenging. To meet this need, in this article, we propose a novel architecture called UaalS, that is, unmanned aerial vehicles (UAVs) as a intelligent service for the air-ground integrated network, featuring the UAV as a key enabler to boost edge intelligence with the help of machine learning (ML) techniques. We envision that the proposed UaalS architecture could intelligently provision wireless communication service, edge computing service, and edge caching service by a network of UAVs, making full use of UAVs' flexible deployment and diverse ML techniques. We also conduct a case study where UAVs participate in the model training of distributed ML among multiple terrestrial users, whose result shows that the model training is efficient with low energy consumption of UAVs. Finally, we discuss the challenges and some open research issues in UaaIS.

computer science, information systems,telecommunications,engineering, electrical & electronic, hardware & architecture

What problem does this paper attempt to address?

This paper primarily addresses the need for seamless, near-instantaneous hyper-connectivity in future sixth-generation (6G) networks by proposing a novel architecture—namely, Unmanned Aerial Vehicles as Intelligent Services (UaaIS)—to solve the problem of providing various intelligent services in integrated air-ground networks.

Specifically, the paper attempts to address the following issues:

1. **Dynamic Service Demand**: With the increasing requirements of 6G networks for global three-dimensional connectivity, massive device connectivity (10^7 devices per square kilometer), user experience data rates (≥1 Tb/s), and ultra-low latency (10-100 milliseconds), there is a need to flexibly respond to the spatiotemporal dynamic changes in communication, computing, and caching service demands.

2. **Large-Scale Complex Decision-Making and Resource Management**: How to effectively handle large-scale connection decision-making and resource allocation issues, such as access and handover decisions, spectrum, computing, and storage resource allocation, which are usually large-scale and complex.

3. **Pervasive Intelligence Needs in the Network**: 6G networks are expected to have built-in intelligent functions, thus requiring solutions for how mobile users can obtain powerful computing capabilities and how UAVs can participate in the training of distributed machine learning models.

To address the above challenges, the paper proposes the UaaIS architecture, which enhances edge intelligence by utilizing UAVs as key components. This architecture envisions UAVs as key providers of wireless communication services, edge computing services, and edge caching services, all of which are intelligent through machine learning technologies. Additionally, the paper discusses the results of case studies, showing that when UAVs participate in distributed machine learning model training, the training efficiency is high and the UAV energy consumption is low. Finally, the paper explores the challenges faced by UaaIS and some open research questions.