[Paper Review] Air Computing: A Survey on a New Generation Computation Paradigm in 6G Wireless Networks

Baris Yamansavascilar, Atay Özgövde+1 more|arXiv (Cornell University)|Sep 10, 2022

UAV Applications and Optimization113 references4 citations

TL;DR

This paper proposes Air Computing, a novel 6G-enabled, 3D computation paradigm integrating Low Altitude Platforms (LAP), High Altitude Platforms (HAP), and Low Earth Orbit (LEO) satellites with terrestrial edge and cloud resources to overcome bottlenecks in latency, capacity, and coverage. By leveraging 6G’s ultra-reliable, low-latency communication and AI-driven orchestration, Air Computing enables dynamic, intelligent task offloading and resource allocation across air and ground layers, significantly enhancing QoS and QoE for next-generation applications.

ABSTRACT

There is an ever-growing race between what novel applications demand from the infrastructure and what the continuous technological breakthroughs bring in. Especially after the proliferation of smart devices and diverse IoT requirements, we observe the dominance of cutting-edge applications with ever-increased user expectations in terms of mobility, pervasiveness, and real-time response. Over the years, to meet the requirements of those applications, cloud computing provides the necessary capacity for computation, while edge computing ensures low latency. However, these two essential solutions would be insufficient for the next-generation applications since computational and communicational bottlenecks are inevitable due to the highly dynamic load. Therefore, a 3D networking structure using different air layers including Low Altitude Platforms, High Altitude Platforms, and Low Earth Orbits in a harmonized manner for both urban and rural areas should be applied to satisfy the requirements of the dynamic environment. In this perspective, we put forward a novel, next-generation paradigm called Air Computing that presents a dynamic, responsive, and high-resolution computation and communication environment for all spectrum of applications using the 6G Wireless Networks as the fundamental communication system. In this survey, we define the components of air computing, investigate its architecture in detail, and discuss its essential use cases and the advantages it brings for next-generation application scenarios. We provide a detailed and technical overview of the benefits and challenges of air computing as a novel paradigm and spot the important future research directions.

Motivation & Objective

• Address the growing computational and communication bottlenecks in 5G and traditional edge/cloud computing due to the dynamic, heterogeneous demands of next-generation IoT and real-time applications.
• Propose a novel 3D computational paradigm—Air Computing—that unifies terrestrial edge computing with aerial platforms (UAVs, HAPs, LEO satellites) to achieve seamless, low-latency, and high-capacity service delivery.
• Investigate the architectural components, system integration, and operational mechanisms required for a fully functional Air Computing infrastructure in 6G networks.
• Identify key challenges such as energy efficiency, dynamic mobility, regulatory compliance, and secure data processing, and propose future research directions.

Proposed method

• Design a 3D network architecture integrating Low Altitude Platforms (LAP), High Altitude Platforms (HAP), and Low Earth Orbit (LEO) satellites with terrestrial edge and cloud servers to form a unified, dynamic computational fabric.
• Utilize 6G wireless networks as the foundational communication infrastructure, enabling ultra-reliable, low-latency (0.1 ms) and high-throughput (up to 1 Tbps) connectivity across all layers.
• Integrate artificial intelligence (AI) and machine learning (ML) techniques—particularly Federated Learning (FL) and Deep Reinforcement Learning (DRL)—to enable decentralized, privacy-preserving, and adaptive decision-making in dynamic environments.
• Propose a standardized, secure, and efficient Air Computing protocol to manage inter-entity communication, ensuring data integrity, low latency, and reliability across heterogeneous aerial and terrestrial nodes.
• Optimize energy consumption and flight trajectories of aerial vehicles (UAVs, HAPs) through collaborative resource allocation and mobility-aware task offloading strategies.
• Ensure regulatory and standardization compliance by aligning Air Computing protocols with national and international regulations for unmanned aerial vehicles and 6G spectrum usage.

Experimental results

Research questions

• RQ1 How can a 3D integrated network of air and terrestrial layers overcome the latency, capacity, and coverage limitations of current 2D edge and cloud computing systems?
• RQ2 What architectural and protocol-level components are essential to enable seamless, secure, and low-latency communication across UAVs, HAPs, LEO satellites, and ground infrastructure in Air Computing?
• RQ3 How can AI and machine learning—especially Federated Learning and Deep Reinforcement Learning—be leveraged to enable intelligent, adaptive, and privacy-preserving resource management in dynamic air computing environments?
• RQ4 What are the key challenges in energy efficiency, mobility management, and regulatory compliance for aerial vehicles in Air Computing, and how can they be addressed?
• RQ5 What future research directions are most promising for enabling the full potential of Air Computing in 6G and beyond?

Key findings

• Air Computing provides a unified, 3D computational framework that integrates UAVs, HAPs, LEO satellites, and terrestrial edge/cloud resources to deliver ultra-low latency and high-capacity computing for next-generation applications.
• 6G wireless networks are essential for Air Computing, as they enable the required 0.1 ms latency, 1 Tbps data rates, and 1000 km/h mobility support that 5G cannot achieve.
• Federated Learning (FL) is a more suitable ML approach than centralized deep learning in Air Computing due to the high volume of data and privacy concerns in heterogeneous, distributed environments.
• Deep Reinforcement Learning (DRL) enables autonomous, real-time decision-making in dynamic air computing environments where labeled data is scarce or unavailable.
• Collaborative trajectory and resource optimization among aerial vehicles can significantly reduce energy consumption and improve system efficiency.
• Standardized, secure, and low-latency communication protocols are critical for reliable coordination between diverse aerial and terrestrial entities in Air Computing.