Title: Blind reconfigurable intelligent surfaces for dynamic offloading in fixed-NOMA mobile edge networks

Authors: Guoliang Zhou; Amin Mohajer

Addresses: Jiangsu Automation Research Institute, Lianyungang, 222061, China ' Department of Communications Technology, ICT Research Institute (ITRC), Tehran, Iran

Abstract: Integrating unmanned aerial vehicles (UAVs) into cellular networks offers a way to boost connectivity and service quality in both cities and remote areas. In this paper, we present a new framework that combines advanced deep learning techniques with backhaul traffic optimisation to manage resources efficiently in UAV-supported communication networks. Using intelligent reflecting surfaces (IRS) and cell-free communication strategies, our approach improves backhaul traffic, ensuring smooth data transmission and better network performance. To deal with diverse resource challenges, we introduce an adaptive offloading framework that enhances mobile edge networks through dynamic, traffic-aware resource allocation and serverless load balancing. Our method uses multi-agent deep learning for prediction of network demands, allowing for dynamic power control. Our simulations show that the proposed framework performs well in different scenarios, leading to higher efficiency and total network throughput. This framework offers a promising approach for managing resources in future UAV-supported mobile edge networks.

Keywords: UAV-aided edge networks; dynamic deep learning; power control; intelligent reflecting surfaces; IRS; serverless load balancing; cell-free communication.

DOI: 10.1504/IJSNET.2024.142517

International Journal of Sensor Networks, 2024 Vol.46 No.3, pp.142 - 160

Received: 06 Aug 2023
Accepted: 12 Jul 2024
Published online: 05 Nov 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article