Supervised Learning Based Real-Time Adaptive Beamforming On-board: Abstract & Introduction

Table of Links

• Abstract & Introduction
• System Model and Problem Statement
• Antenna Design and Training Data Generation
• Supervised Learning for Adaptive Beamforming
• Numerical Results
• Conclusions, Acknowledgment, and References

Abstract

Satellite communications (SatCom) are crucial for global connectivity, especially in the era of emerging technologies like 6G and narrowing the digital divide. Traditional SatCom systems struggle with efficient resource management due to static multibeam configurations, hindering quality of service (QoS) amidst dynamic traffic demands. This paper introduces an innovative solution - real-time adaptive beamforming on multibeam satellites with software-defined payloads in geostationary orbit (GEO). Utilizing a Direct Radiating Array (DRA) with circular polarization in the 17.7 - 20.2 GHz band, the paper outlines DRA design and a supervised learning-based algorithm for on-board beamforming. This adaptive approach not only meets precise beam projection needs but also dynamically adjusts beamwidth, minimizes sidelobe levels (SLL), and optimizes effective isotropic radiated power (EIRP).

Index Terms—antennas, beamforming, multibeam satellite, supervised learning.

I. INTRODUCTION

Satellite communications (SatCom) is a fundamental pillar of modern global connectivity, providing the means to bridge the digital divide and offering ubiquitous coverage in an increasingly connected world. The integration of terrestrial systems such as 6G further underlines the importance of SatCom as it continues to facilitate communication on a global scale [1]. However, the increased data traffic on SatCom systems poses a considerable challenge: effectively managing radio resource allocation while meeting stringent quality of service (QoS) requirements remains a formidable task [2].

Traditionally, SatCom systems have relied on static multibeam configurations with fixed bandwidth and power allocations. However, these configurations fall short of adapting to the dynamic nature of today’s traffic demands, leading to inefficient resource utilization and potential service degradation. Recognizing the temporal and spatial variations in demand, software-defined payloads have emerged as a revolutionary solution. These payloads offer unprecedented flexibility and adaptability in radio resource management (RRM) for SatCom [3].

While software-defined payloads are very promising, their effective utilization requires advanced RRM techniques to optimize resource allocation in real-time. A crucial facet of RRM in SatCom is the adaptive beamwidth, power, and pointing through beamforming control. Conventional optimizationbased approaches, while theoretically sound, often lack the computational efficiency and adaptability needed to cope with the diverse and dynamic traffic patterns encountered in SatCom systems [4] .

Recent studies have proposed schemes to enhance spectral efficiency and user fairness in multi-beam satellite systems through robust beamforming and non-orthogonal multiple access, considering imperfect channel information among terminals [5]. There’s also a shift towards adaptive multibeam planning and demand-based footprinting to cater to dynamic traffic demands, especially in remote areas [6]. While technological advances have spurred the use of all-digital phased arrays, the high computational cost of adaptive beamforming remains a challenge, with some research exploring neural networks for real-time scenarios, albeit without fully addressing service area traffic demands and SatCom system constraints [7].

This paper explores two novel approaches to on-board realtime adaptive beamforming based on supervised learning. Specifically, the use of a Direct Radiating Array (DRA) operating in circular polarization within the frequency band 17.7 - 20.2 GHz is proposed. In addition to addressing the beam requirements, this research delves into other vital parameters such as beamwidth in the azimuthal and elevation planes, sidelobe level (SLL) control, and effective isotropic radiated power (EIRP) control.