Towards a Unified SAGIN System with Evaluation and Optimization: A Spherical Stochastic Geometry Perspective

Abstract

In addressing the challenges of expanding wireless coverage to unconnected regions, the space-air-ground integrated network (SAGIN) emerges as a transformative paradigm designed to meet the demands for high-rate and high-reliability communications. This dissertation introduces SAGIN, comprising satellite, aerial, and terrestrial components, as a two-hop relay network enhanced by cooperative links between space-air and air-ground segments. We explore the role and features of clustering relays by unmanned aerial vehicles (UAVs) within this framework, modeling their spatial distribution as a homogeneous Poisson Point Process (PPP) within a three-dimensional spherical field. By evaluating the ensemble outage performance in this setting, we derive a closed-form approximation of the outage probability, further substantiated by a parametric analysis and numerical results validating our analytical approaches. In advancing this framework, we propose an energy-efficient communication strategy utilizing orthogonal frequency-division multiplexing (OFDM) that circumvents the need for instantaneous channel state information or complex computations at the UAV relay nodes. This strategy involves a novel power allocation method aimed at minimizing the overall transmission power of the SAGIN system while adhering to specific outage probability constraints. Through solving the convex optimization problems involved in the system and potentially integrating artificial neural networks (ANNs) with it, we have refined this approach, demonstrating its computational efficiency and practical viability. Further extending the application of SAGIN, we examine the integration of low Earth orbit (LEO) satellites with geostationary Earth orbit (GEO) satellites to enhance connectivity for IoT devices in remote areas. Utilizing a stochastic geometry model, we assess the dynamic interactions within satellite constellations, providing low-complexity analytical estimates for end-to-end availability and coverage performance. Our study delves into the effects of constellation configuration, transmission setting, and the spatial dynamics between IoT devices and GEO satellites, presenting a comprehensive analysis of the impact on network architecture. The precise approximation of outage probability ensures the robustness of system spatial distribution modeling, while the research on the power allocation model represents a significant advancement toward energy efficiency. The analysis of end-to-end connection availability and coverage further seamlessly integrates the entire SAGIN system. These developments collectively contribute to advancing a unified SAGIN system.

Brief Biography

Jiusi Zhou obtained his Bachelor of Engineering from Communication and Information Engineering Department, University of Electronic Science and Technology of China (UESTC), Chengdu, Sichuan, P. R. China in 2018. Before formally joining KAUST, he worked as a internship student at CTL in KAUST during Aug 2017 to Jan 2018.