loading page

Joint Two-tier User Association and Resource Management for Integrated Satellite-Terrestrial Networks
  • +2
  • Hung Nguyen-Kha ,
  • Vu Nguyen Ha, ,
  • Eva lagunas ,
  • Symeon Chatzinotas ,
  • Joel Grotz
Hung Nguyen-Kha
Interdisciplinary Centre for Security

Corresponding Author:[email protected]

Author Profile
Vu Nguyen Ha,
Author Profile
Eva lagunas
Author Profile
Symeon Chatzinotas
Author Profile
Joel Grotz
Author Profile

Abstract

This paper investigates the uplink transmission of an integrated satellite-terrestrial network, wherein the low-earth-orbit (LEO) satellites provide backhaul services to isolated cellular base stations (BSs) for forwarding mobile user (UE) data to the core network. In this integrated system, the high mobility of LEO satellites introduces significant challenges in managing radio resource allocation (RA), as well as the associations between UEs, BSs, and LEO satellites for supporting users’ demands efficiently, while also dynamically balancing the capacity of UE-BS access and BS-LEO backhaul links. Regarding these critical issues, the paper aims to jointly optimize the two-tier UE-BS-LEO satellite association, sub-channel assignment, bandwidth allocation, and power control to meet users’ demands in the shortest transmission time. This optimization problem, however, falls into the category of mixed-integer non-convex programming, making it very challenging and requiring advanced solution techniques to find optimal solutions. To tackle this complex problem efficiently, we first develop an iterative centralized algorithm by utilizing convex approximation and compressed-sensing-based methods to deal with binary variables. Furthermore, for practical implementation and to offload computation from the central processing node, we propose a decentralized algorithm that can be implemented in parallel at local controllers and achieve efficient solutions. Numerical results are also illustrated to strengthen the effectiveness of our proposed algorithms compared to traditional greedy and benchmark algorithms.