Graph-Based Resource Allocation for Air-Ground Integrated Networks

• Published in: Mobile networks and applications Vol. 27; no. 2; pp. 492 - 501
• Main Authors: Chen, Qian, Meng, Weixiao, He, Chenguang
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.04.2022; Springer Nature B.V
• Subjects: Algorithms; Communications Engineering; Computer architecture; Computer Communication Networks; Electrical Engineering; Engineering; Graph theory; Ground stations; IT in Business; Network latency; Networks; Optimization; Optimization algorithms; Radio equipment; Resource allocation; Satellite communications; Wireless communications; Air-ground integrated network; Graph theory; Fairness; Resource allocation; Quality of service
• ISSN: 1383-469X; 1572-8153
• DOI: 10.1007/s11036-020-01694-1

With the combined advantages of satellite communications, aerial networks and terrestrial systems, a space-air-ground integrated network has gradually become a promising architecture for the next generation wireless communication. Due to heterogeneous characteristics of different layers, it is necessary to perform efficient resource allocation. Motivated by this fact, we propose a novel architecture of air-ground integrated networks (AGIN), which leverages civil aircrafts and ground base stations to support terrestrial users’ service. Aiming at maximizing the overall capacity of downlink transmission in an AGIN, we formulate the resource allocation problem as an optimization problem subject to both quality of service (QoS) and fairness requirements. To address the formulated problem, we propose a graph-based joint optimization algorithm for resource block (RB) and power allocation. Specifically, an improved Kuhn-Munkras (KM) algorithm based on graph theory is proposed for RB allocation, which guarantees the fairness. Meanwhile, a multi-level water-filling method is proposed for power allocation. By leveraging an alternating descent approach, a joint optimal solution can be obtained after a finite number of iterations. It is demonstrated through simulation results that the proposed joint optimization algorithm is converges fast and shows significant improvement in terms of the sum-rate, fairness, access latency, and system capacity.