Skip to content
Open access

Joint AP Selection and Power Allocation in Satellite–Terrestrial Cell-Free Network With Heterogeneous Multi-Agent Learning

2026 · IEEE Transactions on Network and Service Management · Vol 23, pp. 8736-8751 · 0 citations · 48 references

TL;DR

A Heterogeneous Multi-Agent Proximal Policy Optimization (H-MAPPO) algorithm for joint Access Point (AP) selection and power allocation that effectively adapts to dynamic LEO environments and achieves superior User QoS over learning-based benchmarks is developed.

Abstract

While Low Earth Orbit (LEO) satellites are vital for 6G non-terrestrial networks (NTN), applying Cell-Free Massive MIMO (CF-mMIMO) to high-mobility constellations remains challenging despite its capability to resolve conventional handover and interference issues. The traditional centralized processing model becomes impractical due to prohibitive latency, signaling overhead, and outdated Channel State Information (CSI), which degrades system performance. To overcome this challenge, we design a LEO Satellite Cell-Free Massive MIMO (LCF-mMIMO) system. In this system, each satellite independently determines its transmit power within the serving cluster, while statistical CSI is utilized to counter the fast-varying instantaneous CSI resulting from the rapid movement of LEO satellites. Additionally, to solve the minimum UE quality of service(QoS) maximization problem with reduced latency and signaling overhead, we develop a Heterogeneous Multi-Agent Proximal Policy Optimization (H-MAPPO) algorithm for joint Access Point (AP) selection and power allocation. In this algorithm, User Equipments (UEs) and satellites operate as heterogeneous agents: UEs perform AP selection based on local service demands, while satellites allocate power according to local power and interference states. Their decisions are coupled through the AP selection matrix, which converts UE-side access results into satellite-side associated UE sets and limits information exchange to selected UE-satellite associations. Simulation results demonstrate that the proposed scheme effectively adapts to dynamic LEO environments and achieves superior User QoS over learning-based benchmarks. Compared with the centralized SR-JCPA method, H-MAPPO maintains a comparable minimum QoS of 0.52 versus 0.53 under 16 UEs, while reducing the per-episode execution time from 17646.81 s to 96.81 s.

Read PDF

Similar papers

Open access 2026

Interference-Aware Resource Allocation for Direct-to-Cell Satellite–Terrestrial Integrated Networks

Direct-to-Cell (D2C) Satellite-Terrestrial Integrated Networks (STINs) are emerging as a promising technology to provide ubiquitous global connectivity. To support direct access for existing smartphones, D2C satellite systems must reuse spectrum bands originally allocated to terrestrial cellular networks, which may int...

Bo-Yu Yang, Xing-Qiu He, Jin Zhao et al. · 0 citations
2026

Multi-Beam Opportunistic Rate-Splitting Multiple Access in MEC-Assisted Vehicular Networks

The evolution of sixth-generation (6G) communication technologies and intelligent transportation has driven vehicular networks toward low latency and high connectivity. Mobile edge computing (MEC) has been widely recognized as a promising paradigm for supporting the diverse service requirements of vehicular networks. H...

Xi-Ran Zhang, Wen-Bin Sun, Zhao-Lin Zhang et al. · 0 citations
Preprint Oct 2026

Multi-Timescale Context-Aware Radio Resource Optimization for Harmonious NTN-TN Coexistence: A Digital Twin and Open RAN Enabled Framework

Coexistence between non-terrestrial networks (NTNs) and terrestrial networks (TNs) is essential for future wireless systems, but terrestrial downlink transmissions can generate harmful aggregate interference at low-Earth orbit (LEO) satellite receivers. We present DT-STABLE, an Open Radio Access Network (O-RAN)-enabled...

Ecem Izde Kaya, Md. Zoheb Hassan, Georges Kaddoum · 0 citations
Open access 2026

Toward Scalable Centralized Network-Assisted Full-Duplex Cell-Free Massive MIMO Networks

Cell-free massive MIMO (CF-mMIMO) is a key candidate technology for 6G networks, enabling more uniform service quality and higher spectral efficiency (SE) through the coordination of distributed access points (APs). Conventional half-duplex (HD) time-division duplexing (TDD) statically partitions uplink (UL) and downli...

Guillem Femenias, Felip Riera-Palou · 0 citations
Oct 2026

Deep Reinforcement Learning for Downlink User Scheduling in LEO Satellite Massive MIMO Systems

Low earth orbit (LEO) satellite communication system is expected to play a key role in the 6G wireless networks. Due to the limited communication resource and inter-user interference (IUI), the user equipments (UEs) must be well scheduled to maximize spectral efficiency. However, rapid channel variations and high...

Chao Deng, Yun Zhong, Xiang-Xu Meng et al. · 0 citations
Open access Sep 2026

Joint Energy Efficiency and Fairness Optimization for D2D Communications in Aerial-Ground Integrated Heterogeneous Networks

This study investigates an Aerial-Ground Integrated Heterogeneous network (AGIHN) architecture that combines terrestrial macro base stations and unmanned aerial vehicles serving as aerial base stations to enhance uplink access for macrocell users and proposes a low-complexity Multi-Channel Rate-Fair (MCRF) algorithm.

Chuan-Chi Lai, Ang-Hsun Tsai, Shang-Long Wu · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.