Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Deep Reinforcement Learning for Interference Suppression in RIS-Aided Space-Air-Ground Integrated Networks

Pujitha Mamillapalli, Shikhar Verma, Tiago Koketsu Rodrigues, Abhinav Kumar

TL;DR

The paper tackles cross-tier interference in RIS-aided SAGINs under shared-spectrum operation, proposing a RIS-assisted HAPS framework and a Deep Deterministic Policy Gradient (DDPG) approach to learn beamforming that forms nulls toward interference while preserving QoS. By modeling the system as a continuous-control task, the authors train actor–critic networks to adjust the HAPS beamforming and RIS phases in response to rapidly changing channels, achieving up to $11.3\%$ throughput gains for a $4\times4$ RIS over conventional zero-forcing beamforming. The results demonstrate improved spectral efficiency and energy usage in dynamic non-terrestrial networks across different RIS configurations and user distributions, validating the adaptability of DRL-based interference suppression. The work suggests scalable deployment potential and outlines future directions addressing larger networks, hardware impairments, and imperfect CSI to bridge toward real-world non-terrestrial 6G systems.

Abstract

Future 6G networks envision ubiquitous connectivity through space-air-ground integrated networks (SAGINs), where high-altitude platform stations (HAPSs) and satellites complement terrestrial systems to provide wide-area, low-latency coverage. However, the rapid growth of terrestrial devices intensifies spectrum sharing between terrestrial and non-terrestrial segments, resulting in severe cross-tier interference. In particular, frequency sharing between the HAPS satellite uplink and HAPS ground downlink improves spectrum efficiency but suffers from interference caused by the HAPS antenna back-lobe. Existing approaches relying on zero-forcing (ZF) codebooks have limited performance under highly dynamic channel conditions. To overcome this limitation, we employ a reconfigurable intelligent surface (RIS)-aided HAPS-based SAGIN framework with a deep deterministic policy gradient (DDPG) algorithm. The proposed DDPG framework optimizes the HAPS beamforming weights to form spatial nulls toward interference sources while maintaining robust links to the desired signals. Simulation results demonstrate that the DDPG framework consistently outperforms conventional ZF beamforming among different RIS configurations, achieving up to \(11.3\%\) throughput improvement for a \(4\times4\) RIS configuration, validating its adaptive capability to enhance spectral efficiency in dynamic HAPS-based SAGINs.

Deep Reinforcement Learning for Interference Suppression in RIS-Aided Space-Air-Ground Integrated Networks

TL;DR

The paper tackles cross-tier interference in RIS-aided SAGINs under shared-spectrum operation, proposing a RIS-assisted HAPS framework and a Deep Deterministic Policy Gradient (DDPG) approach to learn beamforming that forms nulls toward interference while preserving QoS. By modeling the system as a continuous-control task, the authors train actor–critic networks to adjust the HAPS beamforming and RIS phases in response to rapidly changing channels, achieving up to throughput gains for a RIS over conventional zero-forcing beamforming. The results demonstrate improved spectral efficiency and energy usage in dynamic non-terrestrial networks across different RIS configurations and user distributions, validating the adaptability of DRL-based interference suppression. The work suggests scalable deployment potential and outlines future directions addressing larger networks, hardware impairments, and imperfect CSI to bridge toward real-world non-terrestrial 6G systems.

Abstract

Future 6G networks envision ubiquitous connectivity through space-air-ground integrated networks (SAGINs), where high-altitude platform stations (HAPSs) and satellites complement terrestrial systems to provide wide-area, low-latency coverage. However, the rapid growth of terrestrial devices intensifies spectrum sharing between terrestrial and non-terrestrial segments, resulting in severe cross-tier interference. In particular, frequency sharing between the HAPS satellite uplink and HAPS ground downlink improves spectrum efficiency but suffers from interference caused by the HAPS antenna back-lobe. Existing approaches relying on zero-forcing (ZF) codebooks have limited performance under highly dynamic channel conditions. To overcome this limitation, we employ a reconfigurable intelligent surface (RIS)-aided HAPS-based SAGIN framework with a deep deterministic policy gradient (DDPG) algorithm. The proposed DDPG framework optimizes the HAPS beamforming weights to form spatial nulls toward interference sources while maintaining robust links to the desired signals. Simulation results demonstrate that the DDPG framework consistently outperforms conventional ZF beamforming among different RIS configurations, achieving up to throughput improvement for a RIS configuration, validating its adaptive capability to enhance spectral efficiency in dynamic HAPS-based SAGINs.
Paper Structure (10 sections, 12 equations, 4 figures, 2 tables)

This paper contains 10 sections, 12 equations, 4 figures, 2 tables.

Table of Contents

  1. Introduction
  2. System Model
  3. HAPS Antenna Configuration
  4. RIS-aided Propagation Environment
  5. Null Directivity-Based Interference Suppression
  6. Problem Formulation
  7. Deep Deterministic Policy Gradient (DDPG) Framework
  8. Loss Function
  9. Numerical Results and Analysis
  10. Conclusion

Figures (4)

  • Figure 1: Illustration of the RIS-aided HAPS-based SAGIN system model. The figure illustrates the communication links between the satellite, RIS, high-altitude platform station (HAPS), and ground users.
  • Figure 2: Architecture of the DDPG framework: Red dotted arrows indicates updation; black solid arrows indicates forward‑pass data flow.
  • Figure 3: Average rewards versus time steps under various transit power $P_t$ with $\gamma_\text{min}=0~\text{dB}$
  • Figure 4: Sum rate versus number of ground users for Poisson, Normal, and Uniform user distributions under both DDPG-based and ZF beamforming schemes kawamoto2024interference for a $4\times4$ RIS configuration in the HAPS-based SAGIN framework.