Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

RL based Beamforming Optimization for 3D Pinching Antenna assisted ISAC Systems

Qian Gao, Ruikang Zhong, Yue Liu, Hyundong Shin, Yuanwei Liu

TL;DR

This work addresses joint optimization of 3D pinching antenna positioning, TDMA time allocation, and transmit power to maximize downlink ISAC sum rate under sensing and energy constraints, formalized as $\max_{\{\psi^p_{n_a,t}, q^{\text{com}}_{k,t}, p_{k,t}\}} \sum_{t=1}^{T} \sum_{k=1}^{K} R^{\text{com}}_{k,t}$ with constraints $\Gamma_{l,t} \ge \Gamma_{\min}$ and chaos of power and spacing. It introduces HGRL, a heterogeneous graph neural network based reinforcement learning framework that models the ISAC topology as a time-varying graph and uses an A2C-based policy with a HetGNN encoder to jointly optimize 3D antenna positions, TDMA fractions, and power. The method demonstrates that 3D pinching antenna deployment offers superior spatial diversity and beamforming flexibility, achieving higher sum-rate and faster convergence than 1D/2D baselines. The approach provides scalable, topology-aware optimization for power-limited ISAC systems and highlights the practical benefits of 3D deployment for joint sensing and communication tasks, especially under near-field conditions.

Abstract

In this paper, a three-dimensional (3D) deployment scheme of pinching antenna array is proposed, aiming to enhances the performance of integrated sensing and communication (ISAC) systems. To fully realize the potential of 3D deployment, a joint antenna positioning, time allocation and transmit power optimization problem is formulated to maximize the sum communication rate with the constraints of target sensing rates and system energy. To solve the sum rate maximization problem, we propose a heterogeneous graph neural network based reinforcement learning (HGRL) algorithm. Simulation results prove that 3D deployment of pinching antenna array outperforms 1D and 2D counterparts in ISAC systems. Moreover, the proposed HGRL algorithm surpasses other baselines in both performance and convergence speed due to the advanced observation construction of the environment.

RL based Beamforming Optimization for 3D Pinching Antenna assisted ISAC Systems

TL;DR

This work addresses joint optimization of 3D pinching antenna positioning, TDMA time allocation, and transmit power to maximize downlink ISAC sum rate under sensing and energy constraints, formalized as with constraints and chaos of power and spacing. It introduces HGRL, a heterogeneous graph neural network based reinforcement learning framework that models the ISAC topology as a time-varying graph and uses an A2C-based policy with a HetGNN encoder to jointly optimize 3D antenna positions, TDMA fractions, and power. The method demonstrates that 3D pinching antenna deployment offers superior spatial diversity and beamforming flexibility, achieving higher sum-rate and faster convergence than 1D/2D baselines. The approach provides scalable, topology-aware optimization for power-limited ISAC systems and highlights the practical benefits of 3D deployment for joint sensing and communication tasks, especially under near-field conditions.

Abstract

In this paper, a three-dimensional (3D) deployment scheme of pinching antenna array is proposed, aiming to enhances the performance of integrated sensing and communication (ISAC) systems. To fully realize the potential of 3D deployment, a joint antenna positioning, time allocation and transmit power optimization problem is formulated to maximize the sum communication rate with the constraints of target sensing rates and system energy. To solve the sum rate maximization problem, we propose a heterogeneous graph neural network based reinforcement learning (HGRL) algorithm. Simulation results prove that 3D deployment of pinching antenna array outperforms 1D and 2D counterparts in ISAC systems. Moreover, the proposed HGRL algorithm surpasses other baselines in both performance and convergence speed due to the advanced observation construction of the environment.
Paper Structure (14 sections, 14 equations, 4 figures, 1 table, 1 algorithm)

This paper contains 14 sections, 14 equations, 4 figures, 1 table, 1 algorithm.

Table of Contents

  1. Introduction
  2. System Model and Problem Formulation
  3. Signal Model
  4. Communication and Sensing Metrics
  5. Problem Formulation
  6. Heterogeneous Graph Neural Network based Reinforcement Learning Solution
  7. MDP Formulation
  8. Heterogeneous Graph-Based Representation Learning
  9. HGRL Algorithm
  10. Policy Network $\pi_\theta$
  11. Value Network $V_\psi$
  12. Graph Encoder $\phi_{\text{HetGNN}}$
  13. Simulation Results
  14. Conclusion

Figures (4)

  • Figure 1: Illustration of the 3D pinching antenna assisted downlink ISAC system.
  • Figure 2: Illustration of simulation configurations.
  • Figure 3: Reward performance of the proposed HGRL compared to baselines.
  • Figure 4: ISAC performance comparison of different pinching antenna deployment.