Pinching-Antenna Systems-Enabled Multi-User Communications: Transmission Structures and Beamforming Optimization

Jingjing Zhao, Haowen Song, Xidong Mu, Kaiquan Cai, Yanbo Zhu, Yuanwei Liu

TL;DR

This paper introduces Pinching-Antenna Systems (PASS) as a breakthrough for 6G multi-user communications, where antennas (dielectric particles) physically slide along waveguides to optimize signal paths. The authors propose three novel transmission architectures—Waveguide Multiplexing, Division, and Switching—and develop a sophisticated optimization framework that significantly outperforms conventional fixed-antenna MIMO systems by jointly tuning digital beamforming and physical antenna positions.

What Problem Does This Paper Address?

In the quest for 6G, traditional Multiple-Input Multiple-Output (MIMO) systems face a fundamental bottleneck: their antennas are fixed in space. While technologies like Reconfigurable Intelligent Surfaces (RIS) and "fluid" antennas offer some flexibility, they are often limited by high path loss or restricted movement.

The authors address the challenge of scaling PASS—a technology originally prototyped by NTT DOCOMO—from a single-waveguide setup to a multi-user, multi-waveguide system. PASS uses dielectric waveguides (similar to leaky coaxial cables) where "pinching antennas" (PAs) can be activated at any point. The core problem is how to coordinate these moving parts and baseband signals to ensure "Max-Min Fairness" (MMF)—guaranteeing the best possible performance for the user with the worst connection in a crowded network.

Methods

The paper proposes three distinct transmission structures to handle multi-user interference:

  1. Waveguide Multiplexing (WM): Data streams are digitally beamformed and superimposed across all waveguides. This offers the highest flexibility but the most complex optimization.
  2. Waveguide Division (WD): Each data stream is assigned a dedicated waveguide. This simplifies the baseband to mere power allocation.
  3. Waveguide Switching (WS): Users are served in orthogonal time slots. This eliminates multi-user interference entirely but requires precise time synchronization.

To solve the resulting non-convex optimization problems (where antenna positions and beamforming weights are "coupled"), the authors employ a Penalty Dual Decomposition (PDD) based algorithm. This approach breaks the complex problem into manageable sub-problems, using Successive Convex Approximation (SCA) to handle the tricky exponential and fractional terms associated with physical antenna movement.

Datasets

The paper does not use external real-world datasets. Instead, it relies on numerical simulations based on established electromagnetic models for dielectric waveguides and free-space propagation.

  • Parameters: Carrier frequency of 28 GHz (mmWave), waveguides of 10 meters, and varying numbers of PAs (4 to 12).
  • Benchmarks: The system is compared against conventional Fully-Digital MIMO and Hybrid Beamforming MIMO systems with fixed antenna positions.

Key Insights and Ideas

  • The "Last-Meter" Advantage: Unlike fixed arrays, PASS can physically move the radiation point closer to a user. This doesn't just change the phase of the signal; it fundamentally reduces large-scale path loss.
  • Architecture Matters: The authors found that Waveguide Switching (WS) is superior for unicast (one-to-one) communications because it can tailor antenna positions for one user at a time without worrying about interfering with others. Conversely, Waveguide Multiplexing (WM) is better for multicast (one-to-many) because it leverages the full time-resource to serve groups simultaneously.
  • Geographic Isolation: If users are far apart, the performance gap between the simple WD structure and the complex WM structure disappears, suggesting that simpler hardware might suffice in sparse environments.

Key Results

  • Massive Rate Gains: In a multicast scenario with a 20 dBm transmit power, PASS (using the WM structure) achieved a minimum data rate of approximately 11.5 bps/Hz, compared to only 4.8 bps/Hz for conventional hybrid MIMO and 2 bps/Hz for fully-digital MIMO.
  • Scalability: Increasing the number of pinching antennas from 4 to 12 improved the minimum data rate by up to 27% for the WD structure, demonstrating that PASS can effectively utilize higher spatial degrees of freedom.
  • Robustness to Range: While conventional MIMO performance drops sharply as users move further from the array center (due to increased path loss and weakened array gain), PASS maintains a much flatter performance curve because the antennas "follow" the users along the waveguide.

Limitations and Future Work

  • Hardware Latency: The paper assumes antenna positions can be adjusted perfectly. In reality, the mechanical latency of moving dielectric particles (via actuators or motors) must be reconciled with the fast-changing wireless channel (coherence time).
  • CSI Acquisition: Obtaining Channel State Information (CSI) for every possible position along a continuous waveguide is a massive overhead. The authors assume perfect CSI, but practical, low-overhead estimation remains a hurdle.
  • Synchronization: The Waveguide Switching (WS) model requires extremely tight time synchronization to avoid "slot collisions," which increases implementation complexity.
  • Future Directions: The authors suggest extending this work to Integrated Sensing and Communication (ISAC) and exploring PASS-enabled physical layer security.

Summary updated

Abstract

Pinching-antenna systems (PASS) represent an innovative advancement in flexible-antenna technologies, aimed at significantly improving wireless communications by ensuring reliable line-of-sight connections and dynamic antenna array reconfigurations. To employ multi-waveguide PASS in multi-user communications, three practical transmission structures are proposed, namely waveguide multiplexing (WM), waveguide division (WD), and waveguide switching (WS). Based on the proposed structures, the joint baseband signal processing and pinching beamforming design is studied for a general multi-group multicast communication system, with the unicast communication encompassed as a special case. A max-min fairness problem is formulated for each proposed transmission structure, subject to the maximum transmit power constraint. For WM, to solve the highly-coupled and non-convex MMF problem with complex exponential and fractional expressions, a penalty dual decomposition (PDD)-based algorithm is invoked for obtaining locally optimal solutions. Specifically, the augmented Lagrangian relaxation is first applied to alleviate the stringent coupling constraints, which is followed by the block decomposition over the resulting augmented Lagrangian function. Then, the proposed PDD-based algorithm is extended to solve the MMF problem for both WD and WS. Furthermore, a low-complexity algorithm is proposed for the unicast case employing the WS structure, by simultaneously aligning the signal phases and minimizing the large-scale path loss at each user. Finally, numerical results reveal that: 1) the MMF performance is significantly improved by employing the PASS compared to conventional fixed-position antenna systems; 2) WS and WM are suitable for unicast and multicast communications, respectively; 3) the performance gap between WD and WM can be significantly alleviated when the users are geographically isolated.
BibTeX
@article{zhao2025pinching,
  author = {Jingjing Zhao and Haowen Song and Xidong Mu and Kaiquan Cai and Yanbo Zhu and Yuanwei Liu},
  title = {{Pinching-Antenna Systems-Enabled Multi-User Communications: Transmission Structures and Beamforming Optimization}},
  journal = {IEEE Transactions on Communications},
  year = {2025},
  doi = {10.1109/tcomm.2025.3643993},
}
Figure