Segmented Pinching Antenna System (PASS)

• Segmented PASS is a flexible, reconfigurable antenna array that uses movable dielectric antennas along waveguides to achieve localized beamforming and path-loss control.
• It employs segmentation, precise position control, and selective activation protocols to optimize array gain, outperforming conventional MIMO in efficiency and security.
• Advanced optimization techniques, including alternating optimization and PSO, enable dynamic PA positioning for enhanced multi-user communication and integrated sensing.

A Segmented Pinching Antenna System (PASS) is a flexible, dynamically reconfigurable antenna array architecture wherein small dielectric pinching antennas (“PAs”) are deployed at arbitrarily controllable positions along one or more dielectric waveguides. By exploiting the degrees of freedom in PA location and grouping (“segmentation”), PASS can locally emulate large-scale path-loss control and phase steering while maintaining minimal RF chain count. The system enables ultra-efficient, near-user array gain, surpassing conventional MIMO or distributed antenna architectures in communication, sensing, secrecy, and energy efficiency. The segmentation concept underpins all advanced PASS protocols, from single-user link optimization to fully multi-user, multi-waveguide architectures.

1. System Architecture and Segmentation Principles

Segmented PASS comprises multiple parallel dielectric waveguides, each extending over macroscopic distances (meters to tens of meters) with a movable set of PAs that can be mechanically or electronically clipped at arbitrary positions. Each PA operates as a passive coupler, extracting and radiating energy from the guided mode at its specific location. The key distinctions are:

• Segmentation: Each waveguide or group of PAs is treated as a segment, providing localized beamforming aperture. Segmentation can be physical (discrete waveguide lengths), logical (grouping PAs for coordinated operation), or both. Multi-segment architectures include segmented waveguide structures (e.g., SWANs (Ouyang et al., 12 Sep 2025)) and logical partitioning for protocol adaptation.
• Position Control: PAs can be repositioned with fine (subwavelength) or coarse (meter-scale) granularity, leveraging dual-scale actuation if needed (coarse translation and fine phase tuning) (Gan et al., 31 Oct 2025).
• Activation Protocols: Segmentation enables selective activation, with each segment supporting single or multiple PAs, permitting orthogonal or aggregated operation (e.g., segment selection, segment aggregation, segment multiplexing).

The system's composite channel is the concatenation of independent segment responses, allowing control of path-loss via proximity and phase via in-guide propagation and free-space distance. This segmentation enables “last-meter” array formation directly above or near the intended user, reducing path-loss relative to fixed-corner or ceiling-bound antennas (Ouyang et al., 10 Jan 2025).

2. Analytical Framework: Channel, Array Gain, and Segmented Protocols

The location and activation pattern of the PAs in each segment fundamentally determine the channel matrix and thus the array gain, beam patterns, and protocol capabilities.

• Channel Model: For each PA at location $x_n$ (single waveguide, height $d$), the per-antenna channel is $$h_n = \eta^{1/2} \exp\left(-j\frac{2\pi}{\lambda} \sqrt{d^2 + \Delta_n^2}\right) / \sqrt{d^2 + \Delta_n^2}$$, with $\Delta_n$ the lateral offset. For multi-segment systems (e.g., segmented waveguide architecture), the full channel is block-diagonal; for SWANs each segment has an independent feed and signal path (Ouyang et al., 12 Sep 2025).
• Array Gain Optimization: The array gain $G(N,d)$ as a function of number of PAs $N$ and height $d$ admits a closed-form upper bound (Ouyang et al., 10 Jan 2025). The upper bound increases with $N$ initially, peaks at $N_\star$, then decays (excess PAs reduce per-antenna power and have deleterious far-field effects). The optimal $N_\star$ and inter-PA spacing $\Delta_\star$ are determined by joint analysis of propagation and mutual coupling; e.g., $\Delta_\star \simeq 0.715\lambda$ for two-antenna mutual-coupling maximum.
• Position Refinement Algorithms: Antenna positions (${x_n}$) are refined via coordinate-descent or projected gradient methods, with each update maximizing total array gain (or related system metric) subject to minimal spacing and guide-length constraints (Ouyang et al., 10 Jan 2025, Wang et al., 9 Feb 2025). In practice, this enables practical near-optimal placements.
• Mutual Coupling: For closely spaced PAs ($\Delta < \lambda/2$), mutual coupling is significant, requiring matrix correction of the channel coefficients ($g = C^{-1/2}h$ with $[C]_{mn}=j_0((2\pi/\lambda)|m-n|\Delta)$) (Ouyang et al., 10 Jan 2025).

Segmented protocols include:

Protocol Type	Description	Use Cases
Segment Selection (SS)	Only nearest segment/PAs active	Simple/trivial hardware
Segment Aggregation (SA)	Aggregate output power across segments	Enhanced SNR, single chain
Segment Multiplexing (SM)	Each segment independent (M chains)	Maximum multiplexing, full DoF

Each protocol represents a design tradeoff between complexity, achievable gain, throughput, and channel estimation/feedback requirements (Ouyang et al., 12 Sep 2025, Gan et al., 20 Nov 2025).

3. Optimization, Control, and Algorithms

Segmented PASS supports both single- and multi-user operation, with problem formulations built around the joint optimization of digital beamforming (precoder/combiner matrices) and analog PA position/activation vectors.

• Alternating Optimization: All major design problems (sum-rate, multicast rate, secrecy rate, MSE, energy efficiency) are tackled by block-coordinate descent or AO frameworks, alternating between (i) digital domain (power/precoder update, often convex), and (ii) pinching beamforming (continuous/discrete PA position optimization, non-convex) (Zhao et al., 25 Feb 2025, Bereyhi et al., 5 Mar 2025, Zhao et al., 20 Aug 2025, Zhou et al., 29 Oct 2025).
• PSO and Matching Theory: Non-convex PA placement is often addressed by particle-swarm optimization (PSO) or matching-based algorithms for high-dimensional discrete search; both provide monotonically non-decreasing convergence and approach full-search performance with orders-of-magnitude lower complexity (Zhao et al., 25 Feb 2025, Wang et al., 21 May 2025).
• Penalty/Dual Decomposition: Complex nonlinear constraints (e.g., phase alignment, coupling equality, activation exclusivity) are treated by augmented Lagrangian or penalty dual decomposition techniques, ensuring constraint satisfaction asymptotically (Zhao et al., 20 Aug 2025).
• Specialized Solutions: For single-user or highly structured scenarios, closed-form or iteration-free solutions for PA location (e.g., Chebyshev center for multicast, segmented phase-alignment for SE-EE tradeoff) are available (Shan et al., 31 May 2025, Zhou et al., 29 Oct 2025).
• Dual-Scale Deployment: For fine-grained spatial/phase alignment, dual-scale (coarse plus piezoelectric fine) actuation protocols are defined (e.g., STT, SAT) and jointly optimized with power and digital beamforming for energy-efficient operation (Gan et al., 31 Oct 2025).

These methodologies underpin a unified design paradigm where segmentation provides convex partitioning of non-convex spatial optimization tasks into tractable subproblems.

4. System-Theoretic Performance and Tradeoffs

Segmentation directly impacts the physical-layer metrics central to wireless applications:

• Array Gain and Path Loss: Placing PAs physically closer to intended user projections minimizes path loss, particularly in high-frequency (mmWave/THz) and obstructed environments (Ouyang et al., 10 Jan 2025, Wang et al., 9 Feb 2025).
• Beamforming Degrees of Freedom: Segmentation allows per-user, per-group, or per-target spatial kernel adaptation by deploying, activating, or moving a subset of PAs aligned with channel realizations and propagation geometry (Bereyhi et al., 5 Mar 2025).
• Efficiency and Scalability: The optimal number of segments and PAs per segment are determined by the tradeoff between array gain (which improves with segmentation up to $N_\star$ per segment), hardware cost, and control granularity (Gan et al., 31 Oct 2025).
• Tradeoff Surfaces: Explicit SE–EE, secrecy–throughput, computation–communication, and multicast–fairness tradeoff regimes illustrate segmentation’s benefits. For instance, in multi-user EE-optimized design, segmentation enables moving the system Pareto frontier outward relative to conventional architectures (Zhou et al., 29 Oct 2025, Gan et al., 20 Nov 2025).
• Wideband and OFDM Effects: Segmentation limits intra-segment group-delay dispersion, reducing cyclic-prefix overhead and phase misalignment in wideband OFDM-PASS settings. Systematic segmentation (shorter per-segment waveguides) is essential to avoiding severe efficiency degradation near the waveguide cutoff (Xiao et al., 27 May 2025).

Numerical results uniformly show segmented architectures achieving higher rate, lower energy, better scaling, and improved robustness (e.g., to shadowing, misplacement, and user/eavesdropper density) over both traditional arrays and continuous non-segmented PASS (Ouyang et al., 10 Jan 2025, Shan et al., 31 May 2025, Gan et al., 20 Nov 2025).

5. Segmented PASS Applications Across Communication and Sensing

Segmented PASS underpins advanced paradigms in emerging wireless applications:

• Multi-User Access: Waveguide-division and segment-switching enable WDMA, multiplexing, and interference management with near-orthogonal user separation (Zhao et al., 25 Feb 2025, Zhao et al., 20 Aug 2025).
• Secure and Covert Communications: By optimal PA placement, segmentation maximizes the SNR gap between intended users and eavesdroppers, including spatial nulling in arbitrary geometries. PAs can be placed near the user even when the adversary is far closer to a conventional array (Zhu et al., 18 Apr 2025, Jiang et al., 14 Apr 2025, Shan et al., 19 Sep 2025).
• Multicast/Group Transmission: Segmentation allows coverage of wide, irregular user distributions, supporting both multicast rate maximization and max-min fairness (e.g., element-wise AO/MM) (Shan et al., 31 May 2025, Zhao et al., 20 Aug 2025).
• Integrated Sensing and Communication (ISAC): Joint digital-transmit and pinching beamforming over segmented architectures yields CRB-minimizing position/sensing accuracy under communication constraints, significantly outperforming massive MIMO baselines (Li et al., 27 Aug 2025, Wang et al., 21 May 2025).
• Energy-Efficient and UAV-Driven Links: Dual-scale segmentation reduces both flight path and communication energy in UAV delivery by enabling “beam duplication” at each delivery node with segment-local activation, substantially improving overall system efficiency (Lv et al., 30 Sep 2025, Gan et al., 31 Oct 2025).
• Internet of Things (IoT): Segmentation enables partial, optimized coverage with analytic calculation of outage probability and average rate. The optimal segment length (coverage) is a function of propagation loss and region geometry for practical deployment (Zhang et al., 5 Sep 2025).

Design guidelines arising from these applications emphasize maximizing the number of controllable segments subject to control/hardware cost, and matching per-segment parameters ($N$, $L$, etc.) to propagation and interference environments.

6. Implementation, Protocol Design, and Future Directions

Practical realization of segmented PASS involves joint advances in mechanical actuation (motorized slides, piezoelectric tuning), low-loss materials, power/latency-aware control protocols, and scalable RF chain integration. Protocols such as STT, SAT, STA, SAA define tradeoffs in pre-deployment, actuation delay, and PA count, with empirical results identifying SAT as the energy-optimal protocol (Gan et al., 31 Oct 2025).

Challenges and open lines:

• High-speed, scalable position control and real-time group reconfiguration.
• Wideband, frequency-selective operation with per-segment phase correction and inter-segment recombination.
• Distributed, cell-less architectures using physically separated segments for collaborative cell-free MIMO and ISAC (Bereyhi et al., 5 Mar 2025, Li et al., 27 Aug 2025).
• Analytical optimization for large networks, including robustness to shadowing, mobility, and segment failure.
• Integrating advanced hardware models (e.g., coupled-mode, full-EM solvers) to improve model–hardware correspondence at sub-mm scale (Wang et al., 9 Feb 2025, Xiao et al., 27 May 2025).

Segmented PASS defines the current state-of-the-art for reconfigurable, low-overhead, and highly adaptable antenna array architectures, offering a scalable solution for high-DoF, user-centric, energy-efficient, and secure wireless links in next-generation communication and sensing systems.