Source-linked AI summary
Joint Discrete Pinching-Antenna Activation and Beamforming Design for Efficient PASS-ISAC
Joshua Ogundiya, Animesh Yadav, Octavia Dobre
TL;DR
PASS offers flexible radiating points for ISAC, but much prior work assumes continuous antenna positioning, which may not fit pre-installed sites, and discrete-activation studies have largely left multi-waveguide PASS-ISAC unexplored. This paper jointly optimizes beamforming and sparse TPA/RPA activation with JDPABO. Simulations show that it outperforms fixed activation schemes and approaches the continuous-placement benchmark at low sensing SNR requirements.
Problem
Most PASS designs assume continuous antenna positioning, while discrete-activation designs have largely focused on communication rather than multi-waveguide PASS-ISAC.
Source: §I · p. 1Method
JDPABO jointly optimizes beamforming and sparse TPA/RPA activation for multi-waveguide PASS-ISAC using reweighted ℓ1 minimization and successive convex approximation.
Source: §I · p. 1Results
JDPABO outperforms fixed activation schemes and closely approaches the continuous-location benchmark at low sensing SNR requirements.
Source: §IV · p. 5, §V · p. 6Takeaways & Limitations
The simulations show communication-sensing balance with only a limited number of active TPAs/RPAs.
Source: §I · p. 1Discrete activation is restricted to predefined candidate locations, and its rate gap from the continuous-location benchmark grows at higher sensing SNR requirements.
Source: §IV · p. 5
Abstract
from arXiv · showhide
Pinching-antenna systems (PASS) offer flexible radiation points for integrated sensing and communication (ISAC), but existing designs largely assume continuous antenna placement, which may be impractical with pre-installed hardware. We investigate a discrete-activation multi-waveguide PASS-ISAC system that selects sparse transmit and receive pinching antennas (TPAs/RPAs) from predefined locations. We formulate a communication-rate maximization problem by jointly optimizing beamforming and TPA/RPA activation under transmit-power, activation, and sensing signal-to-noise ratio (SNR) constraints. To solve this nonconvex problem, we develop a low-complexity joint discrete pinching-antenna activation and beamforming optimization (JDPABO) algorithm based on alternating optimization, successive convex approximation, and reweighted $\ell_1$ minimization. Simulations show that JDPABO approaches continuous-placement performance at low sensing SNR requirements and outperforms fixed activation schemes with sparse PA activation.
I. Introduction
The paper addresses the gap between practical, predefined antenna sites and PASS-ISAC designs that largely assume continuous positioning. It introduces a discrete-activation multi-waveguide system and jointly optimizes beamforming and sparse TPA/RPA activation.
I. Introduction
- PASS can strengthen line-of-sight links, reduce path loss, and add positional degrees of freedom by activating radiating points near users or sensing targets.Source: §I · p. 1
- Discrete activation addresses the practical limits of continuous antenna positioning with pre-installed sites, while multi-waveguide PASS-ISAC remains largely unexplored in prior discrete designs –.Source: §I · p. 1
- JDPABO jointly optimizes beamforming and sparse TPA/RPA activation using reweighted ℓ1 minimization and successive convex approximation.The design targets communication-rate maximization under transmit-power, antenna-selection, and sensing constraints.Source: §I · p. 1
- Simulations show that JDPABO balances communication and sensing while requiring only a limited number of active TPAs/RPAs.Source: §I · p. 1
II. System Model
- The system selects sparse TPAs and RPAs from discrete candidate locations on transmit and receive waveguides deployed over a three-dimensional serving area.Candidate sites are uniformly spaced along each waveguide.Source: §II · p. 2
A. Channel Model
The channel model accounts for near-field propagation between waveguide-fed pinching antennas and users or the sensing target, using spherical waves and waveguide coupling.
A. Channel Model
- Near-field propagation is modeled with spherical-wave channels.Source: A. Channel Model · p. 2
- The model includes free-space channels from TPAs to the user and target, and from the target to RPAs.Source: A. Channel Model · p. 2
- Waveguide-to-antenna channels include coupling coefficients and propagation distances from each waveguide feed point to its TPA or RPA.Source: A. Channel Model · p. 2
B. Communication Model
The communication model represents transmit-side antenna selection with binary activation variables and combines the selected channels with waveguide beamforming to describe the received user signal.
B. Communication Model
- Each TPA has a binary activation state, and each transmit waveguide is subject to a maximum number of simultaneously active TPAs.Source: B. Communication Model · p. 2
- The received-user-signal expression uses the aggregate channel and waveguide beamforming vector, with additive white Gaussian noise.Source: B. Communication Model · p. 3
- The user-channel model collects free-space TPA-to-user channels and in-waveguide channels, with the TPA activation encoded by a diagonal selection matrix.Source: B. Communication Model · p. 3
C. Sensing Model
The sensing model represents RPAs through binary activation variables and a selection matrix, then evaluates the target echo after receive combining. It also identifies a receive beamformer that maximizes sensing SNR for fixed activations and transmit beamforming.
C. Sensing Model
- Each RPA has a binary activation variable, and the receive selection matrix is B = diag(b).The variables are stacked into the receive activation vector b.Source: C. Sensing Model · p. 3
- The sensing quality is measured by SNR based on post-combining echo strength relative to noise power.The echo is collected by active RPAs from a point-like target.Source: C. Sensing Model · p. 3
- The receive beamformer is chosen to maximize sensing SNR for fixed transmit beamforming and antenna activations.The maximizing direction is BGrft.Source: C. Sensing Model · p. 3
D. Problem Formulation
The formulation maximizes communication rate by jointly optimizing transmit beamforming and sparse TPA/RPA activations, subject to power, activation-count, and minimum sensing-SNR constraints. The resulting problem is highly nonconvex, with combinatorial antenna selection and coupled communication and sensing terms.
D. Problem Formulation
- Communication-rate maximization jointly chooses transmit beamforming and sparse TPA/RPA activations under power, activation-count, and sensing-SNR constraints.The sensing constraint requires radar SNR to meet γs,min.Source: D. Problem Formulation · p. 3
- The constraints cap normalized transmit power at unity and limit the total numbers of active TPAs and RPAs.Source: D. Problem Formulation · p. 3
- The problem is highly nonconvex because ℓ0 sparsity makes antenna selection combinatorial, while communication and sensing terms are coupled.The passage notes that the antenna-selection problem is NP-hard in general.Source: D. Problem Formulation · p. 3
III. Proposed Algorithm
The proposed algorithm uses alternating optimization to address the nonconvex problem by updating beamforming and antenna activations in turn. These updates repeat until convergence.
III. Proposed Algorithm
- For fixed TPA and RPA activation patterns, the algorithm optimizes the transmit beamforming vector.Source: §III · p. 3
- With the beamforming vector fixed, it optimizes the TPA and RPA activation variables.Source: §III · p. 3
- The beamforming and activation subproblems are solved alternately until convergence.Source: §III · p. 3
A. Subproblem 1
Subproblem 1 optimizes transmit beamforming with TPA and RPA activations fixed. It uses slack variables and successive convex approximation to handle nonconvex terms, then solves a convex approximation to update the beamformer.
A. Subproblem 1
- With binary activation matrices A and B fixed, subproblem 1 optimizes the transmit beamforming vector w.Source: A. Subproblem 1 · p. 4
- Slack variables ρu and ρt reformulate the communication objective and sensing constraint, whose quadratic lower bounds remain nonconvex.Source: A. Subproblem 1 · p. 4
- Successive convex approximation replaces the nonconvex quadratic terms with first-order Taylor lower bounds and affine constraints.Source: A. Subproblem 1 · p. 4
- The resulting convex problem is solved for w, which is then used in the antenna-activation update.Source: A. Subproblem 1 · p. 4
B. Subproblem 2
The activation subproblem fixes the beamforming vector and jointly optimizes TPA/RPA activation. Slack-variable reformulation, successive convex approximation, and reweighted ℓ1 minimization produce a convex problem for the alternating-optimization algorithm.
B. Subproblem 2
- With the transmit beamforming vector fixed, the second subproblem optimizes the TPA and RPA activation patterns.The resulting activation variables are used in the next alternating-optimization step.Source: B. Subproblem 2 · p. 4
- Slack variables reformulate the communication and sensing terms in the challenging activation problem.The problem also includes sparsity constraints and binary activation variables.Source: B. Subproblem 2 · p. 4
- SCA approximates the remaining nonconvex constraints, while reweighted ℓ1 minimization relaxes binary constraints and promotes sparse activation.Sparsity-promoting penalties are added to the objective.Source: B. Subproblem 2 · pp. 4–5
- The resulting problem (28) is convex and can be solved efficiently with standard convex optimization solvers.It is the activation step in the overall AO algorithm for JDPABO.Source: B. Subproblem 2 · p. 5
IV. Simulation Results
JDPABO approaches the continuous-location benchmark at low sensing SNR requirements, while the gap grows at higher requirements, especially when user and target are farther apart. Against fixed-activation baselines, it achieves the highest communication rate and converges within a few iterations in the tested cases.
IV. Simulation Results
- The three representative cases use M = 2 and 81 candidate sites for each antenna type, with 0.5 m spacing.This keeps the candidate-grid resolution consistent with the exhaustive-search setting in.Source: §IV · p. 5
- At γs,min = 4 dB, JDPABO converges within a few AO iterations in both representative cases.The convergence plot tracks communication rate against AO iterations.Source: Figure 2 · p. 5, §IV · p. 5
- JDPABO achieves the highest communication rate as transmit power increases; the target-centric baseline is closest, while fixed activation performs worst.The target-centric scheme improves target illumination, whereas fixed activation cannot adapt to user-target geometry.Source: Figure 4 · p. 6, §IV · p. 6
- At low sensing SNR requirements, JDPABO closely approaches the continuous-location benchmark, but its rate gap grows at higher requirements, especially in Cases 2 and 3.The abrupt rate reductions reflect TPA shifts toward target-favorable locations to satisfy sensing constraints, with a larger communication-rate penalty.Source: §IV · p. 5, Figure 3 · p. 6
V. Conclusion
The paper concludes that JDPABO jointly optimizes waveguide beamforming and sparse TPA/RPA activation for discrete-activation multi-waveguide PASS-ISAC. Simulations show it outperforms fixed activation schemes and approaches the continuous-placement benchmark at low sensing SNR requirements.
V. Conclusion
- The system selects TPAs/RPAs from pre-installed candidate locations and jointly optimizes waveguide beamforming with sparse activation.Source: §V · p. 6
- JDPABO outperforms fixed activation schemes and approaches the continuous-placement benchmark at low sensing SNR requirements.Fig. 5 plots communication rate versus sensing SNR requirement under random user and target locations.Source: §V · p. 6, Figure 5 · p. 6
- The communication-rate maximization is subject to transmit-power, antenna-selection, and sensing SNR constraints.Source: §V · p. 6