STARS-assisted Near-field ISAC: Sensor Deployment and Beamforming Design

Na Xue, Xidong Mu, Yue Chen, Yuanwei Liu

In the race toward 6G, Integrated Sensing and Communication (ISAC) has emerged as a flagship technology. The goal is simple but ambitious: use the same wireless signals and hardware to both transmit data and "see" the world. However, traditional approaches often struggle with signal attenuation over long round-trips.

A new paper by Na Xue and colleagues, titled "STARS-assisted Near-field ISAC: Sensor Deployment and Beamforming Design," proposes a clever architectural shift. By moving the "eyes" of the system directly onto a smart surface and exploiting the physics of the near-field, they’ve found a way to make sensing more accurate and cost-effective.

TL;DR

This paper proposes a "sensing-at-STARS" architecture where radio sensors are mounted directly onto a Simultaneously Transmitting and Reflecting Surface (STARS). By exploiting near-field spherical wavefronts, the system can estimate a target's position with high precision while minimizing hardware costs through a novel joint optimization of sensor deployment and beamforming.

What Problem Does This Paper Address?

Traditional RIS (Reconfigurable Intelligent Surface)-assisted sensing usually follows a "round-trip" path: the Base Station (BS) sends a signal, the RIS reflects it to a target, the target reflects it back to the RIS, and the RIS sends it back to the BS. This multi-hop journey causes massive signal power loss (path loss).

Furthermore, most existing research assumes far-field propagation, where waves are treated as flat planes. As we move to higher frequencies (mmWave/THz) and larger antenna arrays in 6G, we increasingly operate in the near-field. In this region, waves are spherical. While this adds mathematical complexity, spherical waves actually contain "hidden" distance information that planar waves lack.

The authors tackle two main challenges:

  1. Signal Decay: How to reduce the path loss of sensing.
  2. Resource Trade-off: How to balance the number of sensors (hardware cost) against the required positioning accuracy.

Methods

The authors introduce a STARS-assisted NF-ISAC framework. Unlike a standard RIS that only reflects, a STARS can split an incident signal into a reflected component (for one side of the surface) and a transmitted component (for the other), providing 360-degree coverage.

Key Technical Contributions:

  • Sensing-at-STARS: They co-locate sensors with the STARS elements. This bypasses the long trip back to the Base Station, significantly boosting the signal-to-noise ratio (SNR) of the echoes.
  • Near-Field SPEB Derivation: They derive a new expression for the Squared Position Error Bound (SPEB). They prove that in the near-field, the Fisher Information Matrix (which measures how much information a signal carries about a parameter) can be approximated as a diagonal matrix, simplifying the localization math.
  • Joint Optimization: They formulate a problem to minimize a weighted cost of the SPEB and the deployment cost (number of sensors).
  • Algorithms:
    • For Sensor Deployment, they use Successive Convex Approximation (SCA) to determine the optimal number of sensors and their spacing.
  • For Beamforming, they use a penalty-based iterative method to optimize the active signals at the BS and the passive phase shifts at the STARS.

Datasets

As a theoretical and algorithmic study, the paper relies on Monte Carlo simulations rather than external datasets.

  • Setup: A 128-antenna BS operating at 28 GHz.
  • Hardware: A STARS with 32 elements and a variable number of sensors.
  • Environment: Randomly distributed users and targets in the near-field region (within the Rayleigh distance).

Key Insights and Ideas

  • Spherical Waves are a Feature: In the far-field, you can only easily estimate the angle of a target. In the near-field, the curvature of the spherical wavefront allows the sensors to estimate the distance directly from the phase delays across the array.
  • The Aperture Match: One of the most interesting findings (Proposition 2) is that the best sensing performance occurs when the aperture size of the sensor array matches the aperture size of the STARS elements.
  • Diagonal Approximation: The discovery that the FIM is approximately diagonal in the near-field is a major "aha moment" for researchers, as it suggests that angle and distance estimation are essentially decoupled in this regime, simplifying receiver design.

Key Results

  • Accuracy: The derived SPEB expression was validated against exact numerical models, proving it is a highly reliable metric for system design.
  • Performance Gain: The proposed joint optimization achieved at least a 20dB gain in sensing performance compared to random sensor deployment.
  • Cost Efficiency: The proposed scheme outperformed "Conventional RIS" (which only covers one side) and "Conventional ISAC" (sensing at the BS) while maintaining a lower deployment cost.
  • Convergence: The alternating optimization algorithm was shown to converge quickly (typically within 6 iterations), making it computationally feasible.

Limitations and Future Work

  • Perfect CSI Assumption: The study assumes perfect Channel State Information (CSI). In real-world 6G environments, estimating the channel for both users and targets simultaneously is a massive challenge.
  • Static Scenarios: The model focuses on static targets and users. Future work needs to address high-mobility scenarios (e.g., vehicle-to-everything communication).
  • Hardware Impairments: Real STARS elements and sensors have phase noise and quantization errors that weren't fully modeled here.
  • Multi-target Sensing: Extending the framework to resolve multiple targets in a cluttered environment remains an open area for exploration.

Summary updated

Abstract

A simultaneously transmitting and reflecting surface (STARS) assisted near-field (NF) integrated sensing and communication (ISAC) framework is proposed, where the radio sensors are installed on the STARS to directly conduct the distance-domain sensing by exploiting the characteristic spherical wavefront. A new squared position error bound (SPEB) expression is derived to reveal the dependence on beamforming (BF) design and sensor deployment. To balance the trade-off between the SPEB and the sensor deployment cost, a cost function minimization problem, a cost function minimization problem is formulated to jointly optimize the sensor deployment, the active and passive BF, subject to communication and power consumption constraints. For the sensor deployment optimization, a joint sensor deployment algorithm is proposed by invoking the successive convex approximation. Under a specific relationship between the sensor numbers and BF design, we derive the optimal sensor interval in a closed-form expression. For the joint BF optimization, a penalty-based method is invoked. Simulation results validated that the derived SPEB expression is close to the exact SPEB, which reveals the Fisher information Matrix of position estimation in NF can be approximated as a diagonal matrix. Furthermore, the proposed algorithms achieve the best SPEB performance than the benchmark schemes accompanying the lowest deployment cost.
BibTeX
@article{xue2026stars,
  author = {Na Xue and Xidong Mu and Yue Chen and Yuanwei Liu},
  title = {{STARS-assisted Near-field ISAC: Sensor Deployment and Beamforming Design}},
  journal = {IEEE Trans. Wireless Communications},
  year = {2026},
  doi = {10.1109/twc.2025.3616074},
}
Figure