STAR-RIS in Cognitive Radio Networks

Haochen Li, Yuanwei Liu, Xidong Mu, Yue Chen, Zhiwen Pan, Xiaohu YouIEEE Trans. Wireless Communications 2024 · DOI

Introduction

As the industry moves toward 6G, the "spectrum crunch" has become a critical bottleneck. Cognitive Radio (CR) offers a promising solution by allowing secondary users (SUs) to opportunistically access spectrum owned by primary users (PUs), provided interference is strictly managed. However, traditional Reconfigurable Intelligent Surfaces (RIS) are limited by 180-degree coverage, as they can only reflect signals. This geographical constraint limits the flexibility of CR networks.

This paper introduces Simultaneously Transmitting and Reflecting Reconfigurable Intelligent Surfaces (STAR-RIS) into MIMO cognitive radio networks. Unlike conventional RIS, STAR-RIS acts as a "smart window," providing 360-degree coverage by simultaneously transmitting and reflecting signals. This allows for a more robust and flexible deployment of secondary networks. As shown in the literature comparison below, this work uniquely addresses the combination of STAR-RIS, MIMO, and multiple secondary users.

Comparison of the proposed framework with existing literature.

System Model and Methodology

The proposed framework considers a MIMO CR system operating in "underlay" mode, where a Secondary Base Station (SBS) serves multiple SUs while coexisting with a Primary Base Station (PBS) and multiple PUs. The STAR-RIS is deployed to enhance the secondary network's performance while nulling interference at the PUs.

System model of a STAR-RIS aided MIMO cognitive radio network.

The system divides the space into Transmission (T) and Reflection (R) regions. To maximize the sum-rate of the secondary network under interference temperature (IT) constraints, the authors employ a Block Coordinate Descent (BCD) framework to solve a complex non-convex optimization problem:

  1. Active Beamforming: SBS beamformers are optimized via Quadratically Constrained Quadratic Programming (QCQP).
  2. Passive Beamforming (Independent Model): For theoretical scenarios with independent phase shifts, Successive Convex Approximation (SCA) is used.
  3. Passive Beamforming (Coupled Model): For realistic hardware where transmission and reflection phases are linked, a Penalty Dual Decomposition (PDD) algorithm is developed to handle highly non-linear constraints.

Key Insights and Results

The study yields several critical insights for 6G network design:

  • Full-Space Freedom: STAR-RIS removes the "blind spots" of conventional RIS, allowing SUs to be served regardless of their position relative to the surface.
  • Interference Mitigation: The STAR-RIS is leveraged not just for signal enhancement but to create destructive interference at PU locations, effectively "hiding" the secondary network's footprint.
  • Hardware Resilience: While hardware coupling between transmission and reflection phases is a known challenge, the PDD-based optimization demonstrates that the performance loss compared to an ideal independent model is minimal.
  • Performance Gains: Monte Carlo simulations confirm that the STAR-RIS framework consistently outperforms conventional RIS and no-RIS baselines. Both the SCA and PDD algorithms converge rapidly (within ~20 iterations), making them suitable for practical implementation.

Conclusion and Future Work

STAR-RIS significantly enhances the spectral efficiency of cognitive radio networks by providing full-space signal control. While this work establishes a strong foundation, future research is needed to address the overhead of Channel State Information (CSI) estimation and the transition to near-field communication models necessitated by extremely large antenna arrays in 6G.

Summary updated

BibTeX
@article{li2024star,
  author = {Haochen Li and Yuanwei Liu and Xidong Mu and Yue Chen and Zhiwen Pan and Xiaohu You},
  title = {{STAR-RIS in Cognitive Radio Networks}},
  journal = {IEEE Trans. Wireless Communications},
  year = {2024},
  doi = {10.1109/twc.2024.3485819},
}
Figure