English

On Achievable Covert Communication Performance under CSI Estimation Error and Feedback Delay

Information Theory 2026-03-03 v2 math.IT

Abstract

Covert communication's effectiveness critically depends on precise channel state information (CSI). This paper investigates the impact of imperfect CSI on achievable covert communication performance in a two-hop relay system. Firstly, we introduce a two-hop covert transmission scheme utilizing channel inversion power control (CIPC) to manage opportunistic interference, eliminating the receiver's self-interference. Given that CSI estimation error (CEE) and feedback delay (FD) are the two primary factors leading to imperfect CSI, we construct a comprehensive theoretical model to accurately characterize their effects on CSI quality. With the aid of this model, we then derive closed-form solutions for detection error probability (DEP) and covert rate (CR), establishing an analytical framework to delineate the inherent relationship between CEE, FD, and covert performance. Furthermore, to mitigate the adverse effects of imperfect CSI on achievable covert performance, we investigate the joint optimization of channel inversion power and data symbol length to maximize CR under DEP constraints and propose an iterative alternating algorithm to solve the bi-dimensional non-convex optimization problem. Finally, extensive experimental results validate our theoretical framework and illustrate the impact of imperfect CSI on achievable covert performance.

Keywords

Cite

@article{arxiv.2404.05983,
  title  = {On Achievable Covert Communication Performance under CSI Estimation Error and Feedback Delay},
  author = {Jiaqing Bai and Ji He and Yanping Chen and Yulong Shen and Xiaohong Jiang},
  journal= {arXiv preprint arXiv:2404.05983},
  year   = {2026}
}

Comments

After discussion among the authors, we found that some theoretical derivations in the manuscript require further confirmation and validation. To ensure the rigor and accuracy of the work, we respectfully request to withdraw the manuscript

R2 v1 2026-06-28T15:48:16.124Z