Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Cross-Layer Security for Semantic Communications: Metrics and Optimization

Lingyi Wang, Wei Wu, Fuhui Zhou, Zhijin Qin, Qihui Wu

TL;DR

The paper tackles the security of goal-oriented semantic communications by introducing a unified cross-layer metric, the cross-layer semantic security rate $\Omega_u^{S}$, and formulating its maximization as a mixed integer nonlinear program over beamforming, symbol bits, and semantic decoders. To solve the NP-hard problem, it proposes a policy-iteration RL-enhanced resource allocation scheme modeled as an MDP, with a reward that integrates both cross-layer security gains and channel conditions. The authors prove convergence of the RL scheme under finite action/state spaces and demonstrate, through simulations, that the proposed CLSS approach significantly improves CLSSR and task reliability compared to traditional physical-layer semantic security methods, while enabling faster convergence. The work advances cross-layer security design for semantic communications in AI-native networks, offering practical gains in task reliability and secure-rate metrics under realistic wireless conditions.

Abstract

Different from traditional secure communication that focuses on symbolic protection at the physical layer, semantic secure communication requires further attention to semantic-level task performance at the application layer. There is a research gap on how to comprehensively evaluate and optimize the security performance of semantic communication. In order to fill this gap, a unified semantic security metric, the cross-layer semantic secure rate (CLSSR), is defined to estimate cross-layer security requirements at both the physical layer and the application layer. Then, we formulate the maximization problem of the CLSSR with the mixed integer nonlinear programming (MINLP). We propose a hierarchical AI-native semantic secure communication network with a reinforcement learning (RL)-based semantic resource allocation scheme, aiming to ensure the cross-layer semantic security (CL-SS). Finally, we prove the convergence of our proposed intelligent resource allocation, and the simulation results demonstrate that our proposed CLSS method outperforms the traditional physical layer semantic security (PL-SS) method in terms of both task reliability and CLSSR.

Cross-Layer Security for Semantic Communications: Metrics and Optimization

TL;DR

The paper tackles the security of goal-oriented semantic communications by introducing a unified cross-layer metric, the cross-layer semantic security rate , and formulating its maximization as a mixed integer nonlinear program over beamforming, symbol bits, and semantic decoders. To solve the NP-hard problem, it proposes a policy-iteration RL-enhanced resource allocation scheme modeled as an MDP, with a reward that integrates both cross-layer security gains and channel conditions. The authors prove convergence of the RL scheme under finite action/state spaces and demonstrate, through simulations, that the proposed CLSS approach significantly improves CLSSR and task reliability compared to traditional physical-layer semantic security methods, while enabling faster convergence. The work advances cross-layer security design for semantic communications in AI-native networks, offering practical gains in task reliability and secure-rate metrics under realistic wireless conditions.

Abstract

Different from traditional secure communication that focuses on symbolic protection at the physical layer, semantic secure communication requires further attention to semantic-level task performance at the application layer. There is a research gap on how to comprehensively evaluate and optimize the security performance of semantic communication. In order to fill this gap, a unified semantic security metric, the cross-layer semantic secure rate (CLSSR), is defined to estimate cross-layer security requirements at both the physical layer and the application layer. Then, we formulate the maximization problem of the CLSSR with the mixed integer nonlinear programming (MINLP). We propose a hierarchical AI-native semantic secure communication network with a reinforcement learning (RL)-based semantic resource allocation scheme, aiming to ensure the cross-layer semantic security (CL-SS). Finally, we prove the convergence of our proposed intelligent resource allocation, and the simulation results demonstrate that our proposed CLSS method outperforms the traditional physical layer semantic security (PL-SS) method in terms of both task reliability and CLSSR.

Paper Structure

This paper contains 12 sections, 2 theorems, 14 equations, 3 figures.

Table of Contents

  1. Introduction
  2. Secure Semantic Communication
  3. Physical Layer Security and Application Layer Security
  4. Physical Layer Security
  5. Application Layer Security
  6. Cross-Layer Semantic Security
  7. CLSSR Maximization Problem Formulation
  8. Proposed Policy Iteration RL-enhanced Intelligent Resource Allocation Scheme
  9. MDP Model
  10. The Convergence of Proposed RL for CLSSR Maximization
  11. Simulation And Analyses
  12. Conclusion

Key Result

Lemma 1

The policy $\pi$ is iteratively improved.

Figures (3)

  • Figure 1: The Secure semantic communication Model.
  • Figure 2: The reliability convergence of semantic security models.
  • Figure 3: The achievable CLSSR of semantic secure communication models.

Theorems & Definitions (9)

  • Definition 1
  • Remark 1
  • Remark 2
  • Remark 3
  • Remark 4
  • Lemma 1
  • proof
  • Theorem 1
  • proof