跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2305.17924 · v8

First- and Second-Order Bounds for Covert Communication Over AWGN Channels

首次提交 2023-05-29 15:32
版本更新 2026-11-13 09:06
雷达观测 2026-11-17 04:57
arXiv ↗PDF ↗

摘要

This paper investigates the first- and second-order asymptotic behavior of covert communication over additive white Gaussian noise (AWGN) channels under a Kullback--Leibler (KL) divergence covertness constraint. We derive explicit achievable first- and second-order throughput terms and a converse bound in terms of the blocklength, decoding error probability, adversary noise variance, and covertness constraint. The achievability analysis employs truncated Gaussian random coding with an arbitrary fixed parameter μ∈(1/2,1)μ\in(1/2,1), a common decoding threshold accommodating codewords with different energies, and expurgation to meet a maximal-error reliability requirement. We also establish sufficient secret-key sizes and identify the dependence of key consumption on the adversary noise variance and μμ. Both throughput bounds exhibit O(n)O(\sqrt{n}) first-order and O(n1/4)O(n^{1/4}) second-order scalings. Although their coefficients differ for each fixed μ<1μ<1, the ratio of the achievable to converse first-order coefficients equals μμ. Thus, choosing a fixed μμ arbitrarily close to one makes the first-order coefficient gap arbitrarily small. The results quantify the tradeoffs among throughput, reliability, covertness, and secret-key resources in finite-blocklength AWGN covert communication.

AI 论文解读

基于摘要 · 非全文审阅

此版本尚未生成 AI 解读。解读由后台任务生成并按论文版本保存,登录后共享已有结果。

版本差异

本站首次收录为 v8,暂无更早版本可对照。

观测到的版本

相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。

v8 · 2026-11-13 09:06 · 最新观测版本

First- and Second-Order Bounds for Covert Communication Over AWGN Channels

This paper investigates the first- and second-order asymptotic behavior of covert communication over additive white Gaussian noise (AWGN) channels under a Kullback--Leibler (KL) divergence covertness constraint. We derive explicit achievable first- and second-order throughput terms and a converse bound in terms of the blocklength, decoding error probability, adversary noise variance, and covertness constraint. The achievability analysis employs truncated Gaussian random coding with an arbitrary fixed parameter μ∈(1/2,1)μ\in(1/2,1), a common decoding threshold accommodating codewords with different energies, and expurgation to meet a maximal-error reliability requirement. We also establish sufficient secret-key sizes and identify the dependence of key consumption on the adversary noise variance and μμ. Both throughput bounds exhibit O(n)O(\sqrt{n}) first-order and O(n1/4)O(n^{1/4}) second-order scalings. Although their coefficients differ for each fixed μ<1μ<1, the ratio of the achievable to converse first-order coefficients equals μμ. Thus, choosing a fixed μμ arbitrarily close to one makes the first-order coefficient gap arbitrarily small. The results quantify the tradeoffs among throughput, reliability, covertness, and secret-key resources in finite-blocklength AWGN covert communication.

查看此版本解读 →查看此版本原文 ↗