Covert Capacity of AWGN Channels under Average Error Probability

TL;DR

This paper establishes bounds on the covert capacity of AWGN channels considering average error probability and Kullback-Leibler divergence, revealing the absence of a strong converse in this context.

Contribution

It provides the first tight bounds on covert capacity under average error probability and introduces a novel codebook analysis with BPSK and all-zero codewords.

Findings

Bounds confirm no strong converse in reliability and covertness

Codebook analysis with BPSK and all-zero codewords is effective

Characterization of covert capacity limits in AWGN channels

Abstract

We derive upper and lower bounds for the covert capacity of Additive White Gaussian Noise channels when measuring reliability in terms of the average error probability and covertness in terms of Kullback-Leibler divergence. This characterization confirms the absence of strong converse for this setting in both the reliability and covertness parameters. The crux of our approach is to analyze a codebook of BPSK-modulated codewords carefully augmented with "all-zero" codewords.