Exact Throughput-Covertness Relation for Covert Communications

Daizhong Yu,Lin Yang,Jiawei Liu,Guangrong Yue
DOI: https://doi.org/10.1109/globecom48099.2022.10001208
2022-01-01
Abstract:Covert communication is crucial in privacy protection for users of wireless networks, and efforts have been made to improve the system throughput subject to a constraint on a warden's detection error probability. However, most existing research is based on an unclear relation between system throughput and covertness, due to the use of the Kullback-Leibler divergence. To solve this problem, in this paper, we explore the exact throughput-covertness relation by investigating the likelihood variation distance. An algorithm is established to find the minimum channel uses required for the transmission of a given amount of information and a certain covertness constraint. The rigorous proof of the convergence of the algorithm is also given. Our results show that an overly strict constraint on covertness would lead to a large sacrifice of system throughput, and it is not cost-effective to increase the number of channel uses to improve covertness if only a small amount of information is to be transmitted. These results could be very helpful for system designers in achieving a desirable tradeoff between system throughput and covertness.
What problem does this paper attempt to address?