Jamming Aided Covert Communication with Multiple Receivers

TL;DR

This paper investigates covert communication with multiple receivers aided by a friendly jammer, optimizing power and rate strategies under different channel variation scenarios to maximize effective or ergodic rates while maintaining covertness.

Contribution

It introduces optimization methods for power and rate allocation in covert multi-receiver communication with jamming, considering both slow and fast fading channels, and provides solutions for non-convex problems.

Findings

Optimal power and rate allocation strategies enhance covert communication performance.

Methods successfully find global and sub-optimal solutions for complex non-convex problems.

Numerical results demonstrate the effectiveness of the proposed optimization approaches.

Abstract

We consider that a transmitter covertly communicates with multiple receivers under the help of a friendly jammer. The messages intended for different receivers are transmitted in mutually orthogonal frequency bands. An adversary observes all these frequency bands aiming at detecting whether or not communication occurs, while the friendly jammer broadcasts jamming signals to degrade the detection performance of the adversary. We consider a block Rayleigh fading channel model and evaluate the performance of covert communication in two situations: 1) the wireless channels vary slowly such that the transmission ends within one channel coherent time block, and 2) the wireless channels vary fast such that the wireless channels have changed several times before the whole transmission is finished. In the former case, subject to a covertness constraint, we maximize the sum of the effective rates by optimizing the transmit power allocation and the transmission rate for each receiver. In the latter case, we take the channel training process into consideration, and subject to a covertness constraint, we maximize the sum of the ergodic rates by optimizing the power allocation and the pilot length. Though both of the two optimization problems are non-convex, we presented methods to find their global optimal solutions. Besides, we also present methods to find sub-optimal solutions with lower computational complexities. Numerical results are presented to evaluate the performance under the two situations.