Probabilistic Jamming Aided Covert Communication in the Finite Blocklength Regime

Abstract

In this work, the covert communication assisted by a probabilistic jammer is investigated in the finite blocklength regime, in which the jammer emits jamming signals with a certain probability to alleviate its negative effect on legitimate link but still enhance the covertness performance. Specifically, we first derive a tractable expression of covertness constraint based on the average KL divergence. Then, a closed-form expression of average decoding error probability (DEP) involving the effect of probabilistic jamming is presented. Moreover, joint optimization of transmit power and blocklength at the transmitter for maximizing the average effective covert throughput (AECT) is formulated and tackled, which demonstrates the trade-off between the transmit power and blocklength to improve covert communication performance. Our results show that the probabilistic jamming can indeed enhance the covert communication performance, and the jamming probability can be optimized to further improve the AECT.