Proactive Eavesdropping via Jamming in UAV-Enabled Relaying Systems With Statistical CSI

Abstract

We investigate proactive eavesdropping with one half-duplex legitimate monitor (E) in unmanned aerial vehicle (UAV)-enabled suspicious relaying systems, which consist of a suspicious transmitter (ST), a suspicious UAV-based relay (SU) and a suspicious destination (SD). Under this setup, we propose two jamming-assisted eavesdropping strategies, namely, &#x201C;Eavesdrop-Then-Jam&#x201D; (ETJ) and &#x201C;Jam-Then-Eavesdrop&#x201D; (JTE), to maximize the eavesdropping throughput of E. Specifically, for ETJ, E first eavesdrops the suspicious signal of the ST in the ST-SU phase and then jams the SD in the SU-SD phase; for JTE, E first jams the SU in the ST-SU phase and then eavesdrops the suspicious signal of the SU in the SU-SD phase. <i>However, in both strategies, as the jamming power of E (independent variable) changes, the SU can also adaptively adjust its deployment (dependent variable) to maximize the benefit of the suspicious system</i>. Facing this challenge, we can only employ the undesirable exhaustive search over both the independent and dependent variables to achieve the optimal solution. To decrease the complexity, we further derive a tight approximation of the Lambert function and then develop easy-to-implement algorithms to find the sub-optimal solutions. Numerical results demonstrate the effectiveness of our proposed strategies compared to conventional passive eavesdropping.