A Pilot Contamination Attacker-Defender Model for Wireless Networks under Stackelberg Game.

Zhangnan Wang,Yichen Wang
DOI: https://doi.org/10.1109/vtc2022-spring54318.2022.9860696
2022-01-01
Abstract:Existing studies on pilot contamination attacks often assume that the enemy has not made any strategic corrections to the detection plan. In this paper, we assume that the intelligent malicious user can obtain legitimate users' pilot sequence and power in a timely and accurate manner. Based on the acquired pilot information, the attacker adjusts the pilot contamination attack strategy during the reverse training phase and the jamming power during the data transmission phase to improve his eavesdropping performance and reduce the downlink transmission rate of legitimate users. By modeling the defender-attacker interaction as a Stackelberg game, Bob as the leader chooses his pilot training power, while a full-duplex eavesdropper as the follower determines the pilot contamination power according to the observed Bob's ongoing training signals transmission. In addition, we analyze Bob's pilot transmission power and secrecy rate under complete and incomplete information conditions. Simulation results show that the proposed scheme can defend against an intelligent active eavesdropper with a higher secrecy rate and utility.
What problem does this paper attempt to address?