期刊
IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY
卷 71, 期 11, 页码 12315-12320出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TVT.2022.3190382
关键词
Proactive eavesdropping; massive MIMO-OFDM; deep reinforcement learning; beam misleading; spoofing relay
资金
- National Key R&D Programof China [2019YFE0113200]
- National NaturalScience Foundation of China [U1936201, 62072229, 62071220, 61976113]
In this paper, a scheme for monitoring suspicious links in massive MIMO-OFDM systems is proposed. The scheme uses proactive eavesdropping and implements beam misleading and data eavesdropping algorithms. The optimal precoders and power split factors are found using the MADDPG algorithm. Simulation results demonstrate the effectiveness and superiority of the proposed eavesdropping scheme.
It is challenging to monitor suspicious links in massive multiple input multiple output-orthogonal frequency division multiplexing (MIMO-OFDM) systems with directional beamforming. When the monitor and suspicious user are not in the coverage of the same beam, to implement legitimate monitoring, we propose the proactive eavesdropping scheme in which the monitor acts as a spoofing relay to realize beam misleading and data eavesdropping. In the phase of beam sweep, the beam misleading algorithm induces the transmitter to choose the beam which is beneficial to the monitor by optimizing the relay precoding matrices. In the phase of data transmission, the data eavesdropping algorithm elevates the monitoring rate by optimizing the relay power split factors and power gain factors of all subcarriers. Due to the unknown channel state information between suspicious pairs, we propose to find the optimal precoders and power split factors by the multi-agent deep deterministic policy gradient (MADDPG) algorithm. The simulation results verify the effectiveness and superiority of the proposed eavesdropping scheme.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据