.png){kind=icon}
.png){kind=icon}
Return
Toward Reliable Covert Communications: A Pilot-Augmented Intelligent Receiver Based on Hybrid Deep Neural Networks
.png){kind=icon}
PRE
AI
Z
Y
G
J
J
X
T
Y
DOI：10.1109/lcomm.2026.3708571.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In covert communication systems, users hide their transmissions using high-power artificial noise (AN). While this improves covertness, the resulting interference distorts the covert signal at the receiver, degrading demodulation reliability. To address this issue, this letter proposes a deep learning (DL) based covert signal receiving network (CRNet). This network fully considers the impact of pilots on covert signal demodulation, constructs a signal compensation layer via a fully connected neural network (FCNN), and combines a convolutional neural network (CNN) with a bidirectional long short-term memory network (BiLSTM) to extract signal features while suppressing AN jamming. Experimental results demonstrate that when high-power AN is used to enhance covertness, the proposed CRNet ensures a significantly lower bit error rate (BER) compared to conventional demodulation methods.
Keywords:
Reliable covert communication
deep learning (DL)
artificial noise (AN)
Journal
IF:
4.4
Papers:
1.2W
Citations:
2.2W
Organization
Cited Papers