.png){kind=icon}
.png){kind=icon}
Return
Chaotic secure optical communication based on the ESP-LSTM decryption framework
.png){kind=icon}
PRE
AI
DOI：10.1016/j.optcom.2025.132515.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Chaos synchronization is of great importance in chaotic optical communications. However, neural network-based chaos synchronization is prone to chaos offsets. Studies have shown that even extremely small chaos synchronization jitter can significantly affect decryption accuracy at a message transmission rate of 10 gigabits per second (Gbit/s). To maintain decryption accuracy, synchronization offset must be strictly controlled within 25 ps, whereas the response time jitter of neural networks is typically much greater than 25 ps. Addressing this problem, this paper proposes an Embedded Synchronization Persistent Long Short-Term Memory (ESP-LSTM) decryption framework. This framework integrates the chaos synchronization mechanism into the Persistent LSTM (P-LSTM) network and reconstructs the synchronization mechanism through the activation function. Under standard conditions, the bit error rate (BER) of the ESP-LSTM model is 1.7 x 10-5. Across various conditions, ESP-LSTM consistently achieves a lower BER compared to the conventional LSTM and other baseline models.
Keywords:
Security
Chaotic optical communication
Chaos synchronization
Neural network
ESP-LSTM
Journal
IF:
2.5
Papers:
595
Citations:
2.7W