High-security and low-complexity OCDM transmission scheme based on GAN enhanced chaotic encryption

Area of Science:

• Optical Communications
• Cybersecurity
• Artificial Intelligence

Background:

• Traditional communication systems face challenges in balancing security and efficiency.
• Generative Adversarial Networks (GANs) offer potential for advanced encryption techniques.
• Orthogonal Chirp Division Multiplexing (OCDM) is an emerging transmission scheme.

Purpose of the Study:

• To propose a low-complexity, high-security OCDM transmission scheme.
• To enhance chaotic encryption using GANs for improved communication security.
• To evaluate the performance and security of the proposed scheme.

Main Methods:

• Designed novel network architectures and modified loss functions for GAN-enhanced chaotic encryption.
• Implemented a 70 Gb/s transmission system using a 2 km weakly coupled seven-core fiber.
• Compared the proposed scheme against traditional Orthogonal Frequency Division Multiplexing (OFDM) and chaotic schemes.

Main Results:

• Achieved a maximum receiver sensitivity gain of approximately 1.26 dB compared to OFDM.
• Demonstrated a large encryption keyspace of 1x10^202, significantly enhancing security.
• Reduced running time for encryption to only 0.63% compared to traditional chaotic schemes.

Conclusions:

• The GAN-enhanced chaotic encryption scheme offers remarkable reductions in complexity and improvements in security.
• The proposed OCDM transmission scheme is a promising candidate for next-generation passive optical networks (PONs).
• This integration of AI in optical communication security paves the way for more robust and efficient networks.