Applying DFTs-OFDM to QAM-based Quantum Noise Stream Cipher Transmission

Yang, Xiaokun; Wang, Wei; Xu, Cheng; Wang, Kai; Gong, Guangyu; Zhao, Yongli; Hui-bin, Zhang; Gao, Guanjun; Yu, Xiaosong; Zhang, Jie

doi:10.1109/oecc.2018.8729993

Cited by 7 publications

(1 citation statement)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, the mathematical encryption method is having a problem of more vulnerability to computational attacks. In order to solve the mentioned problem, another method used for encryption is physical cryptography which uses the physical phenomenon that is very di cult to decipher [2][3].…”

Section: Introductionmentioning

confidence: 99%

Neural network (ConvNN and CapNN) based joint synchronization of timing and frequency for CO-OFDM system

Kaur

2024

Opt Quant Electron

View full text Add to dashboard Cite

The Quantum Noise Stream Cipher (QNSC) is offering very high security in physical layer of optical ber communication network to meet the today's main requirement of secure networks. However, the problem of timing and carrier frequency offset synchronization was higher for CO-OFDM systems using QNSC for network security. For higher-order modulation systems, the algorithms used to overcome this problem are offering very low accuracy. In addition to this, the chances of attack by an illegal receiver are more due to no encryption used for the preamble creation. Existing algorithms used to create a new preamble is: CNI (conjugate, negative, and image) with PN sequence and CNC (conjugate and negative) with Chu sequence. In this paper, ConvNN (Convolutional Neural Network) and CapNN (Capsule Neural Network) based joint synchronization of timing and frequency for CO-OFDM system is proposed. From the experiments, performed at 10Gbps, it has been illustrated that the proposed networks are giving more robust results. In addition to these proposed networks are offering better and secure performance for both timing and frequency offset synchronization by using less training data.

show abstract

Section: Introductionmentioning

confidence: 99%