Multi-Channel Physical Random Number Generation Based On Broadband Chaotic Entropy Source

Random numbers play an indispensable role in the fields of high-speed communication system and computer technology.The wave of 5G era has promoted the rapid development information communication,Internet of Things(IoT),artificial intelligence and other industries,which has led to a sharp increase in all kinds of information.When the systems are required to transmit information with large capacity and high-speed,corresponding information security risks will also come along.In order to adapt to the situation of faster communication network and ensure the security of information,it is of great significance to generate confidential and reliable random numbers for information encryption.Physical random numbers have attracted the attention of scholars due to their characteristics of high entropy,excellent randomness and difficulty in cracking.However,limited by the bandwidth of the entropy source,the true random number generation rate based on the traditional physical entropy source(such as single photon randomness,electronic noise,ect.),is at the level of Mbps,which is difficult to meet the absolute security requirements of high-speed communication.In recent years,chaotic laser is an ideal entropy source for extracting high-speed true random numbers due to its unique features of ultra-broadband bandwidth,sensitive to initial value and soon.For the existing single channel true random number generation technology based on chaotic laser,it is difficult to be further improve the real-time generation rate affected by the electronic analog-to-digital converter(ADC)jitter bottleneck.Therefore,we propose an experimental scheme for generating multi-channel physical random numbers using multi-mode Fabry-Perot(FP)semiconductor laser combined with filter components.The device requirements for each channel can be reduced by multi-channel parallel processing,at that time,the generation rate of true random numbers can be improved.In addition,we propose a scheme of photon-integrated multi-bit ADC based on nonlinear microring resonators,which provides theoretical support for real-time all-optical quantification of true random numbers.The specific works have been mainly studied from the following aspects:1.We described briefly the importance of random numbers in the field of ultra-high-speed communication,and summarize the single-channel and multi-channel physical random number generation method utilizing chaotic laser.In addition,we introduce emphatically the research status of all-optical quantizer.2.We propose an experimental scheme for multi-channel parallel true random numbers using multi-mode chaotic laser.Specifically,the experimental system of optical feedback multi-mode FP laser combined with three filters for parallel output of true random numbers is structured,and experimentally analyze the radio frequency(RF)spectra of the single-mode and the multi-mode output after and before filtering.The obtained single-mode flat chaotic signals with the sampling rate of 40 GSa/s are quantized into digital bits 0 or 1 by an 8-bit ADC.By selecting 3 LSBs,we can generate true random number with a cumulative rate of 3×120 Gbps(40 GSa/s×3 LSBs).3.We simulate the mathematical model of the optical feedback multi-mode laser,analyze theoretically the spectral characteristics of the multi-mode and single-mode chaotic signals.The simulation results show that the essential reason for the flattening of the single-mode chaotic spectrum caused by the mode-competing in multi-mode lasers.Furthermore,we use the Shannon entropy growth rate to analyze the influence of the selection of LSBs on randomness in multi-bit quantization technology.4.A multi-bit all-optical quantization structure based on nonlinear microring resonator is constructed and the influence of key parameters on the threshold of 1-bit quantizer structure is explored.Paralleling quantizers with transmission characteristics of different periods and high extinction ratios,can achieve 2-bit and 3-bit all-optical quantization and coding.5.We summarize the work of this thesis and discuss the next step.