High Resolution Optical Analog Signal Processing And Its Applications Based On Bandwidth Scaling

Broadband communication networks are rapidly evolving to support higher bandwidth and lower latency applications.The number of devices connected to industrial Internet of Things(IoT)is constantly increasing,while applications such as cloud computing,high-definition video,virtual reality,and augmented displays are experiencing rapid growth.In 5G communication systems,the planned frequency bands have already expanded to several tens of gigahertz,and the forthcoming 6G communication systems are targeting frequency bands from 100 gigahertz to 3 terahertz.Electronic systems struggle to maintain consistent response characteristics across such wide frequency ranges,and as clock frequencies increase,power consumption,size,and weight sharply increase.After more than 60 years of development,Moore’s Law is slowing down,and electronic systems will face numerous challenges in dealing with the explosive growth of data in the next generation of information systems.Analog optical signal processing offers advantages such as wide bandwidth,low loss,and excellent real-time performance.It has the potential to become a highly competitive processing method in the nextgeneration communication systems.This paper introduces bandwidth scaling technology into analog optical signal processing to enhance frequency resolution and reduce processing latency,Using optical frequency combs as a bridge,it fully integrates the advantages of optical broadband,low loss,and electrical fine flexibility,resulting in a multifunctional,programmable,and high-resolution photonic processing core,exploring various applications.The main innovations in this paper are as follows:Microwave photonics spectrum analysishis paper introduces an optical short-time Fourier transform(STFT)based on bandwidth scaling,extending one-dimensional spectral analysis to two-dimensional joint time-frequency analysis.The experiment demonstrates the highest-frequency resolution all-optical short-time Fourier transform currently available.Traditional frequency-time spectrum analysis solutions are limited by the finite dispersion in transparent media,resulting in frequency resolutions typically inferior to GHz.In this paper,we propose using bandwidth-magnified electro-optic conversion to achieve parallel output of the test spectrum in the optical domain,breaking the dispersion medium’s limitations on frequency resolution.By utilizing a wavelength-time mapping(WTM)for both serial and parallel conversion,the joint time-frequency distribution is directly obtained through the collection of temporal waveforms.In experiments,we achieved an optical short-time Fourier transform with a frequency resolution of 60 MHz,a bandwidth of 1.98 GHz,and a frame analysis rate of 160 MHz.This was used for joint time-frequency analysis of various non-stationary signals,including frequency-hopping signals and linear frequency-modulated signals.Programmable high-precision phase responseThis paper significantly improves the precision of optical domain phase modulation by two to three orders of magnitude using bandwidth scalability techniques.It also introduces microwave photonic true-time delay and programmable high-precision optical fractional Fourier transform based on RF phase control.In the microwave photonic true-time delay line,the wideband RF signal is divided into multiple sub-bands,and the delay-bandwidth product is overcome by jointly controlling the phase of these sub-bands.The delay is achieved through RF phase control,ensuring compatibility with existing phased-array antennas.Experimental results demonstrate a true-time delay line with a bandwidth of 1.32 GHz and a delay of 11 ns,resulting in a delay-bandwidth product of approximately 13.The key to achieving wideband adjustable true-time delay lies in phase control for each sub-band.This paper further validates the feasibility of true-time delay using a resonator-coupled structure and proposes tuning the delay by changing the phase of the transmission peak.In addition,this paper achieves tunable effective large dispersion through spectral quadratic phase response,which is employed in optical FRFT to address the issues of low resolution and poor tunability.The study reveals the mathematical consistency between frequency-domain discrete quadratic phase and optical fractional Fourier transform.Experimentally,it demonstrates a tunable FRFT with a resolution of 60 MHz and a realtime bandwidth of 4 GHz.Moreover,it achieves pulse compression of linear frequency modulation signals and signal extraction in noisy backgrounds.Temporal imaging based on spectral recombinationThis paper employs bandwidth scaling technolgy to achieve spectrum recombination and investigates two novel architectures for time-domain imaging,namely,time-domain amplification and time reversal.In the timedomain amplification system,wideband optical spectra are coherently compressed through frequency domain sampling and the Vernier effect,effectively amplifying the time axis.This approach eliminates the need for additional digital signal processing in amplitude and phase measurements and completely avoids dependence on highly dispersive media.This marks the first international report of a real-time full-field observation solution for transient optical signals.In experiments,amplitude and phase measurements were conducted for various complex optical signals,achieving a maximum amplification factor of 224 with a system bandwidth exceeding 300 GHz.Furthermore,this paper introduces a time reversal technique based on frequency inversion.By splitting the input signal into multiple sub-bands and using an optical frequency comb as multiple local oscillators,parallel down-conversion is achieved to perform frequency reversal.This approach eliminates the need for dispersive media,reducing the system’s transmission latency and volume.Experimental results demonstrate an instantaneous bandwidth of 2.4 GHz.