Research On Quantization And Coding Method In Optical Analog-to-Digital Converter

As the key front-end signal processor,the analog-to-digital converter(ADC)is a bridge connecting the analog and digital world.At present,ADCs with higher sampling rate,broader bandwidth and better performance have become an indispensable part of high-speed optical communication systems,ultra-wideband wireless systems,and other real-time signal acquisition and processing systems that require high-speed digitization of input signals.The performance of traditional electronic ADCs is limited in these advanced high-frequency applications due to their aperture jitter and comparator ambiguity.The optical ADC that uses ultra-low jitter optical pulses as the sampling source has the possibility of overcoming the electronic bottleneck due to its high sampling rate and high bandwidth.Thus,it is a promising technology for ultra-high-speed ADCs in the future.In recent years,the research on optical ADCs has mainly focused on the design of optical devices or experimental schemes,and there are few reports on combining or extending them to the field of signal processing algorithms.Considering the importance of signal processing algorithms in the improvement of ADC performance,this thesis is dedicated to studying on the back-end signal processing of optical ADCs to improve the performance of optical ADC systems through algorithm design and implementation.The main research contents and results of this thesis are as follows:1.According to the quantization curve and coding characteristics of the 3.32-bit all-optical quantizer and the actual working conditions of the quantization system,the error coding probability model caused by system noise is analyzed.According to the different output error results obtained by the probability model,this thesis proposed the corresponding error code recognition and error correction algorithm which is given in the form of pseudo-code.The algorithm proposed in this thesis also makes up for the shortcomings of most previous papers that only study single-frequency sinusoidal input.Combined with the latest advances in the field of signal processing,the input signal type is extended to all band-limited signal inputs,making the actual ADC application scenarios wider.The simulation results show that as the noise level gradually increases,the error correction algorithm improves the signal-to-noise ratio of the output digital signal more obviously,and it always maintains a signal-to-noise ratio increase of 3dB to 7dB.Therefore,under different noise levels,the error correction capability of the algorithm has been verified.2.A new type of modulo ADC is proposed,which combines the existing optical ADC with the signal recovery algorithm to overcome the dynamic-range barrier of traditional ADC and realize the high dynamic range in optical ADC.Since the signal recovery algorithm used does not consider the signal quantization process,but the quantization is essential in the ADC,this thesis redefined and proved the sufficient conditions of sampling rate to ensure the signal recovery.The minimum sampling rate required to successfully recover the original analog signal with high-dynamic-range from the low-dynamic-range quantized modulo samples is verified by the numerical simulation results;in addition,this thesis also discussed the recovery conditions of the quantized signal when using different recovery algorithms,compared the pros and cons of different recovery algorithms and gave the simulation results.Finally,an optical ADC system experiment utilising the actual 5-channel cascaded multimode interference coupler was given,which has verified the feasibility of its modulo operation and the capability of realizing a high dynamic range optical ADC.The experimental results show that by increasing the input dynamic range,the effective number of bits of the original optical ADC system increased by 2.8 bits.