A method for designing optical sampling pulse and expanding sampling bandwidth of square-law detection optical receiver

Abstract

The application discloses a light sampling pulse design and sampling bandwidth expansion method suitable for a square detection type optical receiver. The method is based on system transfer function analysis and spectrum convolution characteristics, and is used for optimizing the spectrum distribution of the light sampling pulse in the light spectrum, so that the spectrum is more flat, and the sampling bandwidth of the system is expanded. Different from the prior art, the method controls the light comb amplitude relationship without changing the number of the existing phase-locked light comb, does not need to additionally introduce optical devices, and does not increase the system complexity, thereby providing a feasible scheme for generating the light sampling pulse of a large-scale integrated optical receiver. The application can effectively improve the performance of the square detection type optical receiver, and provides important support for practical application.

Description

Technical Field

[0001] This invention relates to photonic information processing, specifically to a method for designing optical sampling pulses and extending sampling bandwidth in a square-detector type optical receiver. Background Technology

[0002] Traditional electrical receivers face bottlenecks in high-speed signal sampling and processing, particularly in the acquisition and high-precision recovery of ultra-wideband signals, limited by factors such as the bandwidth, noise, and power consumption of electronic components. As signal frequency and bandwidth increase, the performance of electrical receivers struggles to break through further barriers, forming the so-called "electronic bottleneck." Thanks to the ultra-large input bandwidth of electro-optic modulators, optical receivers can overcome this limitation, fully leveraging the advantages of optical systems to achieve efficient sampling and processing of ultra-wideband signals. Among these, square-probe optical receivers have attracted widespread attention due to their simple structure and ease of integration, providing a new solution for high-precision sampling of ultra-wideband signals.

[0003] In square-probe optical receivers, the optical sampling pulse is crucial for high-frequency signal reception. Its high-frequency components can fold signals from multiple Nyquist intervals back into the first Nyquist interval, enabling the low-speed electro-amplification module at the backend to effectively process the signal. Therefore, the design of a broadband optical sampling pulse is essential for extending the sampling bandwidth of square-probe optical receivers. In optical receiver systems, the design of the optical sampling pulse must meet requirements such as simplicity and ease of integration, and consistent sampling frequency comb phase, to achieve stable ultra-wideband signal sampling.

[0004] To this end, researchers have conducted extensive studies on broadband optical sampling pulses. In 2014, Paolo Ghelfi et al. from the Italian National Research Center for Information and Communication Technology (CNIT) proposed a microwave photonic radar based on mode-locked lasers. However, the integration of mode-locked lasers is difficult, and their application in optical sampling pulse generation has gradually decreased in recent years. In 2019, Xu Yirong et al. from Tsinghua University proposed an optical sampling pulse architecture based on intensity modulators and phase modulators and applied it to an optical receiver. However, this scheme requires the introduction of dispersive fiber, which not only increases the difficulty of system integration but also reduces the signal-to-noise ratio. In 2021, Liu Siqi et al. from Shanghai Jiao Tong University proposed an optical sampling pulse generation architecture based on cascaded dual parallel modulators, generating 25 flat optical combs in the optical domain and achieving 8 flat electrical combs within 3.5dB in the electrical domain. However, the impact of the photoelectric conversion process on the sampling frequency combs was not considered. In 2022, Liu Yong's team at the University of Electronic Science and Technology of China optimized the optical sampling pulse generation architecture based on intensity modulator and phase modulator. They made the comb flatter by optimizing low-bias modulation and dispersion compensation, but did not conduct an in-depth analysis of the photoelectric conversion relationship.

[0005] In summary, current research on square-probe optical receivers still lacks a systematic analysis of the optical-to-electrical comb convolution effect during photoelectric conversion. This convolution effect causes the electrical comb to exhibit a triangular envelope, resulting in severe amplitude attenuation. Even when a Nyquist pulse is generated in the optical domain, the corresponding electrical comb remains uneven, ultimately limiting the effective sampling bandwidth of the optical receiver. Furthermore, existing broadband optical sampling pulse designs do not fully consider the impact of photoelectric convolution and fail to optimize the spectrum, leading to a limited number of usable flat electrical combs. Additionally, some schemes require the introduction of extra optical components, further increasing integration difficulty. Therefore, there is an urgent need for an optical sampling pulse design method suitable for square-probe optical receivers. This method should combine the photoelectric conversion convolution relationship to maximize the number of flat electrical combs while ensuring the phase-locked loop and integrability of the sampling frequency comb architecture, thereby effectively expanding the sampling bandwidth of the optical receiver. Summary of the Invention

[0006] This invention addresses the shortcomings of existing technologies by proposing a method for designing optical sampling pulses and extending sampling bandwidth for square-probe optical receivers. This method optimizes the electrical spectrum of the square-probe optical receiver and, based on the convolutional transformation relationship between the optical and electrical spectra, inversely designs the receiver's spectrum, thereby effectively extending the receiver's sampling bandwidth. Unlike traditional optical sampling pulse design methods, this invention quantitatively analyzes the convolutional relationship between the optical sampling pulse spectrum and the electrical spectrum, and compensates for this convolutional effect during spectrum design to achieve a flatter electrical spectrum. Specifically, this invention first establishes the convolutional constraint relationship between the spectral phase-locked comb and the electrical spectrum comb, optimizes the electrical spectrum comb based on the target, and finally inversely determines the required spectral design based on the optimized electrical spectrum. Based on an integrated and mature Mach-Zehnder modulation structure, this invention eliminates the need for additional optical filters, dispersive fibers, or other optical components, thus not increasing system complexity. It achieves a flatter electrical spectrum and effectively extends the sampling bandwidth of the optical receiver simply by optimizing the amplitude relationship between existing optical combs. This method not only improves the performance of optical receivers, but also provides important technical support for the practical application of optical receivers, and has significant engineering value and application prospects.

[0007] The theoretical justification for this invention is as follows:

[0008] In a receiver based on optical sampling pulses, the signal to be sampled, v in (t) is modulated onto the optical sampling pulse by an electro-optic modulator, and the electrical output v can be obtained by a photodetector. out (t). Considering small-signal modulation, the effects of low-pass filtering and demultiplexing can be represented by H. OE (t) indicates that, without considering DC, the relationship between the electrical output signal and the sampled signal can be expressed as:

[0009] vout (t)=H OE (t)*v in (t)·i(t). (1.1)

[0010] Among them, v in (t) is the signal to be sampled, and i(t) is the equivalent sampling pulse of the optical spectrum. Equation (1.1) shows that the output signal in the electrical domain can be mathematically understood as the result of the equivalent sampling pulse of the optical spectrum directly sampling the signal to be sampled. Taking a Fourier transform of equation (1.1) yields the relationship between the optical sampling pulse spectrum and the transfer function, as shown in equation (1.2):

[0011]

[0012] Among them, H OE (f) is a fixed function. The symbol represents the Fourier transform, and * represents convolution. Since H... OE (f) This has no impact on the bandwidth of the optical receiver. Equation (1.2) shows that the electrical spectrum of the optical sampling pulse directly determines the system bandwidth of the optical receiver. Therefore, by directly designing the electrical spectrum response of the optical sampling pulse, and then inversely designing the spectrum based on the convolution relationship between the spectrum and the electrical spectrum, a flatter electrical spectrum response can be obtained, thereby extending the sampling bandwidth of the optical receiver. Next, we analyze the convolution relationship between the spectrum and the electrical spectrum. The optical sampling pulse in the optical domain is a frequency interval of f. m The carrier frequency is f c For an optical frequency comb of order n, the convolution relationship between its spectrum and electrical spectrum can be expressed by equation (1.3):

[0013]

[0014] Based on equation (1.3), we can give the specific constraint relationship between the optical sampling pulse spectrum and the electrical spectrum, and use this to optimize the electrical spectrum, resulting in the optimization problem shown in equation (1.4):

[0015] min|I0(f)-min(I0(f),I1(f),I2(f),...,I n (f))|, (1.4)

[0016] The technical solution of the present invention is as follows:

[0017] This invention provides a method for designing optical sampling pulses and extending sampling bandwidth in a square-probe type optical receiver, characterized in that the method includes the following stages:

[0018] - Determine the constraint relationship between the optical sampling pulse comb and the electrical comb, and solve the electronic spectrum optimization problem.

[0019] Obtain the number M of the phase-locked comb in the optical sampling pulse spectrum and calculate the corresponding number of phase-locked comb lines in the electrical spectrum;

[0020] The constraint relationship between the optical comb and the electric comb for optical sampling pulses is determined based on the convolution effect of the spectrum during photoelectric conversion.

[0021] Determine the optimization problem required to obtain a flat electric comb, and establish the objective function and constraints.

[0022] - Solve the spectral optimization problem and determine the corresponding spectral design.

[0023] For different target numbers of comb lines, solve the optimization problem to obtain the optimal electrical spectrum design;

[0024] The optimal electro-spectral density scheme was selected by considering the number of comb lines and the flatness requirements.

[0025] Based on the photoelectric conversion relationship, the spectral characteristics that meet the optimal electronic spectrum design requirements are deduced.

[0026] Adjust the spectral characteristics of the optical sampling pulse to make it conform to the theoretical design scheme.

[0027] The optical sampling pulse design and sampling bandwidth extension method of the square probe type optical receiver are not limited by the modulator type and modulator driving radio frequency.

[0028] The implementation methods of the spectral design include, but are not limited to, adjusting the modulation depth m and bias voltage θ of the modulator; and using methods such as unequal length interferometers and optical filtering.

[0029] The following provides a detailed explanation of the above steps:

[0030] - Determine the constraint relationship between the optical sampling pulse comb and the electrical comb, and solve the electronic spectrum optimization problem.

[0031] S1, determine the sampling rate NF of the optical receiver. s and single-channel sampling rate F s The optical comb frequency interval of the acquired optical sampling pulse is aligned with F. s Consistent, obtain the number of phase-locked comb lines M of the optical sampling pulse spectrum, and calculate the corresponding number of phase-locked comb lines of the electrical spectrum according to equation (1.3);

[0032] S2, In a square-probe type optical receiver, the optical sampling pulse is a frequency interval of f in the optical domain. m The carrier frequency is f c An optical frequency comb of order n. Since there is a convolution relationship between the spectrum and the electrical spectrum, this convolution relationship can be described by equation (1.3), which clarifies the constraints of the optical comb and the electrical comb.

[0033] S3, determine the optimization problem of the electric comb flattening. According to the photoelectric convolution constraint given by equation (1.3), the zeroth order electric comb amplitude must be the largest. Therefore, the electric comb flatness optimization problem can be mathematically described by equation (1.4) to construct the objective function and constraint conditions.

[0034] - Solve the spectral optimization problem and determine the corresponding spectral design.

[0035] S1, Solve the optimization problem defined by equation (1.4), determine the number of combs n to be optimized, and optimize the flatness of n combs. Here, the value of n is between 0 and M;

[0036] S2, taking into account the system's sampling bandwidth requirements, and while ensuring the flatness of the comb, comprehensively weigh the number of combs and flatness, and select the optimal number of combs.

[0037] S3, based on the optimal solution of the electric spectrum optimization design and the photoelectric conversion relationship described by equation (1.4), the required spectral design is deduced to determine the amplitude distribution of each order optical comb;

[0038] S4. Adjust the spectral parameters to meet the theoretical design requirements, thereby generating an optical sampling pulse that conforms to the optimized scheme, and realizing the construction of the optical sampling pulse module.

[0039] Compared with the prior art, the present invention has the following advantages:

[0040] 1. This invention is based on precise control of the amplitude of the optical sampling pulse comb to obtain electrical comb components with smaller power differences, thereby achieving a flatter electrical spectrum response and effectively expanding the sampling bandwidth of the optical receiver.

[0041] 2. This invention optimizes the spectral flatness by adjusting the amplitude of the optical comb in the optical sampling pulse, without introducing additional optical components or increasing system complexity, and has high engineering feasibility.

[0042] 3. The bandwidth of the optical sampling pulse designed in this invention is limited only by the number of optical combs. When the bandwidth requirements of the square detector type optical receiver increase further, it is only necessary to increase the optical comb frequency of the optical sampling pulse to simultaneously expand the spectral bandwidth, thus exhibiting good scalability.

[0043] 4. This invention is applicable to various square-detection optical receiver architectures, including but not limited to photonic analog-to-digital conversion systems, optical frequency measurement, microwave photonic radar, etc., and can provide a reliable technical solution for future ultra-high bandwidth, large-scale integrated optical receivers. Attached Figure Description

[0044] Figure 1 This is an overall architecture diagram of an embodiment of a square-probe type optical receiver used in the method of the present invention;

[0045] Figure 2 The electrical spectrum design and parameter settings of the cascaded dual parallel modulator proposed in this invention are shown in (a), which represents the frequency settings of the two dual parallel modulators, (b), which represents the electrical spectrum design of each dual parallel modulator, (c), which represents the electrical spectrum interaction relationship between the two balanced modulators, and (d), which represents the electrical spectrum response of the cascaded dual parallel modulator.

[0046] Figure 3 This invention presents the process of generating an electrical comb from an optical comb based on spectral convolution, using a dual parallel modulator as an example. In this paper, (a) represents the corresponding optimized electrical spectrum, (b) represents the process of generating the zeroth-order electrical comb from the corresponding optical comb through convolution, (c) represents the process of generating the first-order electrical comb from the corresponding optical comb through convolution, and (d) represents the process of generating the second-order electrical comb from the corresponding optical comb through convolution.

[0047] Figure 4 The simulation results of a specific embodiment of the present invention demonstrate the improvement effect of the sampling pulse electrical comb flatness in an optical receiver with a cascaded dual parallel modulator. (a) and (b) show the spectrum and electrical spectrum of an optical receiver system using conventional optical sampling pulses, while (c) and (d) show the spectrum and electrical spectrum of the optical receiver system using the spectral convolution-optimized system. Detailed Implementation

[0048] A specific embodiment of the present invention is given below with reference to the accompanying drawings. The optical sampling pulse source module of this embodiment consists of two cascaded dual parallel modulators. It is implemented based on the technical solution of the present invention, and detailed implementation methods and processes are given. However, the scope of protection of the present invention is not limited to the following embodiment.

[0049] This embodiment uses, as follows: Figure 1 The illustrated square-probe type optical receiver architecture includes an optical sampling pulse generation source 1, a sampled signal emission source 2, a photonic sampling gate 3, a demultiplexer array 4, a photodetector array 5, an electronic analog-to-digital converter array 6, and a data integration and processing module 7. The optical sampling pulse generation source 1 consists of two cascaded dual parallel modulators (DPMZMs), with its output connected to the second input of the photonic sampling gate 3. The output of the sampled signal source 2 is connected to the first input of the photonic sampling gate 3. The output of the photonic sampling gate 3 is connected to the input of the demultiplexer array 4. The N outputs of the demultiplexer 4 are connected to the inputs of each unit module of the photodetector array 5. The N outputs of the photodetector array are connected to the inputs of the N electronic analog-to-digital converters in the electronic analog-to-digital converter array 6. The outputs of the N electronic analog-to-digital converters in the electronic analog-to-digital converter array are connected to the N inputs of the data integration and processing module.

[0050] The optical sampling pulse design and sampling bandwidth extension method of the above-mentioned square-probe type optical receiver specifically include the following stages:

[0051] - Determine the constraint relationship between the optical sampling pulse comb and the electrical comb, and solve the electronic spectrum optimization problem.

[0052] S1, determine the sampling rate NF of the optical receiver. s and single-channel sampling rate F s The optical comb frequency interval of the acquired optical sampling pulses is related to the single-channel sampling rate F. s Consistent. In this example, the optical sampling pulse source module is a cascaded dual-parallel modulator. First, we analyze the single dual-parallel modulator, whose optical sampling pulse spectrum phase-locked comb line number is 5. According to equation (1.3), the corresponding number of electrical spectrum phase-locked comb lines is 5. In the cascaded case, the number of comb lines of the optical comb and the electrical comb is the result of multiplying the single modulator. The cascaded dual-parallel modulator has 25 optical sampling pulse spectrum phase-locked comb lines, and the corresponding number of electrical spectrum phase-locked comb lines is 25.

[0053] S2, gives the detailed constraints of equation (1.3) under the dual parallel modulator, and its spectrum E pulse (f) and electric spectrum I pulse The relationship between (f) is shown in equation (1.5):

[0054]

[0055] Where I0(f) to I4(f) represent the 0th to 4th order electrical combs, and E0(f) to E2(f) represent the 0th to 2nd order optical combs. Subsequent analysis will examine the interaction between the two modulators using electrical spectra, and analyze the correspondence between the optical comb and electrical comb for each modulator.

[0056] S3, determine the optimization problem corresponding to obtaining the flat electric comb. Since the constraint relationship between the electric comb and the optical comb has been given by equation (1.5), the zeroth order electric comb amplitude must be the largest. The optimization problem between the flat electric comb can be given by equation (1.6).

[0057] min|I0(f)-min(I0(f),I1(f),I2(f),I3(f),I4(f))|, (1.6)

[0058] Where I0(f) to I4(f) represent the 0th to 4th order electric combs.

[0059] - Solve the spectral optimization problem and determine the corresponding spectral design.

[0060] S1, Solve the optimization problem given in equation (1.6) to find the corresponding optimal electrical spectrum and flatness under the conditions of 5, 4, and 3 target comb numbers. When E1(f) = 4 / 3E0(f) and E2(f) = 8 / 3E0(f), we have I0(f) = 1.375I1(f) = 2.64I2(f) = 2.64I3(f) = 2.64I4(f), and the power difference between I0(f) and I4(f) is 8.4dB. This shows that when the total number of optical combs is 5, i.e., when using a dual parallel modulator, it is difficult to use the 5 generated phase-locked loop combs simultaneously. To reduce the number of target combs to be optimized, we have that when E1(f) = 1 / 2E0(f) and E2(f) = 3 / 4E0(f), I0(f) = 1.5I1(f) = 1.5I2(f), and the power difference between the three combs is 3.5dB. Considering the system's sampling bandwidth requirements and taking into account both the number of combs and flatness, we select 3 combs as the optimal number for a single dual-parallel modulator.

[0061] S2, further, the cascaded dual-parallel modulator is analyzed. Figure 2 (a) illustrates the frequency settings of cascaded dual-parallel modulators, with the frequencies of the two dual-parallel modulators set to a 5-fold ratio. Similarly, regardless of the number of spectral comblines, the method of this invention can be used to solve the optimization problem between the electrical spectral comblines and the corresponding constraints under photoelectric convolution. Figure 2 (b) illustrates the corresponding electrical spectrum design. For a biparallel modulator with an input frequency of one harmonic, the electrical spectrum satisfies I... b,0 (f) = 1.5I b,1 (f) = 1.5I b,2 (f) For a biparallel modulator with an input frequency of five harmonics, the electrical spectrum satisfies I a,0 (f) / 5=I a,1 (f) / 4=I a,2 (f) / 3=I a,3 (f) / 2=I a,4 (f), Figure 2 (c) and (d) illustrate the process and final spectrum generation of the cascaded dual parallel modulator, with a flatness of 3.5 dB between the first 11 frequency combs.

[0062] S3, based on the optimal spectrum design determined in the optical sampling pulse spectrum design stage and the conversion relationship between the optical comb and the electrical comb given in Equation (1.5), the corresponding spectrum design is given. For a dual parallel modulator with an input frequency of one harmonic, the spectrum satisfies I b,0 (f) = 1.5I b,1 (f) = 1.5I b,2(f), whose corresponding spectrum satisfies E1(f)=1 / 2E0(f), E2(f)=3 / 4E0(f); for a dual parallel modulator with an input frequency of five harmonics, its corresponding spectrum satisfies E1(f)=E0(f)=E2(f), Figure 3 Taking a dual parallel modulator with an input frequency of one harmonic as an example, the process of generating an electric comb from an optical comb based on spectral convolution is demonstrated. Figure 3 (a) shows the corresponding optimized electrical spectrum; Figure 3 (b) shows the process of generating the zeroth-order electric comb by convolution with the corresponding optical comb, as shown by I0(f) in Equation (1.5); Figure 3 (c) illustrates the process of generating the first-order electric comb by convolution with the corresponding optical comb, as shown in I1(f) in Equation (1.5). The generation process of the remaining electric combs is the same as described above.

[0063] S4. The modulator architecture for realizing the optical sampling pulse is determined to be a cascaded dual-parallel modulator. Based on the modulator's spectral output expression, the modulation depth *m* and bias voltage *θ* are calculated to achieve the target spectral design. For a dual-parallel modulator with an input frequency of one harmonic, the electrical spectrum satisfies I... b,0 (f) = 1.5I b,1 (f) = 1.5I b,2 (f), whose corresponding spectrum satisfies E1(f)=1 / 2E0(f), E2(f)=3 / 4E0(f), and whose modulation depth and bias voltage satisfy equation (1.7).

[0064] For a biparallel modulator with an input frequency of five harmonics, the electrical spectrum satisfies I a,0 (f) / 5=I a,1 (f) / 4=I a,2 (f) / 3=I a,3 (f) / 2=I a,4 (f), whose modulation depth and bias voltage satisfy equation (1.8).

[0065]

[0066] The above process retains the basic architecture of the receiver based on optical sampling pulses. According to the proposed photoelectric convolution relationship, the amplitude of the optical comb in the optical sampling pulses is controlled to obtain more electrical combs with smaller power differences, thereby achieving a flatter electrical spectrum response. Experiments show that this invention can extend the sampling bandwidth of the optical receiver without adding additional components or increasing system complexity. Furthermore, the bandwidth of the proposed method is only limited by the number of optical combs. When the bandwidth requirement of the square-probe type optical receiver continues to increase, simply increasing the optical comb frequency of the optical sampling pulses will correspondingly increase the electrical spectrum bandwidth based on the proposed method. This invention can provide a more reliable technical solution for realizing large-scale integrated optical receivers with ultra-high bandwidth in the future. Simulation results are shown below when comparing optical sampling pulses designed without the proposed method. Figure 4 As shown. Figure 4 (a) and (b) show the spectral and electrical spectra of an optical receiver system using conventional optical sampling pulses. Figure 4 (c) and (d) show the spectral and electrical spectra of the optical receiver system based on the present invention. Conventional designs exhibit envelope degradation, with a flatness deviation reaching 4.6 dB at the 10th comb line. In contrast, the proposed spectral convolution-based design ensures equal amplitudes for the 5th to 10th comb lines in the electrical domain, 3.5 dB lower than the zeroth comb line. Our method reduces the difference between the 10th and 0th comb lines by 1.1 dB. Therefore, it is clear that the proposed design generates more usable comb lines in the electrical domain and extends the system bandwidth.

Claims

1. A method for optical sampling pulse design and sampling bandwidth extension suitable for square-law detection type optical receivers, characterized in that, The method comprises the following steps: Step 1. Determine the convolution constraint relationship between the optical sampling pulse comb and the electrical comb: obtain the number of optical sampling pulse spectrum phase-locked comb lines M, based on the convolution effect of the spectrum and the electrical spectrum in the photoelectric conversion process, establish the convolution constraint relationship between the optical sampling pulse comb E pulse (f) and the electrical comb I pulse (f) as follows: where I n (f) represents the nth order optical comb, E k-N (f) represents the k-Nth order optical comb, f m represents the frequency interval of the optical comb, and N represents the total order of the optical comb; Step 2. Constructing an optimization problem required for constructing an electrical spectrum flat comb: constructing an electrical spectrum optimization model with the minimum comb amplitude difference as an objective function according to the convolution constraint relationship, as follows: min|I0(f)-min(I0(f),I1(f),I2(f),...,I n (f))|; where I n (f) represents the nth order electrical comb; Step 3. Solving the electrical spectrum optimization problem and backstepping the optical spectrum design; Solving the electrical spectrum optimization model for different target optimized comb numbers; Selecting an optimal electrical spectrum design scheme in combination with the comb number and flatness requirement; Based on the convolution constraint relationship, backstepping the optical comb amplitude distribution corresponding to the optimal comb, adjusting the modulation parameters of the optical sampling pulse, and making the optical spectrum characteristics meet the design requirement.

2. The optical sampling pulse design and sampling bandwidth extension method suitable for square-law detection type optical receivers according to claim 1, characterized in that, The generation module of the optical sampling pulse comprises a single or cascaded multiple intensity modulator, a double parallel modulator.

3. The optical sampling pulse design and sampling bandwidth extension method suitable for square-law detection type optical receivers as claimed in claim 1, wherein, The adjustment of the optical spectrum design of the optical sampling pulse is realized by the following ways: Adjusting the modulation depth m and the bias voltage θ of the modulator; Adjusting the optical comb phase consistency by using an unequal-length interferometer, or suppressing the optical comb components of the non-target frequency band by using an optical filter.

4. The optical sampling pulse design and sampling bandwidth extension method suitable for square-law detection type optical receivers as claimed in claim 1, wherein, The square detection type optical receiver is suitable for a time division channel interleaving type optical analog-to-digital converter, an optical frequency measurement system or a microwave photon radar.

Images