Phase-encoded multi-threshold all-optical quantization system
By combining multi-threshold quantization and phase coding in an all-optical quantization system, the problem of low quantization accuracy in all-optical quantization systems is solved, achieving higher quantization accuracy and more effective quantization states, thereby increasing the number of bits in analog-to-digital conversion.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- UNIV OF ELECTRONICS SCI & TECH OF CHINA
- Filing Date
- 2023-03-16
- Publication Date
- 2026-04-10
AI Technical Summary
Existing all-optical quantization systems cannot identify self-frequency shift at low peak power, and spectral broadening at high peak power leads to aliasing. Wavelength division multiplexing devices have limited resolution, which restricts the number of quantization states and reduces system accuracy.
A method combining multi-threshold quantization and phase coding is adopted, utilizing a high repetition frequency femtosecond pulse light source, an electro-optic modulator, a multi-threshold coding module, and an optical fiber structure. By improving quantization accuracy through phase coding and spectral compression, decision-making at multiple threshold points and spectral splitting are achieved.
It improves the quantization accuracy of the all-optical quantization system, increases the number of effective quantization states, increases the number of bits for analog-to-digital conversion to log2(2MN), solves the code pattern repetition problem, and achieves a higher quantization bit depth.
Smart Images

Figure CN116400547B_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The application belongs to the field of all-optical analog-to-digital conversion, and particularly relates to a multi-threshold all-optical quantization system based on phase coding. BACKGROUND
[0002] The conversion between continuous domain and discrete domain is the core of modern digital signal processing technology. High-speed and high-precision analog-to-digital conversion technology has become a key technology in many optoelectronic systems, such as high-resolution radar systems, high-speed imaging systems, sensor network systems and wideband wireless communication systems. Due to the aperture jitter of the sampling clock and the ambiguity of the comparator, the electronic analog-to-digital converter has many limitations in accurately sampling wideband signals. The introduction of photon technology opens up a new way for analog-to-digital conversion technology. All-optical analog-to-digital conversion technology uses optical means to realize signal sampling and holding and quantization, and injects new vitality into the long-stagnant analog-to-digital conversion technology. The multi-threshold all-optical quantization scheme based on phase coding has important use value in the field of all-optical analog-to-digital conversion. Compared with the existing electronic analog-to-digital conversion system and the optoelectronic hybrid analog-to-digital conversion system, the multi-threshold all-optical quantization system based on phase coding has the advantages of relatively simple system structure and high analog-to-digital conversion precision.
[0003] The existing all-optical quantization system directly uses high nonlinear optical fiber and arrayed waveguide grating, and has the following problems: at a low peak power, the too small self-frequency shift cannot be distinguished by the wavelength division device; at a high peak power, the too large spectral broadening causes spectral aliasing, reducing the system quantization precision; the wavelength resolution of the wavelength division device is limited, which theoretically limits the maximum number of effective quantization states, and in the existing scheme, one channel is one effective quantization state. SUMMARY
[0004] The purpose of the present application is to propose a multi-threshold all-optical quantization system based on phase coding to solve the problems in the background art. The multi-threshold all-optical quantization system can be applied to the field of all-optical analog-to-digital conversion and improve the quantization precision of analog-to-digital conversion.
[0005] To achieve the above purpose, the technical scheme adopted by the present application is as follows:
[0006] A multi-threshold all-optical quantization system based on phase coding comprises one high-repetition-frequency femtosecond pulse light source, a 1xM optical splitter, a phase-shifted optical quantization module, a soliton self-frequency shift module and a multi-threshold coding module. The phase-shifted optical quantization module is an M-channel electro-optical modulator formed by 1xM electro-optical modulators. The soliton self-frequency shift module comprises a first M-channel high nonlinear optical fiber, M-channel single-mode optical fiber and a second M-channel high nonlinear optical fiber. The multi-threshold coding module comprises M wavelength division multiplexers, MxN optical delay line arrays, MxN photodetector arrays and a coding module.
[0007] The high repetition frequency femtosecond pulse light source generates sampling light pulses with a frequency of tens of megahertz or more and a pulse width of femtosecond order, and the sampling light pulses are power-divided into M paths by a 1×M optical splitter and input into M-path electro-optical modulators; the M-path sampling light pulses are modulated by the analog signals to be sampled in the M-path electro-optical modulators, and the bias voltage of the electro-optical modulator is adjusted to work at different operating points, so that the phase shift difference between adjacent channels in the M-path electro-optical modulator is π / M, and phase encoding is completed; the peak power of the modulated M-path sampling light pulses carries the amplitude information of the analog signals to be sampled, and is input into a first M-path high nonlinear optical fiber; the modulated M-path sampling light pulses pass through the first M-path high nonlinear optical fiber, complete linear mapping from the peak power of the light pulses to the self-frequency shift amount, convert the amplitude information of the analog signals to be sampled into the wavelength shift amount of the light pulses, and are input into a cascade structure of an M-path single-mode optical fiber and a second M-path high nonlinear optical fiber; the cascade structure of the M-path single-mode optical fiber and the second M-path high nonlinear optical fiber performs spectrum compression for the spectrum broadening caused by the soliton self-frequency shift effect, improves the quantization precision, and then is transmitted to M wavelength division demultiplexers; threshold quantization is completed through the M wavelength division demultiplexers, and the light pulses are divided according to the spectral components, wherein each wavelength division demultiplexer has N output channels, and the optimal wavelength range of the N output channels is obtained by using a threshold point distribution algorithm; M×N optical power outputs are obtained after quantization; after dispersion compensation by an M×N optical delay line array, the M×N optical power outputs are input into an M×N photodetector array for detection, and M×N power values are obtained; the M×N power values are compared in an encoding module, and the channel corresponding to the maximum power value is encoded as an output.
[0008] Further, the phase shifts of each of the M-path electro-optical modulators are different, and the bias voltage of the electro-optical modulator is controlled so that the phase shift difference between adjacent electro-optical modulators is π / M, wherein the electro-optical modulator can be a lithium niobate electro-optical modulator.
[0009] Further, the materials of the first M-path high nonlinear optical fiber and the second M-path high nonlinear optical fiber are the same, and can be a tellurite, a sulfide or other optical fiber material with a nonlinear refractive index greater than the refractive index of silicon material by more than one order of magnitude; the first M-path high nonlinear optical fiber includes M segments of optical fibers with the same length, the second M-path high nonlinear optical fiber includes M segments of optical fibers with the same length, and the lengths of the optical fibers in the first M-path high nonlinear optical fiber and the second M-path high nonlinear optical fiber are different.
[0010] Further, the M-path single-mode optical fiber and the second M-path high nonlinear optical fiber need to satisfy positive and negative chirp offset.
[0011] Further, the process of the threshold point distribution algorithm is as follows:
[0012] Step 1, setting the number of parallel channels M and the number of output channels N of each wavelength division multiplexer according to requirements;
[0013] Step 2, representing the difference between the maximum wavelength shift and the minimum wavelength shift of the first M high nonlinear fiber output as C O , 0~C O is divided into N segments averagely, and N segments of ranges (0~C O , (C O / N~2C O , (N-1)C O / N~C O ) are obtained;
[0014] Step 3, according to the N segments of ranges obtained in step 2, the optimal wavelength ranges △λ1, △λ2, …, △λ N of the N output channels are determined by using a deep first search (DFS) algorithm and an evaluation model; wherein the evaluation model is:
[0015]
[0016] wherein SNR (dB) represents signal-to-noise ratio, A represents the amplitude of the analog signal to be sampled, N represents the number of quantization states, x(i) and x ’ (i) represent ideal quantization bit width and actual calculated quantization bit width respectively, σ q 2 represents the inherent quantization noise power of an ADC (analog-to-digital conversion) system, generally, the quantization noise can be approximated as white noise uniformly distributed within the bandwidth, σ q 2 = Δ 2 / 12, wherein Δ is the quantization step.
[0017] Compared with the prior art, the present application has the following beneficial effects:
[0018] In the existing nonlinear all-optical quantization scheme, each output in the encoding module only sets a threshold point at half of the maximum input optical power, and does not perform phase encoding, resulting in low efficiency of improving quantization accuracy. The multi-threshold all-optical quantization system based on phase encoding provided by the present application adopts a multi-threshold encoding module with phase encoding, realizes simultaneous judgment of multiple threshold points on the power transmission curve of M electro-optical modulators, solves the code type repetition problem of the all-optical quantization system, realizes more effective quantization states with fewer channel numbers, and effectively improves the quantization accuracy. At the same time, the quantization bit number of the all-optical quantization system of the present application is improved to log2(2MN). BRIEF DESCRIPTION OF DRAWINGS
[0019] Figure 1 A structure schematic diagram of a multi-threshold all-optical quantization system based on phase encoding provided by the present application is shown in the figure.
[0020] Figure 2 An implementation scheme of an all-optical quantization system with 2 parallel channels and 3 threshold points provided by the embodiment is shown in the figure. DETAILED DESCRIPTION
[0021] The technical solutions of the present application are described in detail below in combination with the drawings and embodiments.
[0022] Embodiment
[0023] Taking an all-optical quantization system with parallel channel number M=2 and output channel number N=3 of each wavelength division demultiplexer as an example. Wherein:
[0024] High repetition frequency femtosecond pulse light source: a mode-locked laser is used to generate sampling light pulses with a frequency of 200MHz, a pulse width of 150fs and a wavelength of 1550nm;
[0025] 1x2 electro-optic modulator: a Mach-Zehnder modulator made of lithium niobate material is used, with a half-wave voltage of 8V and bias voltages of 0V and 4V respectively;
[0026] Inputted analog signal to be sampled: a 50MHz sinusoidal signal with an amplitude of 4V;
[0027] First high nonlinearity optical fiber: core layer material is AsSe2, cladding layer material is As2S5, core layer is 6μm, cladding layer diameter is 6 times of the core layer diameter, and the length is 1m; the dispersion coefficient at 1550nm is 50ps / nm / km;
[0028] Second high nonlinearity optical fiber: core layer material is AsSe2, cladding layer material is As2S5, core layer is 6μm, cladding layer diameter is 6 times of the core layer diameter, and the length is 9.2m; the dispersion coefficient at 1550nm is 50ps / nm / km;
[0029] Single-mode optical fiber: the length is 2m;
[0030] Wavelength division demultiplexer: arrayed waveguide grating with channel spacing of 100GHz.
[0031] The working process of the all-optical quantization system of the embodiment is as follows:
[0032] The mode-locked laser generates sampling light pulses with a frequency of 200MHz, a pulse width of 150fs and a wavelength of 1550nm, which are power-divided into two paths by a 1*2 optical splitter and input into a 1*2 electro-optical modulator; the two paths of sampling light pulses are modulated by the analog signal to be sampled in the two paths of electro-optical modulators, and the bias voltage of the electro-optical modulator is adjusted to work at different operating points, so that the phase shift difference between adjacent channels in the two paths of electro-optical modulators is π / 2, the phase encoding is completed, and the peak power of the two paths of modulated sampling light pulses carries the amplitude information of the analog signal to be sampled and is input into a first 2-path high nonlinear optical fiber; the two paths of modulated sampling light pulses pass through the first 2-path high nonlinear optical fiber, complete the linear mapping between the peak power of the optical pulse and the self-frequency shift, convert the amplitude information of the analog signal to be sampled into the wavelength shift of the optical pulse, and input into the cascade structure of the 2-path single-mode optical fiber and the second 2-path high nonlinear optical fiber; the cascade structure of the 2-path single-mode optical fiber and the second 2-path high nonlinear optical fiber performs spectrum compression for the spectrum broadening caused by the soliton self-frequency shift, improves the quantization precision, and then is transmitted to two wavelength division demultiplexers; threshold quantization is completed through the two wavelength division demultiplexers, and the optical pulses are divided according to the spectral components, wherein each wavelength division demultiplexer has three output channels, and the optimal wavelength range of the three output channels is obtained by using a threshold point distribution algorithm; after quantization, 2*3 optical power outputs are obtained; after dispersion compensation by a 2*3 optical delay line array, the 2*3 optical power outputs are input into a 2*3 photodetector array for detection, and 2*3 power values are obtained; the 2*3 power values are compared in the encoding module, and the channel corresponding to the maximum power value is encoded as the output.
[0033] The all-optical quantization system of the embodiment adopts an end jump quantization mode, that is, all the sampled voltage values within 1 quantization bit width between certain quantization levels are attributed to the quantization level, so that the system has a total of 2M*N different code types with M parallel channel numbers and N threshold points. Figure 2 The encoding code types of the all-optical quantization system with 2 parallel channels and 3 threshold points are shown, and the encoding table of each combination of different channel numbers and threshold point numbers is different.
Claims
1. A phase-encoded multi-threshold all-optical quantization system, comprising: The high repetition frequency femtosecond pulse light source comprises a 1×M optical splitter, a phase-shifted light quantization module, a soliton self-frequency shift module and a multi-threshold coding module; the phase-shifted light quantization module is an M-path electro-optical modulator, the soliton self-frequency shift module comprises a first M-path high nonlinearity optical fiber, an M-path single-mode optical fiber and a second M-path high nonlinearity optical fiber, and the multi-threshold coding module comprises M wavelength division demultiplexers, an M×N optical delay line array, an M×N photoelectric detector array and a coding module. The high repetition frequency femtosecond pulse light source generates sampling light pulses with a frequency of tens of megahertz or above and a pulse width of femtosecond order, and the sampling light pulses are evenly divided into M paths by the 1×M optical splitter and input into the M-path electro-optical modulators; the M-path sampling light pulses are modulated by the analog signals to be sampled in the M-path electro-optical modulators, and the phase shift difference between adjacent channels in the M-path electro-optical modulators is π / M by adjusting the bias voltage of the electro-optical modulator; the peak power of the modulated M-path sampling light pulses carries the amplitude information of the analog signals to be sampled and is input into the first M-path high nonlinearity optical fiber; the modulated M-path sampling light pulses pass through the first M-path high nonlinearity optical fiber to complete linear mapping from the peak power of the light pulses to the self-frequency shift amount, convert the amplitude information of the analog signals to be sampled into the wavelength shift amount of the light pulses, and are input into the cascade structure of the M-path single-mode optical fiber and the second M-path high nonlinearity optical fiber; the cascade structure of the M-path single-mode optical fiber and the second M-path high nonlinearity optical fiber is subjected to spectrum compression and is transmitted to the M wavelength division demultiplexers; threshold quantization is completed by the M wavelength division demultiplexers, and the light pulses are divided according to the spectral components, wherein each wavelength division demultiplexer has N output channels, the optimal wavelength range of the N output channels is obtained by using a threshold point distribution algorithm, and M×N optical power outputs are obtained after quantization; the M×N optical power outputs are subjected to dispersion compensation by the M×N optical delay line array and are input into the M×N photoelectric detector array for detection to obtain M×N power values; the M×N power values are compared in the coding module, and the channel corresponding to the maximum power value is coded as the output.
2. The phase-encoded, multi-threshold, all-optical quantization system of claim 1, wherein, The electro-optical modulator is a lithium niobate electro-optical modulator.
3. The phase-encoded, multi-threshold, all-optical quantization system of claim 1, wherein, The first M-path high nonlinearity optical fiber and the second M-path high nonlinearity optical fiber are made of the same material, and the material is an optical fiber material with a nonlinearity refractive index greater than the refractive index of silicon material by more than one order of magnitude. The first M-path high nonlinearity optical fiber comprises M segments of optical fibers with the same length, the second M-path high nonlinearity optical fiber comprises M segments of optical fibers with the same length, and the lengths of the optical fibers in the first M-path high nonlinearity optical fiber and the second M-path high nonlinearity optical fiber are different.
4. The phase-encoded, multi-threshold, all-optical quantization system of claim 1, wherein, The M-path single-mode optical fiber and the second M-path high nonlinearity optical fiber satisfy positive and negative chirp offset.
5. The phase-encoded, multi-threshold, all-optical quantization system of claim 1, wherein, The threshold point distribution algorithm comprises the following steps: Step 1, set the number of parallel channels M and the number of output channels N of each wavelength division demultiplexer according to requirements; Step 2, the difference between the maximum wavelength shift and the minimum wavelength shift of the first M-path high nonlinear fiber output is represented as C O , 0 ~ C O , average segmentation into N segments, get N segments range; Step 3, determine the optimal wavelength range of the N output channels according to the N ranges obtained in step 2 by using a depth-first search algorithm and an evaluation model; wherein the evaluation model is: where SNR (dB) represents the signal-to-noise ratio, A represents the amplitude of the analog signal to be sampled, N represents the number of quantization states, x(i) and x ’ (i) represent the ideal quantization bit width and the actual calculated quantization bit width, respectively, σ q 2 represents the inherent quantization noise power of the ADC system, σ q 2 = Δ 2 / 12, where Δ is the quantization step.
Citation Information
Patent Citations
Full light analog-to-digital converter
CN101311811A
Method for optical phase shift assisted frequency multiplication analog-to-digital conversion
CN108803196A