A chip architecture and high-sensitivity signal detection method for optical communications

Abstract

The present invention discloses a chip architecture for optical communications, including a modulated transmitter and a balanced detector. The modulated transmitter is used to modulate a signal light source and transmit it along a signal optical path X and a reference optical path Y, respectively. The modulated transmitter is provided with a polarization combiner for polarization combining light II and light I that has undergone IQ modulation. In the balanced detector, a polarization beam splitter is used to receive and decompose the combined light. A local oscillator light source is used to perform 90° mixing with the decomposed light and transmit the mixing results to a photodetector array. The photodetector array is used to perform photoelectric conversion on the mixing results and transmit them to a coherent processor for coherent analysis. The present invention adopts a polarization multiplexing route architecture to design the chip, so that local oscillator light amplification can further reduce the bit error rate of the transmitted signal. The present invention also utilizes the second-order correlation of the light field to complete signal detection, while eliminating the effects of noise and low power introduced during local oscillator light amplification on signal detection, thereby broadening the application scenarios.

Description

Technical Field

[0001] The present invention relates to the technical field of optical chips, and in particular to a chip architecture applied to optical communications and a high-sensitivity signal detection method. Background Art

[0002] As we move from an information society to an intelligent society, human activities generate massive amounts of data daily. This data needs to be collected, transmitted, and stored before it can be transformed into productive capacity. Data transmission is the bridge between these two. The proliferation of smart terminals and the explosive generation of information have driven the demand for higher-speed transmission with lower error rates. Especially in the field of artificial intelligence, high-speed data transmission can greatly utilize hardware resources, while efficient data exchange can accelerate model training and reduce training costs. Optical chips use light as a signal carrier and achieve multi-dimensional signal modulation through multi-dimensional multiplexing technology, significantly increasing channel capacity. Due to their advantages such as low transmission loss, minimal time delay, strong anti-interference capabilities, and low power consumption, optical chips are widely used in communications and data centers.

[0003] However, optical communications can face various complex transmission environments. Both attenuation and noise can significantly disrupt the actual signal, increasing the bit error rate (BER) of signal transmission. Furthermore, they must contend with challenges such as optical field phase calibration. Currently, to meet the high-speed and low-BER requirements of optical communication chip design, some technologies are adjusting error correction codes and improving digital signal processing algorithms to reduce BER and compensate for channel impairments. However, these approaches often come with the disadvantages of increasing non-redundant data, sacrificing significant bandwidth, and increasing computational complexity. Other technologies employ high-power light sources or external optical amplifiers to enhance signal strength, ensuring sufficient signal at the output. However, this approach inevitably introduces significant noise, resulting in high-impact and distorted results.

[0004] In other words, the above solution has at least the following two problems when solving the communication problem of optical chips:

[0005] 1) High computational complexity and high amount of redundant data;

[0006] 2) A large amount of noise is introduced, resulting in a lot of impurities in the output. Summary of the Invention

[0007] In response to the above two problems, the purpose of the present invention is to propose a chip architecture and a high-sensitivity signal detection method for optical communications. By adopting a polarization multiplexing route architecture to design an optical chip, local oscillator optical amplification can further reduce the bit error rate of the signal transmitted by the optical chip. The optical chip based on the above architecture uses the second-order correlation of the light field to complete signal detection, while effectively eliminating the noise and low power that may be introduced during local oscillator optical amplification on signal detection, greatly broadening the application scenarios of optical chips.

[0008] This is achieved through the following technical solutions:

[0009] First, a chip architecture for optical communication is proposed. The architecture includes a modulated transmitter and a balanced detector. The modulated transmitter is used to modulate the signal light source into light I and light II. Light I is transmitted along the signal optical path X, and light II is transmitted along the reference optical path Y. The modulated transmitter is provided with a polarization beam combiner, which is used to polarization-combine light II with the IQ-modulated light I and transmit the combined light to the balanced detector. The balanced detector is provided with a local oscillator light source, a polarization beam splitter, a photodetector array, and a coherent processor. The polarization beam splitter is used to receive the combined light and decompose it into light I' and light II'. The local oscillator light source is used to perform 90° mixing with light I' and light II' respectively and transmit the mixing results to the photodetector array. The output wavelengths of the local oscillator light source and the signal light source are equal. The photodetector array is used to perform photoelectric conversion on the mixing results and transmit them to the coherent processor for coherent analysis.

[0010] Preferably, the modulated transmitter is further provided with an amplitude modulator, an arbitrary waveform generator, a first beam splitter, a signal source, and an IQ modulator. The amplitude modulator is driven by the arbitrary waveform generator and is used to amplitude modulate the signal light source. The first beam splitter is used to modulate the amplitude-modulated signal light source into Light I and Light II. The signal source is used to load an electrical signal into the IQ modulator, and the IQ modulator performs IQ modulation on Light I based on the electrical signal. After amplitude modulation, the signal light source is split into two independent optical paths, realizing the combination of IQ modulation and reference signal. IQ modulation can carry actual communication data and reduce crosstalk, and the reference signal can be used for subsequent calibration.

[0011] Preferably, the IQ modulator includes a second beam splitter, a first Mach-Zehnder modulator, a second Mach-Zehnder modulator, a phase shifter, and a beam combiner. The second beam splitter is connected to the beam combiner via the first Mach-Zehnder modulator, and is further connected to the beam combiner via the second Mach-Zehnder modulator and the phase shifter. The combination of the Mach-Zehnder modulator and the phase shifter supports quartz crystal modulation, improving spectral efficiency.

[0012] Preferably, the arbitrary waveform generator is used to generate a pseudo-random signal, which is used to drive the amplitude modulator. The pseudo-random signal is used for driving, which can enhance the ability to resist channel noise.

[0013] Preferably, the first beam splitter adopts non-uniform beam splitting; when the first beam splitter performs beam splitting, the light intensity I of the control light II is Y Light intensity I is less than light I X , I Y ≤I X Controlling the optical intensity, that is, significantly suppressing the power of the reference optical path Y, can reduce crosstalk while ensuring that the reference signal is sufficient for polarization and phase noise calibration.

[0014] Preferably, the second beam splitter and beam combiner employ a multimode interference structure or an adiabatic coupling structure, with the second beam splitter configured to perform a 5:5 splitting and the beam combiner configured to perform a 5:5 beam combining. Using the same splitting and combining ratios ensures consistency of the corresponding signals during transmission, facilitating subsequent computational processing.

[0015] Preferably, before the polarization beam combiner performs polarization beam combining, the light II is first converted from TE mode to TM mode, and then the TM mode light II is polarization combined with the IQ modulated light I; before the polarization beam splitter performs decomposition, the TM mode light in the combined light transmitted by the polarization beam combiner is polarized and rotated to TE mode light II' for output, and the TE mode light in the combined light is also output as light I'.

[0016] Preferably, the balanced detection end is further provided with a first 90° optical mixer and a second 90° optical mixer, both of which adopt a fourth-order MMI structure or a cascade structure of multiple low-order MMIs. When the local oscillator light source is used for mixing, it emits incident light of equal intensity to the first 90° optical mixer and the second 90° optical mixer respectively. The photodetector array includes four low-speed detectors and four high-speed detectors. The four low-speed detectors are used to receive the mixing results of the first 90° optical mixer, perform photoelectric conversion, and transmit them to the coherent processor. The four high-speed detectors are used to receive the mixing results of the second 90° optical mixer, also perform photoelectric conversion, and transmit them to the coherent processor. MMI, or multi-mode interference, and the fourth-order MMI or cascaded MMI structure enable the optical mixer to improve phase resolution accuracy. The division of labor between the low-speed and high-speed detectors takes into account both dynamic range and response speed, adapting to different modulation rate requirements.

[0017] In addition, a high-sensitivity signal detection method is proposed, which includes the following steps: 1) first use a signal source to generate bit transmission signals S1 and S2 with a length of M at a rate G, and load the bit transmission signals S1 and S2 into the corresponding transmission light field through the first Mach-Zehnder modulator and the second Mach-Zehnder modulator respectively; the effective characteristic time of each bit transmission signal is τ1, τ1=1 / G; then use an arbitrary waveform generator to generate a pseudo-random signal with an effective characteristic time of τ2, which is used to drive the amplitude modulator, 2τ2=τ1; the signal light source emits signal light, and the signal light is sequentially transmitted to the photodetector array through amplitude modulation, beam splitting, beam combining, decomposition and mixing; 2) the photodetector array transmits four groups of electrical signals to the coherent processor, and the coherent processor generates four groups of different light fields E based on the four groups of electrical signals. i , i=1, 2, 3, 4; 3) Record and measure four different sets of light fields E i The initial coherent detection light intensity corresponding to each step is obtained; the bit transmission signal generated by the control signal source remains unchanged, and steps 1) to 2) are repeated N times, where N is a positive integer, and the coherent detection light intensity of each repetition is recorded; 4) based on the second-order correlation of the light field, the initial coherent detection light intensity in step 3) and the repeated coherent detection light intensity are calculated, and the bit transmission signals S1 and S2 generated by the signal source are reconstructed to complete the signal detection.

[0018] When using the aforementioned architecture for signal detection, random noise can be effectively suppressed through N repeated measurements and ensemble averaging. Based on the principle of second-order correlation of the light field, the bit transmission signal can be effectively reconstructed.

[0019] Preferably, in step 2) E i The expression is: ; Among them, α represents the attenuation coefficient of light field transmission, E X represents the propagation light field corresponding to the signal light path X, E Y (t) represents the propagation light field corresponding to the reference light path Y, E L represents the output light field of the local oscillator light source; S1(t) and S2(t) represent the bit transmission signals received by the IQ modulator respectively; Δ φ =(ω S -ω L )t+( φ S - φ L ), where ω S and ω L Respectively represent the angular frequencies of the emitted light fields of the signal light source and the local oscillator light source, φ S and φ LRepresent the initial phases of the outgoing light fields of the signal light source and the local oscillator light source respectively; the initial coherent detection light intensity in step 3) is recorded as I i (t), using N represents the ensemble average of N measurement results, and As the reconstructed bit transmission signals respectively, then:

[0020] .

[0021] Compared with the prior art, the present invention has the following beneficial effects:

[0022] The technical solution of the present invention adopts a polarization multiplexing route architecture to design an optical chip, so that local oscillator optical amplification can further reduce the bit error rate of the signal transmitted by the optical chip. The optical chip based on the above architecture also uses the second-order correlation of the light field to complete signal detection, while effectively eliminating the noise and low power that may be introduced during local oscillator optical amplification on signal detection, greatly broadening the application scenarios of the optical chip. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A schematic diagram of a chip architecture for optical communications;

[0024] Figure 2 is a flow chart of a highly sensitive signal detection method;

[0025] Figure 3 is a waveform diagram of bit transmission signals S1 and S2 emitted by a signal source;

[0026] Figure 4 A comparison chart of the waveform of a reference signal and the average light intensity of the reference signal in 1000 measurements;

[0027] Figure 5 This is a waveform diagram of a reconstructed two-bit transmission signal.

[0028] Figure 1: Modulation transmitter 1, signal light source 111, amplitude modulator 112, arbitrary waveform generator 113, first beam splitter 114, signal source 121, IQ modulator 122, second beam splitter 1221, first Mach-Zehnder modulator 1222, second Mach-Zehnder modulator 1223, phase shifter 1224, beam combiner 1225, polarization beam combiner 131, balanced detection end 2, polarization beam splitter 221, local oscillator light source 211, first 90° optical mixer 231, second 90° optical mixer 232, photodetector array 241, coherent processor 251, first low-speed detector 2411, second low-speed detector 2412, third low-speed detector 2413, fourth low-speed detector 2414, first high-speed detector 2415, second high-speed detector 2416, third high-speed detector 2417, fourth high-speed detector 2418. DETAILED DESCRIPTION

[0029] The following will be combined with the present invention Figures 1 to 5 , the technical solutions in the embodiments of the present invention are described in detail.

[0030] like Figure 1 Figure 2 shows a schematic diagram of a chip architecture for optical communications. This architecture is based on the principle that light fields exhibit inherent correlation in statistical terms. Polarization multiplexing is used to construct signal and reference optical paths, enabling adjustments to be made for both the modulated transmitter and the coherent receiver. At the coherent receiver, local oscillator light is used to amplify the transmitted signal. Although this amplification process also synchronously amplifies noise and introduces new phase noise, the second-order correlation of the light field significantly reduces the impact of these factors on the actual signal, improving signal detection sensitivity. Multiple measurements can then be combined to enhance the signal-to-noise ratio. IQ modulation is also employed to further increase information transmission density.

[0031] Modulation transmitter 1 includes a signal light source 111, an amplitude modulator 112, an arbitrary waveform generator 113, a first beam splitter 114, a signal source 121, an IQ modulator 122, and a polarization beam combiner 131. The light beam emitted by signal light source 111 is modulated by amplitude modulator 112 and then split into two paths: a signal light path X and a reference light path Y. Amplitude modulator 112, driven by arbitrary waveform generator 113, amplitude modulates the light beam emitted by signal light source 111. Light I from signal light path X is incident on IQ modulator 122, and signal source 121 applies an electrical signal to IQ modulator 122. The output light beam from IQ modulator 122 is incident on polarization beam combiner 131, as is light II from reference light path Y.

[0032] In this embodiment, the output signal of the arbitrary waveform generator 113 is a series of pseudo-random signals. The amplitude modulator 112 is controlled by the arbitrary waveform generator 113 to randomly modulate the light intensity emitted by the signal light source 111 .

[0033] In this embodiment, the first beam splitter 114 is used to achieve non-uniform beam splitting, so that the incident light intensity of the reference light path Y is much smaller than the incident light intensity of the signal light path X. The specific numerical relationship is as follows: Y ≤I X / 10,I X Represents the intensity of light I incident on the signal optical path X; I Y Indicates the intensity of light II incident on the reference optical path Y. For example, you can select I Y =I X / 10, controlling the light intensity is equivalent to significantly suppressing the optical power in the reference optical path Y, which can reduce crosstalk and also ensure that the reference signal can be used to calibrate polarization and phase noise later.

[0034] The IQ modulator 122 includes a second beam splitter 1221, a first Mach-Zehnder modulator 1222, a second Mach-Zehnder modulator 1223, a phase shifter 1224, and a beam combiner 1225. Specifically, the signal source 121 generates a radio frequency electrical signal, which is applied to the first Mach-Zehnder modulator 1222 and the second Mach-Zehnder modulator 1223. The phase shifter 1224 is used to increase the phase of the propagating light field corresponding to the signal optical path X by π / 2. The second beam splitter 1221 and the beam combiner 1225 adopt an adiabatic coupling structure or a multimode interference structure. The second beam splitter 1221 achieves 5:5 splitting, and the beam combiner 1225 achieves 5:5 splitting and combining, ensuring signal consistency during transmission. The adiabatic coupling structure stabilizes the splitting ratio and enables high-power processing with low loss. The multimode interference structure is compact and compact, and can flexibly adjust the various splitting ratios. The IQ modulator can physically separate the reference optical path and the signal optical path used for measurement. At the same time, under IQ modulation, the optical fields of the reference signal and the measurement signal are orthogonal, effectively reducing interference between the signals.

[0035] In this embodiment, the polarization beam combiner 131 has two functions: first, it rotates the polarization of light II in the reference optical path Y, converting the TE mode beam into the TM mode; second, it combines the two beams with different polarization states into a single beam for output. The TE mode means that the electric field direction lies entirely within the transverse plane of light propagation, that is, perpendicular to the direction of light propagation; the TM mode is the opposite of the TE mode.

[0036] The balanced detection end 2 includes a polarization beam splitter 221, a local oscillator light source 211, a polarization beam splitter 221, a first 90° optical mixer 231, a second 90° optical mixer 232, a photodetector array 241, and a coherent processor 251. The polarization beam splitter 221 receives the optical signal from the polarization beam combiner 131 in the modulated transmitter 1. Using the same structural design as the polarization beam combiner 131, the polarization of the TM mode light in the received light is rotated into TE mode light II' and inputted into the reference optical path Y. The TE mode light I' in the received light is also inputted into the signal optical path X. The signal optical path X runs sequentially from the first beam splitter 114, the IQ modulator 122, the polarization beam combiner 131, the polarization beam splitter 221, to the first 90° optical mixer 231. The signal optical path Y runs sequentially from the first beam splitter 114, the polarization beam combiner 131, the polarization beam splitter 221, to the second 90° optical mixer 232. The reference optical path Y is connected to the first 90° optical mixer 231, and the reference optical path Y is connected to the second 90° optical mixer 232. The first 90° optical mixer 231 and the second 90° optical mixer 232 each have four output ports, and the output light is incident on the photodetector array 241. The photodetector array 241 converts each incident optical signal into a corresponding electrical signal, which is then input into the coherent processor 251.

[0037] The emission wavelengths of the local oscillator light source 211 and the signal light source 111 are equal, λ = 1550nm; expressed in angular frequency, ω S =ω L =193THz; where ω S and ω L They represent the angular frequencies of the output light fields of the signal light source 111 and the local oscillator light source 211. In addition, it is also necessary to control the light intensity of the local oscillator light source 211 incident on the first 90° optical mixer 231 and the second 90° optical mixer 232 to be equal, which facilitates calculation and high-sensitivity control.

[0038] The first 90° optical mixer 231 and the second 90° optical mixer 232 may both be a fourth-order MMI structure of silicon nitride material. MMI stands for multi-mode interference. In this case, the following characteristics are exhibited at 1550 nm:

[0039] ;in, Phase error is the phase error, CMRR It refers to the common mode rejection ratio. The smaller the phase error and common mode rejection ratio, the smaller the impact of the optical device on the signal.

[0040] In this embodiment, the photodetector array 241 includes eight photodetectors, including a first low-speed detector 2411, a second low-speed detector 2412, a third low-speed detector 2413, and a fourth low-speed detector 2414, with a bandwidth of BW1 = 500 MHz, which are used to record the total light intensity in a certain time window; it also includes a first high-speed detector 2415, a second high-speed detector 2416, a third high-speed detector 2417, and a fourth high-speed detector 2418, with a bandwidth of BW2 = 20 GHz. Each high-speed detector needs to match the modulation rate of the arbitrary waveform generator 113 to ensure the accuracy of calibration. The first low-speed detector 2411 and the second low-speed detector 2412 combine to output a set of mixing result signals, generating an initial coherent detection light intensity in the coherent processor 251, which is recorded as I1(t); the third low-speed detector 2413 and the fourth low-speed detector 2414 combine to output a set of mixing result signals, generating an initial coherent detection light intensity in the coherent processor 251, which is recorded as I2(t); the first high-speed detector 2415 and the second high-speed detector 2416 combine to output a set of mixing result signals, generating an initial coherent detection light intensity in the coherent processor 251, which is recorded as I3 (t); the third high-speed detector 2417 and the fourth high-speed detector 2418 are combined to output a set of mixing result signals, and the initial coherent detection light intensity is generated in the coherent processor 251 and recorded as I4(t); this differential merging process can eliminate light intensity noise and common-mode interference, and the high-speed combination is conducive to the rapid separation of instantaneous signals, and the low-speed combination processes slowly varying parameters and provides a time-integrated signal. The time-integrated signal is used to provide a noise reference for the high-speed detector group, ultimately realizing high-sensitivity, low-noise coherent optical communication, while supporting real-time adjustment of complex modulation formats.

[0041] In addition, it should be noted that, for components with the same name in this solution, instead of using labels to distinguish them, they can also be distinguished by using the first or second prefix.

[0042] like Figure 2 As shown in FIG, it is a flow chart of a high-sensitivity signal detection method; Figure 3 As shown in FIG, it is a waveform diagram of the bit transmission signals S1 and S2 emitted by a signal source; Figure 4 As shown in FIG, it is a comparison diagram of the waveform of a reference signal and the average light intensity of the reference signal in 1000 measurements; Figure 5 , which is a waveform diagram of a reconstructed two-bit transmission signal; Figures 3 to 5 The vertical axis in the table represents the light intensity in units of au. Figures 3 to 5 The horizontal axis in the table represents time in ns. Figures 1 to 5 As shown, the signal detection method includes the following steps:

[0043] Step 1): Before signal detection, control the signal source 121 to generate a bit transmission signal S1 with a length of M=20 at a rate of G=10 GHz, as shown in FIG. Figure 3 (a), and loaded onto the corresponding transmission light field through the first Mach-Zehnder modulator 1222; at the same time, a bit transmission signal S2 with the same length of M=20 is generated, as shown in Figure 3 (b) and loaded onto the corresponding transmission light field through the second Mach-Zehnder modulator 1223. Both light fields show amplitude changes, and the effective characteristic time of the two-bit transmission signal is τ1 = 0.1ns.

[0044] Then, an arbitrary waveform generator 113 is used to generate a pseudo-random signal sequence as a reference signal to modulate the amplitude modulator 112, thereby achieving random modulation of the intensity of the light emitted by the signal light source 111. According to the Nyquist sampling theorem, the effective characteristic time τ2 of the pseudo-random signal sequence should not be greater than half of the effective characteristic time τ1 of the above-mentioned bit transmission signal. In the above-mentioned high-sensitivity signal detection method, τ2 = 0.05ns should be selected. In the 2ns time window, the length of the bit transmission signal is M = 20, then the length of the reference signal is 2M = 40. The specific normalized reference signal is as follows: Figure 4 In addition, similar to each bit transmission signal, the reference signal is also a time-discrete sequence.

[0045] Step 2): Use the coherent processor 251 to perform coherent analysis on each detector in the photodetector group: the first low-speed detector 2411\the second low-speed detector 2412, the third low-speed detector 2413\the fourth low-speed detector 2414, the first high-speed detector 2415\the second high-speed detector 2416 and the third high-speed detector 2417\the fourth high-speed detector 2418. Here, \ represents a combination, and the DC component is removed to obtain four different groups of light fields E i , i=1, 2, 3, 4. The output power of the control signal light source 111 is 11 mW. The light field amplitude is normalized, and the signal light field amplitude and the reference light field amplitude are E X =√10 and E Y =1, local oscillator light amplitude E L =2. The light field attenuation coefficients are α1=0.09 and α2=0.085, respectively, including the attenuation of the light beam emitted by the modulated transmitter 1 in actual communication applications. Light field E i The details are as follows:

[0046]

[0047] Wherein, S1(t) and S2(t) respectively represent the bit transmission signals received by the IQ modulator 122, which vary with time. S1(t) corresponds to the In-phase path, i.e., the corresponding path including the first Mach-Zehnder modulator 1222, where In-phase means in-phase, and S2(t) corresponds to the Quadrature path, i.e., the corresponding path including the second Mach-Zehnder modulator 1223, where Quadrature means orthogonal; Δ φ =(ω S -ω L )t+( φ S - φ L ), where ω S and ω L denote the angular frequencies of the light fields emitted by the signal light source 111 and the local oscillator light source 211, respectively. φ S and φ L They represent the initial phases of the outgoing light fields of the signal light source 111 and the local oscillator light source 211 respectively.

[0048] Step 3): Measure and record the coherent detection light intensities I1(t), I2(t), I3(t) and I4(t) at the balanced detection end 2; under the premise that the bit transmission signals S1 and S2 generated by the signal source 121 remain unchanged, repeat steps 1 to 2 multiple times and record the light intensities respectively, that is, achieve N = 1000 bit transmission signal measurements, as shown in FIG. Figure 4 The dotted line in the figure represents the mean value of the random intensity of the reference signal in 1000 measurements, which is a constant value of 0.5.

[0049] Step 4): Reconstruct the bit transmission signal generated by the signal source 121 through the second-order correlation of the light field and , the specific formula is as follows:

[0050]

[0051] in,<> N represents the ensemble average of N measurement results; and To reconstruct the bit transmission signal, it is shown in Figure 5 middle.

[0052] It should be noted that the second-order correlation of the light field is a core concept in quantum optics and statistical optics. It is used to reflect the statistical correlation characteristics of the light field at two different points in space and time. Therefore, based on this principle, the linearly transformed light field can be restored again, thereby restoring and reconstructing the corresponding bit transmission signal.

[0053] In summary, the present invention designs optical chips using a polarization multiplexing architecture, enabling local oscillator optical amplification to further reduce the bit error rate of signals transmitted by the optical chip. Furthermore, based on the optical chip's aforementioned architecture, the second-order correlation of the light field is utilized to perform signal detection, while effectively eliminating the effects of noise and low power that may be introduced during local oscillator optical amplification on signal detection. This significantly broadens the application scenarios of optical chips and represents a significant advancement.

[0054] The above embodiments are only for illustrating the technical idea of ​​the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the present invention.

Claims

1. A highly sensitive signal detection method, operating using a chip architecture for optical communications, characterized in that: The chip architecture includes: a modulation transmitting end (1) and a balanced detection end (2), the modulation transmitting end (1) is used to modulate the signal light source (111) into light I and light II, the light I is transmitted along the signal light path X, and the light II is transmitted along the reference light path Y, the modulation transmitting end (1) is provided with a polarization beam combiner (131) and an IQ modulator (122), the polarization beam combiner (131) is used to polarization-bundle the light II and the IQ-modulated light I and transmit the combined light to the balanced detection end (2), the IQ modulator (122) includes a second beam splitter (1221), a first Mach-Zehnder modulator (1222), a second Mach-Zehnder modulator (1223), a phase shifter (1224) and the beam combiner (1 225); a local oscillator light source (211), a polarization beam splitter (221), a photodetector array (241) and a coherent processor (251) are provided in the balanced detection end (2); the polarization beam splitter (221) is used to receive the combined light and decompose it into light I' and light II'; the local oscillator light source (211) is used to perform 90° mixing with light I' and light II' respectively and transmit the mixing results to the photodetector array (241); the emission wavelengths of the local oscillator light source (211) and the signal light source (111) are equal; the photodetector array (241) is used to perform photoelectric conversion on the mixing results and transmit them to the coherent processor (251) to complete coherent analysis; the signal detection method comprises the following steps: 1) First, a signal source (121) is used to generate bit transmission signals S1 and S2 of length M at a rate G, and the bit transmission signals S1 and S2 are loaded into the corresponding transmission light field through the first Mach-Zehnder modulator (1222) and the second Mach-Zehnder modulator (1223) respectively; the effective characteristic time of each bit transmission signal is ,τ1=1 / G; Then, an arbitrary waveform generator (113) is used to generate a pseudo-random signal with an effective characteristic time of τ2, which is used to drive an amplitude modulator (112), where 2τ2=τ1; a signal light source (111) emits signal light, which is sequentially transmitted to a photodetector array (241) through amplitude modulation, beam splitting, beam combining, decomposition, and frequency mixing; 2) The photodetector array (241) transmits four sets of electrical signals to the coherent processor (251), and the coherent processor (251) generates four sets of different light fields E based on the four sets of electrical signals. i , i=1, 2, 3, 4; 3) Record and measure four different light fields E i The initial coherent detection light intensity corresponding to each of them; the bit transmission signal generated by the control signal source (121) remains unchanged, step 1) to step 2) are repeated N times, N is a positive integer, and the coherent detection light intensity of each repetition is recorded; 4) Based on the second-order correlation of the light field, the initial coherent detection light intensity in step 3) and the repeated coherent detection light intensity are calculated, and the bit transmission signals S1 and S2 generated by the signal source (121) are reconstructed to complete the signal detection.

2. A high-sensitivity signal detection method according to claim 1, characterized in that: The modulation transmitting end (1) is further provided with an amplitude modulator (112), an arbitrary waveform generator (113), a first beam splitter (114) and a signal source (121); the amplitude modulator (112) is driven by the arbitrary waveform generator (113) and is used to perform amplitude modulation on the signal light source (111); the first beam splitter (114) is used to modulate the amplitude-modulated signal light source (111) into light I and light II; the signal source (121) is used to load an electrical signal to the IQ modulator (122), and the IQ modulator (122) performs IQ modulation on the light I based on the electrical signal.

3. A high-sensitivity signal detection method according to claim 2, characterized in that: The second beam splitter (1221) is connected to the beam combiner (1225) via the first Mach-Zehnder modulator (1222), and is further connected to the beam combiner (1225) via the second Mach-Zehnder modulator (1223) and the phase shifter (1224).

4. A high-sensitivity signal detection method according to claim 2, characterized in that: The arbitrary waveform generator (113) is used to generate a pseudo-random signal, and the pseudo-random signal is used to drive the amplitude modulator (112).

5. A high-sensitivity signal detection method according to claim 2, characterized in that: The first beam splitter (114) adopts non-uniform beam splitting; when the first beam splitter (114) performs beam splitting, the light intensity I of the control light II is Y Light intensity I is less than light I X , I Y ≤I X / 10.

6. A high-sensitivity signal detection method according to claim 3, characterized in that: The second beam splitter (1221) and the beam combiner (1225) adopt a multimode interference structure or an adiabatic coupling structure; the second beam splitter (1221) is used for 5:5 light splitting, and the beam combiner (1225) is used for 5:5 beam combining.

7. A high-sensitivity signal detection method according to claim 1, characterized in that: Before the polarization beam combiner (131) performs polarization beam combining, the light II is first converted from the TE mode to the TM mode, and then the TM mode light II is polarization-combined with the IQ modulated light I; before the polarization beam splitter (221) performs decomposition, the TM mode light polarization in the combined light transmitted by the polarization beam combiner (131) is rotated into the TE mode light II' for output, and the TE mode light in the combined light is also output as the light I'.

8. A high-sensitivity signal detection method according to claim 1, characterized in that: A first 90° optical mixer (231) and a second 90° optical mixer (232) are also provided in the balanced detection end (2), and both 90° optical mixers adopt a fourth-order MMI structure or a plurality of low-order MMI cascade structures; when the local oscillator light source (211) is used for mixing, incident light of equal intensity is emitted to the first 90° optical mixer (231) and the second 90° optical mixer (232); The photodetector array (241) includes four low-speed detectors and four high-speed detectors; the four low-speed detectors are used to receive the mixing results of the first 90° optical mixer (231), perform photoelectric conversion, and transmit the results to the coherent processor (251); the four high-speed detectors are used to receive the mixing results of the second 90° optical mixer (232), perform photoelectric conversion, and transmit the results to the coherent processor (251).

9. The high-sensitivity signal detection method according to claim 1, characterized in that: E in step 2) i The expression is: ; Among them, α represents the attenuation coefficient of light field transmission, E X represents the propagation light field corresponding to the signal light path X, E Y (t) represents the propagation light field corresponding to the reference light path Y, E L represents the output light field of the local oscillator light source (211); S1(t) and S2(t) respectively represent the bit transmission signals received by the IQ modulator (122); , where ω S and ω L represent the angular frequencies of the light fields emitted by the signal light source (111) and the local oscillator light source (211), respectively. and Respectively represent the initial phases of the light fields emitted by the signal light source (111) and the local oscillator light source (211); The initial coherent detection light intensity in step 3) is recorded as I i (t), using N represents the ensemble average of N measurement results, and As the reconstructed bit transmission signals, the expressions are: 。

Images