JP5446094B2 - Photon detection device, photon detection method, and photon detection program - Google Patents

Description

  The present invention relates to a single-photon detection circuit in a gate mode using an avalanche photodiode, and more particularly to setting a gate phase and an identification threshold in the circuit.

  In order to perform long-distance single-photon transmission using an optical fiber, a compound-based avalanche photodiode (hereinafter referred to as APD) having sensitivity in the 1.55 μm band is an optimum element as a photon detection element. As described in Non-Patent Document 1, in a single photon detection device using a compound APD, application of a gate mode is essential.

  The photon detection signal output from the APD in the gate mode is the right side of FIG. 3 of Non-Patent Document 3, as shown in the second stage Transient pulse of the waveform time chart, and the photon detection signal is included in the differential waveform of the gate pulse. It is superimposed. Since this differential waveform is caused by the parasitic capacitance of the PN junction of the APD, it is sometimes called a charge pulse.

  The charge pulse is observed as a large amplitude as the circuit band increases, and inhibits the detection of a small-amplitude photon detection signal. Non-Patent Documents 2, 3, and 4 have been devised as high-accuracy charge pulse compensation methods for solving this problem and realizing highly sensitive photon detection.

  Here, FIG. 2 of Non-Patent Document 5, FIG. 2A of Patent Document 1 and FIG. 4 of Non-Patent Document 7 report analysis on the amplitude of the signal output from the APD in the photon detection in the gate mode. ing. These reports reveal that the photon signal and noise amplitude distribution change depending on the threshold voltage, and that the count rate increases due to background noise in the low amplitude region.

  In the methods of Non-Patent Documents 2 to 4, charge pulse compensation is performed on the APD output signal by signal processing as an analog signal, and the signal is identified. For this reason, a compensation signal generation circuit and a delay adjustment are required in accordance with the type of APD and individual differences.

  On the other hand, Non-Patent Document 6 points out that there is a problem of charge pulse compensation error due to individual differences of elements in the charge pulse compensation circuits of Non-Patent Documents 2 to 4, and together with Patent Document 2, The solution is shown. More specifically, charge pulse compensation and signal identification are performed by sampling an APD output waveform using a high-speed A / D converter (ADC) and performing digital signal processing (hereinafter referred to as an ADC system). Since the past APD output waveform is used as the compensation waveform, individual adjustment of APD is not required, and the mass productivity is excellent.

  FIG. 8 is a circuit block diagram showing a configuration of a conventional photon detection circuit 300a described in Non-Patent Documents 5-6. The photon detection circuit 300a includes a gate generation circuit 2 for generating a gate pulse to be applied to the APD 1 and APD, a clock source 3 and a clock processing circuit 31, a sampling circuit 5, a memory 6, a memory control circuit 61, It comprises an average waveform generation circuit 62 and an identification circuit 9.

  The clock processing circuit 31 supplies the gate signal S32 to the gate generation circuit 2 and the sampling clock S35 to the sampling circuit 5 in synchronization with the clock S3 supplied from the clock source 3. The gate generation circuit 2 generates a gate pulse S21 based on the gate signal S32 and applies it to the APD1.

  The APD 1 performs single photon detection with a gate pulse S21 with respect to the incidence of the single photon signal S1, and outputs a detection signal S15. The sampling circuit 5 samples the detection signal S15 at the cycle of the sampling clock S35 and converts it to a sampling waveform S5. The sampling waveform is a multi-bit digital signal.

  The memory control circuit 61 creates an address for recording the sampling waveform S5 on the memory 6 based on the gate signal S32 and the sampling clock S35. The sampling waveform S5 is recorded at an address created by the memory control circuit 61 of the memory 6. In the average waveform generation circuit 62, the sampling waveform S5 recorded in time series in the memory 6 is read in units of gate periods, and the average waveform S6 is generated by calculation.

  The identification circuit 9 matches the phases of the average waveform S6 and the sampling signal S5 based on the gate signal S32, calculates the difference between the two waveforms, identifies the presence or absence of the photon detection signal, and generates the photon detection signal S9.

  FIG. 9 is a conceptual diagram illustrating an operation from generation of an average waveform to identification in the photon detection circuit 300a disclosed in FIG. Arbitrary N times of gate pulses S21 are applied to the APD 1 in the time direction, and the waveform of the detection signal S15 output from the APD 1 is sampled by the sampling circuit 5 in units of the period of the gate pulse S21 and stored in the memory 6 at the same time. , N times of sampled waveforms are averaged.

  In FIG. 9, (A) shows the waveform of the gate pulse S21, (B) shows the waveform of the detection signal S15 output from the APD1 up to the Nth time, and (C) shows the waveform of the detection signal S15 up to the Nth time. The stored contents are sampled by the circuit 5 and stored in the memory 6.

  The average waveform generation circuit 62 averages the waveform of the detection signal S15 up to the Nth time shown in FIG. 9C, whereby an average waveform (N-time averaged waveform) S6 shown in FIG. can get. Further, the (N + 1) th detection signal S15 shown in (E) of FIG. 9 is subtracted from the average waveform S6 by subtraction to compensate the charge pulse, as shown in (F) of FIG. The (N + 1) th differential waveform is obtained.

  In FIG. 9, the single photon signal S1 is not incident on the APD 1 in the (N-3) th to Nth times, whereas the single photon signal S1 is incident in the (N + 1) th time. For this reason, since a waveform having a clear difference is detected as the (N + 1) th differential waveform, the photon can be easily identified in the identification circuit 9.

  10 to 11 are conceptual diagrams for explaining a method for determining an optimum phase and threshold when identifying a photon in the identification circuit 9 disclosed in FIG. With respect to the gate pulse S21 shown in the waveform (A) of FIG. 10, the detection signal S15 output from the APD 1 increases in amplitude in the latter half of the gate pulse period, as shown by the dotted line in the waveform (B) of FIG. . This is due to the extremely large multiplication factor of APD while the gate is applied. Also in the (N + 1) th differential waveform shown in the waveform (C) of FIG. 10, an increase in amplitude is observed in the second half of the gate pulse period.

  Due to this property, the condition for the maximum photon detection efficiency is to match the phase of photon incidence to the front end of the gate pulse S21, as shown in the waveform (D) of FIG. In order to determine the phase, the phase of the gate pulse S21 (hereinafter sometimes referred to as the gate phase) is scanned as shown in the waveform (E) as shown in the waveform (E) with respect to the incident timing of the photon. It is necessary to measure the photon detection efficiency for each phase condition with the (N + 1) th differential waveform.

  Therefore, in the photon detection circuit 300a of FIG. 8, the gate phase is adjusted as follows. The phase of the gate signal S32 is changed by the delay circuit 22 controlled by the gate phase control circuit 23. The photon detection signal S9 from the identification circuit 9 is counted by the counter 96a and recorded in the memory 91 as a value corresponding to the variable amount of the phase.

  The memory control circuit 94 creates an address for recording the photon detection signal S9 on the memory 91 based on the gate signal S32. After the gate phase control circuit 23 finishes scanning a preset phase variable width, the maximum value detection circuit 93 obtains the phase at which the count number of the counter 96a is maximum.

  On the other hand, the identification circuit 9 needs to set an optimum value as a threshold value for identifying the presence or absence of photon detection in the (N + 1) th differential waveform. As shown in FIG. 2 of Non-Patent Document 5, it is necessary to scan the threshold setting and measure the photon detection probability at each threshold.

  FIG. 12 is a circuit block diagram showing a configuration of a photon detection circuit 300b in which a configuration for further scanning a threshold is added to the photon detection circuit 300a disclosed in FIG. In the photon detection circuit 300b, a threshold control circuit 95 is added to the photon detection circuit 300a of FIG. The threshold value of the identification circuit 9 is controlled by a threshold control signal S91 from the threshold control circuit 95. It is assumed that the gate phase is optimized by the above sequence.

  The photon detection signal S9 from the identification circuit 9 is counted by the counter 96a and recorded in the memory 91 as a value corresponding to the variable amount of the threshold. After the threshold control circuit 95 finishes scanning a preset threshold variable width, the maximum value detection circuit 93 obtains the phase at which the count number of the counter 96a is maximum. The threshold scan may be performed in two ways: measurement in the photon incident state and measurement of the number of counts by dark count noise in the absence of photon incidence.

  The memory control circuit 94 creates an address for recording the photon detection signal S9 on the memory 91 based on the gate signal S32. In the above description, the memory 91 is described as being commonly used for the gate phase scan and the threshold value scan. However, a separate memory may be provided for each of them.

B.F.Levine and C.G.Bethea J.C.Campbell, "Near room temperature 1.3um single photon counting with a InGaAs avalanche photodiode" Electronics Letters, vol.20 No.14, p596, 1964 Donald S Bethune, William P. Risk, Gary W. Pabst, "A high-performance integrated single-photon detector for telecom wavelengths" Journal of Modern Optics, Vol. 51, No. 9-10 / 15 June-10, pp1359- 1368, July 2004 GRE´ GOIRE RIBORDY, NICOLAS GISIN, OLIVIER GUINNARD, DAMIEN STUCKI, MARK WEGMULLER and HUGO ZBINDEN, "Photon counting at telecom wavelengths with commercial InGaAs / InP avalanche photodiodes: current performance" journal of modern optics, vol. 51, no. 9- 10, pp1381-1398,15 june-10 july 2004 Akio Yoshizawa, Ryosaku Kaji, and Hidemi Tsuchida, "Gated-mode single-photon detection at 1550 nm by discharge pulse counting" APPLIED PHYSICS LETTERS, VOLUME 84, NUMBER 18, pp3606-3608, 3 MAY 2004 Seigo Takahashi, Takeshi Nakata, Akio Tajima and Akihisa Tomita, "Accurate Evaluation of APD Noise Characteristics for Quantum Cryptosystems" 11th Optoelectronics & Communications Conf (OECC2006), July, 2006 Seigo Takahashi, Akio Tajima, Akihisa Tomita (NEC, JST), "High-efficiency single photon detector combined with an ultra-small APD module and a self-training discriminator for high-speed quantum cryptosystems" The 13th Microoptics conference (MOC'07 ) technical digest, Oct 28, 2007 PAUL L. VOSS, KAHRAMAN G. KO¨ PRU¨ LU¨, SANG-KYUNG CHOIy, SARAH DUGAN and PREM KUMAR, "14MHz rate photon counting with room temperature InGaAs / InP avalanche photodiodes" journal of modern optics, 15 june-10 july 2004 , Vol. 51, no. 9-10, pp1369-1379 JP 2007-147472 A JP 2006-284202 A

  However, setting an optimal gate phase requires a large memory and a circuit scale, and further requires a long time measurement. The reason is that there is no method for efficiently detecting the phase at which the photon signal arrives during the gate period in the digital signal processing circuit. Therefore, the average waveform is calculated and the photon detection is counted. This is because it is necessary to search for the optimum photon detection phase.

  At this time, in the actual system, the amount of delay depending on the fiber length is different, so the photon detection phase is unknown during the gate period. Therefore, it is necessary to search for a signal waveform by a phase adjustment sequence when the system is started up. Here, in order to perform a search with high accuracy, it is necessary to perform sufficient photon detection counting. At this time, in order to perform the averaging process and the identification process for a long time, a large memory area and a large-scale circuit for processing the memory area are required.

  More specifically, assuming that the time width of the gate pulse shown in the waveform (D) of FIG. 11 is 5 nsec, the sampling frequency is 1 GHz, and phase adjustment is performed with a time width three times that of the gate pulse, the waveform (E As shown in (1), sampling per gate period requires 20 points or more. When the ADC resolution is 10 bits, 200-bit data is generated in one gate period.

  For each sampling point, in order to repeat the calculation of the average waveform and the signal identification process with high accuracy, averaging of about 100 gate cycles is required. Therefore, in the averaging process shown in FIG. 9, the memory 6 needs a memory capacity of 20 kilobits or more. At the same time, the average waveform generation circuit 62 that performs the averaging process from these data also becomes large-scale.

  In addition, the gate phase must be determined so that the photon detection efficiency is maximized. As described above, it is considered that the photon detection efficiency is maximized when a photon is incident on the leading portion of the gate, since avalanche multiplication can be expected over the entire time width of the gate. At this time, considering the accuracy of photon detection, when performing single photon transmission using a long-distance fiber, the photon detector detects photons with a probability of 1 / 10,000 to 1 / 100,000.

For this reason, in order to obtain the frequency of photon detection with sufficient accuracy with respect to the sampling points, it is necessary to count photon detection up to the order of 1000 to 10,000 times. For this purpose, it is necessary to count by a gate pulse is applied about 10 ⁹ times. This requires 100 seconds for one gate phase when the gate period is 100 nsec. When a gate sweep of 15 nsec is performed in units of 1 nsec, since this count is performed for 15 gate phases, a very long time of 1500 seconds = 25 minutes is required.

  Similarly, in order to obtain the optimum signal identification threshold for the (N + 1) th differential waveform, a large memory and circuit scale are required, and a longer measurement is required. The reason is that it is difficult to distinguish between the noise of the sampling signal caused by the noise in the apparatus and the dark count noise emitted from the single photon detector by APD.

  In order to identify a minute photon detection signal as shown in the waveform (C) of FIG. 10, it is necessary to lower the identification level (threshold). However, as shown in FIG. 2 of Non-Patent Document 5 or FIG. 4 of Non-Patent Document 7, as the threshold value is lowered to near the baseline, the device noise and the noise of the sampling signal resulting therefrom are misidentified, and the error probability Leads to an increase.

  In other words, there is an optimum point for a phenomenon in which detection errors due to system noise increase, contrary to the fact that lowering the threshold value increases the photon detection efficiency. However, in order to accurately determine the optimum threshold level for each single photon detector, the threshold value is scanned and obtained from the lowest value such as the error rate. It was.

Noise mixed in the identification circuit of the photon detection circuit includes power supply noise and amplitude fluctuation caused by sampling clock jitter. These occur randomly with a statistical distribution. When an error rate of about 10 ⁻⁵ to 10 ⁻⁸ per gate pulse is required for photon detection in single photon transmission, the error rate can be reduced by using a sufficiently high value from the noise level for the threshold of the discrimination circuit. At the same time, a small-amplitude photon detection signal cannot be detected, and the photon detection performance deteriorates.

In order to determine the lowest threshold for obtaining a desired error rate, measurement with an accuracy of 10 ⁻⁸ is required for a stochastic phenomenon. For example, assume that noise generated with a probability of 10 ⁻⁵ is counted about 1000 times in a gate pulse period of 1 μsec. This measurement takes 100 seconds for one threshold. In order to determine the optimum threshold value for each photon detector, it is necessary to repeat the measurement while adjusting the threshold value, and a very large adjustment time is required.

  As described above, setting of the optimum values of the gate phase and the discrimination threshold requires a large memory and a circuit scale, and further requires a long time measurement. This has led to mass production failures and even increased downtime in the calibration sequence during system operation.

  An object of the present invention is to provide a photon detection device, a photon detection method, and a photon detection program that can optimize a photon detection phase and a discrimination threshold with high speed and high accuracy with a small amount of memory.

To achieve the above object, the photon detector according to the present invention comprises a gate voltage generation circuit for generating a gate voltage based on the gate period, and a gate phase control circuit for changing the phase of the gate voltage, avalanche photodiode A sampling circuit that samples a signal output from the digital signal and converts it into discrete data based on digital information, a first memory circuit that stores discrete data for each gate period, and a statistic that calculates a statistical value of discrete data for each gate period and a value operation circuit further includes a second memory circuit for storing statistics for each change gates phase, the phase of the maximum value and the gate voltage at that time the statistical value stored in the second memory circuit An average of the APD output waveform data using the maximum statistical value detection circuit to detect and the discrete data stored in the first memory circuit Possess an average waveform data generation circuit for generating a waveform, compares the signal output from the average waveform data and the avalanche photodiode for each gate period and a decision circuit which performs identification of the photon detection signal, a gate phase control circuit However, the phase of the gate voltage is changed so that the statistical value detected by the maximum statistical value detection circuit becomes the maximum value, and the statistical value calculation circuit causes the noise component from the standard deviation value at each sampling phase in the statistical value. A threshold control circuit that calculates a detection error rate and calculates a threshold value that serves as a reference for identification of the photon detection signal in the identification circuit is provided so that the detection error rate satisfies a required level .

In order to achieve the above object, a photon detection method according to the present invention is a method for detecting a photon in a single-photon detection device using a gate mode using an avalanche photodiode, which generates a gate voltage based on a gate period. A voltage generation step, a gate voltage application step for applying a gate voltage to the avalanche photodiode, a sampling step for sampling a signal output from the avalanche photodiode and converting it into discrete data based on digital information, and discrete data for each gate period A first storage step for storing data, a statistical value calculation step for calculating statistical values of discrete data for each gate period, a second storage step for storing statistical values for each phase of the gate voltage, and a second storage maximum detecting the phase of the maximum value and the gate voltage at that time of the stored statistics in step integration Using the value detection step, a gate phase control step of statistical value detected by the maximum statistical value detection step to change the phase of the gate voltage so that the maximum value, the discrete data stored in the first storage step A level at which a detection error rate is required by calculating an average waveform data generation step for generating an average waveform of APD output waveform data, and calculating a detection error rate due to a noise component from a standard deviation value at each sampling phase in a statistical value. A threshold control step for calculating a threshold so as to satisfy the above, and an identification step for identifying the photon detection signal based on the threshold by comparing the average waveform data and the signal output from the avalanche photodiode for each gate period. It is characterized by having.

In order to achieve the above object, a photon detection program according to the present invention outputs from avalanche photodiodes to which a gate voltage based on a gate period is applied to a computer included in a gate mode single photon detection apparatus using avalanche photodiodes. Sampling processing for sampling the converted signal and converting it into discrete data based on digital information, first storage processing for storing discrete data for each gate period, and statistical value calculation for calculating a statistical value of discrete data for each gate period maximum statistical value detection process for detecting a process, a second storage process of storing statistical value for each phase of the gate voltage, the phase of the maximum value and the gate voltage at the time of the second statistical values stored in the storage process When the phase of the gate voltage so that the statistical value detected by the maximum statistical value detection process is a maximum value variance A gate phase control process of the average waveform data generation process of generating an average waveform of the APD output waveform data by using the discrete data stored in the first storing process, the standard deviation at each sampling phase in statistics The detection error rate due to the noise component is calculated from the value of, and the photon detection is performed by comparing the average waveform data and the signal output from the avalanche photodiode for each gate period so that the detection error rate satisfies the required level. And a threshold value control step of calculating a threshold value that is a reference for identification when the signal is identified .

  In the present invention, the signal output from the avalanche photodiode is sampled and converted into discrete data based on digital information, and the dispersion value of the discrete data is handled by a statistical method. A sample number of about tens of times is sufficient. As a result, it is possible to provide a photon detection device, a photon detection method, and a photon detection program that are excellent in the past, and that can optimize the photon detection phase and the identification threshold with high speed and high accuracy with a small amount of memory.

  FIG. 1 is a circuit block diagram showing a configuration of the photon detection circuit 100a according to the first embodiment of the present invention. The photon detection circuit 100 a includes APD 1, a gate generation circuit 2 for generating a gate pulse to be applied to APD 1, a clock source 3, a clock processing circuit 31, a sampling circuit 5, a memory 6, a memory control circuit 61, and an average waveform generation circuit 62. , A dispersion value calculation circuit 63, and an identification circuit 9.

  The clock processing circuit 31 supplies the gate signal S32 to the gate generation circuit 2 and the sampling clock S35 to the sampling circuit 5 in synchronization with the clock S3 supplied from the clock source 3. The gate generation circuit 2 generates a gate pulse S21 based on the gate signal S32 and applies it to the APD1. The gate phase control circuit 23 controls the delay circuit 22 by the gate phase control signal S23 to change the gate phase delay amount of the gate signal S32. The APD 1 performs single photon detection with a gate pulse S21 with respect to the incidence of the single photon signal S1, and outputs a detection signal S15.

  The sampling circuit 5 samples the detection signal S15 at the cycle of the sampling clock S35, converts it to a sampling waveform S5, and records it in the memory 6. From the sampling waveform S5 recorded in the memory 6, the average waveform generation circuit 62 generates an average waveform S6. At the same time, the variance value calculation circuit 63 calculates a variance value for each sampling phase from the sampling waveform S5.

Here, the variance value is calculated by using the average value X _avr shown in _Equation 1 for the data X _i of the sampling waveform S5 of each sampling phase with the gate period as a unit, and the sample variance V (= σ ² ) or the unbiased variance s ² shown in Equation 3 is calculated.

  Here, the number n of repetitions may be a sample number of about 10 to several tens of times in terms of photon detection. Therefore, a time reduction of 1/10 to 1/100 when using the conventional photon detection count value can be expected.

The obtained sample variance V (= σ ² ) or unbiased variance s ² (hereinafter collectively referred to as variance value) data (variance value data S 63) corresponds to the delay amount of the gate phase given by the delay circuit 22. And stored in the memory 91. When the gate phase control circuit 23 finishes scanning the gate phase, the memory 91 records a two-dimensional dispersion value regarding the sampling phase and the gate phase.

  The maximum dispersion value detection circuit 93 detects the gate phase at which the dispersion is maximum with respect to the entire dispersion value stored in the memory 91. The gate phase at which the detected dispersion takes the maximum is a phase at which a signal based on the photon detection phenomenon is likely to appear, and can be determined as the gate phase at which the photon detection efficiency is optimum.

  The photon detection circuit 100a of FIG. 1 requires a control circuit for controlling the cooperative operation of the entire system in addition to the functions shown in the figure, which is well known to those skilled in the art and is particularly useful in explaining the present invention. Since it is not necessary to mention, description is omitted. Further, this control circuit may be configured by a computer, and processing described below may be realized based on a computer program that operates on the computer.

  FIG. 2 is a schematic diagram of a calculation result at a certain gate phase in the photon detection circuit 100a disclosed in FIG. By applying a gate pulse S21 shown in waveform (A) to APD1, a signal waveform S15 shown in waveform (B) is output from APD. Most of the signal waveform S15 is a waveform shown by a solid line, but becomes a waveform shown by a dotted line due to a rare photon detection event.

  The signal waveform S15 is sampled by the sampling circuit 5, stored in the memory 6, and the variance value shown in the waveform (G) is obtained by the variance value calculation circuit 63. The dispersion in the second half of the gate pulse where the amplitude of the signal waveform 15 of the APD output increases takes a large value. Further, even in the no-signal area at both ends of the waveform (G), the fluctuation of the signal due to the system noise takes a value larger than zero indicated by the dotted line as an offset.

  The phase at which the dispersion shown in the waveform (G) is maximized is determined as the gate phase at which the photon detection efficiency is optimal. The waveform (G) is a schematic diagram of the calculation result in one gate phase, and the gate phase is scanned by the gate phase control circuit 23 to obtain a graph of a plurality of dispersion values corresponding to the waveform (G). Can do.

  In order to detect the phase at which the dispersion shown in the waveform (G) is maximum, for example, as shown in a configuration diagram of FIG. 5 described later, a maximum value selection circuit 97 that selects the maximum value of dispersion in one gate phase. May be added. As a result, the amount of distributed data stored in the memory 91 can be reduced.

  FIG. 3 is a circuit block diagram showing a configuration of the photon detection circuit 100b according to the second embodiment of the present invention. Since the configuration of the photon detection circuit 100b is mostly the same as that of the photon detection circuit 100a according to the first embodiment, common components are denoted by the same reference numerals and description thereof is omitted.

  The photon detection circuit 100b is obtained by adding a threshold control circuit 96 to the photon detection circuit 100a according to the first embodiment. The conventional threshold control circuit 95 in the conventional photon detection circuit 300b described in FIG. 12 searches for the optimum value by scanning the set value of the threshold. The threshold control circuit 96 of the present embodiment is The threshold value is determined from the value of the standard deviation (σ) obtained as the square root of the variance (V).

  FIG. 4 is a schematic diagram of a threshold calculation result in the photon detection circuit 100b disclosed in FIG. By applying a gate pulse S21 shown in waveform (A) to APD1, a signal waveform S15 shown in waveform (B) is output from APD. Most of the signal waveform S15 is a waveform shown by a solid line, but becomes a waveform shown by a dotted line due to a rare photon detection event.

  The signal waveform S15 is sampled by the sampling circuit 5, stored in the memory 6, and the variance value shown in the waveform (G) is obtained by the variance value calculation circuit 63. Since the relationship between the standard deviation and the variance is schematically similar for positive values from Equation 4, the vertical axis of the waveform (G) is read as the standard deviation in this description.

In the regions at both ends of the waveform (G), no signal is changed because no gate is applied, and in principle, the standard deviation value is zero. In practice, however, a non-zero offset appears due to fluctuations in the signal and power supply originating from switching noise in the system. Based on the offset standard deviation value σ _N , the false detection probability (error probability) due to noise with respect to the threshold value can be determined.

The calculation of the error probability when the threshold value is _Vth can be obtained from Equation 5 from the probability density function of the normal distribution. In Equation 5, the average value (μ) of the measurement result samples is assumed to be zero and is omitted.

As an example, the threshold value (V _th ) as an integer multiple of the standard deviation and the error occurrence probability in that case are as follows.

Threshold (V _th ) Error probability 1σ 1.59E-01
2σ 2.28E-02
3σ 1.35E-03
4σ 3.17E-05
5σ 2.87E-07

That is, it can be seen that in order to set the error probability to 10 ⁻⁶ or less, it is only necessary to substitute σ into σ _N and set the threshold value V _th to 5σ.

  As described above, assuming that the noise caused by the system is a normal distribution, the standard deviation derived from the variance can be used to infer the distribution function from a small sample variation. As a result, the threshold value for obtaining the required noise occurrence probability can be obtained as an inverse function of the distribution function, so that the optimum discrimination threshold value can be determined at high speed.

Note that in the determination of the threshold value _Vth , signal fluctuation in the gate region can be taken into consideration. In the waveform (G) of FIG. 4, the standard deviation (σ _GP ) of the head portion of the gate is larger than the standard deviation (σ _N ) of the non-gate region. Since this is considered to be caused by the fluctuation of the gate signal itself, σ _GP is considered to be equal to the standard deviation of the noise of the signal in the gate. Therefore, by selectively reading the standard deviation σ _GP in the gate head region and substituting this σ _GP into σ to obtain the threshold value Vth, it is possible to set the signal in consideration of signal fluctuations in the gate region. .

  FIG. 5 is a circuit block diagram showing a configuration of a photon detection circuit 100c according to the third embodiment of the present invention. Since the configuration of the photon detection circuit 100c is mostly the same as that of the photon detection circuits 100a and 100b according to the first and second embodiments, common constituent elements are denoted by the same reference numerals and description thereof is omitted. To do. In addition, as described above, the maximum value selection circuit 97 that selects the maximum value of dispersion in one gate phase is added to reduce the amount of data stored in the memory 91.

  In the photon detection circuit 100c, an optical switch 11 that controls the incidence of photons on the APD 1 is added to the photon detection circuit 100b according to the second embodiment, and a memory 98 and an optical switch control circuit 99 are added to the threshold control circuit 96. It is supposed to be attached.

  The optical switch 11 is controlled to be opened and closed by the signal S92 from the optical switch control circuit 99 based on an instruction from the threshold control circuit 96. When the optical switch 11 is opened, photons can enter the APD 1, and when the optical switch 11 is closed, the photons are not incident on the APD 1.

  FIG. 6 is a schematic diagram of a threshold calculation result in the photon detection circuit 100c disclosed in FIG. By applying a gate pulse S21 shown in waveform (A) to APD1, a signal waveform S15 shown in waveform (B) is output from APD. Most of the signal waveform S15 is a waveform shown by a solid line, but becomes a waveform shown by a dotted line due to a rare photon detection event.

With the optical switch 11 opened, this signal waveform S15 is sampled by the sampling circuit 5, stored in the memory 6, and the value of the variance σ _PD indicated by the solid line of the waveform (G2) is obtained by the variance value calculation circuit 63. Similarly, with the optical switch 11 closed, this signal waveform S15 is sampled by the sampling circuit 5, stored in the memory 6, and the variance σ _DK indicated by the dotted line of the waveform (G2) is obtained by the variance value calculation circuit 63. obtain.

In this waveform (G2), a standard deviation graph obtained from dark count noise is indicated by a dotted line. At this time, the maximum value σ _DK of the standard deviation of dark count noise is expected to be larger than the standard deviation σ _GP of fluctuation by the gate and smaller than the maximum value σ _PD of the standard deviation of the signal generated from the photon detection. By switching the optical switch 11, σ _PD and σ _DK are measured and stored in the memory 98, and σ _{DK is} converted from σ _DK to σ of _Formula 5 according to the method described in the second embodiment. Can be substituted to determine the threshold value V _th .

  Although dark count noise is generated only in the gate, as shown in Non-Patent Document 5, it has been clarified that it shows a distribution of amplitude different from the signal waveform in photon detection. In particular, the dark count tends to appear smaller in amplitude than the photon detection signal. This embodiment uses this phenomenon to efficiently remove dark count noise and reduce the error rate of photon detection.

  The photon detection phenomenon and the occurrence of dark count noise follow a Poisson distribution, but the amplitude of the output signal is considered to follow the distribution described in FIGS. . For this reason, the relationship illustrated in FIGS. 7 to 8 of Patent Document 1 may be used to determine the threshold value which is the object of the present embodiment.

  FIG. 7 is a circuit block diagram showing a configuration of a photon detection circuit 100d according to the fourth embodiment of the present invention. Since the configuration of the photon detection circuit 100d is mostly the same as that of the photon detection circuits 100a to 100c according to the first to third embodiments, common constituent elements are denoted by the same reference numerals and description thereof is omitted. To do.

  In the photon detection circuit 100d, a maximum value detection circuit 64 is attached to the memory 6 of the photon detection circuit 100b according to the second embodiment, in addition to the average waveform generation circuit 62 and the variance value calculation circuit 63. . Similarly to the average waveform generation circuit 62 and the variance value calculation circuit 63, the maximum value detection circuit 64 detects the maximum value for the same phase in the gate period with respect to the time waveform recorded in the memory for each sampling point. The maximum value data S64 is output to the memory 91. That is, instead of using the variance of the sampled signal, the optimum gate phase is detected by detecting the phase for detecting the maximum amplitude from among those stored in the memory 6.

  In the first to fourth embodiments described above, the dispersion value is used to optimize the photon detection phase and the identification threshold. However, instead of the dispersion value, a standard deviation, a maximum value, an amplitude, etc. It is also possible to perform the same processing using the statistical value.

  Although the present invention has been described with the specific embodiments shown in the drawings, the present invention is not limited to the embodiments shown in the drawings, and is known so far as long as the effects of the present invention are achieved. It goes without saying that any configuration can be adopted.

  The present invention can be applied to devices using a single photon detection circuit, such as a quantum key distribution device, a quantum cryptography device, a photon number detector, an OTDR, a dark field camera, and the like.

1 is a circuit block diagram showing a configuration of a photon detection circuit according to a first embodiment of the present invention. It is a schematic diagram of the calculation result in a certain gate phase in the photon detection circuit disclosed in FIG. It is a circuit block diagram which shows the structure of the photon detection circuit which concerns on the 2nd Embodiment of this invention. It is a schematic diagram of the calculation result of the threshold value in the photon detection circuit disclosed in FIG. It is a circuit block diagram which shows the structure of the photon detection circuit which concerns on the 3rd Embodiment of this invention. FIG. 6 is a schematic diagram of a threshold calculation result in the photon detection circuit disclosed in FIG. 5. It is a circuit block diagram which shows the structure of the photon detection circuit which concerns on the 4th Embodiment of this invention. It is a circuit block diagram which shows the structure of the conventional photon detection circuit described in the nonpatent literatures 5-6. It is a conceptual diagram explaining the operation | movement from average waveform generation to identification in the photon detection circuit disclosed in FIG. FIG. 9 is a conceptual diagram illustrating a method for determining an optimum phase and threshold when identifying photons in the identification circuit disclosed in FIG. 8. It is a continuation of FIG. FIG. 9 is a circuit block diagram showing a configuration of a photon detection circuit obtained by adding a configuration for further scanning a threshold to the photon detection circuit disclosed in FIG. 8.

Explanation of symbols

1 APD
DESCRIPTION OF SYMBOLS 11 Optical switch 2 Gate generation circuit 22 Delay circuit 23 Gate phase control circuit 3 Clock source 31 Clock processing circuit 5 Sampling circuit 6 Memory 61 Memory control circuit 62 Average waveform generation circuit 63 Dispersion value calculation circuit 64 Maximum value detection circuit 9 Identification circuit 91 Memory 93 Maximum value detection circuit 94 Memory control circuit 96, 95 Threshold control circuit 96a Counter 97 Maximum value selection circuit 98 Memory 99 Optical switch control circuit 100a, 100b, 100c, 100d, 300a, 300b Photon detection circuit S1 Photon signal S15 Detection signal S21 Gate pulse S23 Gate phase control signal S3 Clock S32 Gate signal S35 Sampling clock S5 Sampling waveform S6 Average waveform S63 Dispersion value data S64 Maximum value data S9 Photon detection signal S91 Value control signal

Claims (16)

It is a single photon detection device by gate mode using an avalanche photodiode,
And the gate voltage generation circuit for generating a gate voltage based on the gate period, provided with a gate phase control circuit for changing the phase of the gate voltage,
A sampling circuit that samples the signal output from the avalanche photodiode and converts it into discrete data based on digital information, a first memory circuit that stores the discrete data for each gate period, and the discrete circuit for each gate period and a statistical value calculating circuit for calculating a statistical value of the data,
Further, a second memory circuit for storing the statistical value for each changed gate phase, a maximum value of the statistical value stored in the second memory circuit, and a maximum for detecting the phase of the gate voltage at that time A statistical value detection circuit; an average waveform data generation circuit that generates an average waveform of APD output waveform data using the discrete data stored in the first memory circuit; and the average waveform data for each gate period wherein by comparing the signal outputted from the avalanche photodiode and possess an identification circuit that performs identification of the photon detection signal,
The gate phase control circuit changes the phase of the gate voltage so that the statistical value detected by the maximum statistical value detection circuit becomes a maximum value,
In the statistical value calculation circuit, a detection error rate due to a noise component is calculated from a standard deviation value in each sampling phase in the statistical value, and the detection error rate is satisfied in the identification circuit so as to satisfy a required level. A photon detection apparatus, further comprising a threshold control circuit that calculates a threshold value that serves as a reference for identifying the photon detection signal .   The photon detection device according to claim 1, wherein the statistical value includes one or more of a dispersion value, a standard deviation, a maximum value, and an amplitude. 3. The photon detection device according to claim 2 , wherein the maximum statistical value detection circuit obtains the maximum value in the entire gate phase from the maximum value of the amplitude stored in the second memory circuit. The photon detection device according to claim 1, wherein a maximum value selection circuit that detects a maximum value of the statistical value of the discrete data for each of the gate periods is provided in the statistical value calculation circuit. 2. The photon detection device according to claim 1 , wherein the threshold control circuit calculates a detection error rate due to the noise component from the statistical value in a phase range in which a photon signal detection signal does not appear. 2. The photon detection device according to claim 1 , wherein the threshold control circuit calculates the detection error rate based on a mathematical expression assuming that the noise component is a normal distribution. An optical switch for opening and closing a photon signal incident on the avalanche photodiode;
The photon detection device according to claim 1 , further comprising a third memory circuit that stores the statistical value corresponding to opening / closing of the optical switch. The threshold control circuit calculates a standard deviation in a phase range in which a dark count noise signal is detected from the statistical value in a state where the optical switch is closed, and from the standard deviation in a phase range in which the dark count noise signal is detected. 8. The photon detection device according to claim 7 , wherein a detection error rate is calculated, and the threshold is determined so that the detection error rate satisfies a required level. A method for detecting photons in a single-photon detection device in a gate mode using an avalanche photodiode,
A gate voltage generation step for generating a gate voltage based on the gate period;
Applying a gate voltage to the avalanche photodiode;
A sampling step of sampling a signal output from the avalanche photodiode and converting it into discrete data by digital information;
A first storage step of storing the discrete data for each gate period;
A statistical value calculation step of calculating a statistical value of the discrete data for each gate period;
A second storage step of storing the statistical value for each phase of the gate voltage ;
A maximum statistical value detection step of detecting a maximum value of the statistical value stored in the second storage step and a phase of the gate voltage at that time;
A gate phase control step of changing the phase of the gate voltage so that the statistical value detected by the maximum statistical value detection step becomes a maximum value;
An average waveform data generation step of generating an average waveform of APD output waveform data using the discrete data stored in the first storage step ;
A threshold control step of calculating a detection error rate due to a noise component from a standard deviation value at each sampling phase in the statistical value, and calculating a threshold so that the detection error rate satisfies a required level;
A photon detection method comprising: an identification step of identifying the photon detection signal based on the threshold value by comparing the average waveform data and a signal output from the avalanche photodiode for each gate period. The photon detection method according to claim 9 , wherein the statistical value includes one or more of a dispersion value, a standard deviation, a maximum value, and an amplitude. 10. The photon detection method according to claim 9 , wherein the maximum statistical value detection step obtains the maximum value in the entire gate phase from the maximum amplitude value stored in the second storage step. 10. The photon detection method according to claim 9 , wherein the threshold control step calculates the detection error rate based on an equation assuming that the noise component is a normal distribution. In the computer that the single photon detection device by the gate mode using the avalanche photodiode has,
Sampling processing for sampling a signal output from the avalanche photodiode to which a gate voltage based on a gate period is applied and converting it into discrete data by digital information;
A first storage process for storing the discrete data for each gate period;
Statistical value calculation processing for calculating the statistical value of the discrete data for each gate period;
A second storage process for storing the statistical value for each phase of the gate voltage ;
A maximum statistical value detection process for detecting the maximum value of the statistical value stored in the second storage process and the phase of the gate voltage at that time;
A gate phase control process for changing the phase of the gate voltage so that the statistical value detected by the maximum statistical value detection process becomes a maximum value;
An average waveform data generation process for generating an average waveform of APD output waveform data using the discrete data stored in the first storage process ;
A detection error rate due to a noise component is calculated from a standard deviation value at each sampling phase in the statistical value, and the average waveform data and the average waveform data are calculated for each gate period so that the detection error rate satisfies a required level. A threshold value control step of calculating a threshold value which is a reference for identification when comparing a signal output from an avalanche photodiode and identifying a photon detection signal. . The photon detection program according to claim 13 , wherein the statistical value includes one or more of a dispersion value, a standard deviation, a maximum value, and an amplitude. The photon detection program according to claim 13 , wherein the maximum statistical value detection process obtains the maximum value in the entire gate phase from the maximum value of the amplitude stored in the second storage process. 14. The photon detection program according to claim 13 , wherein the threshold control process calculates the detection error rate based on a mathematical formula assuming that the noise component is a normal distribution.

Images