Time synchronization device

Abstract

To execute statistical time synchronization processing with high accuracy.SOLUTION: An information processing part 12 sequentially executes error measurement processing to acquire a timing error and a frequency error about a reference pulse signal and a local pulse signal with the lapse of time. The information processing part 12 calculates a timing control value and a frequency control value to the local pulse signal by statistical processing with a plurality of timing errors by error measurement processing executed over a plurality of times and a plurality of frequency errors by the error measurement processing executed over the plurality of times as a parent population. The information processing part 12 determines the size of the parent population in the statistical processing on the basis of the accuracy of the reference pulse signal or an operation state of the source of the local pulse signal.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a time synchronization device, and more particularly to a device that receives a reference pulse signal and performs time synchronization based on the reference pulse signal. [Background technology]

[0002] Wireless communication systems are widely used, which include multiple base stations and multiple wireless communication terminals, with each wireless communication terminal communicating with one of the base stations. Some wireless communication systems have multiple wireless communication terminals that each communicate with one of the base stations in a time-division manner. Each wireless communication terminal changes the base station with which it communicates as it moves. To smoothly change base stations, time synchronization is performed between the base stations.

[0003] A base station that performs time synchronization receives a signal from a communication system such as a satellite positioning system and extracts a reference signal synchronized with time (such as Coordinated Universal Time) from the received signal. For example, the base station described in Cited Document 1 extracts a 1PPS (1 Pulse Per Second) signal from a signal received from a GPS (Global Positioning System) satellite (GPS satellite) as a reference signal synchronized with time, and performs time synchronization processing using the 1PPS signal. Here, the 1PPS signal is a signal synchronized with Coordinated Universal Time with a period of 1 second. The base station, for example, generates a clock signal synchronized with the 1PPS signal and operates at a timing according to the clock signal. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2010-88071 [Patent Document 2] Japanese Patent Publication No. 2022-167388 [Non-patent literature]

[0005] [Non-Patent Document 1] Toyoizumi, Genda, "Precise 1PPS Signal Output Using GPS," Transactions on Electrical Engineering, Vol. 125, No. 8, 2005, pp. 1217-1222. Summary of the Invention [Problem to be solved by the invention]

[0006] Non-Patent Document 1 describes errors in the period of a 1PPS signal. It is generally known that 1PPS signals have period fluctuations. It is also thought that reference signals obtained from satellite positioning systems other than GPS may also have period fluctuations. In recent years, the operation of each device constituting a wireless communication system has become faster, and if the period (frequency) of the reference signal fluctuates, the performance of the wireless communication system may not be fully demonstrated.

[0007] Therefore, Patent Document 2 proposes a technique in which the timing error between a reference signal and a clock signal is acquired multiple times in the past, and the frequency or phase of the clock signal is controlled according to the statistical value of the timing error acquired multiple times. However, in such statistical synchronization processing, the accuracy of the synchronization processing may be insufficient depending on how the population is selected when calculating the statistical value.

[0008] An object of the present invention is to perform statistical time synchronization processing with high accuracy. [Means for solving the problem]

[0009] The present invention is characterized in that it comprises a receiving unit that extracts a reference pulse signal from a received signal obtained by receiving a radio signal, a local clock that generates a local pulse signal, and an information processing unit that controls the local clock, wherein the information processing unit sequentially executes error measurement processes over time to obtain a timing error based on a difference between the pulse timing indicated by the reference pulse signal and the pulse timing indicated by the local pulse signal, and a frequency error based on a difference between the timing error obtained for each earlier pulse timing and the timing error obtained for each later pulse timing, determines a timing control value and a frequency control value by statistical processing using a population of a plurality of the timing errors obtained by the error measurement processes executed a plurality of times and a plurality of the frequency errors obtained by the error measurement processes executed a plurality of times, and executes an adjustment process that performs at least one of a process of adjusting the pulse timing of the local pulse signal using the timing control value and a process of adjusting the frequency of the local pulse signal using the frequency control value on the local pulse signal, and determines a population size for the statistical processing based on the accuracy of the reference pulse signal.

[0010] Preferably, a temperature sensor, wherein the information processing unit determines a temperature gradient based on a temperature indicated by the temperature sensor, and the smaller the temperature gradient, the more times the statistical processing is performed based on the error measurement processing that has been performed; The size of the population in the statistical processing is determined based on the accuracy of the reference pulse signal as well as the operating status of the source of the local pulse signal.

[0011] The present invention provides a radio communication system including: a receiving unit that extracts a reference pulse signal from a received signal obtained by receiving a radio signal; a local clock that generates a local pulse signal; and an information processing unit that controls the local clock. A temperature sensor;the information processing unit sequentially executes error measurement processes over time to determine a timing error based on a difference between the pulse timing indicated by the reference pulse signal and the pulse timing indicated by the local pulse signal, and a frequency error based on a difference between the timing error determined for each earlier pulse timing and the timing error determined for each later pulse timing, and determines a timing control value and a frequency control value by statistical processing using a population of a plurality of the timing errors obtained by the error measurement processes executed a plurality of times and a plurality of the frequency errors obtained by the error measurement processes executed a plurality of times, and performs an adjustment process on the local pulse signal by performing at least one of a process of adjusting the pulse timing of the local pulse signal using the timing control value and a process of adjusting the frequency of the local pulse signal using the frequency control value, a temperature gradient is calculated based on the temperature indicated by the temperature sensor, and the smaller the temperature gradient is, the more times the statistical processing is performed based on the error measurement processing that has been performed; The size of the population in the statistical processing is determined based on the operating status of the source of the local pulse signal.

[0012] Preferably, the radio signal is a signal transmitted from an artificial satellite, and the receiving unit Quality information indicating the accuracy of the reference pulse signal is generated based on the received signal, and the information processing unit performs the statistical processing based on the error measurement processing that has been performed more times the less accurate the accuracy indicated by the quality information is.

[0014] Preferably, the information processing unit performs the statistical processing by applying a correction process based on the timing control value and frequency control value in the latest adjustment process to a plurality of timing errors resulting from the error measurement process executed a plurality of times in the past and a plurality of frequency errors resulting from the error measurement process executed a plurality of times in the past. [Effects of the Invention]

[0015] According to the present invention, statistical time synchronization processing can be performed with high accuracy. [Brief explanation of the drawings]

[0016] [Figure 1] FIG. 1 is a diagram illustrating a configuration of a time synchronization device. [Figure 2] 3A and 3B are diagrams schematically illustrating examples of the time waveform of a 1PPS signal and the time waveform of a local pulse signal. [Figure 3] FIG. 10 is a diagram conceptually illustrating a process for determining the latest timing control value and frequency control value. [Figure 4] FIG. 10 is a diagram illustrating an example of the relationship between the passage of time and timing error. [Figure 5] 10A and 10B are diagrams illustrating an example of frequency fluctuations over time for a time signal and a local pulse signal generated by a local clock. [Figure 6] 10A and 10B are diagrams illustrating an example of frequency fluctuations when the ambient temperature of the local clock changes. [Figure 7] FIG. 10 is a diagram conceptually illustrating an example of a population determination map. [Figure 8] 10A and 10B are diagrams showing specific examples of population determination maps. DETAILED DESCRIPTION OF THE INVENTION

[0017] FIG. 1 shows the configuration of a time synchronization device 100 according to an embodiment of the present invention. The time synchronization device 100 is installed in, for example, a base station of a wireless communication system, and outputs a clock signal that defines the operation of the base station. The time synchronization device 100 includes a receiving unit 10, an information processing unit 12, a local clock 14, a time signal output unit 16, a memory 18, and a temperature sensor 20. The information processing unit 12 includes a timing comparison unit 22, a control unit 24, and a temperature gradient acquisition unit 26. All or part of the components of the information processing unit 12 (the timing comparison unit 22, the control unit 24, and the temperature gradient acquisition unit 26) may be configured by one or more processors that execute programs to perform arithmetic processing.

[0018] The receiving unit 10 receives a radio signal transmitted from a GPS satellite, which is an artificial satellite. The receiving unit 10 extracts a 1PPS signal from the received signal obtained by receiving the radio signal, and outputs the 1PPS signal to the timing comparison unit 22. The 1PPS signal is a reference pulse signal that includes a pulse that rises every second.

[0019] The local clock 14 generates a time signal that defines timing by pulses rising at equal time intervals, and outputs the time signal to the time signal output unit 16. The time signal output unit 16 outputs a clock signal synchronized with the timing indicated by the time signal. The time signal output unit 16 may also generate a timing signal synchronized with the timing indicated by the time signal. The timing signal is a signal that defines the operation timing of the device in which it is installed. The local clock 14 generates a local pulse signal that includes a pulse that rises every second based on the time signal, and outputs the local pulse signal to the timing comparison unit 22.

[0020] The timing comparison unit 22 calculates a timing error t by subtracting the pulse timing of the local pulse signal from the pulse timing of the 1PPS signal. err is sequentially calculated over time. Here, the pulse timing in this embodiment refers to the time at which the signal rises. When the pulse timing of the local pulse signal is delayed relative to the pulse timing of the 1PPS signal, the timing error is negative, and when the pulse timing of the local pulse signal is advanced relative to the pulse timing of the 1PPS signal, the timing error is positive.

[0021] 2A and 2B are schematic diagrams showing examples of the time waveform of the 1PPS signal and the time waveform of the local pulse signal. In the example shown in FIG. 2, the local pulse signal is delayed from the 1PPS signal. The timing comparator 22 detects a timing error t err (j-1), where j is an integer equal to or greater than 1. The timing comparator 22 calculates a timing error t err Find (j).

[0022] The timing comparator 22 detects a frequency error f err (j) is calculated. Frequency error f err (j) is the timing error t when the j-th pulse of the local pulse signal and the 1PPS signal rises err From (j), the timing error t err (j-1) is subtracted from the err (j)=t err (j)-t err (j-1). The frequency error f err (j) becomes positive when the frequency of the local pulse signal increases, and becomes negative when the frequency of the local pulse signal decreases.

[0023] The timing comparator 22 in FIG. 1 sequentially calculates the timing error t err (j) and frequency error f err (j) is the TF error TF(j)=[t err (j),f err (j)] are calculated and output to the control unit 24 sequentially as time passes.

[0024] In this way, the timing comparison unit 22 compares the timing error based on the difference between the pulse timing indicated by the 1PPS signal as the reference pulse signal and the pulse timing indicated by the local pulse signal with the timing error t err (j-1) and the timing error t obtained for each subsequent pulse timing err Frequency error f based on the difference with (j) err The error measurement process for obtaining (j) is executed sequentially over time.

[0025] The control unit 24 stores the TF errors TF(j) output from the timing comparison unit 22 in the memory 18 sequentially over time. Each TF error TF(j) stored in the memory 18 is corrected by the control unit 24 when performing statistical processing, which will be described later. This correction is performed by an overwriting process in which the value before correction is replaced with the value after correction.

[0026] The control unit 24 calculates the timing control value t by statistical processing of the latest TF error TF(i) and n TF errors TF(in) to TF(i-1) calculated before the latest TF error TF(i). c (i) and frequency control value f c (i) is calculated. Here, the timing control value t c (i) is a value indicating how much the pulse timing of the local pulse signal should be adjusted, and is the frequency control value f c (i) is a value indicating how much the frequency of the local pulse signal, that is, the pulse timing interval, should be adjusted. The statistical processing will be explained below.

[0027] The control unit 24 calculates the latest frequency error f err (i) The frequency control value f c (i) is calculated according to the following equation (1), and the latest timing error t err (i) Timing control value t c (i) is calculated according to (Equation 2).

[0028]

number

[0029]

number

[0030] Here, the corrected timing error t err_c (k) (k=in to i) is calculated according to (Equation 3).

[0031]

number

[0032] The frequency correction value fad(h) and the timing correction value tad(h) in (Equation 3) are calculated according to (Equation 4) and (Equation 5), respectively.

[0033]

number

[0034]

number

[0035] The control unit 24 controls the local clock 14 so that the frequency of the local pulse signal output by the local clock 14 is adjusted to a frequency control value f c (i), and the rising edge of each pulse included in the local pulse signal is lowered by the timing control value t c The pulse timing of the time signal with which the local pulse signal is synchronized is adjusted so that it is delayed by (i).

[0036] Here, the frequency control value f obtained by (Equation 1) c (i) indicates the variation width of the timing error, that is, the average value of the frequency error, when it is assumed that the timing error changes in equal increments as the time variable changes over n stages from in to i. The timing control value t obtained by (Equation 2) c (i) indicates the average value of the timing error from when the time variable changes from i-n+1 to i.

[0037] Furthermore, (Equation 3) to (Equation 5) are the corrected timing error t err_c (k) is the original timing error t err (k) shows that the frequency control value and timing control value obtained at each past time step are cumulatively subtracted.

[0038] Figure 3 shows the latest timing control value t c (i) and frequency control value f c The process of calculating (i) according to (Equation 1) to (Equation 5) is conceptually shown. One rectangle shown in FIG. 3 represents one TF error TF(j)=[t err (j),f err (j)]. The rightmost rectangle on the time axis represents the latest TF error TF(i) = [t err (i),f err (i)].

[0039] In the first step S0, the control unit 24 performs statistical processing on the TF errors TF(-n) to TF(0) to obtain the latest timing control value t c (0) and frequency control value f c The control unit 24 calculates the timing error t(0) constituting the latest TF error TF(0). err (0) and frequency error f err (0) to the timing control value t c (0) and frequency control value f c (0) is subtracted to obtain the corrected TF error TF(0) (correction process SS0).

[0040] The control unit 24 executes the correction process SS0 and also executes an adjustment process for the local clock 14. That is, the control unit 24 controls the local clock 14 so that the frequency of the local pulse signal output by the local clock 14 is adjusted to a frequency control value f c (0), and the pulse timing of the local pulse signal is decreased by the timing control value t c The pulse timing of the time signal with which the local pulse signal is synchronized is adjusted so that it is delayed by (0).

[0041] In step S1, the control unit 24 calculates the timing errors t included in the TF errors TF(-n+1) to TF(0). err (-n+1)~t err (0) to the timing control value t cThe control unit 24 further subtracts the frequency error f (0) included in each of the TF errors TF(-n+1) to TF(0). err (-n+1)~f err (0) to obtain the frequency control value f c (0) is subtracted from the TF error TF(−n+1) to TF(0) after correction (correction process ST0).

[0042] In step S1, the control unit 24 performs statistical processing on the TF errors TF(-n+1) to TF(1) to calculate the latest timing control value t c (1) and frequency control value f c The control unit 24 calculates the timing error t (1) that constitutes the latest TF error TF(1). err (1) and frequency error f err From (1), the timing control value t c (1) and frequency control value f c (1) is subtracted to obtain the corrected TF error TF(1) (correction process SS1).

[0043] The control unit 24 executes the correction process SS1 and also executes an adjustment process for the local clock 14. That is, the control unit 24 controls the local clock 14 so that the frequency of the local pulse signal output by the local clock 14 is adjusted to a frequency control value f c (1), and the pulse timing of the local pulse signal is decreased by the timing control value t c The pulse timing of the time signal with which the local pulse signal is synchronized is adjusted so that it is delayed by (1).

[0044] In step S2, the control unit 24 calculates the timing errors t err (-n+2)~t err From each of (1), the timing control value t c The control unit 24 further subtracts the frequency error f (1) included in each of the TF errors TF(−n+2) to TF(1). err (-n+2)~f err (1) from each of the frequency control values ​​fc (1) is subtracted from the TF error TF(−n+2) to TF(1) after correction (correction process ST1).

[0045] In step S2, the control unit 24 performs statistical processing on the TF errors TF(-n+2) to TF(2) to obtain the latest timing control value t c (2) and frequency control value f c The control unit 24 calculates the timing error t (2) that constitutes the latest TF error TF(2). err (2) and frequency error f err (2), respectively, the timing control value t c (2) and frequency control value f c (2) is subtracted to obtain the corrected TF error TF(2) (correction process SS2).

[0046] The control unit 24 executes the correction process SS2 and also executes an adjustment process for the local clock 14. That is, the control unit 24 controls the local clock 14 so that the frequency of the local pulse signal output by the local clock 14 is adjusted to the frequency control value f c (2) and the pulse timing of the local pulse signal is decreased by the timing control value t c The pulse timing of the time signal with which the local pulse signal is synchronized is adjusted so that it is delayed by (2).

[0047] In this way, in step Sj, the control unit 24 calculates the timing errors t err (-n+j)~t err (j-1), the timing control value t c The control unit 24 further subtracts the frequency error f(j-1) included in each of the TF errors TF(-n+j) to TF(j-1). err (-n+j)~f err (j-1), the frequency control value f c (j-1) is subtracted from the TF error TF(-n+j) to obtain the corrected TF errors TF(-n+j) to TF(j-1) (correction process SSj).

[0048] In step Sj, the control unit 24 performs statistical processing on the TF errors TF(-n+j) to TF(j) to obtain the latest timing control value t c (j) and frequency control value f c The control unit 24 calculates the timing error t(j) constituting the latest TF error TF(j). err (j) and frequency error f err (j) respectively, the timing control value t c (j) and frequency control value f c (j) is subtracted to obtain the corrected TF error TF(j) (correction process SSj).

[0049] The control unit 24 executes the correction process SSj and also executes an adjustment process for the local clock 14. That is, the control unit 24 controls the local clock 14 so that the frequency of the local pulse signal output by the local clock 14 is adjusted to a frequency control value f c (j), and the pulse of the local pulse signal is decreased by the timing control value t c The pulse timing of the time signal with which the local pulse signal is synchronized is adjusted so that it is delayed by (j).

[0050] In step Sj where the TF error TF(j) is generated, each of the TF errors (-n+j) to (j-1) before the TF error TF(j) is calculated based on the timing control value t c (j-1) and frequency control value f c (j-1) (correction process STj-1).

[0051] Then, by statistical processing based on the TF error (-n+j) to TF error (j), the latest timing control value t corresponding to the TF error TF(j) is calculated. c (j) and frequency control value f c (j) is obtained (correction process SSj). Furthermore, the TF error TF(j) is calculated by the timing control value t c (j) and frequency control value f c As amended by (j).

[0052] In this way, the control unit 24 performs statistical processing by applying correction processing based on the timing control value and frequency control value (timing error and frequency error) in the latest adjustment processing to multiple timing errors resulting from error measurement processing executed multiple times in the past and multiple frequency errors resulting from error measurement processing executed multiple times in the past.

[0053] If statistical processing were performed without performing the correction process STj-1, the statistical processing would be performed on a population containing a mixture of TF errors corrected using different timing control values ​​and different frequency control values. As a result, the latest timing control value and frequency control value would not be appropriate. According to the process conceptually shown in Figure 3, statistical processing is performed on a population of TF errors corrected under the same conditions, and the latest timing control value and frequency control value are appropriately determined.

[0054] As mentioned above, the timing control value t c (i) and frequency control value f c (i) is obtained by statistical processing using a population of multiple timing errors obtained by error measurement processes executed multiple times (n+1 times going back in time including the most recent process) and multiple frequency errors obtained by error measurement processes executed multiple times (n+1 times going back in time including the most recent process). More specifically, the timing control value t corresponding to the latest TF error TF(i) is c (i) and frequency control value f c (i) is obtained by statistical processing of a population of n+1 TF errors, including the latest TF error TF(i) and n TF errors acquired and corrected before the TF error TF(i).

[0055] In the time synchronization device 100 according to this embodiment, the population size for statistical processing is determined based on at least one of the accuracy of the 1PPS signal and the operating conditions of the local clock 14, which is the source of the local pulse signal. Specifically, the number n+1 of TF errors used for statistical processing may be determined based on one or both of quality information indicating the accuracy of the 1PPS signal and the ambient temperature of the local clock 14. The quality information is generated by the receiving unit 10 based on, for example, positioning signals (radio signals) transmitted from one or more GPS satellites. In the following description, the number n+1 of TF errors used for statistical processing may be referred to as the population size N (N=n+1) for statistical processing.

[0056] As shown in (Equation 1), the frequency error (frequency control value) is found by subtracting the timing error going back n times from the current time from the most recent timing error and dividing the result into n equal parts. As shown in (Equation 2), the timing error (timing control value) is found by dividing the integral value by n, assuming that the timing error goes back n times from the current time and increases linearly with a constant slope over n times. This constant slope is the frequency error found according to (Equation 1). However, the actual timing error may not increase linearly from the timing error going back n times from the current time.

[0057] FIG. 4 shows an example of the relationship between the passage of time and timing error. The horizontal axis represents discrete time, and the vertical axis represents timing error. Characteristic 40 represents timing error that increases linearly with a constant slope over n periods, starting from the timing error going back n periods from the present time. Characteristic 42 represents timing error whose slope changes over time. If the frequency error were calculated by integrating characteristic 40 over time according to (Equation 1) and (Equation 2) to find the average value, an error would occur between the frequency error calculated from the slope of characteristic 42. This error increases the longer the period for which the frequency error is calculated, i.e., the larger the population size N.

[0058] In general, the larger the population size N of the statistical processing, the more appropriate it is to consider the timing error to change in a curved line over time, and the less appropriate the linear approximation shown in (Equation 2) is. Also, the greater the change in the temperature around the local clock 14, the greater the change in the time derivative of the timing error, and the more appropriate it is to consider the timing error to change in a curved line over time. Therefore, the greater the change in the temperature around the local clock 14, the less appropriate the linear approximation shown in (Equation 2) is.

[0059] Figure 5 shows an example of frequency fluctuation over time for the time signal and local pulse signal generated by the local clock 14. The horizontal axis represents time [sec], and the vertical axis represents frequency fluctuation [ppb]. However, the ambient temperature of the local clock 14 is constant. In the example shown in Figure 5, a frequency fluctuation of 0.02083 ppb is observed over 3600 seconds.

[0060] Fig. 6 shows frequency fluctuations when the ambient temperature of the local clock 14 changes. The horizontal axis represents the temperature gradient [°C / time], and the vertical axis represents the difference between the detected frequency error value and the true frequency error. In the example shown in Fig. 6, the greater the temperature gradient, the greater the detected frequency error value becomes relative to the true frequency.

[0061] To suppress the effects of such frequency error fluctuations, the control unit 24 reduces N as the temperature change per unit time (temperature gradient) around the local clock 14 increases. Such a time synchronization device 100 will be described again with reference to FIG. 1. The temperature sensor 20 sequentially measures the temperature around the local clock 14 over time and outputs the measurement values ​​to the temperature gradient acquisition unit 26 sequentially over time. The temperature sensor 20 may be installed around the local clock 14. The temperature gradient acquisition unit 26 calculates the temperature gradient [°C / time] based on the measurement values ​​sequentially output by the temperature sensor 20 over time, and outputs this to the control unit 24. The control unit 24 reduces the population size N as the temperature gradient increases, and increases the population size N as the temperature gradient decreases.

[0062] As mentioned above, it is known that 1PPS signals have periodic fluctuations. If the TF errors corresponding to the N pulses of a 1PPS signal are considered as a population, the distribution of the timing error on the horizontal axis and the frequency of occurrence on the vertical axis will be a normal distribution, with the timing error with the highest frequency being the average. The larger the population size N, the more likely it is that the timing error with the highest frequency, i.e., the average value of the timing errors, will be the true timing error.

[0063] Therefore, the receiving unit 10 generates quality information indicating the positioning accuracy based on the positioning signals transmitted from the GPS satellites. The quality information may be, for example, information indicating the level of the received positioning signals. The quality information may also be DOP (Dilution of Precision). DOP is a numerical value indicating the arrangement status of multiple GPS satellites. The smaller the DOP, the better the arrangement status and the higher the positioning accuracy. Furthermore, if the receiving unit 10 is placed in an environment where the reception status of the positioning signals transmitted from each GPS satellite is poor, the DOP will be relatively large, and the positioning accuracy and the accuracy of the 1PPS signal will be low.

[0064] The time synchronization device 100 may perform the following processing. The receiving unit 10 generates quality information from received signals based on positioning signals transmitted from GPS satellites, and outputs the quality information to the control unit 24. The control unit 24 increases the population size N as the positioning accuracy indicated by the quality information decreases, and decreases the population size N as the positioning accuracy indicated by the quality information increases. The receiving unit 10 generates a DOP as quality information from each received signal based on positioning signals transmitted from multiple GPS satellites, for example. The smaller the DOP, the higher the positioning accuracy and the smaller the timing error of the 1PPS signal. The larger the DOP, the lower the positioning accuracy and the larger the timing error of the 1PPS signal.

[0065] FIG. 7 conceptually illustrates an example of a population determination map for determining the population size N. The horizontal axis represents the DOP, and the vertical axis represents the temperature gradient obtained based on measurements by the temperature sensor 20. The population determination map is a map in which the population size N is associated with a pair of DOP and temperature gradient. The population determination map may be stored in the control unit 24 or the memory 18. When the population determination map is stored in the memory 18, the control unit 24 refers to the population determination map stored in the memory 18. The control unit 24 obtains the population size N associated with the DOP and temperature gradient, and performs statistical processing according to the population size N.

[0066] An example of a population decision map is shown in Figure 8. In the population decision map, a DOP less than 3 is identified as a small DOP, a DOP between 2 and 6 is identified as a medium DOP, and a DOP greater than or equal to 6 is identified as a large DOP.

[0067] In addition, in the population determination map, if the temperature gradient is less than 2°C / hour, the temperature gradient is identified as small. If the temperature gradient is 2°C / hour or more but less than 8°C / hour, the temperature gradient is identified as medium. If the temperature gradient is 8°C / hour or more, the temperature gradient is identified as large. For each of the three cases of small, medium, and large DOP, there are three cases of small, medium, and large temperature gradients, for a total of nine types of population size N.

[0068] The control unit 24 refers to the population determination map, acquires the population size N corresponding to the temperature gradient output from the temperature gradient acquisition unit 26 and the DOP output from the receiving unit 10, and performs statistical processing according to the population size N.

[0069] According to the population determination map, the lower the accuracy indicated by the DOP as quality information, the more times statistical processing based on error measurement processing is performed. Also, according to the population determination map, the smaller the temperature gradient, the more times statistical processing based on error measurement processing is performed.

[0070] According to the time synchronization device 100 of this embodiment, an appropriate population size N for statistical processing is determined based on the accuracy of the pulse timing of the 1PPS signal transmitted from a GPS satellite and the temperature characteristics of the local clock 14. This allows for highly accurate time synchronization processing to be performed, synchronizing the time signal and local pulse signal generated by the local clock 14 with the 1PPS signal.

[0071] In the above embodiment, an adjustment process is performed on both the timing error and the frequency error of the local pulse signal. The adjustment process may be performed on at least one of the timing error and the frequency error. In this case, the correction process performed on the multiple timing errors resulting from the error measurement process performed multiple times in the past and the multiple frequency errors resulting from the error measurement process performed multiple times in the past is performed in response to one or both of the timing errors and the frequency errors for which the adjustment process has been performed.

[0072] Although the embodiment described above uses GPS as the satellite positioning system, other systems such as GLONASS (Global Navigation Satellite System) may also be used as the satellite positioning system.

[0073] [Configuration of the present invention] Configuration 1: A receiving unit that extracts a reference pulse signal from a received signal obtained by receiving a wireless signal; a local clock that generates a local pulse signal; an information processing unit that controls the local clock, The information processing unit a timing error based on a difference between the pulse timing indicated by the reference pulse signal and the pulse timing indicated by the local pulse signal; Sequentially executing an error measurement process over time to determine a frequency error based on a difference between the timing error determined for each of the earlier pulse timings and the timing error determined for each of the later pulse timings; determining a timing control value and a frequency control value by statistical processing using a population of a plurality of timing errors obtained by the error measurement process executed a plurality of times and a plurality of frequency errors obtained by the error measurement process executed a plurality of times; an adjustment process that performs at least one of a process of adjusting a pulse timing of the local pulse signal using the timing control value and a process of adjusting a frequency of the local pulse signal using the frequency control value on the local pulse signal; A time synchronization device, characterized in that the size of the population in the statistical processing is determined based on the accuracy of the reference pulse signal. Configuration 2: The time synchronization device according to configuration 1, A time synchronization device characterized in that the size of the population in the statistical processing is determined based on the accuracy of the reference pulse signal as well as the operating status of the source of the local pulse signal. Configuration 3: a receiving unit that extracts a reference pulse signal from a received signal obtained by receiving a wireless signal; a local clock that generates a local pulse signal; an information processing unit that controls the local clock, The information processing unit a timing error based on a difference between the pulse timing indicated by the reference pulse signal and the pulse timing indicated by the local pulse signal; Sequentially executing an error measurement process over time to determine a frequency error based on a difference between the timing error determined for each of the earlier pulse timings and the timing error determined for each of the later pulse timings; determining a timing control value and a frequency control value by statistical processing using a population of a plurality of timing errors obtained by the error measurement process executed a plurality of times and a plurality of frequency errors obtained by the error measurement process executed a plurality of times; an adjustment process that performs at least one of a process of adjusting a pulse timing of the local pulse signal using the timing control value and a process of adjusting a frequency of the local pulse signal using the frequency control value on the local pulse signal; A time synchronization device, characterized in that the size of the population in the statistical processing is determined based on the operating status of a source that generates the local pulse signal. Configuration 4: The time synchronization device according to the first or second aspect of the present invention, the radio signal is a signal transmitted from an artificial satellite, The receiving unit generating quality information indicating accuracy of the reference pulse signal based on the received signal; The information processing unit A time synchronization device characterized in that the less accurate the quality information indicates, the more times the statistical processing is performed based on the error measurement processing that has been performed. Configuration 5: The time synchronization device according to the second or third aspect of the present invention, Equipped with a temperature sensor, The information processing unit A temperature gradient is calculated based on the temperature indicated by the temperature sensor; A time synchronization device characterized in that the smaller the temperature gradient, the more times the statistical processing is performed based on the error measurement processing that has been performed. Configuration 6: The time synchronization device according to any one of configurations 1 to 5, The information processing unit A time synchronization device characterized by performing a correction process based on the timing control value and frequency control value in the latest adjustment process on multiple timing errors resulting from the error measurement process that has been performed multiple times in the past and multiple frequency errors resulting from the error measurement process that has been performed multiple times in the past, and then performing the statistical processing. [Explanation of symbols]

[0074] 10 receiving unit, 12 information processing unit, 14 local clock, 16 time signal output unit, 18 memory, 20 temperature sensor, 22 timing comparison unit, 24 control unit, 26 temperature gradient acquisition unit, 40, 42 characteristics, 100 time synchronization device.

Claims

1. a receiving unit that extracts a reference pulse signal from a received signal obtained by receiving a wireless signal; a local clock that generates a local pulse signal; an information processing unit that controls the local clock, The information processing unit a timing error based on a difference between the pulse timing indicated by the reference pulse signal and the pulse timing indicated by the local pulse signal; Sequentially executing an error measurement process over time to determine a frequency error based on a difference between the timing error determined for each of the earlier pulse timings and the timing error determined for each of the later pulse timings; determining a timing control value and a frequency control value by statistical processing using a population of a plurality of timing errors obtained by the error measurement process executed a plurality of times and a plurality of frequency errors obtained by the error measurement process executed a plurality of times; an adjustment process that performs at least one of a process of adjusting a pulse timing of the local pulse signal using the timing control value and a process of adjusting a frequency of the local pulse signal using the frequency control value on the local pulse signal; A time synchronization device, characterized in that the size of the population in the statistical processing is determined based on the accuracy of the reference pulse signal.

2. 2. The time synchronization device according to claim 1, Equipped with a temperature sensor, The information processing unit A temperature gradient is calculated based on the temperature indicated by the temperature sensor; the smaller the temperature gradient, the more times the statistical processing is performed based on the error measurement processing; A time synchronization device characterized in that the size of the population in the statistical processing is determined based on the accuracy of the reference pulse signal as well as the operating status of the source of the local pulse signal.

3. a receiving unit that extracts a reference pulse signal from a received signal obtained by receiving a wireless signal; a local clock that generates a local pulse signal; an information processing unit that controls the local clock; A temperature sensor; Equipped with The information processing unit a timing error based on a difference between the pulse timing indicated by the reference pulse signal and the pulse timing indicated by the local pulse signal; Sequentially executing an error measurement process over time to determine a frequency error based on a difference between the timing error determined for each of the earlier pulse timings and the timing error determined for each of the later pulse timings; determining a timing control value and a frequency control value by statistical processing using a population of a plurality of timing errors obtained by the error measurement process executed a plurality of times and a plurality of frequency errors obtained by the error measurement process executed a plurality of times; an adjustment process that performs at least one of a process of adjusting a pulse timing of the local pulse signal using the timing control value and a process of adjusting a frequency of the local pulse signal using the frequency control value on the local pulse signal; A temperature gradient is calculated based on the temperature indicated by the temperature sensor; the smaller the temperature gradient, the more times the statistical processing is performed based on the error measurement processing; A time synchronization device, characterized in that the size of the population in the statistical processing is determined based on the operating status of a source that generates the local pulse signal.

4. 3. The time synchronization device according to claim 1, the radio signal is a signal transmitted from an artificial satellite, The receiving unit generating quality information indicating accuracy of the reference pulse signal based on the received signal; The information processing unit A time synchronization device characterized in that the less accurate the quality information indicates, the more times the statistical processing is performed based on the error measurement processing that has been performed.

5. The time synchronization device according to any one of claims 1 to 3, The information processing unit A time synchronization device characterized by performing a correction process based on the timing control value and frequency control value in the latest adjustment process on multiple timing errors resulting from the error measurement process that has been performed multiple times in the past and multiple frequency errors resulting from the error measurement process that has been performed multiple times in the past, and then performing the statistical processing.

Images