Error rate measurement device and error rate measurement method

Abstract

To provide an error rate measurement device capable of obtaining a desired reliability level in error rate measurement of an input signal, and an error rate measurement method.SOLUTION: An error rate measurement device includes: an average value calculation section 31 which calculates a sum x of a value resulting from dividing a parameter min_x by 2 and a value resulting from dividing a parameter max_x by 2; a temporary reliability calculation section 32 which calculates a temporary reliability level CLx using the sum x; an upper / lower limit value update section 33 which substitutes the sum x for the parameter min_x in a case where the temporary reliability level CLx is equal to or less than a target reliability level CLS and substitutes the sum x for the parameter max_x in a case where the temporary reliability level CLx is larger than the target reliability level CLS; and a measurement time calculation section 38 which calculates as a measurement time, in a case where a change of a value of the sum x calculated by the average value calculation section 31 is settled within a predetermined range, a value resulting from dividing the settled sum x by a product of a data rate and a target error rate ERS.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to an error rate measuring apparatus and an error rate measuring method for measuring the error rate of a signal input from a device under test. [Background technology]

[0002] For some time now, error rates have been measured using a Bit Error Rate Tester (BERT) (see, for example, Patent Documents 1 to 3). Measuring error rates is a very important analytical tool for evaluating the performance of various communication devices under test.

[0003] Most errors that occur in real systems are due to random noise, which means that errors occur at random times, and in systems that use a Decision Feedback Equalizer (DFE), random errors can appear as burst errors.

[0004] For this reason, the reliability of the Bit Error Rate (BER) is usually evaluated under the condition E≧1, using the confidence level defined by the following formula (1). Here, the confidence level CL in formula (1) indicates the probability that the true BER of the system will be smaller than the target BER.

[0005]

number

[0006] In addition, in the formula (1), N: Bit rate [bit / s] × measurement time [s] (number of measurement bits) BER S :Target BER E: Expected number of errors It is.

[0007] For a predefined target BER, the measurement time required to achieve a desired reliability level CL varies depending on the desired reliability level CL and the bit rate. For example, when E=0 for error-free measurement, the BER S If CL is given, it is easy to solve equation (1) for N, but when E ≥ 1, solving equation (1) for N is mathematically very difficult and unrealistic. [Prior art documents] [Patent documents]

[0008] [Patent Document 1] Patent No. 6818056 [Patent Document 2] JP 2007-155457 A [Patent Document 3] Special Publication No. 2017-538367 Summary of the Invention [Problem to be solved by the invention]

[0009] In other words, conventional error rate measurement devices such as those disclosed in Patent Documents 1 to 3 were unable to automatically evaluate how long the measurement time needed to achieve the required reliability level when the expected number of errors E was 1 or more.

[0010] In reality, the measurement time step set by the user is 1 second or 0.1 second, and there is no need to calculate the measurement time mathematically strictly. For example, bit rate = 32 Gbit / s, BER S When E = 1E-10 and E = 2, the reliability level CL for the measurement time is as shown in the graph in Figure 9. For example, the measurement time when CL = 95% is approximately 1.967 seconds, but in reality an accuracy of 2 seconds (CL = 95.36%) is sufficient.

[0011] However, in the past, the user had to search for a measurement time that would give the desired reliability level CL by, for example, gradually changing the measurement time substituted into equation (1), or had to set the measurement time longer, resulting in a decrease in work efficiency.

[0012] Currently, there are many types of error measurements using error rate measuring devices, and not only the BER of NRZ (Non Return to Zero) signals (hereinafter simply referred to as "NRZ signals"), but also the Symbol Error Rate (SER) is added to the evaluation index for PAM4 (Pulse Amplitude Modulation 4) signals (hereinafter simply referred to as "PAM4 signals"). Furthermore, in standards that assume Forward Error Correction (FEC), such as 400GbE (Gigabit Ethernet) and PCIe (Registered Trademark) (Peripheral Component Interconnect Express) Gen (Generation) 6, not only evaluations such as BER and SER, but also "Uncorrectable Codeword Rate" and "Flit Error Rate", which are indicators of whether error correction by FEC is possible, are important.

[0013] The present invention has been made to solve the above-mentioned problems in the conventional art, and has as its object to provide an error rate measuring apparatus and an error rate measuring method that are capable of calculating a measurement time that provides a desired reliability level in measuring the error rate of an input signal. [Means for solving the problem]

[0014] In order to achieve the above object, an error rate measurement device according to the present invention is an error rate measurement device (1) for measuring an error rate of an input signal from a device under test (200), and includes a processing unit (30) for calculating the measurement time for giving the target reliability level based on a target reliability level, a data rate of the input signal, a target error rate of the input signal, and an assumed number of errors during a measurement time of the input signal, the processing unit including an average calculation unit (31) for calculating a sum x of a value obtained by dividing a variable min_x by 2 and a value obtained by dividing a variable max_x by 2, a provisional reliability level calculation unit (32) for calculating a provisional reliability level CLx by substituting the sum x into the following formula (2), and and an upper and lower limit value updating unit (33) that assigns the sum x to the variable min_x if the provisional reliability level CLx is equal to or lower than the target reliability level, and assigns the sum x to the variable max_x if the provisional reliability level CLx is greater than the target reliability level, and the processing unit further includes a measurement time calculation unit (38) that repeats the processing of the average value calculation unit, the provisional reliability level calculation unit, and the upper and lower limit value updating unit using the latest variable min_x and the latest variable max_x, and when a change in the value of the sum x calculated by the average value calculation unit converges to a predetermined range, calculates a value obtained by dividing the converged sum x by the product of the data rate and the target error rate as the measurement time.

number

[0015] With this configuration, the error rate measurement apparatus of the present invention performs a process of narrowing down the sum x in measuring the error rate of an input signal, thereby making it possible to calculate a measurement time that will obtain a desired target reliability level, without the user being aware that a measurement time calculation process is being performed.

[0016] Furthermore, the error rate measurement device according to the present invention may be configured so that the sum x calculated by the average value calculation unit contains a rounding error, and the measurement time calculation unit calculates, when significant digits of the sum x converge to a constant value, a value obtained by dividing the converged sum x by the product of the data rate and the target error rate, as the measurement time.

[0017] With this configuration, the error rate measurement device according to the present invention uses convergence due to rounding error as the end judgment condition for the narrowing down process of the sum x, so that the user does not have to take the time to consider the range of possible values ​​of the sum x and the number of times the process is to be repeated every time various measurement conditions are changed on the setting screen.

[0018] Moreover, the error rate measurement device according to the present invention may be arranged so that the sum x is data of a double-precision floating-point number, and the significant digit is a mantissa part of the double-precision floating-point number.

[0019] With this configuration, the error rate measurement device according to the present invention can calculate the measurement time using sum x having approximately 14 significant digits when converted to a decimal number when the numeric type of variable min_x, variable max_x, and sum x is a double-precision floating-point number.

[0020] Moreover, the error rate measurement device according to the present invention may further comprise an error rate calculation section (25) that calculates an error rate of the input signal over the measurement time calculated by the measurement time calculation section.

[0021] With this configuration, the error rate measurement device according to the present invention can calculate the error rate of the input signal over a measurement time period that allows a desired target reliability level to be obtained.

[0022] Further, an error rate measurement method according to the present invention is an error rate measurement method for measuring an error rate of an input signal from a device under test (200), and includes an input step (S22) of inputting a target reliability level, a data rate of the input signal, a target error rate of the input signal, and an expected number of errors during a measurement time of the input signal, a step (S24) of setting initial values ​​to variables min_x and max_x, an average calculation step (S25) of calculating a sum x of a value obtained by dividing the variable min_x by 2 and a value obtained by dividing the variable max_x by 2, a provisional reliability level calculation step (S27) of calculating a provisional reliability level CLx by substituting the sum x into the following formula (2), and an upper limit updating step (S30) of substituting the sum x for the variable max_x if the provisional reliability level CLx is greater than the target reliability level; a step (S26) of repeating the processes of the average value calculation step, the provisional reliability level calculation step, the lower limit updating step, and the upper limit updating step using the latest variable min_x and the latest variable max_x; and a measurement time calculation step (S31) of calculating, when a change in the value of the sum x calculated in the average value calculation step converges to a predetermined range, the converged sum x divided by the product of the data rate and the target error rate as the measurement time that gives the target reliability level.

number

[0023] Furthermore, in the error rate measuring method according to the present invention, the sum x calculated in the average value calculating step may include a rounding error, and the measuring time calculating step may be configured to calculate, when significant digits of the sum x converge to a constant value, a value obtained by dividing the converged sum x by a product of the data rate and the target error rate as the measuring time.

[0024] Moreover, the error rate measurement method according to the present invention may further include an error rate calculation step (S34) of calculating an error rate of the input signal over the measurement time calculated in the measurement time calculation step. Effect of the Invention

[0025] The present invention provides an apparatus and method for measuring an error rate, which are capable of calculating a measurement time that provides a desired reliability level in measuring the error rate of an input signal. [Brief description of the drawings]

[0026] [Figure 1] 1 is a block diagram showing a configuration of an error rate measurement device according to an embodiment of the present invention; [Diagram 2] 2 is a block diagram showing the configuration of a processing unit included in the error rate measurement device according to the embodiment of the present invention. FIG. [Diagram 3] 5 is an explanatory diagram for explaining the operation of a processing unit included in the error rate measurement device according to the embodiment of the present invention. FIG. [Figure 4] This is a table showing the change in each value when the numeric type is double type. [Diagram 5] 13 is a graph showing the change in each value when the numeric type is double type. [Figure 6] 5 is an example of a setting screen of the error rate measurement device according to the embodiment of the present invention. [Figure 7] 1 is a table showing units of measurement and data rates for each unit of measurement. [Figure 8] 4 is a flowchart showing the process of an error rate measurement method using the error rate measurement device according to the embodiment of the present invention. [Figure 9] 13 is a graph showing an example of a reliability level versus measurement time. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0027] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of an error rate measurement device and an error rate measurement method according to the present invention will now be described with reference to the drawings.

[0028] As shown in FIG. 1, an error rate measurement device 1 according to an embodiment of the present invention measures the error rate of an input signal output from a device under test (DUT) 200, and includes a signal output section 10, a signal input section 20, a data memory section 40, an operation section 41, a display section 42, and a control section 43.

[0029] Examples of standards supported by DUT 200 include PCIe Gen1 to 6, USB (registered trademark) (Universal Serial Bus) 3.1 to 4, CEI (Common Electrical Interface), IEEE802.3, InfiniBand HDR, and Fibre Channel.

[0030] The data storage unit 40 is configured with a memory such as a RAM (Random Access Memory). The data storage unit 40 stores, as signal data of a known pattern input from a signal output unit 10 (described later) to the DUT 200, for example, bit string data of an NRZ signal (bit string data consisting of 0 or 1) and symbol string data of a PAM4 signal (symbol string data consisting of 0, 1, 2, or 3).

[0031] The data storage unit 40 may also store bit string data of the MSB (Most Significant Bit) and LSB (Least Significant Bit) of the PAM4 signal input to the DUT 200. The symbol string data of the PAM4 signal, the bit string data of the MSB and LSB, and the bit string data of the NRZ signal stored in the data storage unit 40 also serve as reference data for the error rate calculation unit 25 (described later) to compare with the input signal from the DUT 200.

[0032] The signal output unit 10 generates a test signal consisting of data of a known pattern input from the data storage unit 40. Then, the signal output unit 10 outputs the generated test signal to the DUT 200. At this time, the DUT 200 loops back the test signal output from the signal output unit 10 and uses it as an input signal to the signal input unit 20.

[0033] The signal input section 20 receives the input signal output from the DUT 200 , and includes an equalizer 21 and an error rate measuring section 22 .

[0034] The input signal from the DUT 200 is, for example, a test signal such as a PAM4 signal output from the signal output unit 10 and looped back from the DUT 200 .

[0035] The equalizer 21 adjusts the frequency characteristics of an input signal from the DUT 200. The equalizer 21 is configured with, for example, a continuous time linear equalizer (CTLE), a low frequency equalizer (LFE), a DFE, and the like.

[0036] The error rate measuring section 22 includes a clock recovery section 23 , a data extraction section 24 , an error rate calculation section 25 , and a processing section 30 , and is configured to measure the error rate of an input signal from the DUT 200 .

[0037] The clock recovery unit 23 generates a recovered clock signal from the input signal from the DUT 200 that has been adjusted by the equalizer 21 .

[0038] The data extracting section 24 extracts bit string data or symbol string data of the input signal from the DUT 200 adjusted by the equalizer 21 .

[0039] The data extraction unit 24 extracts bit string data or symbol string data of the input signal output from the DUT 200 by punching the input signal adjusted by the equalizer 21 at the rising or falling timing of the clock signal. Here, the above clock signal used by the data extraction unit 24 may be the recovered clock signal output from the clock recovery unit 23, or may be an external clock signal corresponding to the baud rate of the input signal. In this specification, the recovered clock signal and the external clock signal are collectively referred to simply as "clock signal".

[0040] For example, the data extraction unit 24 has a plurality of 0 / 1 decision devices, and a clock signal is input to each of the 0 / 1 decision devices, so that the level of the input signal output from the DUT 200 can be decided at the timing of the clock signal. Note that the recovered clock signal output from the clock recovery unit 23 may be used as an operating clock not only in the data extraction unit 24 but also in each unit constituting the error rate measurement device 1.

[0041] The error rate calculation unit 25 is configured to calculate the error rate of the input signal from the DUT 200 extracted by the data extraction unit 24 by sequentially comparing the bit string data or symbol string data extracted by the data extraction unit 24 with reference data stored in the data storage unit 40. In this specification, the BER, SER, Uncorrectable Codeword Rate, and Flit Error Rate are collectively referred to simply as "error rate".

[0042] The error rate calculation section 25 counts the number of errors in the input signal from the DUT 200 over a measurement time set on a setting screen 50 (to be described later) or a measurement time calculated by a measurement time calculation section 38 (to be described later). Furthermore, the error rate calculation section 25 divides the counted number of errors by the number of measurement data of the input signal over the above-mentioned measurement time, and calculates the result as the error rate of the input signal from the DUT 200.

[0043] The processing unit 30 determines the target reliability level CL S , data rate of the input signal, target error rate of the input signal ER S Based on the measurement conditions such as the expected number of errors E during the measurement time of the input signal, the target reliability level CL S The measurement time for the input signal that gives the target reliability level CL S is the target error rate ER S represents the probability that the true error rate of the input signal from DUT 200 is smaller than

[0044] As shown in FIG. 2, the processing unit 30 includes an average value calculation unit 31, a provisional reliability level calculation unit 32, an upper and lower limit update unit 33, a convergence determination unit 36, a transmission rate conversion unit 37, and a measurement time calculation unit 38.

[0045] The average value calculation unit 31 calculates the sum x of a value obtained by dividing the variable min_x by 2 and a value obtained by dividing the variable max_x by 2.

[0046] The provisional reliability level CLx obtained by substituting the initial value of the variable min_x for x in the formula (2) described later is the target reliability level CL S Similarly, the provisional reliability level CLx obtained by substituting the initial value of the variable max_x for x in the formula (2) described later is smaller than the target reliability level CL S It needs to be larger than

[0047] For this reason, for example, it is recommended that the initial value of the variable min_x be 0, and the initial value of the variable max_x be the maximum valid value of the numeric type used in the calculation.

[0048] The average calculation unit 31 performs rounding to truncate the least significant digits of each value of min_x / 2, max_x / 2, and x so that each value falls within the maximum number of significant digits of the numeric type being used. In other words, min_x / 2, max_x / 2, and x calculated by the average calculation unit 31 all contain rounding errors. Note that the above rounding is not limited to rounding down, and may be any rounding such as rounding up.

[0049] The numeric type of the variables min_x, max_x, and sum x is, for example, a single-precision floating-point number (float type), a double-precision floating-point number (double type), or a fixed-point number. Of these, the double type is 64-bit data consisting of a 1-bit sign part, a 52-bit mantissa part, and an 11-bit exponent part. The maximum valid value of the double type is 1.79769313486232E+308 in decimal conversion.

[0050] The provisional reliability level calculation unit 32 calculates the provisional reliability level CLx by substituting the sum x calculated by the average value calculation unit 31 into the following formula (2). As can be seen from formula (2), the provisional reliability level CLx increases monotonically with respect to the sum x.

[0051]

number

[0052] In addition, in the formula (2), x:N×ER S N: Data rate [data / s] × measurement time [s] ER S : Target Error Rate (ER) E: Expected number of errors That is, equation (2) is the N×BER of equation (1). S N×ER S (=x).

[0053] The data rate is a parameter indicating the number of measurement units contained in the input signal from the DUT 200 per second. The measurement unit is, for example, one bit, one symbol (PAM4 Symbol), one flit, or one codeword. Therefore, N is a parameter indicating the number of measurement units contained in the input signal from the DUT 200 over the measurement time, i.e., the number of measurement data. Target error rate ER S is, for example, any one of BER, SER, Uncorrectable Codeword Rate, or Flit Error Rate, and is the upper limit of the allowable error rate of the input signal.

[0054] The upper and lower limit value update unit 33 updates the provisional reliability level CLx calculated by the provisional reliability level calculation unit 32 to the target reliability level CL S In the following cases, the upper and lower limit value update unit 33 substitutes the latest sum x for the variable min_x. S If the sum is greater than x, the latest sum x is substituted for the variable max_x.

[0055] Since the provisional reliability level CLx increases monotonically with respect to the sum x, it is possible to narrow down the range of the value of the sum x and obtain an approximation of the sum x by repeating the processes by the average value calculation unit 31, the provisional reliability level calculation unit 32, and the upper and lower limit value update unit 33. FIG. 3 is a diagram for explaining the process of narrowing down the sum x by the average value calculation unit 31, the provisional reliability level calculation unit 32, and the upper and lower limit value update unit 33.

[0056] First, the average value calculation unit 31 calculates the sum x using the initial values ​​of the variables min_x and max_x (step S1).

[0057] Next, the provisional reliability level calculation unit 32 calculates the provisional reliability level CLx by substituting the sum x calculated in step S1 into equation (2). The upper and lower limit value update unit 33 checks whether the provisional reliability level CLx calculated by the provisional reliability level calculation unit 32 is equal to or lower than the target reliability level CL S If it is greater than the initial value, the latest sum x is substituted for the variable max_x. The average value calculation unit 31 calculates the sum x again using the initial value of the variable min_x and the updated variable max_x (step S2).

[0058] Next, the provisional reliability level calculation unit 32 calculates the provisional reliability level CLx by substituting the sum x calculated in step S2 into equation (2). The upper and lower limit value update unit 33 checks whether the provisional reliability level CLx calculated by the provisional reliability level calculation unit 32 is equal to or lower than the target reliability level CL S In the following cases, the latest sum x is substituted for the variable min_x: The average value calculation unit 31 recalculates the sum x using the updated variables min_x and max_x (step S3).

[0059] Next, the provisional reliability level calculation unit 32 calculates the provisional reliability level CLx by substituting the sum x calculated in step S3 into the formula (2). The upper and lower limit value update unit 33 checks whether the provisional reliability level CLx calculated by the provisional reliability level calculation unit 32 is equal to or lower than the target reliability level CL S If it is greater than the sum x, the latest sum x is substituted for the variable max_x. The average value calculation unit 31 recalculates the sum x using the variable min_x and the updated variable max_x (step S4).

[0060] That is, the average value calculation unit 31 calculates the latest sum x using the latest variable min_x and the latest variable max_x. Furthermore, the provisional reliability level calculation unit 32 calculates the provisional reliability level CLx by substituting the latest sum x into equation (2). Then, the upper and lower limit value update unit 33 substitutes the latest sum x into the variable min_x or the variable max_x according to the latest provisional reliability level CLx. In this way, the processing unit 30 repeats the narrowing down process of the sum x by the average value calculation unit 31, the provisional reliability level calculation unit 32, and the upper and lower limit value update unit 33 using the latest variable min_x and the latest variable max_x.

[0061] The convergence determination unit 36 ​​is configured to determine whether or not the change in the value of the sum x calculated by the average value calculation unit 31 has converged to a predetermined range. As an example, the convergence determination unit 36 ​​may determine whether or not the significant digits of the sum x calculated by the average value calculation unit 31 have converged to a constant value. When the numeric types of the variables min_x, max_x, and sum x are double type, the significant digits of the variables min_x, max_x, and sum x are mantissa parts of double type.

[0062] The convergence determination unit 36 ​​determines that the significant digits of the sum x have converged to a constant value when the latest sum x is equal to the previous sum x, for example. Alternatively, the convergence determination unit 36 ​​may determine that the significant digits of the sum x have converged to a constant value when the sum x calculated by the average value calculation unit 31 is equal three or more times in a row.

[0063] Fig. 4 is a table showing the change in each value when the numeric type of the variables min_x, max_x, and sum x is double type. Fig. 5 is a graph showing data in the vicinity of where sum x converges in the changes in the variables min_x, max_x, and sum x shown in Fig. 4. Here, the input signal from the DUT 200 is an NRZ signal, the data rate is 2.4 Gbps, and the target error rate ER, which is the standard for the reliability level, is 1.0 Gbps. S is 1E-14, the expected number of errors E as the reliability level standard is 3, and the target reliability level CL S is said to be 0.98.

[0064] In the example of Figure 4, we can see that the value of min_x has changed in the 1071st and 1072nd calculation results, but the value of x has not changed. This indicates that the value of x has converged due to rounding error.

[0065] In addition, in the 1069th to 1072nd calculation results, CL S and CLx are the same, but depending on the input value to equation (2), CL Sx may converge before CLx coincides with CL. S It is not desirable to use whether or not CLx matches as a condition for determining whether the narrowing down process of sum x is to be terminated.

[0066] When the variable min_x, the variable max_x, and the sum x are of double type, for example, a result with about 14 significant digits in decimal conversion can be obtained. The number of repetitions of the narrowing down process of the sum x is about 1000, and the time required for 1000 repetitions is about several tens of ms. Therefore, it is considered that the narrowing down process of the sum x is not likely to impair user convenience.

[0067] In addition, in this narrowing process, the end judgment condition for the process is the convergence of the sum x due to the rounding error, not the number of times the process is repeated. Therefore, the user does not need to consider the possible range of the value of the sum x or the number of times the process is repeated every time various measurement conditions are changed on the setting screen 50.

[0068] Alternatively, the convergence determination unit 36 ​​may determine whether the change in the value of the sum x calculated by the average calculation unit 31 is equal to or less than a predetermined value having fewer digits than the maximum number of significant digits of the numeric type being used, rather than waiting for the convergence of the sum x due to the rounding error. In this case, although the number of significant digits decreases, it is possible to further speed up the narrowing down process.

[0069] The transmission speed conversion unit 37 converts the baud rate inputted on a setting screen 50 (described later) into a transmission speed (data rate) corresponding to the measurement unit.

[0070] As shown in the following formula (3), the measurement time calculation unit 38 calculates the sum x calculated by the average calculation unit 31 and determined to have converged by the convergence determination unit 36, and calculates the sum x by multiplying it by the data rate converted by the transmission rate conversion unit 37 and the target error rate ER S The measurement time is calculated by dividing the measured value by the product of

[0071] Measurement time [s] = x / (data rate [data / s] × ER S ) ···(3)

[0072] The operation unit 41 is for accepting operation input by a user, and is composed of user interfaces such as operation knobs, various keys, switches, buttons, and soft keys on the display screen of the display unit 42, which are provided in the error rate measurement device 1 shown in Fig. 1. The operation unit 41 also performs various settings related to error rate measurement of the error rate measurement device 1 and various measurement condition settings on the setting screen 50.

[0073] The display unit 42 is comprised of a display device such as an LCD (Liquid Crystal Display) or a CRT (Cathode Ray Tube) provided in the error rate measurement device 1 shown in Fig. 1, and displays the setting screen 50, measurement results, etc. based on a display control signal from the control unit 43. The display unit 42 may also have the operation functions of the operation unit 41, such as soft keys on the display screen.

[0074] The control unit 43 controls the signal output unit 10, the signal input unit 20, the data storage unit 40, the operation unit 41, and the display unit 42. The control unit 43 is configured with a control device such as a computer including a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), an FPGA (Field Programmable Gate Array), a ROM (Read Only Memory), a RAM, and an HDD (Hard Disk Drive). The control unit 43 can also configure at least a part of the error rate measurement unit 22 in software form by executing a predetermined program using the CPU or GPU.

[0075] As shown in FIG. 6, the display unit 42 displays a setting screen 50 for setting the measurement conditions for the error rate of the input signal from the DUT 200.

[0076] The setting screen 50 includes, in a display area 51 ("Gating" in the figure), a "Target" pull-down menu 52, "ER" text boxes 53a, 53b, an "EC" text box 54, text 55 indicating the unit of the expected number of errors E, a "Cycle" pull-down menu 56, a "Unit" pull-down menu 57, measurement time text boxes 58a, 58b, a "Baud Rate" text box 59, and a "Confidence Level" text box 60.

[0077] The "Target" pull-down menu 52 constitutes a measurement unit input section for inputting the measurement unit of the error rate. The "Target" pull-down menu 52 allows the user to select, for example, one of the measurement units "Bit", "PAM4 Symbol", "Flit", or "Codeword". The measurement unit input section and a baud rate input section, which will be described later, constitute a transmission speed input section for inputting the transmission speed of the input signal from the DUT 200.

[0078] FIG. 7 is a table showing measurement units selectable in the "Target" pull-down menu 52 and the data rates for each measurement unit converted by the transmission speed converter 37.

[0079] "Bit" is a measurement unit selected when measuring the BER of the input signal from the DUT 200. When the input signal from the DUT 200 is an NRZ signal, the transmission speed conversion unit 37 outputs the baud rate input in a "Baud Rate" text box 59 (described later) as the data rate. When the input signal from the DUT 200 is a PAM4 signal, the transmission speed conversion unit 37 outputs a rate that is twice the baud rate input in the "Baud Rate" text box 59 as the data rate.

[0080] "PAM4 Symbol" is a measurement unit selected when measuring the SER of the input signal from the DUT 200. When the input signal from the DUT 200 is a PAM4 signal, the transmission speed conversion unit 37 outputs the baud rate input in the "Baud Rate" text box 59 as the data rate without any change.

[0081] "Flit" is a measurement unit selected when measuring the Flit Error Rate of the input signal from the DUT 200. The transmission speed conversion unit 37 doubles the baud rate entered in the "Baud Rate" text box 59 and divides the result by the Flit length, and outputs the result as the data rate. Here, the Flit length is 2048 bits.

[0082] "Codeword" is a measurement unit selected when measuring the Uncorrectable Codeword Rate of the input signal from the DUT 200. The transmission speed conversion unit 37 doubles the baud rate entered in the "Baud Rate" text box 59 and divides the result by the CW (Codeword) length, and outputs the result as the data rate. Here, the CW length is 5440 bits in the case of RS-FEC (Reed-Solomon Forward Error Correction) (544,514) specified in IEEE802.3.

[0083] The “ER” text boxes 53a and 53b are used to input a desired target error rate ER of the input signal from the DUT 200. s The text boxes 53a and 53b form a target error rate input section for inputting the target error rate ER s can be input in exponential notation, with the mantissa being input in text box 53a and the exponent being input in text box 53b. For example, values ​​in the range of 1E-3 to 1E-15 can be input in text boxes 53a and 53b.

[0084] The "EC" text box 54 constitutes an expected error number input section for inputting the expected error number E during the measurement time of the input signal from the DUT 200. For example, a value in the range of 0 to 10 can be input into the text box 54. Furthermore, the display of the text 55 indicating the unit of the expected error number E input into the text box 54 changes according to the measurement unit selected in the "Target" pull-down menu 52.

[0085] For example, when "Bit" is selected in the "Target" pull-down menu 52, the text 55 displays "Bit." When "PAM4 Symbol" is selected in the "Target" pull-down menu 52, the text 55 displays "Symbol." When "Flit" is selected in the "Target" pull-down menu 52, the text 55 displays "Flit." When "Codeword" is selected in the "Target" pull-down menu 52, the text 55 displays "Codeword."

[0086] The "Cycle" pull-down menu 56 allows the user to select, for example, one of the measurement operations, "Repeat (CL)", "Single (CL)", "Repeat", "Single", or "Untimed".

[0087] "Repeat (CL)" is a measurement operation that repeats error rate measurement for a measurement time calculated by the measurement time calculation unit 38. "Single (CL)" is a measurement operation that performs an error rate measurement once for a measurement time calculated by the measurement time calculation unit 38. "Repeat" is a measurement operation that repeats error rate measurement for a measurement time according to a measurement period selected in the "Unit" pull-down menu 57 described later. "Single" is a measurement operation that performs an error rate measurement once for a measurement time according to a measurement period selected in the "Unit" pull-down menu 57. "Untimed" is a measurement operation that continues to execute error rate measurement from a measurement start instruction given by pressing a measurement start button (not shown) to a measurement end instruction given by pressing a measurement stop button (not shown).

[0088] The "Unit" pull-down menu 57 allows the user to select the unit of the measurement cycle from, for example, "Time," "Clock Count," or "Error count." When "Repeat (CL)" or "Single (CL)" is selected in the "Cycle" pull-down menu 56, the "Unit" pull-down menu 57 is fixed to "Time."

[0089] "Time" is an item for inputting or displaying the measurement time, which is the unit of the measurement cycle, in the measurement time text boxes 58a, 58b. When "Repeat" or "Single" is selected in the "Cycle" pull-down menu 56, the measurement time per measurement cycle can be input in 1-second increments in the measurement time text boxes 58a, 58b, for example, in the range of 1 second to 99 days, 23 hours, 59 minutes, and 59 seconds. In the text box 58a, the number of days can be input, for example, in the range of 0 days to 99 days. In the text box 58b, the time can be input, for example, in the range of 1 second to 23 hours, 59 minutes, and 59.9 seconds.

[0090] When "Repeat (CL)", "Single (CL)", or "Untimed" is selected in the "Cycle" pull-down menu 56, values ​​cannot be entered into the measurement time text boxes 58a and 58b.

[0091] When "Repeat (CL)" or "Single (CL)" is selected in the "Cycle" pull-down menu 56, the measurement time text boxes 58a and 58b form a measurement time display section that displays the measurement time calculated by the measurement time calculation section 38. The text box 58a can display the number of days, for example, in the range of 0 to 99 days. The text box 58b can display the time, for example, in the range of 1 second to 23 hours 59 minutes 59.9 seconds.

[0092] If there is a change in any of the selection made in the "Target" pull-down menu 52, the input in the "ER" text boxes 53a and 53b, the input in the "EC" text box 54, the input in the "Baud Rate" text box 59, or the input in the "Confidence Level" text box 60, the display of the measurement time calculation results in the measurement time text boxes 58a and 58b is also automatically updated.

[0093] Here, the text box 58b displays, for example, a value rounded up to the first or second decimal place for the number of seconds of the measurement time calculated by the measurement time calculation unit 38. In the example of Fig. 4, the value of the sum x of the 1071st and 1072nd times is 9.08411538241318E+00, and the measurement time calculated from the formula (3) at this time is 3.78504807600549E+05 [s]. At this time, the number of days "4" is displayed in the text box 58a, and the time "09:08:24.9" (rounded up to the second decimal place) or "09:08:25" (rounded up to the first decimal place) is displayed in the text box 58b.

[0094] "Clock Count" is an item for setting the unit of the measurement cycle in clock counts in the measurement time text boxes 58a, 58b when "Repeat" or "Single" is selected in the "Cycle" pull-down menu 56. In other words, the error rate measurement ends when a time equal to or greater than the clock count set in the measurement time text boxes 58a, 58b has elapsed.

[0095] "Error Count" is an item for setting the unit of the measurement cycle in the measurement time text boxes 58a, 58b with the number of errors detected by the error rate measurement section 22 when "Repeat" or "Single" is selected in the "Cycle" pull-down menu 56. In other words, the error rate measurement ends when the error rate measurement section 22 detects the number of errors set in the measurement time text boxes 58a, 58b.

[0096] The "Baud Rate" text box 59 constitutes a baud rate input section for inputting the baud rate of the transmission speed of the input signal from the DUT 200. "Baud Rate" is an item for allowing the transmission speed of the input signal from the DUT 200 to be input in 1 kBaud steps within the range of 2.4 GBaud to 64.2 GBaud, for example.

[0097] The "Confidence Level" text box 60 is used to enter the desired target confidence level CL when "Repeat (CL)" or "Single (CL)" is selected in the "Cycle" pull-down menu 56. S The target reliability level input unit is configured to input the target reliability level CL S The case where the measurement time is 0 seconds or E in formula (2) is infinite is when the reliability becomes 0%, which is not possible in reality. S The value of 100% would be reached if the measurement time were infinite, which is also impossible in reality. For this reason, the text box 60 is designed so that values ​​of 0% and 100% cannot be set.

[0098] For example, when 0% is entered in the text box 60, the display unit 42 automatically switches the input value in the text box 60 to 0.1%. When 100% is entered in the text box 60, the display unit 42 automatically switches the input value in the text box 60 to 99.9%. Alternatively, when 0% or 100% is entered in the text box 60, the display unit 42 displays an error dialog box to inform the user of a target reliability level CL in the range of 0.1% to 99.9%. S The user may be prompted to enter:

[0099] Furthermore, when "Repeat" or "Single" is selected in the "Cycle" pull-down menu 56, the "Confidence Level" text box 60 can display the confidence level CL calculated by the processing unit 30 using the following formula (4) rounded off to the first decimal place in the range of 0% to 100% in 0.1% steps. When "Repeat" or "Single" is selected in the "Cycle" pull-down menu 56, the "Confidence Level" text box 60 cannot be used to input a value.

[0100]

number

[0101] When "Untimed" is selected in the "Cycle" pull-down menu 56, the display related to "Confidence Level" is erased from the setting screen 50. Also, when anything other than "Time" is selected in the "Unit" pull-down menu 57, the display related to "Confidence Level" is erased from the setting screen 50.

[0102] An example of the process of an error rate measurement method using the error rate measurement device 1 of this embodiment will be described below with reference to the flowchart in Fig. 8. Note that descriptions that overlap with the description of the configuration of the error rate measurement device 1 described above will be omitted as appropriate. Here, the process will be described when "Repeat (CL)" or "Single (CL)" is selected in the "Cycle" pull-down menu 56.

[0103] First, the control unit 43 displays on the display unit 42 the setting screen 50 for setting the measurement conditions for the error rate of the input signal from the DUT 200 (setting screen display step S21).

[0104] Next, the user inputs a target reliability level CL S, unit of measurement for error rate, baud rate of input signal (data rate information), target error rate of input signal ER s and the expected number of errors E during the measurement time of the input signal are input to the setting screen 50 (input step S22).

[0105] Next, the transmission speed conversion unit 37 converts the baud rate entered in the "Baud Rate" text box 59 into a transmission speed (data rate) corresponding to the measurement unit selected in the "Target" pull-down menu 52 (transmission speed conversion step S23).

[0106] Next, upper / lower limit update unit 33 sets 0 as an initial value to variable min_x, and sets the maximum valid value of the numeric type (for example, double type) used in the calculation as an initial value to variable max_x (step S24).

[0107] Next, the average value calculation unit 31 calculates the sum x of the latest variable min_x divided by 2 and the latest variable max_x divided by 2 (average value calculation step S25). This sum x includes a rounding error.

[0108] Next, the convergence determination unit 36 ​​determines whether the change in the value of the sum x calculated in the average calculation step S25 has converged to a predetermined range (step S26). For example, in step S26, the convergence determination unit 36 ​​determines whether the significant digits of the sum x have converged to a constant value.

[0109] If the change in the value of the sum x has converged to the predetermined range (step S26: YES), the processing unit 30 executes the processes of step S31 and after. If the change in the value of the sum x has not converged to the predetermined range (step S26: NO), the processing unit 30 executes the processes of step S27 and after.

[0110] That is, step S26 is a step of repeating the processes of average value calculation step S25, provisional reliability level calculation step S27 described later, lower limit value update step S29, and upper limit value update step S30 using the latest variable min_x and the latest variable max_x.

[0111] In step S27, the provisional reliability level calculation unit 32 substitutes the sum x calculated in the average value calculation step S25 into equation (2) to calculate the provisional reliability level CLx (provisional reliability level calculation step S27).

[0112] Next, the processing unit 30 determines whether the provisional reliability level CLx calculated in the provisional reliability level calculation step S27 is equal to or smaller than the target reliability level CL S It is determined whether or not it is equal to or less (step S28).

[0113] Provisional reliability level CLx is the target reliability level CL S In the following case (step S28: YES), the upper and lower limit value update unit 33 assigns the latest sum x to the variable min_x (lower limit value update step S29). Then, the average value calculation unit 31 calculates the sum x again in step S25 using the variable min_x updated in step S29.

[0114] Provisional reliability level CLx is the target reliability level CL S If it is greater than (step S28: NO), the upper / lower limit value update unit 33 assigns the latest sum x to the variable max_x (upper limit value update step S30). Then, the average value calculation unit 31 calculates the sum x again in step S25 using the variable max_x updated in step S30.

[0115] In step S31, the measurement time calculation unit 38 calculates the sum x determined to have converged in step S26, the data rate converted in the transmission rate conversion step S23, and the target error rate ER s Substituting and into equation (3), the target reliability level CL S A measurement time that gives the above is calculated (measurement time calculation step S31).

[0116] Next, the display unit 42 displays the measurement time calculated in the measurement time calculation step S31 in the text boxes 58a and 58b (measurement time display step S32).

[0117] Next, the processing unit 30 sets the information on the measurement time displayed in the measurement time display step S32 in the error rate calculation unit 25 (step S33).

[0118] Next, the error rate calculation section 25 calculates the error rate of the input signal from the DUT 200 over the measurement time set by the processing section 30 in step S33 (error rate calculation step S34).

[0119] As described above, the error rate measurement apparatus 1 according to this embodiment narrows down the sum x in the error rate measurement of the input signal, thereby enabling the user to measure the desired target reliability level CL without being aware that the measurement time is being calculated. S The measurement time at which the above can be obtained can be calculated.

[0120] Furthermore, the error rate measurement device 1 according to this embodiment uses convergence due to rounding error as the end determination condition for the narrowing down process of the sum x, so that it is possible to save the user the trouble of having to consider the possible range of values ​​of the sum x and the number of times the process is to be repeated every time various measurement conditions are changed on the setting screen 50.

[0121] Furthermore, when the numeric types of the variables min_x, max_x, and sum x are double type, the error rate measurement apparatus 1 according to this embodiment can calculate the measurement time using sum x having approximately 14 significant digits when converted into a decimal number.

[0122] Moreover, the error rate measurement apparatus 1 according to this embodiment has a desired target reliability level CL S Over the measurement time over which the error rate of the input signal is obtained, the error rate can be calculated. [Explanation of symbols]

[0123] 1. Error rate measurement device 25 Error rate calculation section 30 Processing section 31 Average value calculation section 32 Provisional Reliability Level Calculation Unit 33 Upper and lower limit update section 36 Convergence judgment section 37 Transmission speed conversion unit 38 Measurement time calculation section 41 Operation section 42 Display section 43 Control Unit 50 Setting screen 52,56,57 Pull-down menu 53a, 53b, 54, 58a, 58b, 59, 60 Text Box 55 Text 200 DUT

Claims

1. An error rate measuring device (1) for measuring an error rate of an input signal from an object to be measured (200), comprising: a processing unit (30) for calculating the measurement time that gives the target reliability level based on a target reliability level, a data rate of the input signal, a target error rate of the input signal, and an expected number of errors during a measurement time of the input signal; The processing unit includes: an average value calculation unit (31) that calculates a sum x of a value obtained by dividing a variable min_x by 2 and a value obtained by dividing a variable max_x by 2; a provisional reliability level calculation unit (32) for calculating a provisional reliability level CLx by substituting the sum x into the following formula (2); an upper / lower limit value update unit (33) that assigns the sum x to the variable min_x when the provisional reliability level CLx is equal to or less than the target reliability level, and assigns the sum x to the variable max_x when the provisional reliability level CLx is greater than the target reliability level, The processing unit includes: repeating the processes of the average value calculation unit, the provisional reliability level calculation unit, and the upper and lower limit value update unit using the latest variable min_x and the latest variable max_x; and a measurement time calculation unit (38) for calculating, when a change in the value of the sum x calculated by the average value calculation unit converges to a predetermined range, a value obtained by dividing the converged sum x by a product of the data rate and the target error rate as the measurement time. [0010]

2. the sum x calculated by the average value calculation unit includes a rounding error, 2. The error rate measurement device according to claim 1, wherein the measurement time calculation unit calculates, when a significant digit of the sum x converges to a constant value, a value obtained by dividing the converged sum x by a product of the data rate and the target error rate as the measurement time.

3. The sum x is double-precision floating-point number data, 3. The error rate measuring device according to claim 2, wherein the significant digit is a mantissa of a double-precision floating-point number.

4. 4. The error rate measuring device according to claim 1, further comprising an error rate calculation section (25) for calculating an error rate of the input signal over the measurement time calculated by the measurement time calculation section.

5. An error rate measurement method for measuring an error rate of an input signal from a device under test (200), comprising: an input step (S22) of inputting a target reliability level, a data rate of the input signal, a target error rate of the input signal, and an expected number of errors during a measurement time of the input signal; A step (S24) of setting initial values ​​to variables min_x and max_x; an average value calculation step (S25) of calculating a sum x of a value obtained by dividing the variable min_x by 2 and a value obtained by dividing the variable max_x by 2; a provisional reliability level calculation step (S27) of calculating a provisional reliability level CLx by substituting the sum x into the following formula (2); a lower limit updating step (S29) of substituting the sum x for the variable min_x when the provisional reliability level CLx is equal to or lower than the target reliability level; an upper limit updating step (S30) of substituting the sum x for the variable max_x when the provisional reliability level CLx is greater than the target reliability level; a step of repeating the average value calculation step, the provisional reliability level calculation step, the lower limit value update step, and the upper limit value update step by using the latest variable min_x and the latest variable max_x (S26); and a measurement time calculation step (S31) of calculating, when a change in the value of the sum x calculated by the average value calculation step has converged to a predetermined range, a value obtained by dividing the converged sum x by a product of the data rate and the target error rate, as the measurement time that gives the target reliability level. [0025]

6. the sum x calculated in the average calculation step includes a rounding error, 6. The error rate measurement method according to claim 5, wherein said measurement time calculation step calculates, when a significant digit of said sum x converges to a constant value, a value obtained by dividing said converged sum x by a product of said data rate and said target error rate as said measurement time.

7. 7. The error rate measuring method according to claim 5, further comprising an error rate calculating step (S34) of calculating an error rate of the input signal over the measurement time calculated in the measurement time calculating step.

Images