Test and measurement instrument and method for providing user interface
A user-friendly interface for constructing tensor images addresses the challenge of efficient data input to machine learning networks, improving signal measurement accuracy and efficiency in test and measurement systems by enabling precise data structuring for deep learning applications.
Patent Information
- Application Number
- JP2024084935
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2024-05-15
- Filing Date
- 2024-05-24
- Publication Date
- 2024-12-05
AI Technical Summary
Existing systems lack an efficient user interface for constructing tensor images to be input to machine learning networks for signal measurement, particularly in test and measurement applications, which hinders accurate data input for training and runtime processes.
A novel user interface and test and measurement apparatus that allows users to select and construct tensor images, including waveform vectors, waveform images, and S-parameter images, using interactive menus and panels to define parameters and formats for input to deep learning neural networks.
Enables efficient construction of tensor images for machine learning applications, improving the accuracy and efficiency of signal measurement by allowing users to precisely define and structure data inputs, enhancing the performance of deep learning networks in test and measurement tasks.
Smart Images

Figure 2024169418000001_ABST
Abstract
Description
[Technical field]
[0001] The present disclosure relates to test and measurement instruments and associated methods, and more particularly to a test and measurement instrument and its user interface for efficiently constructing tensor images for input to a machine learning network used to measure signals from a device under test. [Background technology]
[0002] U.S. Patent Publication No. 2022 / 0373598, filed May 18, 2022, describes a system and method for applying a trained machine learning system to make predictions of measurements of signals received from a device under test (DUT). One example of a signal measurement is the Transmitter and Dispersion Eye Closure Quaternary (TDECQ) when the DUT is, for example, an optical transceiver. The input data to the machine learning system described in U.S. Patent Publication No. 2022 / 0373598, both for training and run-time, is in the form of tensor images, which are constructed using graphical representations of specific short pattern segments within a longer waveform acquired from the DUT.
[0003] The specification of U.S. Patent No. 11,923,895 (Patent Document 2), filed on March 24, 2022, describes using a trained machine learning system to predict optimal tuning parameters for an optical transceiver (DUT).
[0004] U.S. Patent Publication No. 2023 / 0398694, filed May 19, 2023, describes a system and method that utilizes a trained machine learning system to predict optimal locations of tuning elements in a tunable device under test (e.g., a cavity filter, etc.). Input data to the machine learning system described in U.S. Patent Publication No. 2023 / 0398694, both for training and run-time, is in the form of tensor images, which are constructed using graphical representations of S-parameters or other parameters that characterize the DUT.
[0005] Tensors are an efficient way to provide information to machine learning systems at training or runtime. U.S. Patent Application No. 18 / 510,234, filed November 15, 2023, describes a system and method for creating a highly efficient 3D RGB tensor by combining three-channel reference parameters (e.g., S-parameters or other types of data). [Prior art documents] [Patent documents]
[0006] [Patent Document 1] US Patent Publication No. 2022 / 0373598 [Patent Document 2] U.S. Patent No. 1,192,3895 [Patent Document 3] US Patent Publication No. 2023 / 0398694 [Patent Document 4] U.S. Patent Application Serial No. 18 / 510,234 [Non-patent literature]
[0007] [Non-Patent Document 1] "Tektronix oscilloscope" introduction site, Tektronix, [online], [searched May 22, 2024], Internet<https: / / www.tek.com / ja / products / oscilloscopes> Summary of the Invention [Problem to be solved by the invention]
[0008] As mentioned above, tensor images are an efficient way to provide information to machine learning systems, both for training and for runtime. Therefore, to efficiently construct the tensor images that users intend, an improved user interface is needed to input the data to construct the tensor images. [Means for solving the problem]
[0009] According to embodiments of the present disclosure, numerous applications can be constructed using waveforms or S-parameters as input data to deep learning networks in machine learning applications. Broadly speaking, embodiments of the present disclosure include test and measurement instruments and associated user interfaces and methods for constructing tensor images for input to deep learning networks to make predictions in test and measurement applications. [Brief description of the drawings]
[0010] [Figure 1] FIG. 1 illustrates an embodiment of a test and measurement device. [Diagram 2] Figure 2 shows an embodiment of a user interface for the tensor builder to allow the user to select the tensor format and associated parameters. [Diagram 3] Figure 3 shows another embodiment of a user interface for the tensor builder to allow the user to select the tensor format and associated parameters. [Figure 4] Figure 4 shows another embodiment of a user interface for the tensor builder to allow the user to select the tensor format and associated parameters. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0011] FIG. 1 illustrates an embodiment of a test and measurement instrument that provides one or more user interfaces. The embodiment may be a test and measurement instrument having one or more processors, which may be configured to execute code (programs) that cause the one or more processors to perform various tasks. The machine learning system may take the form of a programmed model running on the one or more processors. FIG. 1 illustrates an embodiment of a test setup in the case of an optical transmitter (Tx) 30 as a device under test (DUT). The test setup includes a test and measurement system that may include a test and measurement instrument 10, such as an oscilloscope. The test and measurement instrument 10 receives signals from the DUT 30 via a test probe 32. In the case of an optical transmitter, the test probe 32 typically consists of a test optical fiber coupled to an optical-to-electrical converter (O / E) 34 that provides signals to the test and measurement instrument via one or more ports 14. Two ports may be used for differential signaling and one port may be used for single-ended signaling. The signals are sampled and digitized by test and measurement instrument 10 into waveform data (hereafter simply referred to as "waveform"). This waveform data may be stored in memory 20. A clock recovery unit (CRU) 18 may recover a clock signal from the data signal if test and measurement instrument 10 comprises, for example, a sampling oscilloscope. In a real-time oscilloscope, software clock recovery may be used.
[0012] The test and measurement instrument 10 includes one or more processors, represented by processor 12, a memory 20, and a user interface 16. The memory 20 stores executable instructions in the form of code (programs) that, when executed by the processor, cause the processor to perform tasks. The user interface 16 of the test and measurement instrument 10 allows a user to interactively operate the test and measurement instrument 10, such as entering settings, configuring tests, and the like. The test and measurement instrument 10 may also include a reference equalizer and analysis module 24. The user interface 16 may include a display and one or more controls that allow a user to configure the test and measurement instrument. The controls may include one or more switches, knobs, or other types of input devices, and may also include touch buttons on the display.
[0013] Embodiments of the present application utilize machine learning in the form of a machine learning network 26, such as a deep learning neural network. The machine learning network 26 may include either a processor programmed with the machine learning network as part of the test and measurement instrument 10, or a processor to which the test and measurement instrument has access via a connection. As the capabilities and processors of the test and measurement instrument 10 evolve, one or more processors, such as processor 12, may include both.
[0014] FIG. 2 illustrates one embodiment of a user interface according to some embodiments of the present disclosure. One or more processors render the user interface on a display. The user interface, in this example, displays a tensor builder menu tab 38, within which a deep learning network input panel 40 is displayed, allowing the user to select from various options for the format of the tensor. As shown, the tensor formats include a waveform (WFM) vector, a waveform image, and an S-parameter image. The input panel 40 of the user interface allows the user to select to create a waveform vector to be used as an input to the application of the deep learning neural network. In the particular example of FIG. 2, the user has selected a waveform vector, and in response, among other display areas (panels), a short pattern definition input panel 42 is displayed to allow the user to define a short pattern input. As used herein, the term "short pattern" comprises a short pattern waveform segment, as described in the example of U.S. Patent No. 6,399,323 referenced above.
[0015] The tensor consists of a one-dimensional waveform data vector, where multiple user-defined short pattern waveform sequences are concatenated to form a one-dimensional waveform data vector. The user interface also provides an averaging switch 52 that, when turned on, allows all similar short patterns found in the waveform to be averaged together, rather than concatenated into one.
[0016] The bar graph input panel 44 allows the user to convert the bar graphs into waveform segments and concatenate them into a waveform data tensor vector. The bar graph input panel 44 includes panels 46, 48, and 50. The user input parameter panel 46 provides options for user-defined input parameters that may be utilized for a particular test measurement application, such as temperature, fiber length, or any other desired parameter. The measurement noise panel 48 allows the user to tailor the creation of noise bar graphs that are created by making noise measurements on the waveform. This allows the user to have the option to control low pass filtering on each defined group of short patterns. The user can also select from among the measurements as shown in the measurements panel 50 that are used to create the bar graphs. The measurement options may be different than those shown in the example.
[0017] FIG. 3 illustrates a user interface when a user selects a waveform image as a tensor format to create a waveform image to be used as input to the application of a deep learning network. In the user interface in this situation, there is an image builder panel 60. The user can select from a two-dimensional or three-dimensional waveform image. In a first embodiment, the tensor is constructed from a two-dimensional waveform data image whose Z-axis contains histogram data of how many counts each point occurred in the short pattern waveform data. In a second embodiment, the user can also select a three-dimensional image as the image, whose Z-axis represents the amplitude of each waveform, and many more waveforms can be accommodated in multiple rows in the XY plane. This makes the image more efficient and allows more data to be processed. All of the user-defined short pattern waveform segments can be accommodated in one image.
[0018] As mentioned above, bar graphs can also be created and incorporated into the tensor image.
[0019] A tensor image can be an R (red), G (green), or B (blue) image, with a different waveform in each of these color channels. Which waveform goes into which color channel is controlled in the background by the application, but you can add a menu of controls to the menu display that allow you to specify which data goes into which color channel.
[0020] Another feature is that bar graphs may be converted into waveform segments and concatenated to the waveform data tensor vector using the bar graph user input panel 44, also shown in FIG. 2. Bar graphs may be created based on user defined input parameters such as temperature, fiber length or any other desired parameters as shown in panel 46, also shown in FIG. 2. The user may choose to include noise bar graphs by specifying a noise measurement on the waveform as shown in panel 48. The user has the option to control low pass filtering on each group of short patterns that are defined. Bar graphs may also be created from a list of user selected measurements present in the particular application that contains this tensor builder menu, as shown in panel 50.
[0021] The interface includes a waveform image builder panel 60 that is used to control various ways to structure the data that goes into the image tensor. Note that what is shown is merely an example and there may be additional controls in the menu. This user interface allows the user to control how the test and measurement instrument structures the data in the waveform image.
[0022] Figure 4 shows an embodiment of the Tensor Builder menu when the user selects to create an S-parameter image as input to a deep learning neural network. The tensor is composed of a three-dimensional S-parameter data image, which is an XY image composed of real or imaginary parts, or amplitude and phase parts. Each row of the XY image represents the real, imaginary, amplitude or phase of one of the S-parameters in the set. Bar graphs can also be created as described above and incorporated into the tensor image.
[0023] The tensor image may be constructed from an RGB image, with three different reference S-parameter results measured from the DUT, each in one of the RGB color channels. The test and measurement instrument may also allow the user to view an animation of the RGB tensor image, allowing the user to more easily see how the tensor image is structured. The user interface may display a user input parameter panel 46 within the bar graph user input panel 44, allowing the user to specify the bar graph to be converted from either measurements, user-defined input parameters such as temperature, or other. Note that the measurement noise panel 48 and measurement value panel 50 shown in Figures 2 and 3 are not used in the tensor builder menu for S-parameter images in Figure 4.
[0024] The S-parameter image user interface includes an S-parameter image builder panel 70 that is used to control various ways to configure the structure of the data that goes into the S-parameter image tensor. Additional operations may be available in the menus in this user interface beyond those shown below.
[0025] The above user interface elements are described in more detail below. As noted above, the Tensor Builder menu tab 38 comprises the top level menu in this embodiment and contains the menu items necessary to build either a waveform vector, a waveform image, or an S-parameter image for use as input to the application of a deep learning neural network. This may include one or more of the following menu items:
[0026] The Deep Learning Neural Network Input Panel 40 has three options for the format of the tensor data that is input to the neural network. The user interface does not present to the user of this application the algorithms for creating these tensors.
[0027] One of the three options is a waveform vector, which allows the creation of a one-dimensional waveform vector containing many short pattern waveform segments, discarding a large amount of the waveform and keeping only the parts that are important for training the network. The waveform vector may further include a bar graph containing the input parameters, which are converted into waveform segments that represent the bar graph.
[0028] The second of the three options, Waveform Imaging, specifies that the test and measurement instrument will create 2D or 3D image tensors that contain bar graphs and multiple short pattern waveform data segments. These waveform images can be monochrome or RGB.
[0029] The last of the three options is S-parameter images. This option takes the S-parameters acquired from the device under test (DUT) and builds RGB or monochrome images from them to use as input to the neural network. These S-parameter images can also use other parameters such as time domain S-parameters, Z-parameters (open circuit or impedance parameters), H-parameters (hybrid parameters), and X-parameters (generalized versions of S-parameters).
[0030] The Short Pattern Definition Entry Panel 42 for Waveform Images shown in Figures 2 and 3 allows the user to define the code sequence for each of the short pattern waveform segments contained in the image tensor. There are three groups of pattern formats. There are up to eight short patterns that may be used in each group. It supports signals from NRZ to PAM3, PAM4, and PAM8 formats. The number representing the PAM level number defines the signal level of the signal per unit interval. Thus, the levels for NRZ will be 0 and 1, the levels for PAM3 will be 0, 1, 2, etc. The Level Pattern Panel 64 contains the number of consecutive unit intervals where the signal level remains at the same level for each unit interval in the sequence. However, the user can define any kind of pattern sequence in this menu tab 38 as needed. The Pulse Pattern Panel 66 definitions are intended to create pulse patterns where several unit intervals in a sequence are set to the same level except for one unit interval that is set to a different level. The user can define up to eight different pulse patterns and whether each one is enabled for inclusion in the tensor image. In this example, checking the check box associated with each pulse pattern definition enables the corresponding pulse pattern for inclusion in the tensor image. The other Pattern Panel 68 allows the user to define up to eight additional short pattern waveforms for inclusion in the tensor input data.
[0031] In these menu panels, the "number of signs" specifies the number of signal unit intervals (UIs) contained in the total length of the short pattern. Note that this menu item may be optional, since the number of UIs for each pattern can be determined separately by the number of unit intervals the user enters into the edit boxes that define the short pattern. As mentioned above, the user interface has an averaging switch, such as 52 in Figure 2. When the averaging switch is off, the application searches for all individual instances of all included short patterns and places them in a tensor according to a set of internal rules for constructing the selected pattern. Although not shown here, other versions of the Tensor Builder may have menus for selecting additional rules for constructing the tensor. When the averaging switch is set to on, all similar patterns found in the waveform are averaged together before the Tensor Builder allows them to be placed in a tensor according to the Tensor Builder's internal rules. Some of these rules are determined by the menu system of this Tensor Builder. Some of these rules may be internal to a specific machine learning application, while others may be defined in the future by adding new menus to the Tensor Builder menu system.
[0032] The menu of the image builder panel in Figures 3 and 4 is displayed when the user selects a waveform image or an S-parameter image in the menu of the deep learning neural network input panel 40. The menu of the image builder panel has different operation menu items depending on whether the user selects a waveform image or an S-parameter image.
[0033] If the user selects a waveform image, one of the options is Dimensions, which allows the user to select the format of the image. If two-dimensional (2D) is selected, the test and measurement instrument will build a two-dimensional tensor image where the XY plane is the time vs. amplitude for each waveform pattern, and the Z axis is the sample count for all instances of the various patterns contained in the waveform. However, if a three-dimensional (3D) tensor image is selected, more waveforms and data will be stored in the image, a longer record length of the patterns will be stored, and the vertical resolution of the data input will be improved. The time and amplitude of the waveforms in each pattern will be stored in the XZ plane. Each waveform will be stored in a row in the XY plane of the image. These images are also RGB, with internal control over which waveform data goes into each color channel. This Image Builder menu may incorporate additional menu items to control which waveforms go into which color channels.
[0034] The Resolution Enhancement panel of the Image Builder panel 60 of FIG. 3 allows the user to determine which vertical waveform resolution enhancement algorithms are applied when creating the waveform image. The algorithms include "Single Point Compression," which controls the remapping of each individual point in the waveform with respect to the minimum and maximum of all instances of the particular short pattern in which the point resides. This increases the gain of low level signals and reduces the gain of high level signals, improving the vertical resolution of low level signals. In some embodiments, additional menu items, not shown, may be displayed to control the details of the algorithm. The "Linear Segment" algorithm creates a piecewise linear segment gain vector that is multiplied to the short pattern waveform. This algorithm is intended to be applied primarily to pulse waveforms, and is intended to increase the gain of the low level decaying portions of the pulse as a function of time. This boosts the low level amplitude portions of the pulse, improving the resolution of the pulse in the tensor image. If this option is set to "None," no gain compression or expansion algorithms are used before placing the short pattern into the tensor.
[0035] The "row spacing" selection determines the spacing of the various waveform segments across the rows of the XY image plane. When "adjacent" is selected, each waveform pattern or folded pattern segment is placed in adjacent rows across the XY plane of the image. When "space by group" is selected, blank rows are placed between each waveform pattern row or adjacent groups of waveforms on the XY plane. For example, all instances of one pattern type are placed adjacent to each other. However, there is a space of several blank rows between one short pattern group and the next short pattern group. The number of rows is equal to the dimension of the convolution filter in the deep learning input layer.
[0036] Selecting "Space each pattern" will space each pattern in the waveform a distance equal to the dimensions of the convolution filter, or another menu of operations may be added (not shown) to determine the spacing.
[0037] When the user selects an S-parameter image, the menu of the image builder 70 displays the following: First, the characteristics (Dimension) include the following items: S1P: a set of 1-port S-parameters, S2P: a set of 2-port S-parameters, S3P: a set of 3-port S-parameters, S4P: a set of 4-port S-parameters, RI: uses real and imaginary amplitudes in the frequency domain for each S-parameter waveform, MP: uses logarithmic (log) amplitude and phase in the frequency domain for each S-parameter waveform, Frequency: represents S-parameters as either real and imaginary or amplitude and phase in the frequency domain, Time: represents time-domain S-parameters as a single waveform of amplitude versus time.
[0038] Second, the menus in the Image Builder Panel 70's Include S-Parameters panel allow you to enable only those S-parameters you wish to enter into the tensor image. Check the box next to each S-parameter to include it in the tensor image. Leaving the box blank will exclude the corresponding S-parameter from the tensor image. The number of check boxes and S-parameters available will vary depending on the selection in the Properties menu. The Row Spacing menu specifies how S-parameter waveforms are spaced in rows in the XY plane of the tensor image, either in the time domain or the frequency domain. Selecting Adjacent will result in no space being placed between the individual S-parameter data vectors that are placed in rows of the image in the XY plane of the tensor image. Space by Group specifies that the tensor image will have an empty set of rows between each pair of frequency domain S-parameter waveforms. The spacing rules are determined by an internal algorithm. The Resolution Enhancement menu applies the same rules as described above for improving the vertical resolution of the S-parameter data.
[0039] The bar graph input panel 44 menus of Figures 2-4 allow the definition of various types of bar graphs to be incorporated into the tensor image, however, as noted above, the measurement noise panel 48 and measurement values panel 50 shown in Figures 2 and 3 are not used in the tensor builder menu for S-parameter images of Figure 4.
[0040] The User Input Parameters panel 46 allows the user to define their own parameter names to be included in the metadata labels for training the neural network. In the example menu shown, four such user input parameters can be created, although there may be more or less than four.
[0041] The measurement noise panel 48 allows the user to specify how noise is measured on the short pattern waveform and then converted to a bar graph and included in the tensor. The test and measurement instrument 10 removes the noise on the measured spectrum from the short pattern before placing the short pattern into the tensor image. This improves the accuracy of the neural network in identifying features of interest in the pattern.
[0042] Another option is related to the filtering in the Measurement Noise Panel 48. When a "Low Pass Level" filter is applied to a constant amplitude (Level) waveform, as much noise as possible is removed. The filtered waveform is used to create the Level (constant amplitude waveform) image. The cutoff frequency can be fixed or controlled by menu or PI (Programming Interface) commands. The typical cutoff frequency of this filter is 1 GHz. The user can change this frequency at will. When a "Low Pass Pulse" filter is applied to a pulse short pattern waveform, as much noise as possible is removed while the main shape of the pulse waveform is preserved. The filtered waveform is used to create the Pulse Waveform image. The cutoff frequency can be fixed or controlled by the user by menu or PI commands when training the neural network. A "Low Pass Other" filter is applied to the "Other" pattern and its cutoff frequency can be changed at will by the user.
[0043] The measurements panel 50 allows the user to select measurements. The test and measurement instrument 10 converts the selected measurements into a bar graph for each selected measurement, which is then included in the tensor. Examples of measurements include, but are not limited to, TDECQ, Optical Modulation Amplitude (OMA), Level Separation Mismatch Ratio (RLM), Extinction Ratio, Bit Error Rate (BER), and other measurements.
[0044] In summary, embodiments of the disclosed technology present a novel user interface menu for the Tensor Builder. The Tensor Builder defines how to construct input tensor data for training a neural network and for use in a final run-time environment. The embodiments incorporate several novel features, such as the ability to include parameter bar graphs in the tensor image. The parameters may be user defined, measured noise or other measurements. Additionally, some embodiments may have parameters converted to bar graphs represented by segments of a waveform vector. The user interface allows the selection of three different input tensor formats: waveform vector, waveform image, and S-parameters. The user interface has several controls for creating short pattern waveform segments that go into the tensor, and an image builder menu that controls how the tensor image is assembled.
[0045] Aspects of the disclosed technology can operate on specially created hardware, firmware, digital signal processors, or specially programmed general-purpose computers, including processors that operate according to programmed instructions. The term "controller" or "processor" in this application contemplates microprocessors, microcomputers, ASICs, and dedicated hardware controllers, among others. Aspects of the disclosed technology can be implemented in computer-available data and computer-executable instructions, such as one or more program modules, executed by one or more computers (including a monitoring module) or other devices. Generally, program modules include routines, programs, objects, components, data structures, and the like, which, when executed by a processor in a computer or other device, perform particular tasks or implement particular abstract data formats. The computer-executable instructions may be stored in computer-readable storage media, such as hard disks, optical disks, removable storage media, solid-state memory, RAM, and the like. As will be appreciated by those skilled in the art, the functionality of the program modules may be combined or distributed as desired in various embodiments. Furthermore, such functionality may be embodied in whole or in part in firmware or hardware equivalents, such as integrated circuits, field programmable gate arrays (FPGAs), etc. Certain data structures may be used to more effectively implement one or more aspects of the disclosed techniques, and such data structures are considered to be within the scope of the computer-executable instructions and computer-usable data described herein.
[0046] The disclosed aspects may, in some cases, be implemented in hardware, firmware, software, or any combination thereof. The disclosed aspects may also be implemented as instructions carried by or stored on one or more computer-readable media that may be read and executed by one or more processors. Such instructions may be referred to as computer program products. A computer-readable medium as described herein means any medium that can be accessed by a computing device. By way of example, and not limitation, computer-readable media may include computer storage media and communication media.
[0047] Computer storage media means any medium that can be used to store computer-readable information. By way of example and not limitation, computer storage media may include random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital video disc (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, and any other volatile or non-volatile removable or non-removable media implemented in any technology. Computer storage media excludes signals themselves and transitory forms of signal transmission.
[0048] A communication medium refers to any medium capable of communicating computer readable information. By way of example, and not limitation, communication media may include coaxial cables, fiber optic cables, air, or any other medium suitable for communicating electrical, optical, radio frequency (RF), infrared, acoustic, or other types of signals. Working Example
[0049] In the following, examples useful for understanding the technology disclosed in the present application are presented. The embodiment of the technology may include one or more of the examples described below and any combination thereof.
[0050] Example 1 is a test and measurement instrument comprising one or more ports for enabling the test and measurement instrument to receive data from a device under test (DUT), a connection to a machine learning network, a display configured to display a user interface, one or more controls for enabling the test and measurement instrument to receive input from a user, and one or more processors, the one or more processors configured to execute a program that causes the one or more processors to perform the following processes: rendering a menu on the display showing options for various tensor formats; receiving a user selection from the one or more controls identifying a tensor of the selected format; constructing a tensor of the selected format from the data from the DUT; and sending the tensor of the selected format to the machine learning network.
[0051] Example 2 is the test and measurement instrument of Example 1, wherein the data includes one or more of waveform data and bar graph data.
[0052] Example 3 is a test and measurement instrument of either example 1 or 2, wherein the program that causes the one or more processors to render on the display a menu showing options of various types of tensors causes the one or more processors to display options including waveform vectors, waveform images, and S-parameter images.
[0053] Example 4 is the test and measurement instrument of any of Examples 1 to 3, wherein the one or more processors: receiving a user selection for selecting a waveform vector; rendering a second user interface having user definable options for short pattern input, bar graph input, measurement noise or measurement value; The one or more processors may be further configured to execute a program that causes the one or more processors to:
[0054] Example 5 is a test and measurement instrument of Example 4, wherein the program that causes the one or more processors to process the construction of a tensor of the selected type from the data from the DUT causes the one or more processors to process the construction of a one-dimensional waveform vector by concatenating a plurality of user-defined short pattern waveform sequences.
[0055] Example 6 is a test and measurement instrument of Example 5, wherein the program that causes the one or more processors to render the user interface causes the one or more processors to render a user interface having an option that allows the user to average the short pattern waveform sequences before concatenation.
[0056] Example 7 is a test and measurement instrument of any of Examples 1 to 6, wherein the one or more processors are further configured to execute a program that causes the one or more processors to receive a user selection to select a waveform image and render a second user interface having options for the user to define a short pattern input, an image builder, a bar graph input, measurement noise, or a measurement value.
[0057] Example 8 is the test and measurement instrument of Example 7, wherein the program that causes the one or more processors to construct a tensor of the selected type includes a program that causes the one or more processors to construct a two-dimensional waveform image from the short pattern waveform sequence including histogram data of how many counts occur at each point in the short pattern waveform sequence.
[0058] Example 9 is the test and measurement instrument of Example 7, wherein the program that causes the one or more processors to process the construction of a tensor of the selected type causes the one or more processors to process the construction of a three-dimensional image having a Z-axis, the Z-axis representing the amplitude of each waveform included in the three-dimensional image.
[0059] Example 10 is the test and measurement instrument of any one of Examples 1 to 9, wherein the program causing the one or more processors to perform a process of constructing a tensor of the selected type from the data from the DUT comprises: receiving a user selection for selecting an S-parameter image; Rendering a second user interface having options for a property, an image builder, and a bar graph input; The one or more processors are caused to perform the above.
[0060] Example 11 is a test and measurement instrument of Example 10, wherein the program that causes the one or more processors to process the construction of a tensor of the selected type from the data from the DUT causes the one or more processors to process the construction of an S-parameter image having rows in the XY image, each of the rows in the XY image representing either the real or imaginary part of the S-parameter, or the amplitude or phase of the S-parameter.
[0061] Example 12 is the test and measurement instrument of example 10, wherein the program that causes the one or more processors to process the construction of the tensor of the selected type from the data from the DUT causes the one or more processors to process the construction of an S-parameter image including red-green-blue channels, each having a different S-parameter.
[0062] Example 13 is a method for providing a user interface, comprising: Rendering a user interface on a display of the test and measurement device to display a selection of different formats of tensors on the user interface; receiving a user selection from one or more interfaces of the test and measurement instrument specifying a selected type of tensor; constructing a tensor of the selected form from the data from the DUT; Sending the tensor in the selected format to a machine learning network; Equipped with.
[0063] Example 14 is the method of example 13, wherein rendering the user interface on the display and displaying options of various types of tensors on the user interface includes displaying options including waveform vectors, waveform images, and S-parameter images.
[0064] Example 15 is a method of any of Examples 13 or 14, further comprising a process of receiving a user selection to select a waveform vector, and a process of rendering a second user interface having options to enable the user to define a short pattern input, a bar graph input, a measured noise, and a measurement value.
[0065] Example 16 is the method of example 15, wherein constructing a tensor of the selected type from the data from the DUT includes constructing a one-dimensional waveform vector by concatenating a plurality of user-defined short pattern waveform sequences.
[0066] Example 17 is a method of any of Examples 13 or 14, further comprising a process of receiving a user selection of a waveform image and rendering a second user interface having options for the user to define a short pattern input, an image builder, a bar graph input, measurement noise or a measurement value.
[0067] Example 18 is the method of example 17, wherein the process of constructing a tensor of the selected type from the data from the DUT includes a process of constructing a two-dimensional waveform image including histogram data from a short pattern waveform sequence.
[0068] Example 19 is a method of either example 13 or 14, further comprising a process of receiving a user selection of an S-parameter image and a process of rendering a second user interface having options for properties, image builder and bar graph input.
[0069] Example 20 is the method of Example 19, wherein the process of constructing an S-parameter image includes a process of constructing an S-parameter image having rows in an XY image, each of the rows in the XY image representing either the real or imaginary part of the S-parameter, or the magnitude or phase of the S-parameter.
[0070] In addition, the present description refers to specific features. It should be understood that the disclosure herein includes all possible combinations of those specific features. When a specific feature is disclosed in connection with a particular aspect or embodiment, that feature can also be used in connection with other aspects and embodiments, to the extent possible.
[0071] Furthermore, when the application refers to a method having two or more defined steps or processes, those defined steps or processes may be performed in any order or simultaneously, unless the context precludes such a possibility.
[0072] All features disclosed in the specification, claims, abstract and drawings, and all steps in any disclosed method or process, may be combined in any combination, except where at least some of such features or steps are mutually exclusive combinations. Each feature disclosed in the specification, abstract, claims and drawings may be replaced by an alternative feature serving the same, equivalent or similar purpose, unless expressly stated otherwise.
[0073] For convenience of illustration, specific embodiments of the invention have been illustrated and described, but it will be understood that various modifications can be made therein without departing from the spirit and scope of the invention. Accordingly, the invention should not be limited, except as by the appended claims. [Explanation of symbols]
[0074] 10 Test and measurement equipment 12 processors 14 Ports 16 User Interface 18 Clock Recovery Unit 20 Memory 24 Reference Equalizer and Analysis Module 26 Machine Learning Networks 30 Device under test (optical transmitter) 32 Test Fiber Optics / Probes 34 Photoelectric converter 38 Tensor Builder Menu Tab 40 Deep Learning Neural Network Input Panel 42 Short Pattern Definition Input Panel 44 Bar Graph Input Panel 46 User Input Parameter Panel 48 Measurement Noise Panel 50 Measurements Panel 60 Image Builder Panel 64 Level Pattern Panel 66 Pulse Pattern Panel 68 Other Pattern Panels 70 Image Builder Panel
Claims
1. 1. A test and measurement device comprising: one or more ports for enabling the test and measurement instrument to receive data from a device under test (DUT); A connection to the machine learning network; and a display configured to display a user interface; one or more controls for enabling the test and measurement instrument to receive input from a user; One or more processors Equipped with The one or more processors rendering on said display a menu showing a selection of various formats of tensors; receiving a user selection identifying a tensor of a selected type from the one or more controls; constructing a tensor of the selected type from the data from the DUT; sending the tensors in the selected format to the machine learning network; a test and measurement instrument configured to execute a program that causes the one or more processors to:
2. 2. The test and measurement instrument of claim 1, wherein the program causing the one or more processors to render on the display a menu showing options of various formats of tensors causes the one or more processors to display options including waveform vectors, waveform images, and S-parameter images.
3. the one or more processors receiving a user selection of a waveform vector as the format of the tensor; rendering a second user interface having options for allowing the user to define short pattern input, bar graph input, measurement noise or measurement value; 2. The test and measurement instrument of claim 1, further configured to execute a program that causes the one or more processors to:
4. 4. The test and measurement instrument of claim 3, wherein the program that causes the one or more processors to construct a tensor of the selected type from the data from the DUT causes the one or more processors to construct a one-dimensional waveform vector by concatenating a plurality of user-defined short pattern waveform sequences.
5. the one or more processors receiving a user selection of a waveform image as the format of the tensor; rendering a second user interface having options for the user to define short pattern input, image builder, bar graph input, measurement noise or measurements; 2. The test and measurement instrument of claim 1, further configured to execute a program that causes the one or more processors to:
6. 6. The test and measurement instrument of claim 5, wherein the program that causes the one or more processors to construct the tensor of the selected type includes a program that causes the one or more processors to construct a two dimensional waveform image from the short pattern waveform sequence including histogram data of how many counts occur at each of the points in the short pattern waveform sequence.
7. 6. The test and measurement instrument of claim 5, wherein the program that causes the one or more processors to construct a tensor of the selected type causes the one or more processors to construct a three-dimensional image having a Z-axis, the Z-axis representing the amplitude of each waveform included in the three-dimensional image.
8. a program for causing the one or more processors to process the selected type of tensor from the data from the DUT, receiving a user selection of an S-parameter image as the format of the tensor; rendering a second user interface having options for properties, an image builder, and a bar graph input; 2. The test and measurement instrument of claim 1, further comprising one or more processors configured to perform the steps of:
9. 9. The test and measurement instrument of claim 8, wherein the program that causes the one or more processors to construct the selected type of tensor from the data from the DUT causes the one or more processors to construct an S-parameter image having rows in the XY image, each of the rows in the XY image representing either a real or imaginary part of the S-parameter, or an amplitude or phase of the S-parameter.
10. Rendering a user interface on a display of the test and measurement instrument to display a selection of different formats of tensors on the user interface; receiving a user selection from one or more interfaces of the test and measurement instrument specifying a selected type of tensor; constructing a tensor of the selected type from the data from the DUT; Sending the tensor in the selected format to a machine learning network; A method for providing a user interface comprising:
Citation Information
Patent Citations
Optical transmitter tuning using machine learning and reference parameters
US11923895B2
Short pattern waveform database based machine learning for measurement
US20220373598A1
Automated cavity filter tuning using machine learning
US20230398694A1
Methods for 3D tensor builder for input to machine learning
US20240169210A1
