ADAPTIVE CLOSED-LOOKING GENERATIVE FILTER WITH NEGATIVE NETWORK

Abstract

A test and measurement system contains a plurality of nodes, wherein one or more nodes comprise sensors configured to receive signals, one or more generative adversarial network (GAN) models, one or more signal generators configured to generate and transmit signals, and one or more processors configured to execute code that causes the one or more processors to receive a request for a signal with a signal profile from one node of the plurality of nodes, send the signal profile and the request to one of the one or more GAN models, receive a matching signal profile that matches the signal profile, and transmit the matching signal profile to one of the one or more signal generators to generate and transmit a matching signal.

Description

CROSS-REFERENCE TO RELATED REGISTRATIONS

[0001] This disclosure is a non-provisional application and claims the benefits of the provisional US application No. 63 / 699,259 entitled “CLOSED LOOP ADAPTIVE GENERATIVE ADVERSARILY NETWORK FILTER”, which was filed on September 26, 2024, and the disclosure of which is incorporated herein in full by reference. AREA OF TECHNOLOGY

[0002] This disclosure relates to test and measurement equipment, and in particular to artificial intelligence for generating waveforms for test and measurement equipment and environments. BACKGROUND

[0003] Waveforms, the curve of a signal over time, have many different functions in electronics testing. Devices under test (DUTs) generate signals that form waveforms and allow for analysis of the devices' performance. The DUT can generate signals in response to a signal applied to it. Some test equipment, such as arbitrary waveform generators and arbitrary function generators, generate and apply these signals.

[0004] In many cases, it is necessary to update the embedded software in these devices so that these waveform generators can produce different types of waveforms with specific characteristics. Furthermore, some aspects of the generated signals may not account for the effects of an unknown channel or environment on the signal.

[0005] The ability to replicate signals from DUTs and generate signals that exhibit specific environmental and channel conditions can offer a level of flexibility and precision that was previously impossible. BRIEF DESCRIPTION OF THE DRAWINGS Fig. Figure 1 shows a system diagram of an artificial intelligence-based system for testing and measurement. Fig. Figure 2 shows a flowchart for an embodiment of a closed-loop test and measurement system for optimization. Fig. Figure 3 shows an embodiment of a sensor node with a test and measuring device. Fig. Figure 4 shows an embodiment of a node during a receiving process. Fig. Figure 5 shows an embodiment of a node during a transmission process. Fig. Figure 6 shows a flowchart of an embodiment of a method for generating a suitable signal. Fig. Figure 7 shows an embodiment of an artificial intelligence-based system for testing and measurement in a specific environment. DESCRIPTION

[0006] The embodiments described here are test and measurement systems comprising a network of nodes, one or more central servers, and a database. Some embodiments may include an interface for a large language model (LLM) to the database. Some embodiments have nodes that contain test and measurement devices. Most, if not all, nodes have processing capabilities that allow the node to operate a generative adversarial network (GAN). The nodes with processing capabilities may also be configured to employ federated learning.

[0007] Fig. Figure 1 shows an embodiment of the test and measurement system. The system 10 comprises a network of nodes, such as node 12, one or more servers, such as server 14, and one or more databases, such as database 22. Server 14 can be referred to as a "centralized server," meaning that server 14 communicates with many, if not all, nodes 12, regardless of server 14's physical location. Server 14, as well as any nodes 12 comprising a test and measurement system, can include one or more processors 16, a memory 24, a port 18 that allows server 14 or node 12 to communicate with a device under test (DUT), and a machine learning system 20. The machine learning system 20 can include a processor separate from a main processor, programmed to perform a machine learning process, both of which are represented by processor 16.

[0008] For the purposes of this discussion, the term “artificial intelligence” (AI) refers to machine learning, i.e., a neural network or other machine learning architecture that has undergone supervised or unsupervised training to produce an output, as well as to “generative AI.” Generative AI may involve a large language model (LLM) and include a transformer. Generative AI may also refer to GANs, which are most likely to operate without a transformer.

[0009] GANs typically consist of two neural networks. One network, the generative network or generator, produces an output that mimics an input. The other network, the discriminative network or discriminator, evaluates the output and determines whether it truly mimics the input. The two networks compete until the discriminator "loses" because the output mimics the input so closely that the discriminator can no longer distinguish it, which is usually detected through an error process. The term "GAN model" or "GAN" as used here refers to a pair of neural networks. Multiple GAN models, i.e., multiple pairs of neural networks, can exist in the AI.

[0010] System 10 of Fig. 1 uses federated learning, where the base or "bootstrap" model of the GAN is implemented on most, if not all, parts of the system. Some nodes 12 in system 10 may be transport nodes or other types of nodes that do not handle inputs and outputs. These nodes forward data or perform other tasks unrelated to receiving or transmitting signals outside of system 10. In this discussion, nodes that receive signals are referred to as receive nodes, and nodes that transmit signals outside the system are referred to as transmit nodes. Many nodes can be both receive and transmit nodes.

[0011] For example, to illustrate this point: Node 12 in system 10 can be a sensor node receiving a signal. Since a waveform results from signal extension, the terms "signal" and "waveform" are used interchangeably. The signal can be from a device under test (DUT) or a "threat signal," such as an incoming radar signal. Node 12 receives the signal and generates a response. This response can include replicating the signal using a GAN and deriving the environment and channel properties from the replication process.

[0012] Fig. Figure 2 shows an embodiment of the various elements of the AI ​​components within AI component 20 and their interactions. As mentioned earlier, AI component 20 can reside on server 14 and on multiple nodes within system 10. Fig. The process has a transmission side 30 and a sending side 40. On the receiving side 30, the input signal data received by a sensor becomes training data 34, which may be coupled with the validation data 32. The validation data 32 can be received through direct data samples from the DUT environment (referred to here as online) or through file uploads (referred to here as offline). The training data 34 and the context information 36 become inputs for the GAN learning discriminator 42. The context information 36 can include the environment in which the receiving node is located, specifics of the node, such as the node's processing capabilities, its role, etc., and the type of signal received, e.g., from a DUT or an external source. The GAN learning discriminator 42 receives the training data 34, the context information 36, and the parameters 39 extracted from the validation data 32.The GAN learning discriminator 42 then sets how the simulations of the received signal should look.

[0013] The data received from the receiving side 30 can pass through the parameter extraction module 38 and be stored in the waveform library 44. The incoming data 32 and / or 34 can be digitally attached or encoded, e.g., as a watermark. The waveform library 44 can also contain contextual information 36. The GAN generator 48 can receive the received waveform from the waveform library 44 if the replication / transmission process is sequential with the received signal. If the received signal does not require a response, the system can store the signal for later use, relying on the encoding for systematic, fast retrieval.

[0014] If a response is required, the GAN generator 48 can use a stored waveform or, in the case of a sequential response, the actual response waveform. The GAN generator 48 generates a candidate signal based on the incoming, received signal. The candidate signal generated by the GAN generator 48 can be a replica of the incoming signal, a replica with the incoming signal but with variations in some of its features, or an inversion of the signal, as used in electronic warfare to "cancel" the incoming signal. The features of the signal can also be referred to as the signal profile. The generated signal is then compared in the comparator 47 with the validation data 32 of the received signal, which has undergone parameter extraction 38 without modification.

[0015] For comparison, an error measurement such as the mean squared error (MSE) or another error measurement can be used. An optimization function applied to comparator 47 controls the error measurement. Optimization functions generally find the best possible solution, typically a maximum or a minimum. The optimization function defines an objective function that evaluates the proposed solutions and finds the input that yields the most desirable output. In this case, the objective function minimizes the error between the "real" and the synthesized waveform. If the synthesized waveform has too high a measurement error, the "real" waveform can be selected from the received signal.If the real waveform is insufficient because the desired waveform is an inversion or another variation of the real waveform, the data can be mixed between the real and the synthesized waveform until the desired performance indicator is achieved.

[0016] The error measurement may include, among others, the mean squared error, the mean absolute error, the square root of the mean squared error, R-squared, the fitted R-squared, the mean log-squared error, and the mean absolute deviation. The procedure can use any of these error measurements to evaluate the accuracy of the GAN model. The generated or "synthesized" signal is also filtered in filter 49 to compensate for the errors detected during the comparison. The unfiltered synthesized signal or the filtered signal is then selected at selector 46 depending on the error measurement and output.

[0017] The output signal can be transmitted if the generated signal requires transmission. Alternatively, the output signal can also be applied to a device under test (DUT). The signal fed to the DUT can be generated when the DUT is undergoing testing and initially no signal was received, but the DUT has a test requirement that necessitates a waveform with specific characteristics.

[0018] The signal synthesized by the GAN generator 48 can be stored in the waveform library 44 with the associated error measurement if the synthesized signal has been selected at the selector 46. The error measurement may be appended to the waveform that is to be used as a selection criterion in the waveform library 44.

[0019] Fig. Figure 3 shows an example of a node located in the Fig. The network shown in Figure 1 can be located within a test and measurement device. In one embodiment, the node can use a GAN to replicate a signal within the device. The test and measurement device can be a "composite" device, meaning that the device comprises two different devices. In the embodiment of Fig. 3 The composite instrument comprises a mixed-signal oscilloscope (MSO) 54 and an arbitrary waveform generator (AWG) 58. The term AWG, as used here, includes both arbitrary waveform generators and arbitrary function generators.

[0020] The DUT 50 is connected to the MSO 54 via one or more channels 52. In this embodiment, channel 52 is unknown in that it has unknown properties. The DUT 50 generates a signal that is transmitted to the MSO 54 via channel 52. The MSO 54 receives the signal on channel 52 and copies it to channel 53. The signal received on the first channel 52 undergoes signal processing and conditioning in the MSO 54. This conditioning can include analog-to-digital conversion, signal conditioning, timing adjustment, etc. The signal from the first channel 52 is then filtered in the filter 56 and subsequently replicated by the GAN 57. The GAN 57 iterates until the discriminator determines that the generator has produced a replica that is indistinguishable from the first signal. The iterations of GAN 57 result in a signal for the "calibrated" channel 59.The AWG 58 then takes the signal from the GAN 57 and generates another signal on channel 59. The signal on channel 59 is compared with the "original" signal that entered on channel 52, which passes through the MSO to channel 53, leading to the comparator 55. The results can be used to adjust the filter 56 to eliminate the differences or to modify the original signal in other ways, if desired. The results of the comparison can also be used to determine the characteristics of channel 52 from the calibration process. This channel characterization and the original signal can then be stored in the waveform library 22. Fig. 1 or Fig. 44 of Fig. 2 will be saved.

[0021] As mentioned earlier, nodes in the system capable of running a GAN can participate in federated learning. In federated learning, each node, when the GAN model(s) on that node are tuned and modified by the discriminator process, can send updates to the bootstrap model on that node. Every other node in the system does the same, ensuring that all nodes running GANs have at least the ability to update themselves to bring all nodes to the same level with respect to their models. In some embodiments, different nodes may have different federated learning strategies, accepting only selected updates and rejecting all others, or conversely, rejecting only certain updates but accepting all others.

[0022] The updated nodes also transmit their updates to the central or main server(s) to update the models located on the server(s). When a new node is brought into the system, the update process on the server(s) allows that node to receive the latest version of the model. This node can also receive the bootstrap model and then the updates, allowing the node to filter out the updates that are not relevant to it.

[0023] Fig. 4 and Fig. Figure 5 shows generalized versions of the nodes for both the receiving process and the receiving process. Fig. 4 as well as for the transmission process in Fig. 5. In the reception process shown in Figure 4, a sensor node 60 receives the incoming signal and captures the characteristics of the DUT, the environment under test (EUT), and the channel in which the signal was received, and makes these available to the model in the node. The signal is processed by a signal conditioner 62, and the GAN 64 ensures that the replica of the signal is as close as possible to the original. The results 66 are then made available for analysis to the person who needs to see them, and the resulting replication, which may contain inverse signals, signals with variants of the original signal, etc., is then stored in the waveform database 68. The waveform database 68 can represent a library, as will be explained further below.The Measurement and Metrology Module 70 enables the Sensor Node 60 to take measurements of the incoming signal, decode the signal's components and / or characteristics as required, and monitor the performance of the GAN 64. Depending on the measurement results, the Measurement and Metrology Module can interact with other modules, for example, by communicating with the sensor if the information was not received completely or if other signal problems occur. Module 70 can also communicate with the Signal Conditioning Module if the signal type requires adjustments during conditioning. During GAN 64 operation, the results from Module 70 may change, necessitating an update. Modules 70 and 66 then exchange results until they can be saved.

[0024] Fig. Figure 5 shows the transmission process. When a signal request is received, either in response to an incoming signal or to test a DUT, EUT, or channel, the node or central server of Fig. 1 accesses the database / library 68 and provides the desired version of the signal via the results 66. The signal is then provided to the GAN 64 with the required special characteristics, and the GAN 64 operates to generate the desired signal. The signal is processed in the signal conditioning unit 62, e.g., from digital back to analog, and the signal is provided to the sensor node 60 for use by an AWG at the sensor for signal generation and transmission.

[0025] As mentioned previously, any resulting changes or updates to the GAN model can be distributed to the other nodes as part of the federated learning environment.

[0026] With the development and setup of the waveform database, waveform database 22 could be used like a library by large language models. Back to Fig. 1: An interface (I / F) 26 for large language models (LLM) allows a user to input a prompt to generate a specific required waveform. The test environment from which the prompt originates, as well as the characteristics of the user-specified waveform, can be converted into an LLM-specific vector. The LLM vector is then used to access the waveform library to find a suitable signal, or to find signals that approximate the signal that the GAN models can work with to generate the desired waveform.

[0027] This disclosure covers the concept of generating waveforms using a learned large waveform model (LWM). In this implementation, artificial intelligence or machine learning is used to train a model based on uploaded data relevant to the intended experiment. The generated waveform is based on contextual knowledge of the communicated test environment. The query interface can be various, not just a keyboard.

[0028] The distributed learning that enables the implementations is based on principles of distributed learning, including but not limited to federated learning, with an application programming interface (API) that allows direct translation, loading, or sequencing of waveforms to deliver streamed physical data that interacts with a communication medium, e.g., air, water, etc.

[0029] This disclosure solves the problem of rapidly prototyping waveforms for stimuli on targeted devices under test (DUTs) by providing a unified format for running agnostic waveform translators used to export signals to the real world. Furthermore, users benefit from other participants, who allow them to share their learned waveforms, derived from their shared measurement data and uploaded to the platform, to further train the associated learning waveform translator (LWM). The LWM can be public or private and specific enough to be connected only to industry and its participants.

[0030] Dynamic waveform prototyping with direct storage via the LLM interface assists the platform in learning optimal waveforms based on the context provided through the LLM interface. Based on the provided context data, the learned waveforms can be tagged and grouped based on the customer's unique DUT environments. This capability is not yet available at the time of this disclosure.

[0031] In one embodiment, the user interface of the test and measurement system includes an interface to a large language model. In the embodiment that uses the interface of the large language model, the provided waveforms are enriched with metadata to enable the library to be updated with the new model.

[0032] In this way, a test and measurement system can receive and generate signals / waveforms as needed, using machine learning to ensure accuracy and conformity with the desired signal profile.

[0033] As an example of an application of this system, the Fig. 6 and Fig. 7. A flowchart of an embodiment of a method for generating an adaptation signal from a received signal and a diagram of the nodes in a particular embodiment of a threat environment. For example, a sensor node (90, 92, or 94 in Fig. 7) Receives a test signal representing a threat signal. The system receives the signal and determines a response. Depending on the capabilities of the node (e.g., sensor node 90, 92, 94), the response may come from the central server 14 if the node lacks the necessary capabilities.

[0034] In each of the above situations, the node generates a request (80 in Fig.6) for a signal with a signal profile that includes information about the signal, the environment in which the node operates, metadata information relating to the signal, possibly the destination of the matching signal to be sent, and the information captured above, as examples without limitation. The request is sent to one or more GAN models, as described above in section 82, which may be models on the node, from the node to the central server, which in turn sends the request to the GAN models, to neighboring nodes with this capability, all of which are referred to here as requesters. The requester then receives a matching signal profile from the one or more GAN models in section 84. The requester sends the matching signal profile to one or more signal generators (86).In an optional process, a signal generator can then generate the signal at 88 and transmit it to the node, so that the node can transmit the appropriate signal to the received signal as needed.

[0035] One aspect of the response is a countermeasures signal for electronic warfare to jam the incoming signal. The other aspect can be a radar signal transmitted to determine the location of the signal source. The received signal is processed, and two signals can be generated, both of which are in some way replicas of the received signal. The system generates replicas from the base signal and by accessing the waveform library / database, and the resulting signals are transmitted by an AWG within the system. The above discussion is an example, without limitation to any particular embodiments used here.

[0036] Aspects of the disclosure may function on specially designed hardware, firmware, digital signal processors, or on a specially programmed general-purpose computer with a processor that operates according to programmed instructions. The terms controller or processor, as used herein, are intended to include microprocessors, microcomputers, application-specific integrated circuits (ASICs), and specialized hardware controllers. One or more aspects of the disclosure may be embodied in computer-usable data and computer-executable instructions, such as one or more program modules executed by one or more computers (including monitoring modules) or other devices. In general, program modules include routines, programs, objects, components, data structures, and so on.These are computer-executable instructions that perform specific tasks or implement certain abstract data types when executed by a processor in a computer or other device. The computer-executable instructions may be stored on a non-transient, computer-readable medium such as a hard disk, optical disk, removable storage medium, solid-state memory, random access memory (RAM), etc. As will be clear to those skilled in the art, the functionality of the program modules can be arbitrarily combined or distributed in various aspects. Furthermore, the functionality may be embodied wholly or partially in firmware or hardware equivalents such as integrated circuits, FPGAs, and the like.Certain data structures can be used to implement one or more aspects of the disclosure more effectively, and such data structures are considered within the context of the computer-executable instructions and computer-usable data described here.

[0037] The disclosed aspects may, in some cases, be implemented in hardware, firmware, software, or a combination thereof. The disclosed aspects may also be implemented in the form of instructions stored on one or more non-transitory, computer-readable media that can be read and executed by one or more processors. Such instructions may be referred to as a computer program product. Computer-readable media, as described herein, are all media that a computer can access. Computer-readable media may include, for example, computer storage media and communication media, but are not limited to these.

[0038] Computer storage media are all media that can be used to store computer-readable information. Examples of computer storage media include RAM, ROM, EEPROM (Electrically Erasable Programmable Read-Only Memory), flash memory or other storage technologies, CD-ROM (Compact Disc Read Only Memory), DVD (Digital Video Disc) or other optical disc storage, magnetic cartridges, magnetic tapes, magnetic disk storage or other magnetic storage devices, and all other volatile or non-volatile, removable or non-removable media implemented in any technology. Computer storage media exclude signals as such and temporary forms of signal transmission.

[0039] Communication media are defined as all media that can be used to transmit computer-readable information. Examples of communication media include coaxial cables, fiber optic cables, air, or any other medium suitable for transmitting electrical, optical, radio frequency (RF), infrared, acoustic, or other signals. EXAMPLES

[0040] Examples of the disclosed technologies are described below. An embodiment of the technologies may include one or more, and any combination of, the examples described below.

[0041] Example 1 is a test and measurement system comprising: a plurality of nodes, wherein one or more nodes comprise sensors configured to receive signals; one or more generative adversarial network (GAN) models; one or more signal generators configured to generate and transmit signals; and one or more processors configured to execute code to cause the one or more processors to: receive a request for a signal with a signal profile from one node of the plurality of nodes; send the signal profile and the request to one of the one or more GAN models; receive a matching signal profile that corresponds to the signal profile; and transmit the matching signal profile to one of the one or more signal generators to generate and transmit a matching signal.

[0042] Example 2 is the test and measurement system from Example 1, where the signal profile includes one or more of the following elements: an environment in which a node of the multitude of nodes operates, metadata of the requirement, and a potential target of the matching signal.

[0043] Example 3 is the test and measurement system from one of Examples 1 or 2, wherein the one or more processors are further configured to store the appropriate signal in a database.

[0044] Example 4 is the test and measurement system from Example 3, wherein the one or more processors are further configured to access the database to determine whether the signal profile is present in the database and send the signal profile from the database to the one or more GAN models.

[0045] Example 5 is the test and measurement system according to one of Examples 1 to 4, wherein the one or more processors are further configured to operate the one or more GAN models to generate the appropriate signal profile.

[0046] Example 6 is the test and measurement system according to any one of Examples 1 to 5, wherein the one or more processors configured to operate the one or more GAN models comprise one or more processors on a node of the plurality of nodes that sent the request.

[0047] Example 7 is the test and measurement system from one of Examples 1 to 6, wherein the one or more processors trained to operate the one or more GAN models comprise one or more processors on a central server in the test and measurement system.

[0048] Example 8 is the test and measurement system from Example 7, wherein the one or more processors on the central server are further configured to distribute an updated GAN model resulting from the requirement to the multitude of nodes.

[0049] Example 9 is the test and measurement system from one of Examples 1 to 8, wherein one or more of the sensors receive an incoming signal from a device under test (DUT).

[0050] Example 10 is the test and measurement system from Example 9, wherein one or more of the sensors further comprises a test and measurement device comprising: a port for receiving the signal incoming from the DUT on a channel of the test and measurement device; one or more analog-to-digital converters (ADCs) for converting the incoming signal from the DUT into a digital representation of the incoming signal; and one or more processors configured to execute code to cause the one or more processors to: characterize the incoming signal; apply a local GAN ​​model residing on the test and measurement device to replicate the incoming signal with appropriate features as a replica signal; and send any updates to the local GAN ​​model to a central server.

[0051] Example 11 is the test and measurement device from Example 10, wherein the one or more processors are further configured to execute code to cause the one or more processors to send the replica signal to one or more databases.

[0052] Example 12 is the test and measurement apparatus from one of Examples 1 to 11, wherein one or more signal generators include an arbitrary waveform generator.

[0053] Example 13 is a procedure that includes: receiving a request for a signal with a signal profile from a node in a test and measurement system; sending the signal profile and the request to one or more generative adversarial network (GAN) models; receiving a matching signal profile that matches the signal profile from one or more GAN models; and transmitting the matching signal profile to one or more signal generators to generate and transmit a matching signal.

[0054] Example 14 is the procedure of Example 13, wherein the signal profile contains information about one or more of the following elements: an environment in which the node operates, metadata of the request, and a potential target of the matching signal.

[0055] Example 15 is the method according to one of Examples 13 or 14, which further includes storing the appropriate signal in a database.

[0056] Example 16 is the procedure of Example 15, where the database is accessed to determine whether the signal profile is present in the database and to send the signal profile from the database to the one or more GAN models.

[0057] Example 17 is the procedure of Example 16, which further includes operating one or more GAN models to generate the appropriate signal profile.

[0058] Example 18 is the procedure according to one of Examples 13 to 17, which further includes distributing an updated GAN model resulting from the requirement to other nodes in a system where the updated GAN model is located.

[0059] Example 19 is the method according to any of Examples 13 to 18, wherein receiving the request for the signal includes receiving an incoming signal from a device under test (DUT) at a sensor node.

[0060] Example 20 is the procedure from Example 19, wherein receiving the incoming signal at the sensor node further includes: characterizing the incoming signal; applying a local GAN ​​model located on the node to replicate the incoming signal with matching features as a replica signal; and sending all updates of the local GAN ​​model to a central server.

[0061] Example 21 is the procedure of Example 20, which further includes sending the replica signal to one or more databases.

[0062] All features disclosed in the description, including the claims, the abstract, and the drawings, and all steps in each disclosed method or process may be combined in any combination, except for combinations in which at least some of these features and / or steps are mutually exclusive. Any feature disclosed in the description, including the claims, the abstract, and the drawings, may be replaced by alternative features that serve the same, equivalent, or similar purpose, unless expressly stated otherwise.

[0063] Furthermore, this written description refers to certain characteristics. It is to be understood that the disclosure in this description encompasses all possible combinations of these particular characteristics. For example, if a particular characteristic is disclosed in connection with a specific aspect, this characteristic may, to the extent possible, also be used in connection with other aspects.

[0064] Where this application refers to a procedure with two or more defined steps or operations, the defined steps or operations may be carried out in any order or simultaneously, provided that the context does not preclude such possibilities.

[0065] Although certain aspects of the revelation have been illustrated and described for the purpose of demonstration, various changes can be made without deviating from the spirit and scope of the revelation. QUOTES INCLUDED IN THE DESCRIPTION

[0000] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0000] US 63 / 699,259

[0001]

Claims

[1] A test and measurement system comprising the following: a multitude of nodes, wherein one or more nodes comprise sensors designed to receive signals; one or more generative adversarial network (GAN) models; one or more signal generators designed to generate and transmit signals; and one or more processors trained to execute code to cause the one or more processors to do the following: Receiving a request for a signal with a signal profile from one node of the plurality of nodes; sending the signal profile and the request to one of the one or more GAN models; Receiving a suitable signal profile that matches the signal profile; and transmitting the suitable signal profile to one or more signal generators to generate and transmit a suitable signal. [2] The test and measurement system according to claim 1, wherein the signal profile comprises one or more of the environments in which a node of the plurality of nodes operates, metadata of the requirement and a potential target of the matching signal. [3] The test and measurement system according to claim 1 or 2, wherein the one or more processors are further configured to store the appropriate signal in a database. [4] The test and measurement system according to claim 3, wherein the one or more processors are further configured to access the database to determine whether the signal profile is present in the database and to send the signal profile from the database to the one or more GAN models. [5] The test and measurement system according to any one of claims 1 to 4, wherein the one or more processors are further configured to operate the one or more GAN models to generate the appropriate signal profile. [6] The test and measurement system according to any one of claims 1 to 5, wherein the one or more processors configured to operate the one or more GAN models comprise one or more processors on a node of the plurality of nodes that sent the request. [7] The test and measurement system according to any one of claims 1 to 6, wherein the one or more processors configured to operate the one or more GAN models comprise one or more processors on a central server in the test and measurement system. [8] The test and measurement system according to claim 7, wherein the one or more processors on the central server are further configured to distribute an updated GAN model resulting from the requirement to the plurality of nodes. [9] The test and measurement system according to any one of claims 1 to 8, wherein one or more of the sensors receives an incoming signal from a device under test (DUT). [10] The test and measurement system according to claim 9, wherein one or more of the sensors further comprises a test and measurement device comprising the following: a connection for receiving the input signal from the DUT on a channel of the test and measurement device; one or more analog-to-digital converters (ADCs) to convert the signal arriving from the DUT into a digital representation of the incoming signal; and one or more processors trained to execute code that causes the one or more processors to do the following: to characterize the incoming signal; to apply a local GAN ​​model located on the test and measurement equipment to replicate the incoming signal with suitable properties as a replica signal; and to send all updates of the local GAN ​​model to a central server. [11] The test and measuring device according to claim 10, wherein the one or more processors are further configured to execute code to cause the one or more processors to send the replica signal to one or more databases. [12] The test and measuring device according to any one of claims 1 to 11, wherein one or more signal generators comprise an arbitrary waveform generator. [13] A procedure comprising the following: Receiving a request for a signal with a signal profile from a node in a test and measurement system; sending the signal profile and the request to one or more generative adversarial network (GAN) models; Receiving a suitable signal profile that matches the signal profile of one or more GAN models; and Transferring the appropriate signal profile to one or more signal generators to generate and transmit a suitable signal. [14] The method according to claim 13, wherein the signal profile contains information about one or more of the following elements: one or more of the environments in which the node operates, metadata of the request and a potential target of the matching signal. [15] The method according to claim 13 or 14, further comprising storing the appropriate signal in a database. [16] The method according to claim 15, wherein the database is accessed to determine whether the signal profile is present in the database and to send the signal profile from the database to the one or more GAN models. [17] The method according to claim 16, further comprising operating one or more GAN models to generate the appropriate signal profile. [18] The method according to any one of claims 13 to 17, further comprising distributing an updated GAN model resulting from the requirement to other nodes in a system in which the updated GAN model is located. [19] The method according to any one of claims 13 to 18, wherein receiving the signal request comprises receiving an incoming signal from a device under test (DUT) at a sensor node. [20] The method according to claim 19, wherein receiving the incoming signal at the sensor node further comprises: a characterization of the incoming signal; an application of a local GAN ​​model located on the node to replicate the incoming signal with matching features as a replica signal; and Sending all updates to the local GAN ​​model to a central server. [21] The method according to claim 20, further comprising sending the replica signal to one or more databases.

Images