A radar active jamming device capable of real-time dynamic updating

Abstract

In order to solve the problem of how to realize real-time dynamic updating of a radar active jamming system, the application discloses a radar active jamming device capable of real-time dynamic updating, which comprises a hardware resource module, an operating system module, a middleware module, a system monitoring and management module, a reconstruction implementation module and an application development module; the operating system module is connected with the middleware module; and the system monitoring and management module is connected with the reconstruction implementation module.The reconstruction implementation module matches reconstruction demand information with algorithm components to obtain matching service information; dynamic loading component information is determined by using the matching service information, relevant algorithm components are dynamically loaded by using the dynamic loading component information, and dynamic loading calling information is obtained.The application can realize upgrading of the active jamming capability by upgrading the software without changing the hardware structure of the system, thereby reducing the system implementation cost and ensuring the real-time performance and effectiveness of the jamming effect.

Description

Technical Field

[0001] This invention relates to the field of active radar jamming, and more particularly to an active radar jamming device that can be dynamically updated in real time. Background Technology

[0002] Current active radar jamming systems, while employing a modular design approach, lack a fixed standard for dividing the overall system architecture, and complex dependencies exist between modules. This limits the ability of active jamming systems to static reconfiguration, making real-time dynamic updates a challenge. Currently, the field of active radar jamming lacks a method for achieving dynamic, real-time updates targeting new threat targets. Due to the different implementation architectures of radar systems and active radar jamming systems, existing radar system technologies cannot be directly applied to active radar jamming systems. Summary of the Invention

[0003] To address the problem of how to achieve real-time dynamic updates for radar active jamming systems, this invention discloses a radar active jamming device capable of real-time dynamic updates, comprising a hardware resource module, an operating system module, a middleware module, a system monitoring and management module, a reconfiguration implementation module, and an application development module. The hardware resource module is connected to the operating system module, the operating system module is connected to the middleware module, the middleware module is connected to both the system monitoring and management module and the reconfiguration implementation module, the system monitoring and management module is connected to the reconfiguration implementation module, and the application development module is connected to both the system monitoring and management module and the reconfiguration implementation module.

[0004] The hardware resource module includes DSP computing circuits, FPGA computing circuits, CPU computing circuits, GPU computing circuits, storage devices, display and control devices, input / output devices, a high-speed bus, and a software bus. Each pair of DSP computing circuits, FPGA computing circuits, CPU computing circuits, GPU computing circuits, storage devices, display and control devices, and input / output devices is connected via a high-speed bus. The software bus runs on the high-speed bus; it is used to convert the output signal formats of the different hardware components within the hardware resource module to obtain a unified format of high-speed digital signals, which are then sent to the operating system module. The hardware resource module connects to the operating system module through various I / O interfaces.

[0005] The operating system module is used to implement real-time operating system functions. The operating system module includes a device management submodule, a storage management submodule, a process management submodule, and an I / O management submodule.

[0006] The device management submodule is used to abstract and describe various types of hardware in the hardware resource module, obtain description models for various types of hardware, and use the description models to manage various types of hardware.

[0007] The storage management submodule is used to monitor the status of storage units of various hardware components in the hardware resource module and obtain the status monitoring results; based on the status monitoring results and the storage requirements of the middleware, the storage units are allocated.

[0008] The process management submodule is used to monitor and manage various processes of the real-time operating system;

[0009] The I / O management submodule is used to monitor and manage various I / O interfaces of the hardware resource module.

[0010] The middleware module is used to uniformly abstract various types of hardware in the hardware resource module to obtain abstract models of various types of hardware; to adapt the communication protocols of various types of hardware to obtain unified communication protocol data, and to realize data communication between various types of hardware using the unified communication protocol data; to construct service interfaces using the abstract models of various types of hardware; and to construct data resource interfaces using the unified communication protocol data. The service interfaces and data resource interfaces are used by the system monitoring and management module and the refactoring implementation module for use.

[0011] The middleware module includes communication middleware, computing middleware, reflection middleware, and artificial intelligence middleware.

[0012] The communication middleware adapts to the communication protocols of various hardware to obtain unified communication protocol data; and uses the unified communication protocol data to construct data resource interfaces.

[0013] The computing middleware is used to provide a common basic function call interface for reflection middleware and artificial intelligence middleware;

[0014] The reflection middleware is used to optimize and configure the data resource interface and basic function call interface according to the dynamic environment in which the application development module is located, so as to obtain dynamically updated data resource interface and basic function call interface; and to feed back the dynamically updated data resource interface and basic function call interface to the system monitoring and management module and the refactoring implementation module.

[0015] The artificial intelligence middleware is used to obtain artificial intelligence computing requirements from the system monitoring and management module and the refactoring implementation module; select the corresponding artificial intelligence algorithm or artificial intelligence model to perform calculations based on the artificial intelligence computing requirements, and obtain artificial intelligence computing results; and feed back the artificial intelligence computing results to the system monitoring and management module and the refactoring implementation module.

[0016] The reflection middleware includes a base layer and base objects, a meta-communication model, a meta-layer, and meta-objects. The reflection middleware acquires the dynamic environment in which the application development module is located, represents the dynamic environment as the base layer, and represents the data resource interface and basic function call interface as base objects. The base layer is input into the meta-communication model, and the output of the meta-communication model is used to modify the base layer and base objects to obtain the meta-layer and meta-objects respectively. The meta-objects serve as dynamically updated data resource interfaces and basic function call interfaces.

[0017] The meta-communication model includes data resource interfaces and basic function call interfaces, and the optimal correspondence between them and the dynamic environment in which the application development module is located.

[0018] The meta-communication model receives the input base object and, using the optimal correspondence method, outputs the meta-layer and meta-object that optimally correspond to the input base object; the optimally corresponding meta-layer and meta-object refer to the optimal correspondence method between the dynamic environment and data resource interface and the basic function call interface.

[0019] The dynamic environment in which the application development module exists includes the dynamic loading and calling information generated by the refactoring implementation module and the computational task information and communication task information obtained by the task scheduling submodule.

[0020] The artificial intelligence middleware obtains artificial intelligence computing requirements from the system monitoring and management module and the refactoring implementation module, performs category judgment on the artificial intelligence computing requirements, and obtains the category judgment result;

[0021] If the category judgment result is an object recognition problem, determine whether the artificial intelligence computing requirement includes data labels; if it includes data labels, use a supervised learning method to calculate the artificial intelligence computing requirement and obtain the artificial intelligence computing result; if it does not include data labels, use an unsupervised learning method to calculate the artificial intelligence computing requirement and obtain the artificial intelligence computing result.

[0022] If the category determination result is an active interference generation problem, the complexity of radar active interference generation is judged according to the artificial intelligence computing requirements. If the complexity of radar active interference generation is greater than a set threshold, a deep reinforcement learning algorithm based on deep neural networks is used to calculate the artificial intelligence computing requirements and obtain the artificial intelligence computing result. If the complexity of radar active interference generation is less than the set threshold, a reinforcement learning algorithm based on Q-tables is used to calculate the artificial intelligence computing requirements and obtain the artificial intelligence computing result. The artificial intelligence computing requirements include the dynamically loaded call information generated by the reconstruction implementation module and the computing task information and communication task information obtained by the task scheduling submodule.

[0023] The system monitoring and management module is used to monitor and manage the software and hardware resources of the radar active jamming device in real time, which can be dynamically updated in real time.

[0024] The system monitoring and management module includes a strategy generation submodule, a resource allocation submodule, a task scheduling submodule, a status monitoring submodule, and an effect evaluation submodule.

[0025] The strategy generation submodule is used to generate a radar active jamming strategy based on the instruction information from the application development module and the interference evaluation results from the effect evaluation submodule; and to send the radar active jamming strategy to the resource allocation submodule and the task scheduling submodule.

[0026] The resource allocation submodule is used to receive the radar active jamming strategy and allocate the software and hardware resources of the radar active jamming device, which can be dynamically updated in real time, according to the requirements of the radar active jamming strategy and the status monitoring results.

[0027] The task scheduling submodule is used to receive the radar active jamming strategy, decompose the radar active jamming strategy to obtain the corresponding computational task information and communication task information; and send the computational task information and communication task information to the middleware module according to the status monitoring results.

[0028] The status monitoring submodule is used to monitor the software and hardware resources of the radar active jamming device in real time, which can be dynamically updated in real time, and obtain the status monitoring results; and send the status monitoring results to the task scheduling submodule, the resource allocation submodule and the radar active jamming development environment.

[0029] The effect evaluation submodule is used to receive electromagnetic environment monitoring data sent by the communication middleware, evaluate the effect of radar active interference using the electromagnetic environment monitoring data, obtain interference evaluation results, and send the interference evaluation results to the strategy generation submodule.

[0030] The reconstruction implementation module is used to receive and integrate user development requirement information from the radar active jamming development environment, radar active jamming strategies generated by the strategy generation submodule, and jamming evaluation result information from the effect evaluation submodule to obtain reconstruction requirement information; match the reconstruction requirement information with algorithm components to obtain matching service information; use the matching service information to determine dynamic loading component information, use the dynamic loading component information to dynamically load relevant algorithm components to obtain dynamic loading call information; and send the dynamic loading call information to the reflection middleware and the artificial intelligence middleware.

[0031] The refactoring implementation module includes an event listening and response submodule, a service registration and lookup submodule, and a component lifecycle management submodule.

[0032] The service registration lookup submodule queries and processes the reconstruction requirement information, extracts the service component registration information from the reconstruction requirement information, and determines whether the service component corresponding to the service component registration information has been registered.

[0033] If the service component has been registered, the service component registration information of the service component is used as the matching service information and the matching service information is output to the component lifecycle management submodule;

[0034] If the service component is not registered, the service component registration information is used to match and search all algorithm components. The algorithm component with the highest matching degree with the service component registration information is found and determined as the matching algorithm component. The service corresponding to the matching algorithm component is registered, the service component registration information corresponding to the service is obtained, and a service registration interface and a service reference interface are established. The service component registration information corresponding to the service is output to the component lifecycle management submodule.

[0035] The application development module is used to provide users with a development environment for active radar jamming;

[0036] The application development module includes a radar active jamming development environment, a standard algorithm component library, and a user-defined algorithm component library. The radar active jamming development environment provides users with a software development interface and software component interfaces for radar active jamming, and provides a compilation and debugging environment for user-developed radar active jamming programs.

[0037] The application development module receives interference request information from users; converts and processes the interference request information to obtain instruction information; and sends the instruction information to the policy generation submodule of the system monitoring and management module.

[0038] The beneficial effects of this invention are as follows:

[0039] 1. This invention can upgrade the active interference capability without changing the system hardware structure by upgrading the software, and has the ability to be dynamically upgraded, thus reducing the system implementation cost;

[0040] 2. When faced with new threats, this system can dynamically generate corresponding interference signals in real time, and adjust the interference scheme and interference generation rules in real time according to changes in the threat and the interference effect. It has the ability to dynamically iterate, ensuring the real-time performance and effectiveness of the interference effect. Attached Figure Description

[0041] Figure 1 A block diagram illustrating the composition principle of an existing radar active jamming system;

[0042] Figure 2This is a schematic diagram of the composition of the radar active jamming device that can be dynamically updated in real time according to the present invention.

[0043] Figure 3 This is a schematic diagram of the service-oriented component model of the present invention. Detailed Implementation

[0044] To better understand the content of this invention, an embodiment is provided here.

[0045] Figure 1 This is a block diagram illustrating the composition and operating principle of an existing active radar jamming system. For Figure 1 All active jamming systems are unable to dynamically adjust their jamming strategies based on the jamming effect and electromagnetic environment perception results. They cannot adapt to the dynamic feedback function requirements of cognitive electronic warfare, and therefore cannot fully realize the cognitive countermeasure function of cognitive electronic warfare. Furthermore, they cannot have the ability to be rapidly upgraded and iterated and dynamically reconfigured in response to the dynamic development of threat targets.

[0046] The composition of a radar active jamming device that can be dynamically updated in real time is as follows: Figure 2 As shown. Figure 3 This is a schematic diagram of the service-oriented component model of the present invention.

[0047] This embodiment discloses a radar active jamming device capable of real-time dynamic updates, including a hardware resource module, an operating system module, a middleware module, a system monitoring and management module, a reconfiguration implementation module, and an application development module. The hardware resource module is connected to the operating system module, the operating system module is connected to the middleware module, the middleware module is connected to both the system monitoring and management module and the reconfiguration implementation module, the system monitoring and management module is connected to the reconfiguration implementation module, and the application development module is connected to both the system monitoring and management module and the reconfiguration implementation module.

[0048] The hardware resource module includes DSP computing circuits, FPGA computing circuits, CPU computing circuits, GPU computing circuits, storage devices, display and control devices, input / output devices, a high-speed bus, and a software bus. The hardware resource module is primarily implemented in hardware, while other modules are implemented in software. Each pair of DSP computing circuits, FPGA computing circuits, CPU computing circuits, GPU computing circuits, storage devices, display and control devices, and input / output devices is connected via a high-speed bus. The software bus runs on the high-speed bus; it is used to convert the output signal formats of the different hardware components within the hardware resource module to obtain a unified high-speed digital signal, which is then sent to the operating system module. The hardware resource module connects to the operating system module through various I / O interfaces. The DSP computing circuit is used to perform real-time mathematical calculations; the CPU computing circuit is used to perform logical operations, loop operations, and control operations; the GPU computing circuit is used to perform floating-point calculations, multi-task calculations, and digital beamforming calculations; and the FPGA computing circuit is used to perform fixed-point calculations, received data preprocessing calculations, and digital beamforming calculations.

[0049] To achieve dynamic reconfiguration of system functions, software and hardware decoupling is necessary. In active radar jamming systems, the hardware computing chips used exhibit diverse heterogeneous characteristics, including DSPs, FPGAs, and CPUs. A general-purpose software platform needs to be designed for these various computing resources to achieve software-hardware decoupling and enable dynamic reconfiguration of system functions through software reconfiguration.

[0050] The operating system module, middleware module, system monitoring and management module, refactoring implementation module, and application development module together constitute the system's software platform. Within this platform, common radar active jamming algorithms are developed in a standardized and modular manner. Unified component interfaces improve the standardization of algorithm components and the system's portability. Building an open, efficient, and standardized algorithm component library enhances component reusability and avoids repetitive development. Furthermore, updating, upgrading, and refactoring algorithm components through software programming allows for rapid system function reconstruction, enhancing the system's ability to respond to unknown threats.

[0051] The hardware resource layer also includes drivers and interface description programs for various types of hardware.

[0052] The operating system module implements real-time operating system functions, and the real-time operating system can be the SylixOS embedded hardware real-time operating system. The operating system module is connected to the hardware resource module and the middleware module respectively. Specifically, the operating system module is connected to a high-speed bus, and also to communication middleware and computing middleware. The operating system module includes a device management submodule, a storage management submodule, a process management submodule, and an I / O management submodule.

[0053] The device management submodule is used to abstract and describe various types of hardware in the hardware resource module, obtain description models for various types of hardware, and use the description models to manage various types of hardware.

[0054] The storage management submodule is used to monitor the status of storage units of various hardware components in the hardware resource module and obtain the status monitoring results; based on the status monitoring results and the storage requirements of the middleware, the storage units are allocated.

[0055] The process management submodule is used to monitor and manage various processes of the real-time operating system;

[0056] The I / O management submodule is used to monitor and manage various I / O interfaces of the hardware resource module.

[0057] Middleware is an abstraction module of hardware and software within a radar active jamming system. It covers various processor architectures and communication methods across all signal processing platforms, serving as the foundation for shielding the underlying hardware and decoupling hardware and software. It acts as a compatibility layer, masking underlying differences and providing standardized interfaces. The various components in the application development module call acceleration algorithm functions, perform data transmission and reception operations, the system monitoring and management module control various components, and the development interface collects various statistical information—all of which require middleware to shield hardware differences. The quality of a middleware implementation largely determines the quality of the software architecture's software ecosystem. Only by providing reasonable and high-quality middleware interfaces for component and application developers can middleware standards be promoted more quickly, thereby popularizing this dynamically upgradeable radar active jamming system software architecture. Therefore, implementing a complete, reliable, and standardized middleware for radar active jamming applications is urgently needed in the current technological field. This paper, considering the dynamic upgrade and iteration requirements of radar active jamming, presents an overall middleware design, planning the middleware modules as computational middleware, communication middleware, reflection middleware, and artificial intelligence middleware.

[0058] The middleware module is used to uniformly abstract various types of hardware in the hardware resource module to obtain abstract models of various types of hardware; to adapt the communication protocols of various types of hardware to obtain unified communication protocol data, and to realize data communication between various types of hardware using the unified communication protocol data; to construct service interfaces using the abstract models of various types of hardware; and to construct data resource interfaces using the unified communication protocol data. The service interfaces and data resource interfaces are used by the system monitoring and management module and the refactoring implementation module for use.

[0059] The middleware module includes communication middleware, computing middleware, reflection middleware, and artificial intelligence middleware.

[0060] The communication middleware adapts to the communication protocols of various hardware to obtain unified communication protocol data; and uses the unified communication protocol data to construct data resource interfaces.

[0061] The computing middleware is used to provide a common basic function call interface for reflection middleware and artificial intelligence middleware;

[0062] The reflection middleware is used to optimize and configure the data resource interface and basic function call interface according to the dynamic environment in which the application development module is located, so as to obtain dynamically updated data resource interface and basic function call interface; and to feed back the dynamically updated data resource interface and basic function call interface to the system monitoring and management module and the refactoring implementation module.

[0063] The artificial intelligence middleware is used to obtain artificial intelligence computing requirements from the system monitoring and management module and the refactoring implementation module; select the corresponding artificial intelligence algorithm or artificial intelligence model to perform calculations based on the artificial intelligence computing requirements, and obtain artificial intelligence computing results; and feed back the artificial intelligence computing results to the system monitoring and management module and the refactoring implementation module.

[0064] The reflection middleware is implemented using reflection technology in object-oriented systems. The reflection middleware includes a base layer and base objects, a meta-communication model, meta-layers, and meta-objects. The reflection middleware acquires the dynamic environment of the application development module, represents this dynamic environment as the base layer, and represents the data resource interface and basic function call interface as base objects. The base layer is input into the meta-communication model, and the output of the meta-communication model is used to modify the base layer and base objects, obtaining the meta-layers and meta-objects respectively. The meta-objects serve as dynamically updated data resource interfaces and basic function call interfaces.

[0065] The meta-communication model includes data resource interfaces and basic function call interfaces, and the optimal correspondence between them and the dynamic environment in which the application development module is located.

[0066] The meta-communication model receives the input base object and, using the optimal correspondence method, outputs the meta-layer and meta-object that optimally correspond to the input base object; the optimally corresponding meta-layer and meta-object refer to the optimal correspondence method between the dynamic environment and data resource interface and the basic function call interface.

[0067] The dynamic environment in which the application development module exists includes the dynamic loading and calling information generated by the refactoring implementation module and the computational task information and communication task information obtained by the task scheduling submodule.

[0068] The artificial intelligence middleware obtains artificial intelligence computing requirements from the system monitoring and management module and the refactoring implementation module, performs category judgment on the artificial intelligence computing requirements, and obtains the category judgment result;

[0069] If the category judgment result is an object recognition problem, determine whether the artificial intelligence computing requirement includes data labels; if it includes data labels, use a supervised learning method to calculate the artificial intelligence computing requirement and obtain the artificial intelligence computing result; if it does not include data labels, use an unsupervised learning method to calculate the artificial intelligence computing requirement and obtain the artificial intelligence computing result.

[0070] If the category determination result is an active interference generation problem, the complexity of radar active interference generation is judged according to the artificial intelligence computing requirements. If the complexity of radar active interference generation is greater than a set threshold, a deep reinforcement learning (DRL) algorithm based on a deep neural network (DNN) is used to calculate the artificial intelligence computing requirements and obtain the artificial intelligence computing result. If the complexity of radar active interference generation is less than the set threshold, a reinforcement learning (RL) algorithm based on a Q-table is used to calculate the artificial intelligence computing requirements and obtain the artificial intelligence computing result. The artificial intelligence computing requirements include the dynamically loaded call information generated by the reconstruction implementation module and the computing task information and communication task information obtained by the task scheduling submodule.

[0071] The system monitoring and management module is used to monitor and manage the software and hardware resources of the radar active jamming device in real time, which can be dynamically updated in real time.

[0072] The system monitoring and management module includes a strategy generation submodule, a resource allocation submodule, a task scheduling submodule, a status monitoring submodule, and an effect evaluation submodule.

[0073] The strategy generation submodule is used to generate a radar active jamming strategy based on the instruction information from the application development module and the interference evaluation results from the effect evaluation submodule; and to send the radar active jamming strategy to the resource allocation submodule and the task scheduling submodule.

[0074] The resource allocation submodule is used to receive the radar active jamming strategy and allocate the software and hardware resources of the radar active jamming device, which can be dynamically updated in real time, according to the requirements of the radar active jamming strategy and the status monitoring results.

[0075] The task scheduling submodule is used to receive the radar active jamming strategy, decompose the radar active jamming strategy to obtain the corresponding computational task information and communication task information; and send the computational task information and communication task information to the middleware module according to the status monitoring results.

[0076] The status monitoring submodule is used to monitor the software and hardware resources of the radar active jamming device in real time, which can be dynamically updated in real time, and obtain the status monitoring results; and send the status monitoring results to the task scheduling submodule, the resource allocation submodule and the radar active jamming development environment.

[0077] The effect evaluation submodule is used to receive electromagnetic environment monitoring data sent by the communication middleware, evaluate the effect of radar active interference using the electromagnetic environment monitoring data, obtain interference evaluation results, and send the interference evaluation results to the strategy generation submodule.

[0078] The reconstruction implementation module is used to receive and integrate user development requirement information from the radar active jamming development environment, radar active jamming strategies generated by the strategy generation submodule, and jamming evaluation result information from the effect evaluation submodule to obtain reconstruction requirement information; match the reconstruction requirement information with algorithm components to obtain matching service information; use the matching service information to determine dynamic loading component information, use the dynamic loading component information to dynamically load relevant algorithm components to obtain dynamic loading call information; and send the dynamic loading call information to the reflection middleware and the artificial intelligence middleware.

[0079] The refactoring implementation module is implemented using a service-based refactoring method.

[0080] The refactoring implementation module includes an event listening and response submodule, a service registration and lookup submodule, and a component lifecycle management submodule.

[0081] The event monitoring and response submodule is used to receive and integrate user development requirement information from the radar active jamming development environment, radar active jamming strategies generated by the strategy generation submodule, and jamming evaluation results from the effect evaluation submodule to obtain reconstruction requirement information.

[0082] The event listening and response submodule includes a framework event listener, a component event listener, a service event listener, and a listener integrator.

[0083] The event listening and response submodule uses a framework event listener to receive user development requirement information from the radar active jamming development environment, a component event listener to receive radar active jamming strategy information generated by the strategy generation submodule, and a service event listener to receive jamming evaluation result information from the effect evaluation submodule. It then uses a listener integrator to integrate and process the information received from various event listeners to obtain reconstruction requirement information, and finally sends the reconstruction requirement information to the service registration and lookup submodule.

[0084] The service interfaces of the event listener response submodule include the framework event listener interface, the component event listener interface, and the service event listener interface.

[0085] The event listening and response submodule utilizes the event listening and response mechanism to listen for various types of events.

[0086] The service registration and lookup submodule matches the reconstruction requirement information with the algorithm components to obtain the matching service information;

[0087] The service registration lookup submodule queries and processes the reconstruction requirement information, extracts the service component registration information from the reconstruction requirement information, and determines whether the service component corresponding to the service component registration information has been registered.

[0088] If the service component has been registered, the service component registration information of the service component is used as the matching service information and the matching service information is output to the component lifecycle management submodule;

[0089] If the service component is not registered, the service component registration information is used to match and search all algorithm components. The algorithm component with the highest matching degree with the service component registration information is found and determined as the matching algorithm component. The service corresponding to the matching algorithm component is registered, the service component registration information corresponding to the service is obtained, and a service registration interface and a service reference interface are established. The service component registration information corresponding to the service is output to the component lifecycle management submodule.

[0090] The matching degree between the registration information of the algorithm component and the service component is obtained by converting the metadata information of the service attributes of the algorithm component and the registration information of the service component to obtain two corresponding vector values, calculating the quantized distance between the two vector values, and using the quantized distance as the matching degree between the registration information of the algorithm component and the service component.

[0091] The quantized distance is calculated using Euclidean distance or KL distance.

[0092] The component lifecycle management submodule uses matching service information to determine dynamically loaded component information, uses the dynamically loaded component information to dynamically load relevant algorithm components, obtains dynamic loading call information, and sends the dynamic loading call information to the reflection middleware and artificial intelligence middleware.

[0093] The component lifecycle management submodule uses matching service information to determine dynamically loaded component information, uses the dynamically loaded component information to establish corresponding framework factory objects, framework objects, and framework context objects, uses the framework context objects to dynamically load algorithm components, generates algorithm component objects, uses the algorithm component objects to manage the dynamic start, stop, update, and unload operations of components, obtains dynamic loading call information, and sends the dynamic loading call information to the reflection middleware and artificial intelligence middleware.

[0094] The application development module is used to provide users with a development environment for active radar jamming;

[0095] The application development module includes a radar active jamming development environment, a standard algorithm component library, and a user-defined algorithm component library. The radar active jamming development environment provides users with a software development interface and software component interfaces for radar active jamming, and provides a compilation and debugging environment for user-developed radar active jamming programs.

[0096] The application development module receives interference request information from users; converts and processes the interference request information to obtain instruction information; and sends the instruction information to the policy generation submodule of the system monitoring and management module.

[0097] The standard algorithm component library is a library of algorithm components corresponding to classic algorithms in the radar active jamming signal processing process. Following the relevant procedures of radar active jamming signal processing, some algorithms possess a certain degree of universality in the signal processing of radar active jamming systems of different systems and functions; these algorithms constitute the classic algorithms.

[0098] The standard algorithm component library specifically includes the following categories:

[0099] 1. Basic signal processing algorithm components, including nonlinear operation components, vector operation components, matrix operation components, window function design components, signal generation components, FFT processing and filter components, etc.;

[0100] 2. Signal reconnaissance and processing algorithm components, including signal receiving and detection algorithm components, parameter measurement algorithm components, pulse sorting algorithm components, threat identification algorithm components, etc.;

[0101] 3. Active interference processing algorithm components, including interference source generation algorithm components, interference guidance algorithm components, interference decision-making algorithm components, deception interference algorithm components, and suppression interference algorithm components.

[0102] The user-defined algorithm component library is an algorithm component library composed of algorithm components developed by users according to their specific needs.

[0103] Algorithm components in both the standard algorithm component library and the user-defined algorithm component library include a core algorithm logic layer, a cache space management layer, and a middleware interface layer. The core algorithm logic layer is used to implement the algorithm logic of the component, the cache space layer is used to store and manage the variables used in the implementation of the algorithm component, and the middleware interface layer is used to enable communication between the algorithm component and the system monitoring and management module, the refactoring implementation module, and the radar active jamming development environment.

[0104] The middleware interface layer for each algorithm component in the application development module is implemented using a service-oriented component model.

[0105] The service-oriented component model includes component activators, service interfaces, and service properties.

[0106] The service-oriented component model is designed and implemented using service-oriented programming techniques. The various component models communicate loosely through a service registration and lookup mechanism and an event publish-subscribe mechanism, which reduces the coupling between components.

[0107] The component activator is the core of the service-oriented component model, used to register and deregister component services. The component activator inherits from an abstract class and overrides the `start()` and `stop()` methods of the abstract class to enable dynamic starting and stopping of components.

[0108] Service interfaces are the external interfaces provided by the service-oriented component model in the form of services. They utilize a service registration and lookup mechanism to indirectly access various interfaces of the service-oriented component model, reducing coupling between components. Service interfaces include algorithm interfaces and communication interfaces. Algorithm interfaces provide input and output interfaces for the implementation algorithms of the service-oriented component model; communication interfaces employ an event publish-subscribe mechanism to achieve loosely coupled data communication between components.

[0109] Service attributes include metadata information of the service-oriented component model and algorithm-private data. The component's metadata information is represented in the form of key-value pairs. The component's metadata information is used to describe the unique identifier of the service-oriented component model, the dependencies of the component model, and other information, and is an important identifier of the component; the algorithm-private data is the input and output data used to implement the algorithm interface.

[0110] The service-oriented component model uses a service registration and lookup mechanism for indirect access to service interfaces, and an event publish-subscribe mechanism for indirect data access to communication interfaces, reducing coupling between components. Service attributes are described using metadata information, uniquely identifying components and supporting independent development and deployment. These design principles of the service-oriented component model solve the problems existing in the traditional radar active jamming system component model, enabling real-time dynamic updates of the functional modules of the radar active jamming system.

[0111] The component activator manages the activation state of service-oriented component models using component activation strategies. There are two activation strategy values ​​for service-oriented component models: Lazy and Eager. When the activation strategy is Lazy, the service-oriented component model does not immediately enter the running state upon dynamic startup. Instead, it transitions to the running state when its component service is first invoked. When the activation strategy is Eager, the service-oriented component model immediately enters the running state upon dynamic startup. This activation strategy increases component flexibility and reduces component memory overhead.

[0112] The dynamic loading of service-oriented component models is achieved through the normal resolution of dependencies in the service-oriented component models.

[0113] To ensure proper resolution of dependencies in a service-oriented component model (components), the component resolution process is divided into two phases: the dynamic loading phase and the dynamic startup phase.

[0114] During the component dynamic loading phase, the refactoring management system loads all required components, reads their locations, and saves them.

[0115] During the dynamic component startup phase, the refactoring management system uniformly starts all components. At this time, the system resolves each component to be started. If a component has mandatory dependencies, the system searches for the dependent components in its stored information. If a dependent component is found, it is started first. Once the dependent component is started, the component's dependencies are satisfied, and the component can start normally. If no dependent component is found, component resolution and startup fail, throwing an exception, and startup fails. If a component has optional dependencies or no dependencies, component resolution is normal, and startup is successful.

[0116] To enable components to be dynamically loaded across systems, it is necessary to encapsulate the system call interfaces of different operating systems to form a unified cross-system interface.

[0117] The radar active jamming development environment provides functional programming, object programming, and dataflow programming modes for users to develop corresponding radar active jamming methods for various heterogeneous radar signals.

[0118] Functional programming is the programming model with the lowest logical level and the easiest implementation. It includes programming using heterogeneous accelerated computing models such as OpenCL, CUDA, and MPI. In its implementation, it does not change the basic programming model that uses a single, serial overall control flow. The advantage of functional programming lies in its simplified programming complexity. In non-real-time high-performance computing, it allows for easier debugging and optimization at a relatively low cost. Its disadvantages are obvious: the serial control flow introduces performance bottlenecks, and expanding to heterogeneous, multi-tasking environments requires complex programming logic.

[0119] Object-oriented programming treats each instance of a class as an operable unit, allowing different nodes to interact with local or remote instances, with all operations encapsulated by middleware.

[0120] Dataflow programming divides the radar active jamming method under development into multiple sub-modules, each implemented independently. During programming, each sub-module only focuses on its own independent function, while the infrastructure maintains state information such as data packets. Existing dataflow programming models include ZeroMQ, RocketMQ, and Spark. Because each independent sub-module is not limited by a single serial logic, the parallelism of the data flow is greatly improved, and the system operates in a pipelined structure. The advantage of dataflow programming is that the pipeline control logic is implemented by the infrastructure, enabling better hardware-software decoupling on heterogeneous processing platforms.

[0121] The above description is merely an embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of this application should be included within the scope of the claims of this application.

Claims

1. A radar active jamming device capable of real-time dynamic updates, characterized in that, It includes a hardware resource module, an operating system module, a middleware module, a system monitoring and management module, a refactoring implementation module, and an application development module; the hardware resource module is connected to the operating system module, the operating system module is connected to the middleware module, the middleware module is connected to both the system monitoring and management module and the refactoring implementation module, the system monitoring and management module is connected to the refactoring implementation module, and the application development module is connected to both the system monitoring and management module and the refactoring implementation module. The middleware module includes communication middleware, computing middleware, reflection middleware, and artificial intelligence middleware; The communication middleware adapts to the communication protocols of various hardware types to obtain unified communication protocol data. Utilize unified communication protocol data to construct data resource interfaces; The computing middleware is used to provide a common basic function call interface for reflection middleware and artificial intelligence middleware; The reflection middleware is used to optimize the configuration of data resource interfaces and basic function call interfaces according to the dynamic environment in which the application development module is located, so as to obtain dynamically updated data resource interfaces and basic function call interfaces; and to feed back the dynamically updated data resource interfaces and basic function call interfaces to the system monitoring and management module and the refactoring implementation module. The artificial intelligence middleware is used to obtain artificial intelligence computing requirements from the system monitoring and management module and the refactoring implementation module; select appropriate artificial intelligence algorithms or artificial intelligence models to perform calculations based on the artificial intelligence computing requirements, and obtain artificial intelligence computing results; and feed back the artificial intelligence computing results to the system monitoring and management module and the refactoring implementation module. The artificial intelligence middleware obtains artificial intelligence computing requirements from the system monitoring and management module and the refactoring implementation module, performs category judgment on the artificial intelligence computing requirements, and obtains the category judgment result; If the category judgment result is an object recognition problem, determine whether the artificial intelligence computing requirements include data labels; if they include data labels, use supervised learning methods to calculate the artificial intelligence computing requirements and obtain the artificial intelligence computing results. If no data labels are included, an unsupervised learning method is used to calculate the artificial intelligence computing requirements and obtain the artificial intelligence computing results. If the category judgment result is an active interference generation problem, the complexity of radar active interference generation is judged according to the artificial intelligence computing requirements. If the complexity of radar active interference generation is greater than the set threshold, a deep reinforcement learning algorithm based on deep neural networks is used to calculate the artificial intelligence computing requirements and obtain the artificial intelligence computing result. If the complexity of generating active radar interference is less than a set threshold, a reinforcement learning algorithm based on Q-tables is used to calculate the artificial intelligence computation requirements and obtain the artificial intelligence computation results. The artificial intelligence computation requirements include the dynamic loading and calling information generated by the reconstruction implementation module and the computation task information and communication task information obtained by the task scheduling submodule.

2. The radar active jamming device capable of real-time dynamic updates as described in claim 1, characterized in that, The hardware resource module includes DSP computing circuits, FPGA computing circuits, CPU computing circuits, GPU computing circuits, storage devices, display and control devices, input / output devices, a high-speed bus, and a software bus. Each pair of the DSP computing circuits, FPGA computing circuits, CPU computing circuits, GPU computing circuits, storage devices, display and control devices, and input / output devices is connected via a high-speed bus. The software bus runs on the high-speed bus. The software bus is used to convert the output signal formats of the different hardware components included in the hardware resource module to obtain a unified format of high-speed digital signal, and then sends the high-speed digital signal to the operating system module. The hardware resource module is connected to the operating system module through various I / O interfaces. The application development module is used to provide users with a development environment for active radar jamming. The application development module includes a radar active jamming development environment, a standard algorithm component library, and a user-defined algorithm component library. The radar active jamming development environment provides users with a software development interface and software component interface for radar active jamming, and provides a compilation and debugging environment for user-developed radar active jamming programs. The application development module receives interference request information from users; converts and processes the interference request information to obtain instruction information; and sends the instruction information to the policy generation submodule of the system monitoring and management module.

3. The radar active jamming device capable of real-time dynamic updates as described in claim 2, characterized in that, The operating system module is used to implement real-time operating system functions, and the operating system module includes a device management submodule, a storage management submodule, a process management submodule, and an I / O management submodule; The device management submodule is used to abstract and describe various types of hardware in the hardware resource module, obtain description models for various types of hardware, and use the description models to manage various types of hardware. The storage management submodule is used to monitor the status of storage units of various hardware components in the hardware resource module and obtain the status monitoring results. Based on the status monitoring results and the middleware's storage requirements, storage units are allocated; The process management submodule is used to monitor and manage various processes of the real-time operating system; The I / O management submodule is used to monitor and manage various I / O interfaces of the hardware resource module.

4. The radar active jamming device capable of real-time dynamic updates as described in claim 2, characterized in that, The middleware module is used to uniformly abstract various types of hardware in the hardware resource module to obtain abstract models of various types of hardware; to adapt the communication protocols of various types of hardware to obtain unified communication protocol data, and to realize data communication between various types of hardware using the unified communication protocol data; to construct service interfaces using the abstract models of various types of hardware; and to construct data resource interfaces using the unified communication protocol data. The service interfaces and data resource interfaces are used by the system monitoring and management module and the refactoring implementation module for use.

5. The radar active jamming device capable of real-time dynamic updates as described in claim 1, characterized in that, The reflection middleware includes a base layer and base objects, a meta-communication model, a meta-layer, and meta-objects. The reflection middleware acquires the dynamic environment in which the application development module is located, represents the dynamic environment as the base layer, and represents the data resource interface and basic function call interface as base objects. The base layer is input into the meta-communication model, and the output of the meta-communication model is used to modify the base layer and base objects to obtain the meta-layer and meta-objects respectively. The meta-objects serve as dynamically updated data resource interfaces and basic function call interfaces. The meta-communication model includes data resource interfaces and basic function call interfaces, and the optimal correspondence between them and the dynamic environment in which the application development module is located. The meta-communication model receives the input base object and, using the optimal correspondence method, outputs the meta-layer and meta-object that optimally correspond to the input base object; the optimally corresponding meta-layer and meta-object refer to the optimal correspondence method between the dynamic environment and data resource interface and the basic function call interface. The dynamic environment in which the application development module exists includes the dynamic loading and calling information generated by the refactoring implementation module and the computational task information and communication task information obtained by the task scheduling submodule.

6. The radar active jamming device capable of real-time dynamic updates as described in claim 2, characterized in that, The system monitoring and management module is used to monitor and manage the software and hardware resources of the radar active jamming device in real time, which can be dynamically updated in real time. The system monitoring and management module includes a strategy generation submodule, a resource allocation submodule, a task scheduling submodule, a status monitoring submodule, and an effect evaluation submodule. The strategy generation submodule is used to generate a radar active jamming strategy based on the instruction information from the application development module and the interference evaluation results from the effect evaluation submodule; and to send the radar active jamming strategy to the resource allocation submodule and the task scheduling submodule. The resource allocation submodule is used to receive the radar active jamming strategy and allocate the software and hardware resources of the radar active jamming device, which can be dynamically updated in real time, according to the requirements of the radar active jamming strategy and the status monitoring results. The task scheduling submodule is used to receive the radar active jamming strategy, decompose the radar active jamming strategy, and obtain the corresponding calculation task information and communication task information. Based on the status monitoring results, the computing task information and communication task information are sent to the middleware module; The status monitoring submodule is used to monitor the software and hardware resources of the radar active jamming device in real time, which can be dynamically updated in real time, and obtain the status monitoring results; and send the status monitoring results to the task scheduling submodule, the resource allocation submodule and the radar active jamming development environment. The effect evaluation submodule is used to receive electromagnetic environment monitoring data sent by the communication middleware, evaluate the effect of radar active interference using the electromagnetic environment monitoring data, obtain interference evaluation results, and send the interference evaluation results to the strategy generation submodule.

7. The radar active jamming device capable of real-time dynamic updates as described in claim 2, characterized in that, The reconstruction implementation module is used to receive and integrate user development requirement information of the radar active jamming development environment, radar active jamming strategy generated by the strategy generation submodule, and jamming evaluation result information of the effect evaluation submodule to obtain reconstruction requirement information; match the reconstruction requirement information with algorithm components to obtain matching service information; use the matching service information to determine dynamic loading component information, and use the dynamic loading component information to dynamically load relevant algorithm components to obtain dynamic loading call information; Dynamically load the call information and send it to the reflection middleware and the artificial intelligence middleware.

8. The radar active jamming device capable of real-time dynamic updates as described in claim 7, characterized in that, The refactoring implementation module includes an event listening and response submodule, a service registration and lookup submodule, and a component lifecycle management submodule; The service registration lookup submodule queries and processes the reconstruction requirement information, extracts the service component registration information from the reconstruction requirement information, and determines whether the service component corresponding to the service component registration information has been registered. If the service component has been registered, the service component registration information of the service component is used as the matching service information and the matching service information is output to the component lifecycle management submodule; If the service component is not registered, the service component registration information is used to match and search all algorithm components. The algorithm component with the highest matching degree with the service component registration information is found and determined as the matching algorithm component. The service corresponding to the matching algorithm component is registered, the service component registration information corresponding to the service is obtained, and a service registration interface and a service reference interface are established. The service component registration information corresponding to the service is output to the component lifecycle management submodule.

Images