Radar signal distributed processing method, system and equipment based on resource calculation

Abstract

The invention provides a radar signal distributed processing method, system and equipment based on resource computing, and the method comprises the steps: a main node determines the priority corresponding to each task and the range of a time window according to the event types of a plurality of to-be-distributed signal processing tasks; on the basis of the constructed influence function of the task complexity of the signal processing task on the consumed time duration, the corresponding consumed time duration is determined according to the priority and complexity corresponding to each task; based on the established radar distributed signal processing task scheduling model, pre-allocating the tasks to the slave nodes according to the priorities, the consumed time lengths and the time window ranges corresponding to the tasks and the computing resources of the slave nodes; based on the established mathematical model for constraining the task allocation of the master node, the allocation condition of each task is dynamically adjusted according to the local CPU utilization rate and the memory use condition fed back by each slave node according to the pre-allocated task, and the method can improve the load balance of the slave nodes and is relatively high in portability.

Description

Technical Field

[0001] The present invention relates to the field of radar technology, and in particular to a radar signal distributed processing method, system and device based on resource calculation. Background Art

[0002] With the increasing complexity of the electromagnetic situation environment, the single processing module of radar reconnaissance equipment can no longer meet the needs of the system, and multi-module distributed processing is the direction of future development. Due to the continuous development of computer hardware technology and software technology, especially the popularization of multi-threaded parallel computing and multi-core processors, radar systems can use more computing resources for signal processing. Since different types of signal processing tasks may have different computing requirements and complexities, in order to achieve efficient signal processing, it is necessary to comprehensively consider factors such as the nature of the task, computing resources, load balancing, etc., and carry out architectural design in terms of task allocation strategy, communication protocol, node management and resource scheduling to achieve optimal performance and efficiency.

[0003] In the existing technical solutions, the Xi'an Institute of Electronic Engineering applied for a "Software-based Radar Signal Processing Method Based on a Central Processing Unit (CPU) Blade Server", with patent number CN116560732A. This method can be run on multiple computing nodes, one of which is a master node responsible for external communication and management of slave nodes; the remaining nodes are slave nodes responsible for completing the tasks assigned by the master node. After receiving the data, the slave nodes use a pipeline method to implement parallel computing. Each module receives the parameters and data output by the upper-level module, and sends it to the next-level module after the calculation is completed. After the slave node completes the calculation, the detection result is sent to the master node, and the master node organizes the final result and passes it to the display control and data processing subsystem.

[0004] In the existing technical solutions, although a master-slave multi-node architecture design is adopted, and the inherent strong coupling of software and hardware in the Field Programmable Gate Array (FPGA) + Digital Signal Processing (DSP) architecture, which currently occupies the mainstream in the field of radar signal processing, is solved, the problems of difficult development and debugging, poor scalability, and poor consistency are caused. However, the disadvantage of this technical solution is that the master node only monitors and manages the number of threads of idle nodes, the constraints are single, and the memory resource utilization rate of each node is not high. Therefore, this method is prone to problems such as task accumulation, event processing time expiration, and scheduling failure when the task volume is heavy. Summary of the invention

[0005] In order to solve the above problems existing in the prior art, the present invention provides a radar signal distributed processing method, system and device based on resource calculation, which specifically include: In a first aspect, the present invention provides a radar signal distributed processing method based on resource calculation, which is applied to a radar signal distributed processing system, wherein the radar signal distributed processing system includes a master node and multiple slave nodes; The method comprises: The master node determines the priority and time window range corresponding to each signal processing task to be assigned according to the event type of the multiple signal processing tasks to be assigned; Based on the influence function of the task complexity of the constructed signal processing task on the time consumption, the master node determines the time consumption corresponding to each signal processing task to be assigned according to the priority and complexity corresponding to each signal processing task to be assigned; Based on the established radar distributed signal processing task scheduling model, the master node pre-allocates each signal processing task to be assigned to each slave node according to the priority, time duration, time window range and computing resources of each slave node corresponding to each signal processing task to be assigned; Based on the established mathematical model for constraining the master node task allocation, the master node dynamically adjusts the allocation of each signal processing task to be allocated according to the local CPU utilization and memory usage fed back by each slave node based on the pre-allocated signal processing tasks.

[0006] In a second aspect, the present invention further provides a radar signal distributed processing system, comprising: One master node and multiple slave nodes; The master node is used to determine the priority and time window range corresponding to each signal processing task to be assigned according to the event type of the multiple signal processing tasks to be assigned; The master node is further used to determine the time consumption corresponding to each signal processing task to be assigned based on the influence function of the task complexity of the constructed signal processing task on the time consumption and according to the priority and complexity corresponding to each signal processing task to be assigned; The master node is further used to pre-allocate each signal processing task to be allocated to each slave node based on the established radar distributed signal processing task scheduling model according to the priority, time duration, time window range and computing resources of each slave node corresponding to each signal processing task to be allocated; Each slave node is used to feedback the local CPU utilization and memory usage to the master node based on the pre-assigned signal processing tasks; The master node is also used to dynamically adjust the allocation of each signal processing task to be allocated based on the established mathematical model for constraining the master node task allocation and according to the local CPU utilization and memory usage fed back by each slave node based on the pre-allocated signal processing tasks.

[0007] In a third aspect, the present invention further provides a radar signal distributed processing device based on resource calculation, comprising: A processing module is used to determine the priority and time window range of each signal processing task to be assigned according to the event type of multiple signal processing tasks to be assigned; based on the constructed influence function of the task complexity of the signal processing task on the processing time, determine the time consumption corresponding to each signal processing task to be assigned according to the priority and complexity corresponding to each signal processing task to be assigned; based on the established radar distributed signal processing task scheduling model, pre-allocate each signal processing task to be assigned to each slave node according to the priority, time consumption, time window range and computing resources of each slave node corresponding to each signal processing task to be assigned; The acquisition module is used to obtain the local CPU utilization and memory usage feedback from each slave node based on the pre-assigned signal processing tasks; The processing module is also used to dynamically adjust the allocation of each signal processing task to be allocated based on the established mathematical model for constraining the master node task allocation and according to the local CPU utilization and memory usage fed back by each slave node based on the pre-allocated signal processing tasks.

[0008] In a fourth aspect, the present invention further provides an electronic device, comprising a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; Memory, used to store computer programs; The processor is used to implement the steps executed by the master node or the steps executed by the slave node in any method provided in the first aspect when executing the program stored in the memory.

[0009] Beneficial effects of the present invention: The radar signal distributed processing method, system and device based on resource calculation provided by the present invention determine the priority and time window range corresponding to each signal processing task to be assigned according to the event type of multiple signal processing tasks to be assigned by the master node; based on the influence function of the task complexity of the constructed signal processing task on the time duration, the master node determines the time duration corresponding to each signal processing task to be assigned according to the priority and complexity corresponding to each signal processing task to be assigned; based on the established radar distributed signal processing task scheduling model, the master node determines the time duration corresponding to each signal processing task to be assigned according to the priority, time duration, time window range and calculation of each slave node Resources are used to pre-allocate each signal processing task to be allocated to each slave node; based on the established mathematical model for constraining the task allocation of the master node, the master node dynamically adjusts the allocation of each signal processing task to be allocated according to the local CPU utilization and memory usage fed back by each slave node based on the pre-allocated signal processing task, so that tasks are allocated to appropriate slave nodes for processing according to different task complexity and priority, the load balancing of the slave nodes is improved, and the portability is better. By introducing constraints on computing resources such as CPU utilization and memory usage, tasks are allocated to appropriate slave nodes for execution, the load balancing of the slave nodes is improved, and the portability is better.

[0010] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 A schematic diagram of the structure of a radar signal distributed processing system provided by the present invention; Figure 2 A schematic diagram of a flow chart of a radar signal distributed processing method based on resource calculation provided by the present invention; Figure 3 A schematic diagram of the structure of a radar signal distributed processing device based on resource calculation provided by the present invention; Figure 4 A schematic diagram of a radar signal distributed processing process based on resource calculation provided by the present invention. DETAILED DESCRIPTION

[0012] The present invention is further described in detail below with reference to specific embodiments, but the embodiments of the present invention are not limited thereto.

[0013] In order to solve the problems existing in the prior art, the present invention provides a radar signal distributed processing method based on resource calculation. The master node adds a computing resource constraint when executing event scheduling and task allocation. First, the master node pre-allocates tasks to other slave nodes according to the complexity and priority of the tasks, and then dynamically adjusts the task allocation in a timely manner according to the current computing resource usage feedback from each slave node, such as CPU utilization, memory usage and other information, so as to allocate the tasks to appropriate processing nodes for execution, thereby maximizing the processing capacity of each signal processing node.

[0014] This method is applied to Figure 1 The radar signal distributed processing system shown in the figure includes a master node and multiple slave nodes, wherein the master node is used as the radar master control to complete data processing and resource scheduling, and the slave node completes signal generation and signal processing. Accordingly, before the method provided by the present invention is executed, it is necessary to pre-deploy the collaborative relationship between the master node and each slave node in the system to establish the control word of the communication mode transmission task.

[0015] like Figure 2 As shown, the radar signal distributed processing method based on resource calculation provided by the present invention includes: S201. The master node determines the priority and time window range corresponding to each signal processing task to be assigned according to the event type of multiple signal processing tasks to be assigned.

[0016] In one possible implementation, the master node determines the priority corresponding to each signal processing task to be assigned according to the event type of the multiple signal processing tasks to be assigned, including: based on the correspondence between the preset event type and priority, the master node determines the priority corresponding to each signal processing task to be assigned according to the event type of the multiple signal processing tasks to be assigned.

[0017] Among them, the event types included in the correspondence between preset event types and priorities are, in descending order of priority, tracking events, target capture, target confirmation, low-altitude search, high-altitude search and imaging events.

[0018] Exemplarily, the correspondence between the preset event types and priorities is shown in Table 1, and it is defined that the smaller the value of P is, the higher the priority of the event is.

[0019] Table 1 Correspondence between preset event types and priorities

[0020] Furthermore, a time window range is set for the task according to different event types to constrain the task execution time range. Optionally, the time window range corresponding to each signal processing task to be assigned is expressed as: , in, Indicates The expected execution time of the signal processing tasks to be assigned, Indicates The time window corresponding to the signal processing tasks to be assigned, Indicates The actual execution time of the signal processing task to be assigned.

[0021] Optionally, before step S201, the master node also allocates the multiple signal processing tasks to be allocated to different slave nodes according to the queue order corresponding to the multiple signal processing tasks to be allocated.

[0022] S202. Based on the influence function of the constructed signal processing task complexity on the time consumption, the master node determines the time consumption corresponding to each signal processing task to be assigned according to the priority and complexity corresponding to each signal processing task to be assigned.

[0023] Optionally, the complexity corresponding to the signal processing task to be assigned is expressed as: , in, Indicates the event type corresponding to the signal processing task to be assigned. Indicates the type of waveform. Indicates the number of sampling points, Indicates the number of pulses. Indicates the length of stay; The influence function of the task complexity of the signal processing task on the time duration is expressed as: , in, Indicates The time consumption corresponding to the signal processing tasks to be assigned is Indicates The complexity of the signal processing tasks to be assigned, A function that represents the impact of the task complexity of the signal processing task on the processing time.

[0024] It should be noted that a task is a control word transmitted during the communication process. One task corresponds to one event, and a task includes events.

[0025] S203. Based on the established radar distributed signal processing task scheduling model, the master node pre-allocates each signal processing task to be allocated to each slave node according to the priority, time duration, time window range and computing resources of each slave node corresponding to each signal processing task to be allocated.

[0026] Specifically, when tasks compete, the expected execution time of tasks is moved within the corresponding time window to solve the event competition problem. For the task request list, the task allocation order is first arranged based on the priority principle, that is, high-priority tasks are obtained and arranged first, and then low-priority tasks are arranged; when task competition occurs, the task allocation order is arranged in combination with the time window range corresponding to each task to solve the task competition problem.

[0027] For signal processing tasks to be assigned and The essence of task scheduling is to process the radar signal distributed processing system of the slave nodes. The tasks are scheduled in a reasonable way Executed on processing nodes, The total completion time of the simulation tasks is the shortest. Therefore, task scheduling can be described as a mapping of task sets to processing node sets under certain constraints.

[0028] The result of the mapping can be described by the scheduling matrix: , The efficiency matrix is ​​composed of There are signal processing tasks to be assigned in A total of 1000 slave nodes are composed of the execution time The matrix: , Each row of the efficiency matrix represents the execution time of a signal processing task to be assigned on different slave nodes, and each column represents the time required for a slave node to process different signal processing tasks to be assigned.

[0029] Optionally, the radar distributed signal processing task scheduling model is expressed as: , , in, Indicates the index of the signal processing task to be assigned, Indicates the index of the slave node number, Indicates the total number of signal processing tasks to be assigned, Indicates the total number of slave nodes. Indicates Whether the signal processing task to be assigned is assigned to The indicator variable of the slave node, Indicates The signal processing task to be assigned is The execution time on each slave node, represents the constraints, and , Indicates that the The signal processing tasks to be assigned are assigned to slave nodes, Indicates that the The signal processing tasks to be assigned are assigned to slave nodes, Indicates The expected execution time of the signal processing tasks to be assigned, Indicates the start time of the current scheduling interval, Indicates The duration of the scheduling interval for the signal processing tasks to be assigned, represents the time window, Indicates the dwell time.

[0030] S204. Based on the established mathematical model constraining the master node task allocation, the master node dynamically adjusts the allocation of each signal processing task to be allocated according to the local CPU utilization and memory usage fed back by each slave node according to the pre-allocated signal processing tasks.

[0031] Specifically, under the feedback mechanism, the slave node sends its local CPU utilization and memory usage back to the master node, and the master node dynamically adjusts the allocation of signal processing tasks and continues to execute the next scheduling. The master node will adjust the allocation of signal processing tasks in a timely manner according to the CPU utilization and memory usage sent back by the slave node in real time, and assign the signal processing tasks to the appropriate slave node. After the slave node processes the signal processing task, it will first put the signal processing result into the priority queue and send it back to the data processing part of the master node. The master node analyzes the data processing results, thereby generating various event requests, and then performs the next resource scheduling.

[0032] Optionally, based on the established mathematical model for constraining the master node task allocation, it is expressed as: , , in, Indicates the total number of signal processing tasks to be assigned, Indicates CPU utilization of slave nodes, Indicates the average CPU utilization of all slave nodes. Indicates Memory usage of slave nodes, Indicates the average memory usage of all slave nodes.

[0033] The method provided by the present invention is explained in detail below with reference to an example, as follows: Under the condition of linear frequency modulation waveform, the more sampling points of the signal, the longer the signal processing time. Then, from the function of the influence of task complexity on the time consumption, it can be concluded that the time consumption of each event is from large to small: imaging event> search event> confirmation event> tracking event> interception event. Assuming that there are multiple slave nodes, and the number of cores configured for the signal processing part of each slave node is the same, the master node first pre-allocates according to the task complexity and priority, and then combines the CPU utilization and memory usage of each slave node in real time to allocate other unallocated signal processing tasks in the task queue to the appropriate slave node for processing, so as to achieve the shortest processing time for all signal processing tasks, and avoid unreasonable task allocation, resulting in task accumulation, node overload and other problems.

[0034] Assume that there are 3 slave nodes and there are 13 signal processing tasks in the task queue, among which Task 2 is an imaging event, Task 1, Task 3, Task 5, Task 6, Task 9, Task 10, and Task 13 are low-altitude search events, Task 4, Task 7, and Task 8 are interception events, and Task 11 and Task 12 are tracking events. First, the master node arranges a task for each slave node at time t0 according to the scheduling order determined by the task queue. The tasks are processed in the slave nodes. Due to the differences in the complexity of each task, the corresponding completion time of the tasks is different. For the first scheduling, slave node 2 first completes Task 1 and becomes idle, but other slave nodes are still processing tasks. At this time, each slave node feeds back the resource situation to the master node. The master node performs a complexity analysis on the remaining tasks in the task queue combined with the event type, waveform type, number of sampling points, number of pulses, dwell time and other parameters of the task, and detects the current resource status of the slave node at the same time, so as to prepare in advance for the next task scheduling. For slave nodes with relatively idle resources, they will complete the current task in a shorter time, so it is possible to consider assigning them another task; while for slave nodes with more resource utilization, they cannot complete the current task in a short time. If they are assigned another task, it may cause too much pressure on the client, thereby reducing its efficiency. Therefore, it is necessary to further use the radar signal distributed processing method based on resource calculation provided by the present invention to assign tasks. The task assignment process is as follows: Figure 4 shown.

[0035] The radar signal distributed processing method based on resource calculation provided by the present invention determines the priority and time window range corresponding to each signal processing task to be assigned according to the event type of multiple signal processing tasks to be assigned by the master node; based on the influence function of the task complexity of the signal processing task on the time-consuming duration constructed, the master node determines the time-consuming duration corresponding to each signal processing task to be assigned according to the priority and complexity corresponding to each signal processing task to be assigned; based on the established radar distributed signal processing task scheduling model, the master node determines the time-consuming duration corresponding to each signal processing task to be assigned according to the priority, time-consuming duration, time window range and computing resources of each slave node. Pre-allocate each signal processing task to be allocated to each slave node; based on the established mathematical model of constrained master node task allocation, the master node dynamically adjusts the allocation of each signal processing task to be allocated according to the local CPU utilization and memory usage fed back by each slave node based on the pre-allocated signal processing task, thereby allocating tasks to appropriate slave nodes for processing according to different task complexity and priority, improving the load balancing of slave nodes, and having better portability. By introducing constraints on computing resources such as CPU utilization and memory usage, tasks are allocated to appropriate slave nodes for execution, improving the load balancing of slave nodes, and having better portability.

[0036] The present invention also provides a radar signal distributed processing system, such as Figure 1 As shown, the system includes: One master node and multiple slave nodes.

[0037] The master node is used to determine the priority and time window range corresponding to each signal processing task to be assigned according to the event type of the multiple signal processing tasks to be assigned.

[0038] The master node is also used to determine the time consumption corresponding to each signal processing task to be assigned based on the influence function of the task complexity of the constructed signal processing task on the processing time and according to the priority and complexity corresponding to each signal processing task to be assigned.

[0039] The master node is also used to pre-allocate each signal processing task to be assigned to each slave node based on the established radar distributed signal processing task scheduling model, according to the priority, time consumption, time window range and computing resources of each slave node corresponding to each signal processing task to be assigned.

[0040] Each slave node is used to feedback the local CPU utilization and memory usage to the master node based on the pre-assigned signal processing tasks.

[0041] The master node is also used to dynamically adjust the allocation of each signal processing task to be allocated based on the established mathematical model for constraining the master node task allocation and according to the local CPU utilization and memory usage fed back by each slave node based on the pre-allocated signal processing tasks.

[0042] The present invention also provides a radar signal distributed processing device based on resource calculation, such as Figure 3 As shown, the device comprises: The processing module 31 is used to determine the priority and time window range corresponding to each signal processing task to be assigned according to the event type of multiple signal processing tasks to be assigned; based on the constructed influence function of the task complexity of the signal processing task on the processing time, determine the time consumption corresponding to each signal processing task to be assigned according to the priority and complexity corresponding to each signal processing task to be assigned; based on the established radar distributed signal processing task scheduling model, pre-allocate each signal processing task to be assigned to each slave node according to the priority, time consumption, time window range and computing resources of each slave node corresponding to each signal processing task to be assigned.

[0043] The acquisition module 32 is used to obtain the local CPU utilization and memory usage fed back by each slave node according to the pre-assigned signal processing task.

[0044] The processing module 31 is also used to dynamically adjust the allocation of each signal processing task to be allocated based on the established mathematical model for constraining the master node task allocation and according to the local CPU utilization and memory usage fed back by each slave node based on the pre-allocated signal processing tasks.

[0045] The present invention also provides a structure of an electronic device, including a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus. Memory, used to store computer programs; The processor is used to implement the steps executed by the master node or the steps executed by the slave node in the above method embodiment when executing the program stored in the memory.

[0046] The communication interface is used for communication between the above electronic device and other devices.

[0047] The method provided in the embodiment of the present invention can be applied to electronic devices. Specifically, the electronic device can be: a desktop computer, a portable computer, an intelligent mobile terminal, a server, etc. This is not limited here, and any electronic device that can implement the present invention belongs to the protection scope of the present invention.

[0048] As for the system / device / electronic device embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the specific contents and beneficial effects and other related matters can be referred to the partial description of the method embodiments.

[0049] The terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0050] The above contents are further detailed descriptions of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, several simple deductions or substitutions can be made without departing from the concept of the present invention, which should be regarded as falling within the protection scope of the present invention.

Claims

1. A radar signal distributed processing method based on resource calculation, characterized in that: Applied to a radar signal distributed processing system, the radar signal distributed processing system comprising a master node and multiple slave nodes; The method comprises: The master node determines the priority and time window range corresponding to each of the signal processing tasks to be assigned according to the event types of the signal processing tasks to be assigned; Based on the influence function of the task complexity of the constructed signal processing task on the time consumption, the master node determines the time consumption corresponding to each of the signal processing tasks to be assigned according to the priority and complexity corresponding to each of the signal processing tasks to be assigned; Based on the established radar distributed signal processing task scheduling model, the master node pre-allocates each of the signal processing tasks to be allocated to each of the slave nodes according to the priority, time duration, time window range and computing resources of each of the slave nodes corresponding to each of the signal processing tasks to be allocated; Based on the established mathematical model for constraining the master node task allocation, the master node dynamically adjusts the allocation of each signal processing task to be allocated according to the local CPU utilization and memory usage fed back by each slave node based on the pre-allocated signal processing tasks.

2. The method according to claim 1, characterized in that The master node determines the priority corresponding to each of the signal processing tasks to be assigned according to the event type of the signal processing tasks to be assigned, including: Based on the preset correspondence between event types and priorities, the master node determines the priority corresponding to each of the signal processing tasks to be assigned according to the event types of the multiple signal processing tasks to be assigned; Among them, the event types included in the correspondence between the preset event types and priorities are, in descending order of priority, tracking events, target capture, target confirmation, low-altitude search, high-altitude search and imaging events.

3. The method according to claim 2, characterized in that The range of the time window corresponding to each of the signal processing tasks to be assigned is expressed as: ， in, Indicates The expected execution time of the signal processing tasks to be assigned, Indicates The time window corresponding to the signal processing tasks to be assigned, Indicates The actual execution time of the signal processing task to be assigned.

4. The method according to claim 3, characterized in that The complexity corresponding to the signal processing task to be assigned is expressed as: ， in, Indicates the event type corresponding to the signal processing task to be assigned. Indicates the type of waveform. Indicates the number of sampling points, Indicates the number of pulses. Indicates the length of stay; The influence function of the task complexity of the signal processing task on the processing time is expressed as: ， in, Indicates The time consumption corresponding to the signal processing tasks to be assigned is Indicates The complexity of the signal processing tasks to be assigned, A function that represents the impact of the task complexity of the signal processing task on the processing time.

5. The method according to claim 4, characterized in that The radar distributed signal processing task scheduling model is expressed as: ， ， in, Indicates the index of the signal processing task to be assigned, Indicates the index of the slave node number, Indicates the total number of signal processing tasks to be assigned, Indicates the total number of slave nodes. Indicates Whether the signal processing task to be assigned is assigned to The indicator variable of the slave node, Indicates The signal processing task to be assigned is The execution time on each slave node, represents the constraints, and , Indicates that the The signal processing tasks to be assigned are assigned to slave nodes, Indicates that the The signal processing tasks to be assigned are assigned to slave nodes, Indicates The expected execution time of the signal processing tasks to be assigned, Indicates the start time of the current scheduling interval, Indicates The duration of the scheduling interval for the signal processing tasks to be assigned, represents the time window, Indicates the dwell time.

6. The method according to claim 5, characterized in that The mathematical model based on the established constraint master node task allocation is expressed as: ， ， in, Indicates the total number of signal processing tasks to be assigned, Indicates CPU utilization of slave nodes, Indicates the average CPU utilization of all slave nodes. Indicates Memory usage of slave nodes, Indicates the average memory usage of all slave nodes.

7. The method according to any one of claims 1 to 6, characterized in that: The communication method between the master node and each slave node includes: one or more of UDP or TCP network, optical fiber, and Ethernet.

8. A radar signal distributed processing system, characterized in that: include: One master node and multiple slave nodes; The master node is used to determine the priority corresponding to each of the signal processing tasks to be assigned and the range of the time window corresponding to each of the signal processing tasks to be assigned according to the event type of the multiple signal processing tasks to be assigned; The master node is further used to determine the time consumption corresponding to each of the signal processing tasks to be assigned based on the influence function of the task complexity of the constructed signal processing task on the time consumption and according to the priority and complexity corresponding to each of the signal processing tasks to be assigned; The master node is further used to pre-allocate each of the signal processing tasks to be allocated to each of the slave nodes based on the established radar distributed signal processing task scheduling model and according to the priority, time duration, time window range and computing resources of each of the slave nodes corresponding to each of the signal processing tasks to be allocated; Each of the slave nodes is used to feed back local CPU utilization and memory usage to the master node according to the pre-assigned signal processing tasks; The master node is also used to dynamically adjust the allocation of each signal processing task to be allocated based on the established mathematical model for constraining the master node task allocation and according to the local CPU utilization and memory usage fed back by each slave node based on the pre-allocated signal processing tasks.

9. A radar signal distributed processing device based on resource calculation, characterized in that: include: A processing module, used to determine the priority of each signal processing task to be assigned and the range of the time window corresponding to each signal processing task to be assigned according to the event type of multiple signal processing tasks to be assigned; based on the constructed influence function of the task complexity of the signal processing task on the processing time, determine the time consumption corresponding to each signal processing task to be assigned according to the priority and complexity corresponding to each signal processing task to be assigned; based on the established radar distributed signal processing task scheduling model, pre-allocate each signal processing task to be assigned to each slave node according to the priority, time consumption, range of the time window corresponding to each signal processing task to be assigned and the computing resources of each slave node; An acquisition module is used to obtain the local CPU utilization and memory usage fed back by each slave node according to the pre-assigned signal processing task; The processing module is also used to dynamically adjust the allocation of each signal processing task to be allocated based on the established mathematical model for constraining the master node task allocation and according to the local CPU utilization and memory usage fed back by each slave node based on the pre-allocated signal processing tasks.

10. An electronic device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, used to implement the steps executed by the master node or the steps executed by the slave node in any one of the methods described in claims 1-7 when executing the program stored in the memory.

Images