CN111651278B - Dynamic reconstruction method and platform based on software radar - Google Patents
Dynamic reconstruction method and platform based on software radar Download PDFInfo
- Publication number
- CN111651278B CN111651278B CN202010603849.3A CN202010603849A CN111651278B CN 111651278 B CN111651278 B CN 111651278B CN 202010603849 A CN202010603849 A CN 202010603849A CN 111651278 B CN111651278 B CN 111651278B
- Authority
- CN
- China
- Prior art keywords
- node
- real
- fault
- component
- time
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
- G06F9/505—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals considering the load
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
Abstract
The invention discloses a dynamic reconstruction method based on a software radar, which comprises the following steps: the real-time dispatcher receives the system information sent by the resource manager; the real-time dispatcher receives node fault information or component fault information sent by the real-time fault diagnosis device; and the real-time scheduler performs component scheduling, and the real-time scheduler performs fault node scheduling. The invention can utilize the residual resources to carry out real-time dynamic reconstruction when the radar back-end system locally fails, ensures the basic operation capability of the system and improves the stability.
Description
Technical Field
The invention relates to the field of software radars, in particular to a dynamic reconstruction method and platform based on the software radars.
Background
The radar back-end system is required to be expandable in task-oriented functions, easy to upgrade in processing system algorithm and capable of continuously improving system performance, so that new combat requirements can be responded timely.
The functions and combat tasks undertaken by the radar are also becoming diversified, and the radar is developing towards the integration of detection, electronic combat, communication and the like. These requirements have led radar equipment to develop towards features such as demand customizable, hardware reconfigurable and software reconfigurable, to meet the reconfigurability of systems in multi-functional, multi-tasked, complex and diverse environments towards dynamic environments and task demands, towards the scalability and maintainability of longer equipment lifecycles. However, the conventional radar has the problems of binding application functions with hardware, difficult function reconstruction and the like, so that a technology capable of realizing real-time dynamic reconstruction of radar components is urgently needed, thereby improving the upgrading and maintenance efficiency of a radar system, ensuring that the system can rapidly and adaptively operate under different hardware configuration conditions, and improving the stability of the system.
Disclosure of Invention
In order to solve the above problems, the present invention provides a dynamic reconfiguration method based on a software radar, comprising the following steps:
the real-time dispatcher receives the system information sent by the resource manager;
the real-time dispatcher receives node fault information or component fault information sent by the real-time fault diagnosis device;
the real-time scheduler performs component scheduling, wherein the component scheduling specifically comprises the following steps:
traversing all nodes, preselecting nodes meeting the operation requirement of the component according to the system information,
performing a preferred policy on the pre-selected node traversal to obtain the best scheduling node,
the real-time dispatcher issues executable programs and running files of the components to the optimal dispatching node and executes the executable programs and the running files;
the real-time scheduler performs fault node scheduling, wherein the fault node scheduling comprises:
traversing all nodes, preselecting nodes meeting the operation requirements of all components on the fault node according to the system information,
performing a preferred policy on the pre-selected node traversal to obtain the best scheduling node,
and the real-time scheduler issues executable programs and running files of all components of the fault node to the optimal scheduling node and executes the executable programs and the running files, and if the optimal scheduling node is not found, the real-time scheduler performs component scheduling on each component on the fault node.
Further, the method comprises the steps of,
when the component is required to be updated or a component fault occurs, the real-time scheduler performs component scheduling;
when node faults occur, the real-time scheduler performs fault node scheduling.
Further, the method comprises the steps of,
the node fault evaluation method comprises the steps that each node sends heartbeat information to a real-time fault diagnosis device according to a preset period, and if the real-time fault diagnosis device continuously receives the heartbeat information of a certain node for three times, the node is judged to be in a node fault state;
the component fault message is obtained by the real-time fault diagnosis device in a thread monitoring mode.
Further, the method comprises the steps of,
the resource manager comprises a resource management server and resource management clients deployed on all nodes of the radar back-end system, wherein the resource management clients acquire node state information and task state information from all the nodes and send the node state information and the task state information to the resource management server, the resource management server processes the node state information and the task state information to obtain system information, and the system information comprises task types, task resource requirements, task memory requirements, node core resource utilization rate and node memory resource utilization rate of components.
Further, the method comprises the steps of,
the optimization strategy is specifically that scoring is carried out according to the node core resource utilization rate and the node memory resource utilization rate weighting, and the node with the highest score is the optimal scheduling node.
The dynamic reconfiguration platform based on the software radar comprises a resource manager, a real-time scheduler and a real-time fault diagnosis device, wherein the resource manager comprises a client and a server, the client acquires node state information and task state information in a radar back-end system, the server receives the node state information and the task state information sent by the client, and the server processes the node state information and the task state information to obtain system information and sends the system information to the real-time scheduler; the real-time fault diagnosis device judges whether node faults or component faults occur, and if the node faults or component faults occur, the node fault message or component fault message is sent to the real-time dispatcher; the real-time scheduler performs component scheduling or fault node scheduling according to the node fault message, the component fault message or the system information; the dynamic reconfiguration platform performs dynamic reconfiguration according to the dynamic reconfiguration method.
Compared with the prior art, the invention has the following beneficial effects:
the invention can utilize the residual resources to carry out real-time dynamic reconstruction when the radar system locally breaks down, thereby ensuring the basic operation capability of the system and improving the stability;
the invention fully utilizes the resources by real-time scheduling based on the existing resources, avoids partial full load and partial idle caused by uneven resource allocation, and improves the utilization efficiency of the resources;
the invention provides the function of updating the application in real time, greatly improves the expandability and reduces the upgrade maintenance cost.
Drawings
FIG. 1 is a block diagram of a software radar-based component dynamic reconfiguration platform.
Fig. 2 is a block diagram of an embodiment of the present invention.
Detailed Description
The following describes in detail a dynamic reconfiguration platform and a method based on a software radar according to the present invention with reference to the accompanying drawings.
Example 1
As shown in fig. 1, the dynamic reconfiguration platform based on the software radar provided by the invention comprises:
the resource manager is in charge of acquiring node state information in the radar back-end system and providing the node state information to an upper service global resource view; and counting task state information on each node in the system, and providing the task state information for the upper layer service global task view. Specifically, the resource management module comprises a resource management client and a server, the resource management client is deployed on each node of the radar back-end system, acquires node state information and task state information through an operating system call interface according to a configurable beat (default is 500 ms), and sends the node state information and the task state information to the resource management server through TCP according to a fixed format, and the resource management server collects the node state information and the task state information of all the nodes and integrates the node state information and the task state information to obtain system information, and provides the system information to the real-time fault diagnosis device and the real-time dispatcher; the system information comprises the task type, task resource requirement, task memory requirement, node core resource utilization rate, node memory resource utilization rate and the like of the component; as shown in table 1, the node status information specifically includes a node ID, a node IP, a CPU core number, a CPU occupancy rate, a memory size, a memory usage amount, an operating system type, and a platform type; as shown in table 2, the task state information on each node includes a task name, an entry function name, a task ID, a system priority of a task, a task occupation core number, a task CPU utilization rate, a task memory usage amount, a task running state, a task running error number and a task type; the system information comprises the task type, task resource requirement, task memory requirement, node core resource utilization rate and node memory resource utilization rate of the component.
Sequence number | Name of the name | Type (C++) | Remarks |
1 | Node ID | int | |
2 | Node IP | string | |
3 | CPU core number | unsigned int | MB |
4 | CPU occupancy rate | double | Percentage of |
5 | Memory size | double | MB |
6 | Memory usage | double | MB |
7 | Operating system type | string | Such as Linux, vxworks, windows, etc |
8 | Platform type | string | Such as x86, PPC, etc |
TABLE 1 node State information
Sequence number | Name of the name | Type (C++) | Remarks |
1 | Task name | string | |
2 | Inlet function name | string | |
3 | Task ID | int | |
4 | System priority of tasks | int | |
5 | Task number of occupied cores | int | |
6 | Task CPU utilization | double | Percentage of |
7 | Task memory usage | double | MB |
8 | Task running status | string | |
9 | Task running error number | int | |
10 | Task type | string | Such as Linux task, vxworks task, windows task, etc |
TABLE 2 task state information
The real-time fault diagnosis device evaluates the node fault or the component fault in real time, and if the node fault or the component fault occurs, the real-time dispatcher is informed to carry out dynamic reconstruction; the node fault is to judge whether the node is faulty or not through a node heartbeat mechanism, each node sends heartbeat information to a real-time fault diagnostor according to a period T (T is configurable and defaults to 500 ms), and if the heartbeat information is not received three times continuously, the node is judged to be in a node fault state; the component fault is to acquire the abnormal state of the task in real time by a mode of monitoring threads;
and the real-time scheduler comprises component scheduling and fault node scheduling, and real-time deploys component tasks to complete component scheduling under the condition that component updating or component fault is required. When the radar back-end system has node faults, the whole component scheduling of the fault node is preferentially carried out, and if the whole component scheduling fails, the individual component scheduling is carried out on each component on the fault node.
The component scheduling is specifically that when the radar component completes the initial deployment of an executable program of a certain function of the radar or component updating is required or component faults occur, a new dynamic reconstruction scheme is generated according to the task resource requirement of the component, the task memory requirement and the available resources in the current system. Firstly, acquiring system information through a resource manager; secondly, preselecting the nodes by utilizing the task types (shown in a table 2) of the components, task core resource requirements and task memory requirements; thirdly, traversing a preferred strategy for executing scheduling on the pre-selected nodes, namely scoring according to the node core resource utilization rate and node memory resource utilization rate weighting (50 percent each), wherein the highest scoring node is the optimal scheduling node of the component; and finally, issuing the executable program and related running files of the component to the optimal scheduling node according to the formed deployment scheme, and executing the component.
The fault node scheduling is specifically that when the node of the radar back-end system fails, real-time dynamic reconstruction is carried out according to system information, and the method specifically comprises the following steps:
1) Traversing all nodes, and finding out a node capable of supporting the operation of all components on the fault node (meeting the core resources and memory resources required by the task operation of the components) through the pre-selection and optimization strategies for deployment;
2) And if the node searching in the step 1) fails, respectively carrying out component scheduling on each component on the failed node.
The method for realizing the dynamic reconfiguration based on the software radar comprises the following steps:
the resource management client is deployed on each node of the system, periodically invokes a resource acquisition interface adapted to the node operating system, acquires node state information of the node and task state information corresponding to the component, reports the node state information and the task state information to the resource management server through TCP according to a configurable beat, and the reporting beat is set to be 500ms by default, so that the real-time performance of state update is ensured, and the system is not obviously burdened; the real-time fault diagnosis module diagnoses the possible node fault or component fault of the radar back-end system and informs the real-time dispatcher through the message queue; the real-time scheduler generates and executes a dynamic reconfiguration scheme in real time, directly deploys the newly added component, schedules the running component to other nodes, stops running the component and exits the component, and then performs corresponding deployment operation.
Example two
The embodiment specifically describes a case when the real-time fault diagnosis device monitors that a certain node has a fault. When the real-time fault diagnosis device monitors that a certain node has faults, the real-time fault diagnosis device sends a node fault message to the real-time dispatcher; the real-time scheduler firstly tries to integrally migrate the components on the fault node to other nodes based on the bottom hardware computing resource information obtained by the resource manager, and if no single node has resources supporting the operation of all the components on the fault node, the real-time scheduler respectively and independently performs component scheduling on each component of the fault node.
Example III
The present embodiment specifically describes a case where component update is required. As shown in fig. 2, the radar back-end system generates a mapping scheme of each component and node of the radar signal processing workflow corresponding to the configuration 1 through the real-time scheduler based on the radar signal processing workflow corresponding to the configuration 1, and starts each component to operate based on the scheme. The generation of the mapping scheme comprises the following steps: a) The workflow corresponding to configuration 1 of fig. 2 comprises 8 components and the data flow topological relation among them, traversing from the root node (side lobe cancellation), traversing all 8 components according to the adjacent relation (specifically, side lobe cancellation, anti-narrow pulse, pulse compression, MTI, CFAR, hidden shadow, trace processing and trace processing shown in fig. 2); b) For each traversed component, scheduling the traversed component to a certain node by using a real-time scheduler; c) When all 8 components are scheduled, the real-time scheduler uniformly starts the components to run. In the radar operation process, a new component update application appears, the new component update application corresponds to the configuration 2 in fig. 2, at this time, the operation is stopped and the component corresponding to the configuration 1 is closed, then each component of the radar signal processing workflow corresponding to the configuration 2 is scheduled to a hardware node, and the scheduling function is consistent with the scheduling process of the configuration 1.
Compared with the prior art, the invention has the following beneficial effects:
the invention can utilize the residual resources to carry out real-time dynamic reconstruction when the radar back-end system locally fails, thereby ensuring the basic operation capability of the system and improving the stability;
the invention fully utilizes the resources by real-time scheduling based on the existing resources, avoids partial full load and partial idle caused by uneven resource allocation, and improves the utilization efficiency of the resources;
the invention provides the function of updating the application in real time, greatly improves the expandability and reduces the upgrade maintenance cost.
The foregoing description of the preferred embodiments of the invention is not intended to limit the invention to the precise form disclosed, and any such modifications, equivalents, and alternatives falling within the spirit and scope of the invention are intended to be included within the scope of the invention.
Claims (4)
1. The dynamic reconstruction method based on the software radar is characterized by comprising the following steps of:
the real-time dispatcher receives the system information sent by the resource manager;
the real-time dispatcher receives node fault information or component fault information sent by the real-time fault diagnosis device;
the real-time scheduler performs component scheduling, wherein the component scheduling specifically comprises the following steps:
traversing all nodes, preselecting nodes meeting the operation requirement of the component according to the system information,
performing a preferred policy on the pre-selected node traversal to obtain the best scheduling node,
the real-time dispatcher issues executable programs and running files of the components to the optimal dispatching node and executes the executable programs and the running files;
the real-time scheduler performs fault node scheduling, wherein the fault node scheduling comprises:
traversing all nodes, preselecting nodes meeting the operation requirements of all components on the fault node according to the system information, traversing the preselected nodes to execute a preferred strategy to obtain optimal scheduling nodes,
the real-time scheduler issues executable programs and running files of all components of the fault node to the optimal scheduling node and executes the executable programs and the running files, and if the optimal scheduling node is not found, each component on the fault node is respectively subjected to component scheduling;
the resource manager comprises a resource management server and resource management clients deployed on all nodes of the radar back-end system, wherein the resource management clients acquire node state information and task state information from all the nodes and send the node state information and the task state information to the resource management server, the resource management server processes the node state information and the task state information to acquire system information, and the system information comprises task types, task resource requirements, task memory requirements, node core resource utilization rate and node memory resource utilization rate of components;
the optimization strategy is specifically that scoring is carried out according to the node core resource utilization rate and the node memory resource utilization rate weighting, and the node with the highest score is the optimal scheduling node.
2. The method for dynamic reconfiguration based on a software radar according to claim 1, wherein,
when the component is required to be updated or a component fault occurs, the real-time scheduler performs component scheduling;
when node faults occur, the real-time scheduler performs fault node scheduling.
3. The method for dynamic reconfiguration based on a software radar according to claim 2, wherein,
the node fault evaluation method comprises the steps that each node sends heartbeat information to a real-time fault diagnosis device according to a preset period, and if the real-time fault diagnosis device continuously receives the heartbeat information of a certain node for three times, the node is judged to be in a node fault state;
the component fault message is obtained by the real-time fault diagnosis device in a thread monitoring mode.
4. The dynamic reconfiguration platform based on the software radar is characterized by comprising a resource manager, a real-time scheduler and a real-time fault diagnosis device, wherein the resource manager comprises a client and a server, the client acquires node state information and task state information in a radar back-end system, the server receives the node state information and the task state information sent by the client, and the server processes the node state information and the task state information to obtain system information and then sends the system information to the real-time scheduler; the real-time fault diagnosis device judges whether node faults or component faults occur, and if the node faults or component faults occur, the node fault message or component fault message is sent to the real-time dispatcher; the real-time scheduler performs component scheduling or fault node scheduling according to the node fault message, the component fault message or the system information; the dynamic reconfiguration platform performs dynamic reconfiguration according to the method of any one of claims 1 to 3.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010603849.3A CN111651278B (en) | 2020-06-29 | 2020-06-29 | Dynamic reconstruction method and platform based on software radar |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010603849.3A CN111651278B (en) | 2020-06-29 | 2020-06-29 | Dynamic reconstruction method and platform based on software radar |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111651278A CN111651278A (en) | 2020-09-11 |
CN111651278B true CN111651278B (en) | 2023-08-01 |
Family
ID=72343085
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010603849.3A Active CN111651278B (en) | 2020-06-29 | 2020-06-29 | Dynamic reconstruction method and platform based on software radar |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111651278B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110361961B (en) * | 2019-07-10 | 2022-08-23 | 中国船舶工业综合技术经济研究院 | Dynamic reconfiguration design method for networked weapon control system |
Family Cites Families (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104765024A (en) * | 2015-04-22 | 2015-07-08 | 芜湖航飞科技股份有限公司 | Onboard radar jamming automatic detection system |
CN106932764A (en) * | 2017-04-01 | 2017-07-07 | 中国电子科技集团公司第三十八研究所 | The index test of radar HF receiving subsystem module and fault location system and its method |
CN107315168B (en) * | 2017-07-11 | 2020-09-04 | 电子科技大学 | Software radar signal data processing system and method |
TWI686696B (en) * | 2018-08-14 | 2020-03-01 | 財團法人工業技術研究院 | Compute node, failure detection method thereof and cloud data processing system |
CN110659045B (en) * | 2019-08-30 | 2022-03-15 | 电子科技大学 | Software radar real-time dynamic updating system |
-
2020
- 2020-06-29 CN CN202010603849.3A patent/CN111651278B/en active Active
Also Published As
Publication number | Publication date |
---|---|
CN111651278A (en) | 2020-09-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Sun et al. | Rose: Cluster resource scheduling via speculative over-subscription | |
CN109347974B (en) | Hybrid scheduling system for improving online service quality and cluster resource utilization rate | |
CN113051075B (en) | Kubernetes intelligent capacity expansion method and device | |
CN112214280B (en) | Cloud method and system for electric power system simulation | |
CN109343939B (en) | Distributed cluster and parallel computing task scheduling method | |
CN111930493A (en) | NodeManager state management method and device in cluster and computing equipment | |
CN111651278B (en) | Dynamic reconstruction method and platform based on software radar | |
CN112559159A (en) | Task scheduling method based on distributed deployment | |
EP3274828B1 (en) | Methods and nodes for scheduling data processing | |
CN103827831A (en) | Live module diagnostic testing | |
CN111767145A (en) | Container scheduling system, method, device and equipment | |
CN110096339A (en) | A kind of scalable appearance configuration recommendation system and method realized based on system load | |
CN112395052B (en) | Cluster resource management method and system based on container for hybrid load | |
CN109936471B (en) | Multi-cluster resource allocation method and device | |
CN112636990A (en) | Resource management system | |
CN117435324A (en) | Task scheduling method based on containerization | |
Dai et al. | Towards scalable resource management for supercomputers | |
CN114035919A (en) | Task scheduling system and method based on power distribution network layered distribution characteristics | |
CN109634749B (en) | Distributed unified scheduling method and device | |
CN113986662A (en) | Edge cluster monitoring method and system | |
CN112231223A (en) | Distributed automatic software testing method and system based on MQTT | |
Favarim et al. | Adaptive Fault-Tolerant CORBA Components. | |
Liu et al. | Distributed ale in rfid middleware | |
CN112148462A (en) | Jenkins-based CICD process processing method | |
CN115550371B (en) | Pod scheduling method and system based on Kubernetes and cloud platform |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |