A scheduling method and device for resource allocation management of electronic information system

Abstract

The present application provides a scheduling method and device for resource allocation management of an electronic information system, belonging to the field of electronic information technology. The method includes: sorting task requests arriving at an avionics system within a time slot; determining a candidate set of task start execution times for all sorted task requests; allocating resources to corresponding task requests based on the selected candidate set of task start execution times; when resource allocation fails, releasing resources and allocating resources to the corresponding task request again. The present application does not pre-allocate fixed resources, so the success rate of task request execution is improved, while reducing the workload and difficulty of the initial design stage and improving resource utilization, and is used for avionics systems.

Description

Technical Field

[0001] The present application relates to the field of electronic information technology, and in particular to a scheduling method and device for resource allocation management of an electronic information system. Background Art

[0002] Thanks to the rapid development of electronic information, aircraft real-time information processing systems have achieved certain results in modularization and integration. Based on mission requirements, information networks are used to manage mission execution in an integrated manner. Through information sharing, mission progress can be monitored in real time, enabling better decision-making.

[0003] Integrated Modular Avionics (IMA) is the mainstream avionics system architecture for current aircraft platforms. Functional application software is no longer tied to hardware devices. Instead, software is deployed to hardware devices that meet the requirements based on resource availability, greatly increasing system flexibility. Distributed Integrated Modular Avionics (DIMA) is a new type of avionics system. It adopts a distributed system architecture, distributing information processing hardware resources. Each device communicates via a network, and functional software is deployed locally on demand. DIMA not only shares the modularity of IMA but also utilizes a network to connect individual modules. Placing devices near sensors improves signal source data preprocessing capabilities, while the network facilitates data sharing between modules and reduces hardware redundancy. Building on DIMA's foundation, avionics systems are gradually evolving towards a systematized approach, leveraging information networks. This system leverages resources from different flight platforms to build a multi-platform avionics system encompassing the avionics systems of various aircraft. This allows flight formations to leverage the resource strengths of different platforms and leverage synergy to enhance their ability to execute complex missions.

[0004] However, in order to complete the execution of mission requests, the DIMA avionics system uses pre-allocated fixed resources, which can easily lead to mission request execution failures. At the same time, this method makes the initial design phase more labor-intensive and difficult, and the pre-allocated resource utilization rate is low. Summary of the Invention

[0005] In order to solve the problems in related technologies that task requests are prone to failure, the design phase is labor-intensive and difficult, and the pre-allocation method has low resource utilization, the present application provides a scheduling method and device for resource allocation management of an electronic information system. The technical solution is as follows:

[0006] In a first aspect, a scheduling method for resource allocation management of an electronic information system is provided, the method comprising:

[0007] Sequencing mission requests arriving at the avionics system within a time slot;

[0008] Determine the candidate set of task start execution time for all sorted task requests;

[0009] Allocate resources to the corresponding task request based on the candidate set of task start execution time;

[0010] When resource allocation fails, the resources are released and allocated again to the corresponding task request.

[0011] The step of sorting the task requests arriving at the avionics system within a time slot includes:

[0012] Sort multiple task requests by priority;

[0013] For task requests of the same priority, task requests with execution duration greater than or equal to an adaptive threshold are regarded as long task requests, and task requests with execution duration less than an adaptive threshold are regarded as short task requests. The adaptive threshold is the average execution duration of all task requests received in the time slot.

[0014] For the long task request set, the long task requests are sorted in descending order according to the density; when the density of the long task requests is the same, the long task requests are sorted in descending order according to the number of sensor types to obtain the first set;

[0015] For the short task request set, the short task requests are sorted in descending order according to the density; when the density of the short task requests is the same, the short task requests are sorted in descending order according to the number of sensor types to obtain a second set;

[0016] Merging the first set and the second set to obtain a third set, wherein the long task requests and the short task requests in the third set are arranged alternately;

[0017] The closeness is the ratio of the execution duration of the task request to the length of the time window.

[0018] The candidate set of task start execution times determined for all sorted task requests includes:

[0019] A candidate set of task start execution times is determined according to the type of each task request, the required execution duration, and the time window, wherein the type of the task request includes a short task request and a long task request.

[0020] The step of determining a candidate set of task execution start times based on the type, required execution duration, and time window of each task request includes:

[0021] For short task requests, Δ k As a candidate set of task start execution time for this short task request, Indicates the time interval for the short task request to run;

[0022] For a long task request, all time slots in the candidate set of the task start execution time of the long task request are traversed. The method of determining the candidate set of the task start execution time of the long task request is the same as that of the candidate set of the task start execution time of the short task request. For each time slot i, calculate the time slot [i,d k +i) The average number of tasks μ i And the time slot [i,d k +i) the standard deviation σ of the number of tasks i , when the average number of tasks μ i and standard deviation σ i When the preset conditions are met, time slot i is used as an element in the candidate set of the task start execution time of the long task request. The preset conditions are: μ i <μ||(μ i =μ&σ i <σ);

[0023] Among them, d k The execution duration requested for the task.

[0024] The allocating resources to the corresponding task request according to the candidate set of the selected task start execution time includes:

[0025] Establishing a weighted directed graph corresponding to each time slot in the candidate set according to available nodes and links in the avionics system network, wherein the weighted directed graph corresponding to the time slot represents available resource information of the avionics system resource model of the time slot;

[0026] Select a weighted directed graph with the least available resources from the weighted directed graphs corresponding to all time slots as the target weighted directed graph, where the target weighted directed graph includes the least number of nodes and links;

[0027] Resources are allocated to task requests based on a target weighted directed graph and a task request resource graph, which includes computer resources required by each function when the task is executed, as well as bandwidth requirements for communication between functions.

[0028] The allocating resources to the task request according to the target weighted directed graph and the task request resource graph includes:

[0029] According to the bottom-up order of available nodes in the target weighted directed graph, combined with the mission request resource graph, the availability of the avionics system is checked and resources are allocated to the mission request.

[0030] Furthermore, the method further comprises:

[0031] upon receiving a first task request, releasing resources allocated by an unexecuted second task request, the priority of the first task request being higher than the priority of the second task request;

[0032] The released resources are allocated to the first task request.

[0033] Furthermore, the method further comprises:

[0034] When resource allocation is successful, the task request is executed and the resource allocation information is recorded.

[0035] In a second aspect, a scheduling device for resource allocation management of an electronic information system is provided, the device comprising: a sorting module, a time selection module, a resource allocation module, and a resource release module.

[0036] a sequencing module for sequencing mission requests arriving at a multi-platform avionics system within a time slot;

[0037] The time selection module is used to determine the candidate set of task start execution time for all sorted task requests;

[0038] A resource allocation module is used to allocate resources to corresponding task requests based on the candidate set of task start execution times;

[0039] The resource release module is used to release resources when resource allocation fails and allocate resources to the corresponding task request again.

[0040] The present invention provides a scheduling method and device for resource allocation management of an electronic information system, which is used to solve the problem of allocating effective system resources to functional application software, improve the reliability and security of the application system under the system; design a collaborative scheduling strategy for tasks on shared resources to improve resource utilization; and assist in carrying out system design work to improve design efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 A flowchart of a scheduling method for resource allocation management of an electronic information system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0042] The present application is further described in detail below through specific implementation methods and drawings.

[0043] This application provides a scheduling method for resource allocation management of an electronic information system, such as Figure 1 Shown, including:

[0044] Step 110: sorting the task requests arriving at the avionics system within a time slot;

[0045] Step 120: Determine a candidate set of task execution start times for all sorted task requests;

[0046] Step 130: Allocate resources to the corresponding task request according to the candidate set of task start execution times;

[0047] Step 140: When resource allocation fails, release the resources and allocate resources to the corresponding task request again.

[0048] Compared with related technologies, this application does not pre-allocate fixed resources, so the success rate of task request execution is improved, while reducing the workload and difficulty of the initial design stage and improving resource utilization.

[0049] The present application also provides a scheduling method for resource allocation management of an electronic information system, comprising:

[0050] Step 210: sorting the task requests arriving at the avionics system within a time slot;

[0051] The task requests arriving at the avionics system within a time slot are sequenced, including:

[0052] Categorize multiple task requests according to priority and sort them separately;

[0053] For task requests of the same priority, task requests with execution duration greater than or equal to an adaptive threshold are regarded as long task requests, and task requests with execution duration less than an adaptive threshold are regarded as short task requests. The adaptive threshold is the average execution duration of all task requests received in the time slot.

[0054] For the long task request set, the long task requests are sorted in descending order according to the density; when the density of the long task requests is the same, the long task requests are sorted in descending order according to the number of sensor types to obtain the first set;

[0055] For the short task request set, the short task requests are sorted in descending order according to the density; when the density of the short task requests is the same, the short task requests are sorted in descending order according to the number of sensor types to obtain a second set;

[0056] Merging the first set and the second set to obtain a third set, wherein the long task requests and the short task requests in the third set are arranged alternately;

[0057] The closeness is the ratio of the execution duration of the task request to the length of the time window.

[0058] Put the task requests arriving at the avionics system in a time slot into the set K t middle.

[0059] First, traverse the set K t , according to different priorities, put the task requests of the same priority into the set K t，p middle.

[0060] Then, for each set K with priority t，p Sort them separately. If there are pending task requests of this priority, sort the task request set. Otherwise, no operation is performed and an empty set is directly returned.

[0061] Specifically, when sorting the task request set, the task execution duration is compared with the adaptive threshold, and the set K is sorted. t，p Divide into long task request set K t，p，long and the short task request set K t，p，short Then, the long task request set K t，p，long and the short task request set K t，p，short To sort, first calculate the closeness a of each task request k , tightness a k It is given by the ratio of the duration of the task request to the length of the time window, i.e. Sort all long task request sets K in descending order of density t，p，long and the short task request set K t，p，short The reason for descending order is that the probability of rejecting a request increases with the increase of closeness. In addition, in order to increase the acceptance rate of the system, tasks with large sensor requirements will be completed first. Therefore, when the closeness is a k If they are the same, sort them in descending order by the number of sensor types. Then, in the long task request set K t，p，long and the short task request set K t，p，short After the sorting is completed, the task requests are sorted in an alternating manner according to the long task requests and the short task requests to form a set K of task requests with the same priority. t，p，sequenc .

[0062] Finally, a set of task requests K with different priorities will be t，p，sequenc Arrange in descending order of priority to form a sequence set K of task requests arriving at the avionics system within the time slot t，sequenc .

[0063] Step 220: Determine a candidate set of task execution start times for all sorted task requests;

[0064] The candidate set of task start execution times for all sorted task requests is determined, including:

[0065] A candidate set of task start execution times is determined according to the type of each task request, the required execution duration, and the time window, wherein the type of the task request includes a short task request and a long task request.

[0066] The candidate set of task start execution time is determined based on the type of each task request, the required execution duration, and the time window, including:

[0067] For short task requests, Δ k As a candidate set of task start execution time for this short task request, Indicates the time interval for the short task request to run;

[0068] For a long task request, all time slots in the candidate set of the task start execution time of the long task request are traversed. The method of determining the candidate set of the task start execution time of the long task request is the same as that of determining the candidate set of the task start execution time of the short task request. For each time slot i in the candidate set of the task start execution time of the long task request, calculate the time slot [i,d k +i) the average number of tasks μ i And the time slot [i,d k +i) the standard deviation σ of the number of tasks i , when the average number of tasks μ i and standard deviation σ i When the preset conditions are met, time slot i is taken as an element in the candidate set of the task start execution time requested by the long task. The preset conditions are: μ i <μ||(μ i =μ&σ i <σ);

[0069] Among them, d k The execution duration requested for the task.

[0070] Since short task requests need to start as early as possible to improve the acceptance rate of tasks and ensure full utilization of resources in recent time slots, the candidate start time of short task requests directly adopts Δ k .

[0071] For long task requests, they need to be placed as evenly as possible on the time slots to improve resource utilization. Therefore, all time slots in Δk are traversed to find the time slot with the minimum average number of task requests and the minimum standard deviation of its subsequent time slots, which is used as the start time of the long task request.

[0072] Step 230: Allocate resources to the corresponding task request according to the candidate set of task start execution times selected;

[0073] Allocating resources to the corresponding task request based on the candidate set of task start execution times includes:

[0074] Establishing a weighted directed graph corresponding to each time slot in the candidate set according to available nodes and links in the avionics system network, wherein the weighted directed graph corresponding to the time slot represents available resource information of the avionics system resource model of the time slot;

[0075] Select a weighted directed graph with the least available resources from the weighted directed graphs corresponding to all time slots as the target weighted directed graph, where the target weighted directed graph includes the least number of nodes and links;

[0076] Resources are allocated to task requests based on a target weighted directed graph and a task request resource graph, which includes computer resources required by each function when the task is executed, as well as bandwidth requirements for communication between functions.

[0077] The allocation of resources to the task request according to the target weighted directed graph and the task request resource graph includes:

[0078] According to the bottom-up order of available nodes in the target weighted directed graph, combined with the mission request resource graph, the availability of the avionics system is checked and resources are allocated to the mission request.

[0079] Resource allocation has two main tasks: one is to obtain the system's available resource information based on the start time and duration; the other is to reasonably allocate resources to tasks so that the system's task acceptance rate is as high as possible.

[0080] First, we use a weighted directed graph G t (V t ,E t ) represents the state of the available resources of the avionics system belonging to time slot t, where V represents the set of nodes in the network and E represents the set of links. In graph G, node v∈V represents a switch or computer; directed edge 〈o,p〉∈E represents the link from node o to p, and its bandwidth is denoted by B<o,p> When t and d are given k , you can get a sequence Use the set G deal Representation. Using the set G dealA new weighted directed graph G* is constructed, whose node set is the intersection of the nodes of all weighted directed graphs in the sequence, and whose link set is the intersection of all links between the node sets. Furthermore, the resource capacity of the nodes and links in G* is the minimum residual capacity among all weighted directed graphs. This is used to represent the available resource information of all avionics system resource models in different time slots within the duration of the mission request.

[0081] Secondly, the request is placed at the lowest possible level in the avionics system by finding a set of computers and communication paths with sufficient resources. Specifically, if the task can be completed by computers within a cabinet switch, aggregation switches are not used. Prioritizing the lowest level increases the likelihood of using shorter paths while maximizing bandwidth availability at higher levels. Given a request k and the weighted directed graph G* provided by the resource allocation module, the avionics system availability is checked in a bottom-up order (first at the cabinet switch, then at the aggregation switch, and finally at the core switch) until resources are allocated or the request is rejected after exhausting all levels. Specifically, the system first attempts to complete all the requested functional modules using computers within the cabinet switch. This means determining whether the remaining resources and network bandwidth within a cabinet switch can meet the requirements of all the requested functional modules. The feasibility of placing the requested functional modules within the cabinet switch is then checked. If no cabinet switch has sufficient resources to accommodate all the requested functional modules, the system continues to check the availability of resources within all aggregation switches within the avionics system. If no aggregation switch with sufficient capacity is found, the system checks the final layer of core switches. If it is determined that the resources under the avionics system can meet the task requirements, the resource allocation information is saved in the task allocation list so that resources can be applied according to the allocation information when the start time slot arrives.

[0082] Step 240: When resource allocation fails, release the resources and allocate resources to the corresponding task request again.

[0083] When resource allocation is successful, the task request is executed and the resource allocation information is recorded.

[0084] Furthermore, the method further comprises:

[0085] upon receiving a first task request, releasing resources allocated by an unexecuted second task request, the priority of the first task request being higher than the priority of the second task request;

[0086] The released resources are allocated to the first task request.

[0087] The purpose of resource release is to release the resources requested by unexecuted low-priority tasks, thereby increasing the acceptance rate of high-priority tasks. First, task allocation information with a lower priority than the current task request is obtained from the task allocation list. Then, based on the obtained allocation information and the duration of the task, all resources in the resource model sequence related to it in the avionics system are released. The released low-priority tasks will be placed in the pending task set so that these released tasks can be placed again when the priority task is processed. The resource release module will then call the resource allocation module again to determine the availability of system resources after the low-priority task resources are released. If the released system resources can meet the task requirements, the task request will be accepted. If the released system resources still cannot meet the task requirements, the task will be rejected.

[0088] The present invention provides a scheduling method for resource allocation management of an electronic information system, which solves the problem of allocating effective system resources to functional application software, improving the reliability and security of the application system under the system; designing a collaborative scheduling strategy for tasks on shared resources to improve resource utilization; and assisting in system design work to improve design efficiency.

[0089] The present application also provides a scheduling device for resource allocation management of an electronic information system, which includes: a sorting module, a time selection module, a resource allocation module and a resource release module.

[0090] a sequencing module for sequencing mission requests arriving at a multi-platform avionics system within a time slot;

[0091] The time selection module is used to determine the candidate set of task start execution time for all sorted task requests;

[0092] A resource allocation module is used to allocate resources to corresponding task requests based on the candidate set of task start execution times;

[0093] The resource release module is used to release resources when resource allocation fails and allocate resources to the corresponding task request again.

[0094] Compared with the DIMA avionics system in the related art, the present application does not pre-allocate fixed resources, thereby improving the success rate of task request execution, while reducing the workload and difficulty of the initial design stage and improving resource utilization.

[0095] This application provides a scheduling device for resource allocation management of electronic information systems to solve the problem of allocating effective system resources to functional application software, improving the reliability and security of application systems under the system; designing collaborative scheduling strategies for tasks on shared resources to improve resource utilization; and assisting in system design work to improve design efficiency.

[0096] The above merely describes the embodiments of the present application, which are described in a relatively specific and detailed manner, but should not be construed as limiting the scope of the invention. It should be noted that a person skilled in the art may make various modifications and improvements without departing from the concept of the present application, and these modifications and improvements are all within the scope of protection of the present application.

Claims

1. A scheduling method for resource allocation management of an electronic information system, characterized in that: The method comprises: Sequencing mission requests arriving at the avionics system within a time slot; Determine a candidate set of task start execution times for all sorted task requests: determine a candidate set of task start execution times according to the type, required execution duration, and time window of each task request, wherein the type of the task request includes a short task request and a long task request; Allocate resources to the corresponding task request based on the candidate set of task start execution time; When resource allocation fails, release the resources and allocate them to the corresponding task request again; The step of determining a candidate set of task execution start times according to the type, required execution duration, and time window of each task request includes: For short task requests, Δk is used as the candidate set of the task start execution time of the short task request. Indicates the time interval for the short task request to run; For a long task request, all time slots in the candidate set of the task start execution time of the long task request are traversed. The method of determining the candidate set of the task start execution time of the long task request is the same as that of the candidate set of the task start execution time of the short task request. For each time slot i, calculate the time slot [i,d k +i) the average number of tasks μ i And the time slot [i,d k +i) the standard deviation σ of the number of tasks i , when the average number of tasks μ i and standard deviation σ i When the preset conditions are met, time slot i is used as an element in the candidate set of the task start execution time of the long task request. The preset conditions are: μ i <μ||(μ i =μ&σ i <σ); Among them, d k The execution duration requested for the task.

2. The method according to claim 1, characterized in that The step of sequencing the task requests arriving at the avionics system within a time slot comprises: Sort multiple task requests by priority; For task requests of the same priority, task requests with execution duration greater than or equal to an adaptive threshold are regarded as long task requests, and task requests with execution duration less than an adaptive threshold are regarded as short task requests. The adaptive threshold is the average execution duration of all task requests received in the time slot. For the long task request set, the long task requests are sorted in descending order according to the density; when the density of the long task requests is the same, the long task requests are sorted in descending order according to the number of sensor types to obtain the first set; For the short task request set, the short task requests are sorted in descending order according to the density; when the density of the short task requests is the same, the short task requests are sorted in descending order according to the number of sensor types to obtain a second set; Merging the first set and the second set to obtain a third set, wherein the long task requests and the short task requests in the third set are arranged alternately; The closeness is the ratio of the execution duration of the task request to the length of the time window.

3. The method according to claim 1, characterized in that Allocating resources to the corresponding task request according to the candidate set of selected task execution start times includes: Establishing a weighted directed graph corresponding to each time slot in the candidate set according to available nodes and links in the avionics system network, wherein the weighted directed graph corresponding to the time slot represents available resource information of the avionics system resource model of the time slot; Select a weighted directed graph with the least available resources from the weighted directed graphs corresponding to all time slots as the target weighted directed graph, where the target weighted directed graph includes the least number of nodes and links; Resources are allocated to task requests based on a target weighted directed graph and a task request resource graph, which includes computer resources required by each function when the task is executed, as well as bandwidth requirements for communication between functions.

4. The method according to claim 3, characterized in that Allocating resources to the task request according to the target weighted directed graph and the task request resource graph includes: According to the bottom-up order of available nodes in the target weighted directed graph, combined with the mission request resource graph, the availability of the avionics system is checked and resources are allocated to the mission request.

5. The method according to any one of claims 1 to 4, characterized in that: The method further comprises: upon receiving a first task request, releasing resources allocated by an unexecuted second task request, the priority of the first task request being higher than the priority of the second task request; The released resources are allocated to the first task request.

6. The method according to any one of claims 1 to 4, characterized in that: The method further comprises: When resource allocation is successful, the task request is executed and the resource allocation information is recorded.

7. A scheduling device for resource allocation management of an electronic information system, characterized in that: The device includes: a sorting module, a time selection module, a resource allocation module and a resource release module. a sequencing module for sequencing mission requests arriving at a multi-platform avionics system within a time slot; a time selection module, configured to determine a candidate set of task start execution times for all sorted task requests: determining the candidate set of task start execution times according to the type, required execution duration, and time window of each task request, wherein the type of task request includes short task request and long task request; A resource allocation module is used to allocate resources to corresponding task requests based on the candidate set of task start execution times; The resource release module is used to release resources when resource allocation fails and allocate resources to the corresponding task request again; The time selection module determines a candidate set of task start execution times based on the type of each task request, the required execution duration, and the time window, specifically: For short task requests, Δk is used as the candidate set of the task start execution time of the short task request. Indicates the time interval for the short task request to run; For a long task request, all time slots in the candidate set of the task start execution time of the long task request are traversed. The method of determining the candidate set of the task start execution time of the long task request is the same as that of the candidate set of the task start execution time of the short task request. For each time slot i, calculate the time slot [i,d k +i) the average number of tasks μ i And the time slot [i,d k +i) the standard deviation σ of the number of tasks i , when the average number of tasks μ i and standard deviation σ i When the preset conditions are met, time slot i is used as an element in the candidate set of the task start execution time of the long task request. The preset conditions are: μ i <μ||(μ i =μ&σ i <σ); Among them, d k The execution duration requested for the task.

Images