CN113938816B - Computing power resource scheduling method and device - Google Patents
Computing power resource scheduling method and device Download PDFInfo
- Publication number
- CN113938816B CN113938816B CN202111062337.1A CN202111062337A CN113938816B CN 113938816 B CN113938816 B CN 113938816B CN 202111062337 A CN202111062337 A CN 202111062337A CN 113938816 B CN113938816 B CN 113938816B
- Authority
- CN
- China
- Prior art keywords
- computing
- domain
- service
- node
- force
- 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
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/02—Services making use of location information
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/02—Arrangements for optimising operational condition
Abstract
The application discloses a computing power resource scheduling method and a computing power resource scheduling device, and relates to the field of resource scheduling. The method comprises the following steps: and acquiring the geographic position of the user equipment according to the service request information of the first service from the user equipment. According to the distance between the geographical position of the user equipment and the geographical position of each computing domain in the computing network, sequencing each computing domain in the computing network according to the sequence from the near to the far to obtain a first sequencing result, wherein each computing domain comprises a plurality of computing nodes. And determining a target computing node bearing the first service according to the service request information and the first sequencing result. The target computing force node is one of at least one computing force node in the target computing force domain that meets the requirements of the first business. The target computational domain is the computational domain of the computational node that first includes the computational node that satisfies the requirements of the first business in the first ordering result. The method can be quickly scheduled to the target computing node which can bear the first service.
Description
Technical Field
The present disclosure relates to the field of resource scheduling, and in particular, to a method and an apparatus for scheduling computing resources.
Background
Artificial intelligence (artificial intelligence, AI) services refer to services that utilize AI technology, which refers to the technology of presenting human intelligence through a common computer program. In AI technology, computing power (resources), algorithms, and data are three important elements. Among the three elements, computing power can be regarded as a basic platform of the AI technology, and the application of the AI technology is directly influenced. The deployment of the AI service depends on the application of the AI technology, so that the computing power directly influences the deployment of the AI service.
At present, the deployment of the AI business is mainly performed through an algorithm power network, and the algorithm power network can perform unified management and scheduling on the algorithm power resources and network resources from idle scattered algorithm power of individuals or enterprises so as to meet the requirements of different AI businesses. While different AI service requirements may have different methods of network scheduling.
Disclosure of Invention
The power computing resource scheduling method and the power computing resource scheduling device can rapidly schedule power computing nodes meeting the requirements of the first service, and the scheduling speed is high.
In a first aspect, the present application provides a method for scheduling computing power resources, the method including:
receiving service request information of a first service from user equipment; acquiring the geographic position of user equipment according to service request information of a first service; according to the distance between the geographical position of the user equipment and the geographical position of each calculation domain in the calculation network, sequencing each calculation domain in the calculation network according to the sequence from the near to the far to obtain a first sequencing result; wherein each computing domain in the computing network comprises a plurality of computing nodes; determining a target computing power node bearing the first service according to the service request information of the first service and the first sequencing result; the target computing force node is one of at least one computing force node which is included in the target computing force domain and meets the requirement of the first service; the target computational domain is the computational domain of the computational node that first includes the computational node that satisfies the requirements of the first business in the first ordering result.
In a possible implementation manner, determining a target computing power node carrying the first service according to service request information of the first service and the first ordering result includes: starting from a first computing force domain in a first sequencing result, adopting a first algorithm, and calculating an available value of each computing force node in the computing force domain for a first service according to service request information of the first service; when the computing force node with the largest available value for the first service in the first computing force domain meets the requirement of the first service, determining the computing force node with the largest available value for the first service in the first computing force domain as a target computing force node; when the computing force node with the largest available value in the first computing force domain does not meet the requirement of the first service, determining a target computing force node from the next computing force domain until the target computing force node is obtained.
In another possible implementation manner, determining the target computing power node carrying the first service according to the service request information of the first service and the first ordering result includes: starting from a first computing force domain in a first sequencing result, adopting a first algorithm, and calculating an available value of each computing force node in the computing force domain for a first service according to service request information of the first service; when the first computing domain comprises a plurality of computing nodes meeting the requirement of the first service, sequencing the computing nodes according to the sequence from small to large of the available value of the first service to obtain a second sequencing result; and according to the second sequencing result, determining one of a plurality of computing force nodes with the available value of the first service meeting the requirement of the first service as a target computing force node.
In another possible implementation manner, when the first computing domain includes a plurality of computing nodes meeting the requirement of the first service, the computing nodes are ordered according to the order from small to large of the available values of the first service, so as to obtain a second ordering result; according to the second ordering result, determining one of a plurality of computing nodes for which the available value of the first service meets the requirement of the first service as a target computing node comprises: from a plurality of force nodes in the first force domain that satisfy the requirements of the first service, any one of the force nodes other than the force node having the largest value available to the first service is determined as the target force node.
In yet another possible implementation, determining, from among a plurality of computing nodes in the first computing domain that satisfy the requirement of the first service, any one of the computing nodes other than the computing node having the largest value available for the first service as the target computing node includes: and determining a computing power node with the minimum available value for the first service as a target computing power node from a plurality of computing power nodes meeting the requirements of the first service in the first computing power domain.
In another possible implementation manner, using a first algorithm, according to service request information of a first service, calculating an available value of each computing force node in a computing force domain for the first service includes: acquiring the geographic position of the user equipment, the priority of the first service, the calculation type of the first service and the calculation power requirement of the first service according to the service request information of the first service; and calculating the available value of each computing power node in the computing power domain for the first service according to the geographic position of the user equipment, the priority of the first service, the computing type of the first service, the computing power size requirement of the first service, the computing type provided by the preset computing power node and the computing power size provided by the preset computing power node.
In another possible implementation manner, calculating the available value of each computing node in the computing power domain for the first service according to the geographic location where the user equipment is located, the priority of the first service, the computing type of the first service, the computing power size requirement of the first service, the computing type that can be provided by the preset computing power node, and the computing power size that can be provided by the preset computing power node includes:
for each computing force node in the computing force domain: according to the geographic position of the user equipment and the geographic position of the computing domain of the computing node, calculating the network transmission time delay between the user equipment and the computing node; calculating the ratio of the first weight coefficient to the network transmission delay; calculating the sum of the second weight coefficient and the priority of the first service; determining a first value according to the calculation type of the first service and the calculation type which can be provided by the calculation force node; determining a second value according to the calculation force demand of the first service and the calculation force which can be provided by the calculation force node; and calculating the available value of the computing power node to the first service according to the ratio of the first weight coefficient to the network transmission delay, the sum of the priority of the second weight coefficient and the first service, the first value, the second value and the compensation coefficient.
In another possible implementation manner, according to a distance between a geographical location of the user equipment and a geographical location of each computing domain in the computing network, ordering each computing domain in the computing network in a sequence from near to far includes: calculating Euclidean distance between the geographic position of the user equipment and the geographic position of each computing domain in the computing network; and ordering the computing power domains in the computing power network according to the sequence from the near to the far according to the Euclidean distance between the geographic position of the user equipment and the geographic position of each computing power domain in the computing power network.
In yet another possible implementation, before calculating the euclidean distance between the geographical location of the user device and the geographical location of each computing domain in the computing network, the method further includes:
for each computational domain: the geographic location of the computing force domain is determined based on the geographic locations of the respective computing force nodes included in the computing force domain.
In yet another possible implementation, for each computational domain: determining the geographic location of the computing force domain according to the geographic locations of the computing force nodes included in the computing force domain, including:
for each computational domain: and obtaining the average geographic position of each computing force node according to the geographic position of each computing force node included in the computing force domain, and taking the average geographic position of each computing force node as the geographic position of the computing force domain.
In the power computing resource scheduling method provided by the application, when facing the request information of the first service sent by the user equipment, the power computing resource scheduling module 21 can analyze the geographical position where the user equipment is located, the priority of the first service, the computing type matched with the first service and the power computing requirement of the first service according to the request information of the first service.
And then, selecting the computing power node in the computing power domain closest to the user equipment according to the Euclidean distance between the geographic position of the user equipment and the central position of the preset computing power domain to screen, so that the screening range can be reduced from a plurality of computing power domains to one computing power domain, and the speed of screening the computing power node capable of meeting the first service is increased as a whole.
Then, according to the preset calculation type provided by each calculation node, the preset calculation strength provided by each calculation node, the priority of the first service, the calculation type matched with the first service, the calculation strength requirement of the first service, and the preset first algorithm, a priority calculation node set in a first priority domain is obtained, whether the first priority calculation node in the first priority domain can meet the requirement of the first service or not is judged, whether the first priority domain can meet the requirement of the first service or not can be represented, the speed of screening the calculation nodes capable of meeting the first service is increased as a whole, the factors such as the priority, the calculation type, the requirement and the like are comprehensively considered, and the possibility that the screened target calculation node can be matched with the requirement of the first service is higher.
In addition, after determining the power computing node which can meet the requirement of the first service in a certain priority domain, the power computing resource scheduling method provided by the application can further judge and select whether the power computing node with a smaller available value of the first service can meet the requirement of the first service, so that the waste of power computing resources of the first preferred power computing node in the preferred domain is avoided, and the best power computing node matched with the first service is screened out.
In a second aspect, the present application provides an apparatus for scheduling computing resources, the apparatus comprising:
a receiving unit, configured to receive service request information of a first service from a user equipment;
the processing unit is used for acquiring the geographic position of the user equipment according to the service request information of the first service; the method is also used for sorting all the calculation domains in the calculation network according to the sequence from the near to the far according to the distance between the geographical position of the user equipment and the geographical position of each calculation domain in the calculation network, so as to obtain a first sorting result; wherein each computing domain in the computing network comprises a plurality of computing nodes; the target computing node is used for bearing the first service according to the service request information of the first service and the first sequencing result; the target computing force node is one of at least one computing force node which is included in the target computing force domain and meets the requirement of the first service; the target computational domain is the computational domain of the computational node that first includes the computational node that satisfies the requirements of the first business in the first ordering result.
In a possible implementation manner, the processing unit is specifically configured to calculate, by using a first algorithm from a first computing force domain in the first ordering result, an available value of each computing force node in the computing force domain for the first service according to service request information of the first service; when the computing force node with the largest available value for the first service in the first computing force domain meets the requirement of the first service, determining the computing force node with the largest available value for the first service in the first computing force domain as a target computing force node; when the computing force node with the largest available value in the first computing force domain does not meet the requirement of the first service, determining a target computing force node from the next computing force domain until the target computing force node is obtained.
In another possible implementation manner, the processing unit is specifically configured to calculate, by using a first algorithm from a first computing force domain in the first ordering result, an available value of each computing force node in the computing force domain for the first service according to service request information of the first service; when the first computing domain comprises a plurality of computing nodes meeting the requirement of the first service, sequencing the computing nodes according to the sequence from small to large of the available value of the first service to obtain a second sequencing result; and according to the second sequencing result, determining one of a plurality of computing force nodes with the available value of the first service meeting the requirement of the first service as a target computing force node.
In yet another possible implementation manner, the processing unit is specifically configured to determine, from a plurality of computing nodes in the first computing domain that meet the requirement of the first service, any computing node other than the computing node with the largest value available for the first service as the target computing node.
In yet another possible implementation manner, the processing unit is specifically configured to determine, from a plurality of computing nodes in the first computing domain that meet the requirement of the first service, a computing node with a smallest available value for the first service as the target computing node.
In another possible implementation manner, the processing unit is specifically configured to obtain, according to service request information of the first service, a geographic location where the user equipment is located, a priority of the first service, a calculation type of the first service, and a calculation power requirement of the first service; and calculating the available value of each computing power node in the computing power domain for the first service according to the geographic position of the user equipment, the priority of the first service, the computing type of the first service, the computing power size requirement of the first service, the computing type provided by the preset computing power node and the computing power size provided by the preset computing power node.
In yet another possible implementation, the processing unit is specifically configured to, for each computing force node in the computing force domain: according to the geographic position of the user equipment and the geographic position of the computing domain of the computing node, calculating the network transmission time delay between the user equipment and the computing node; calculating the ratio of the first weight coefficient to the network transmission delay; calculating the sum of the second weight coefficient and the priority of the first service; determining a first value according to the calculation type of the first service and the calculation type which can be provided by the calculation force node; determining a second value according to the calculation force demand of the first service and the calculation force which can be provided by the calculation force node; and calculating the available value of the computing power node to the first service according to the ratio of the first weight coefficient to the network transmission delay, the sum of the priority of the second weight coefficient and the first service, the first value, the second value and the compensation coefficient.
In yet another possible implementation manner, the processing unit is specifically configured to calculate a euclidean distance between a geographic location where the user equipment is located and a geographic location of each computing domain in the computing network; and ordering the computing power domains in the computing power network according to the sequence from the near to the far according to the Euclidean distance between the geographic position of the user equipment and the geographic position of each computing power domain in the computing power network.
In yet another possible implementation, the processing unit is specifically configured to, before calculating the euclidean distance between the geographical location of the user equipment and the geographical location of each computing domain in the computing network, further be configured to, for each computing domain: the geographic location of the computing force domain is determined based on the geographic locations of the respective computing force nodes included in the computing force domain.
In yet another possible implementation, the processing unit is specifically configured to, for each computing domain: and obtaining the average geographic position of each computing force node according to the geographic position of each computing force node included in the computing force domain, and taking the average geographic position of each computing force node as the geographic position of the computing force domain.
In a third aspect, the present application provides an electronic device, comprising: a processor and a memory; the memory stores instructions executable by the processor; the processor is configured to execute instructions to cause the electronic device to implement the method as described in the first aspect above.
In a fourth aspect, the present application provides a computer-readable storage medium comprising: computer software instructions; the computer software instructions, when run in an electronic device, cause the electronic device to implement the method as described in the first aspect above.
The advantages of the second to fourth aspects described above may be referred to the advantages described in the first aspect, and will not be described here again.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a functional schematic of a computing network;
FIG. 2 is a schematic diagram of a computing power resource scheduling system according to an embodiment of the present disclosure;
fig. 3 is a flow chart of a computing power resource scheduling method according to an embodiment of the present application;
FIG. 4 is a schematic flow chart of another method for scheduling computing power resources according to an embodiment of the present disclosure;
FIG. 5 is a schematic diagram of another embodiment of a computing power resource scheduling system according to the present disclosure;
FIG. 6 is another flow chart of a method for scheduling computing power resources according to an embodiment of the present disclosure;
FIG. 7 is a schematic diagram illustrating the division of the computing domain according to an embodiment of the present disclosure;
FIG. 8 is a schematic diagram of an overall architecture of a computing resource scheduling system according to an embodiment of the present disclosure;
fig. 9 is a schematic diagram of a composition of a computing power resource scheduling device according to an embodiment of the present application;
fig. 10 is a schematic structural diagram of an electronic device according to an embodiment of the present application.
Detailed Description
The following description of the embodiments of the present application will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all, of the embodiments of the present application. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present disclosure.
It should be noted that, in the embodiments of the present application, words such as "exemplary" or "such as" are used to mean serving as an example, instance, or illustration. Any embodiment or design described herein as "exemplary" or "for example" should not be construed as preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary" or "such as" is intended to present related concepts in a concrete fashion.
In order to clearly describe the technical solutions of the embodiments of the present application, in the embodiments of the present application, the terms "first", "second", and the like are used to distinguish the same item or similar items having substantially the same function and effect, and those skilled in the art will understand that the terms "first", "second", and the like are not limited in number and execution order.
Artificial intelligence (artificial intelligence, AI) services refer to services that utilize AI technology, which refers to the technology of presenting human intelligence through a common computer program. In AI technology, computing power (resources), algorithms, and data are three important elements. Among the three elements, computing power can be regarded as a basic platform of the AI technology, and the application of the AI technology is directly influenced. The deployment of the AI service depends on the application of the AI technology, so that the computing power directly influences the deployment of the AI service.
At present, AI services are deployed mainly through a power computing network, and fig. 1 is a functional schematic diagram of the power computing network. As shown in fig. 1, the computing power network can perform unified management and scheduling on computing power resources and network resources from idle scattered computing power of individuals or enterprises, and provide computing power services to meet the requirements of different AI services. While different AI business requirements may require different computing services, different computing services may have different methods of network scheduling.
Under the background technology, the embodiment of the application provides a computing power resource scheduling method which can be rapidly scheduled to meet the AI service requirement. The method may be performed by a server or a computer.
Taking a server as an example, in some embodiments, the server may include a power resource scheduling module that may be part of a power network (or referred to as a power resource scheduling system). The server can schedule the computational power resources in the computational power network through the computational power resource scheduling module, so that the computational power resource scheduling method provided by the embodiment of the application is realized.
The computing power resource scheduling module may include a processor and a memory, where the memory stores program codes corresponding to the computing power resource scheduling method provided by the embodiment of the application, and the processor may execute the program codes stored in the memory when running, so as to implement the computing power resource scheduling method provided by the embodiment of the application.
The following describes an exemplary embodiment of the present application by taking a server scheduling, by using a computing resource scheduling module, computing resources in a computing network as an example.
Fig. 2 is a schematic diagram of a computing power resource scheduling system according to an embodiment of the present application. As shown in fig. 2, the computing resource scheduling system 20 may include: the computational resource scheduling module 21.
Fig. 3 is a flow chart of a computing power resource scheduling method according to an embodiment of the present application.
As shown in fig. 3, the method may include S301 to S3010.
S301, the computing power resource scheduling module 21 acquires service request information of a first service from user equipment.
For example, the user device may send service request information for the first service to the computing resource scheduling system 20.
Wherein the first service may be an AI service. The service request information of the first service may include information such as priority of the first service, calculation type of the first service matching, and calculation power requirement of the first service.
By way of example, the user device may be a cell phone, tablet computer, wearable device, vehicle-mounted device, augmented reality (augmented reality, AR)/Virtual Reality (VR) device, notebook computer, ultra-mobile personal computer (UMPC), netbook, personal digital assistant (personal digital assistant, PDA), etc., and the embodiments of the present application are not limited to the specific type of user device.
S302, the computing power resource scheduling module 21 obtains the geographic position of the user equipment according to the service request information of the first service.
In some possible embodiments, the power resource scheduling module 21 obtains the geographic location of the user equipment according to the service request information of the first service, specifically, the power resource scheduling module 21 obtains the geographic location of the user equipment according to the network protocol (internet protocol, IP) address of the service request information of the first service.
In other possible embodiments, the power resource scheduling module 21 obtains the geographic location of the user equipment according to the service request information of the first service, specifically, the service request information of the first service includes geographic location information when the user equipment sends the service request information of the first service, and the power resource scheduling module 21 determines the geographic location of the user equipment according to the geographic location information in the service request information of the first service.
For example, the geographic location of the user device may refer to latitude and longitude coordinates of the location of the user device.
S303, the computing power resource scheduling module 21 determines Euclidean distances from the user equipment to each computing power domain according to the geographical position of the user equipment and the central position of the preset computing power domain.
Taking the coordinates of the geographic position of the user equipment as (x, y, z) and the coordinates of the central position of the preset computing domain as (a, b, c) as examples, the euclidean distance d between the user equipment and the computing domain can be calculated by the following formula (1):
And S304, the power resource scheduling module 21 sequentially arranges the power domains according to the order of the Euclidean distance from the near to the far to obtain a priority domain set corresponding to the first service, and determines the power domain with the nearest Euclidean distance as the first priority domain corresponding to the first service.
The first service may correspond to a first priority domain set, which may be referred to as a first ordering result, where the first ordering result may include a first priority domain, a second priority domain, …, an nth priority domain, and the like, where N is an integer greater than or equal to 1, and the first priority domain is a first computing domain in the first ordering result.
Illustratively, the computing force domain includes S 1 ，S 2 ，...，S N The method comprises the steps of carrying out a first treatment on the surface of the And d 1 Representing user equipment to S 1 Euclidean distance d of (2) 2 Representing user equipment to S 2 Is the Euclidean distance of …, d N Representing user equipment to S N For example, assume d 1 ＜d 2 ＜...＜d N The priority domain set corresponding to the first service is (S 1 ，S 2 ，...，S N ) Wherein S is 1 A first priority domain, which may be called a first service correspondence, S 2 A second priority domain, …, S, which may be called a first traffic correspondence N May be referred to as an nth priority domain corresponding to the first service.
S305, the computing power resource scheduling module 21 obtains a computing type that each computing power node in the first priority domain can provide, a computing power size that each computing power node in the first priority domain can provide, a priority of the first service, a computing type of the first service, and a computing power size requirement of the first service.
In some possible embodiments, the computing power resource scheduling module 21 obtains the computing type and the computing power size that can be provided by each computing power node in the first priority domain, which may be that the technician manually sets the computing type and the computing power size that can be provided by each computing power node in each priority domain on the computing power resource scheduling module 21; alternatively, the computing power resource scheduling module 21 calculates the type of calculation and the magnitude of the computing power that can be provided by each computing power node in each priority domain; alternatively, after the computing power resource scheduling module 21 receives the computing type and the computing power size that can be provided by each computing power node in each priority domain sent by other modules in the computing power resource scheduling system 20, the computing power resource scheduling module 21 selects the first priority domain, and determines the computing type and the computing power size that can be provided by each computing power node in the first priority domain from the computing type and the computing power size that can be provided by each computing power node in each priority domain. The specific manner in which the computing power resource scheduling module 21 obtains the computing type and the computing power size that each computing power node in the first priority domain can provide is not limited in the embodiments of the present application.
S306, the computing power resource scheduling module 21 obtains an available value of each computing power node in the first priority domain for the first service according to the geographical position of the user equipment, the computing type provided by each computing power node in the first priority domain, the computing power size provided by each computing power node in the first priority domain, the priority of the first service, the computing type matched with the first service, the computing power size requirement of the first service and a preset first algorithm.
In some possible embodiments, the preset first algorithm may be as shown in the following formula (2).
In equation (2), q represents the computational power node in the preferred domain, W q Representing the available value of the force node q in the priority domain. L (L) Ri Indicating the geographic location where the user device is located. L (L) Rq The position of the priority domain (calculation domain) where the calculation node q is located is indicated. L Ri -L Rq The i represents the network between the geographical location where the user device is located and the computing node qTransmission delay. K (K) q And K jq Respectively representing the weight coefficients. P is p i Indicating the priority of the first service. T (T) ij Representing the type of computation of the first service. T (T) qj Representing the type of computation that the computing force node q can provide. T (T) ij &&T qj May be referred to as a first value, which represents the calculation type of the first service and the calculation type that can be provided by the calculation node q, and if the two calculation types are consistent, the value is 1; if not, the value is 0.C (C) q Indicating the amount of computing power that the computing power node q can provide. C (C) i Representing the computational power size requirements of the first business. C q -C i The i may be referred to as a second value. V (V) q Representing the compensation coefficient.
S307, the power calculation resource scheduling module 21 sequentially arranges the power calculation nodes in the first priority domain according to the order of the available values of the first service from small to large, obtains a priority power calculation node set corresponding to the first priority domain, and determines a first priority power calculation node in the first priority domain.
The set of priority force-calculating nodes corresponding to the first priority domain may also be referred to as a second ordering result, and the first priority force-calculating node may be a force-calculating node with the largest available value for the first service in the priority domain.
Exemplary, the first priority domain corresponding to the first service is also referred to as S 1 For example, assume S 1 The system comprises a force calculation node 1, force calculation nodes 2 and …, a force calculation node M-1 and a force calculation node M; w (W) 1 Representing the available value of the computing node 1 for the first service, W 2 Representing the availability of the computing node 2 to the first service, …, W M-1 Representing the available value, W, of the computing node M-1 for the first service M Representing an available value of a computing force node M for a first service, wherein M is an integer greater than or equal to 1; and W is 1 ＞W 2 ＞…＞W M-1 ＞W M The computing node 1 is the first priority domain S 1 The first priority force calculation node in the plurality of the first priority domains is the force calculation node 2 which is the first priority domain S 1 The second priority force calculation node, …, the force calculation node M is the first priority domain S 1 The mth priority force node in (a).
S308, the power calculation resource scheduling module 21 detects whether a first priority power calculation node in a first priority domain can meet the requirement of a first service; if yes, then execute S309; if not, S3010 is executed.
S309, the power resource scheduling module 21 determines the first priority power node in the first priority domain as a target power node for carrying the first service.
Exemplary, the first priority domain corresponding to the first service is also referred to as S 1 And the computing node 1 is the first priority domain S 1 For example, if the computing power node 1 can meet the requirement of the first service, the computing power resource scheduling module 21 determines that the computing power node 1 is a target computing power node for bearing the first service.
S3010, the power calculation resource scheduling module 21 calculates available values of each power calculation node in the second priority domain, the third priority domain, … and the Nth priority domain for the first service respectively, and detects whether the first priority power calculation node in each priority domain after the first priority domain can meet the requirement of the first service or not until a power calculation node which can meet the requirement of the first service is selected.
That is, starting from a first computing force domain in the first sequencing result, adopting a first algorithm, and calculating an available value of each computing force node in the computing force domain for the first service according to service request information of the first service; when the computing force node with the largest available value for the first service in the first computing force domain meets the requirement of the first service, determining the computing force node with the largest available value for the first service in the first computing force domain as a target computing force node; when the computing force node with the largest available value in the first computing force domain does not meet the requirement of the first service, determining a target computing force node from the next computing force domain until the target computing force node is obtained.
Illustratively, the priority domain set corresponding to the first service is also defined as (S 1 ，S 2 ，...，S N ) The first priority domain corresponding to the first service is S 1 The computing node 1 in the first preferred domain is the first preferred domain S 1 In the calculation of the maximum value available for the first serviceFor example, assuming that the force node 1 cannot meet the requirement of the first service, and that the second priority domain includes the force node m+1, the force nodes m+2, …, and the force node m+e, where E is an integer greater than or equal to m+1, the force resource scheduling module 21 may calculate to obtain the second priority domain S according to the above formula (2) 2 The method comprises the steps of determining whether a first priority force node in a second priority domain can meet the requirement of a first service or not according to the available value of the first service of the force node 1, the force node 2 and … and the force node A, if yes, determining the first priority force node in the second priority domain as a target force node bearing the first service, if no, continuing to calculate whether the first priority force node in each priority domain after the second priority domain can meet the requirement of the first service or not by analogy until the force node which can meet the requirement of the first service is selected.
In the method for scheduling the computing power resources provided in the embodiment of the present application, when the computing power resource scheduling module 21 faces to the request information of the first service sent by the user equipment, the geographic location where the user equipment is located, the priority of the first service, the computing type of the first service matching, and the computing power requirement of the first service may be resolved according to the request information of the first service.
And then, selecting the computing power node in the computing power domain closest to the user equipment according to the Euclidean distance between the geographic position of the user equipment and the central position of the preset computing power domain to screen, so that the screening range can be reduced from a plurality of computing power domains to one computing power domain, and the speed of screening the computing power node capable of meeting the first service is increased as a whole.
Then, according to the preset calculation type provided by each calculation node, the preset calculation strength provided by each calculation node, the priority of the first service, the calculation type matched with the first service, the calculation strength requirement of the first service, and the preset first algorithm, a priority calculation node set in a first priority domain is obtained, whether the first priority calculation node in the first priority domain can meet the requirement of the first service or not is judged, whether the first priority domain can meet the requirement of the first service or not can be represented, the speed of screening the calculation nodes capable of meeting the first service is increased as a whole, the factors such as the priority, the calculation type, the requirement and the like are comprehensively considered, and the possibility that the screened target calculation node can be matched with the requirement of the first service is higher.
In some possible embodiments, when the first priority domain (the first computing domain in the first ordering result) includes a plurality of computing nodes that meet the requirement of the first service, the computing nodes are ordered in order from the smaller available value to the larger available value of the first service, and after the priority computing node set (the second ordering result) is obtained, in order to save computing resources, any computing node other than the computing node that meets the requirement of the first service in the plurality of computing nodes in the first priority domain is determined to be the target computing node.
Fig. 4 is another flow chart of a method for scheduling computing power resources according to an embodiment of the present application.
In one possible design, as shown in fig. 4, S309 of the method may be replaced with S401 to S403.
S401, detecting whether a second priority power node in a first priority domain can meet the requirement of a first service by a power resource scheduling module 21; if yes, executing S402; if not, S403 is performed.
S402, the power resource scheduling module 21 continues to detect whether the third priority power node, the fourth priority power node, … and the G priority power node in the first priority domain can meet the requirement of the first service until detecting the H priority power node which cannot meet the requirement of the first service, and determines the H-1 priority power node as a target power node for bearing the first service.
Where G is a positive integer greater than 2, the H-th priority force node may be any one of the priority force nodes in the first priority domain other than the first priority force node and the second priority force node.
S403, the power resource scheduling module 21 determines that the first priority power node in the first priority domain is a target power node carrying the first service.
It can be understood that when the power resource scheduling module 21 in S3010 detects whether the priority power nodes in the second priority domain, the third priority domain, …, and the nth priority domain can meet the requirement of the first service, the similar methods in S401 to S402 may also be used, which is not described herein.
That is, starting from a first computing force domain in the first sequencing result, adopting a first algorithm, and calculating an available value of each computing force node in the computing force domain for the first service according to service request information of the first service; when the first computing domain comprises a plurality of computing nodes meeting the requirement of the first service, selecting the computing node with the smallest available value for the first service from the computing nodes meeting the requirement of the first service as a target computing node; when the first computing domain does not include computing force nodes that meet the requirements of the first business, determining a target computing force node from the next computing force domain until the target computing force node is obtained.
After determining the power computing node which can meet the requirement of the first service in a certain priority domain, the power computing resource scheduling method provided by the embodiment of the invention can further judge and select whether the power computing node with smaller available value of the first service can meet the requirement of the first service, so that the waste of power computing resources of the first preferred power computing node in the preferred domain is avoided, and the best power computing node matched with the first service is screened out.
It can be appreciated that S309-S3010 and S401-S403 described above, that is, determine, according to the service request information of the first service and the first ordering result, the target computing node carrying the first service; the target computing force node is one of at least one computing force node which is included in the target computing force domain and meets the requirement of the first service; the target computational domain is the computational domain of the computational node that first includes the computational node that satisfies the requirements of the first business in the first ordering result.
Fig. 5 is another schematic diagram of a computing power resource scheduling system according to an embodiment of the present application.
In some possible embodiments, as shown in fig. 5, the computing resource scheduling system 20 may further include a computing resource unified quantitative modeling module 22, where the computing resource unified quantitative modeling module 22 is connected to the computing resource scheduling module 21.
Fig. 6 is another flow chart of a method for scheduling computing power resources according to an embodiment of the present application.
In some possible embodiments, as shown in fig. 6, the method may further include S601 before S303.
S601, dividing a preset computing force node into a plurality of computing force domains by the computing force resource unified quantitative modeling module 22, and respectively determining the average position of all computing force nodes in each computing force domain as the central position of the computing force domain.
Wherein the computing node may be composed of one or more terminal devices. For example, the computing node may be composed of a mobile phone, a data center, a small data center, a notebook computer, and other terminal devices. The average location may be an average geographic location.
In some possible embodiments, the unified quantitative modeling module 22 for computing power resources may divide computing power nodes that are geographically close into one computing power domain.
In one possible design, after dividing the computing nodes with similar geographic locations into computing domains, if the number of computing nodes in the computing domain is greater than a preset threshold F, the computing resource unified quantitative modeling module 22 may further divide the computing domain into different computing domains until the number of computing nodes in the computing domain is less than the preset threshold F.
Optionally, after dividing the computing power nodes with similar geographic positions into computing power domains, if the number of computing power nodes in the computing power domain is equal to a preset threshold F, the computing power resource unified quantization modeling module 22 may divide the computing power domain into different computing power domains; alternatively, the computational domain is not further partitioned. The embodiment of the application does not limit the division situation when the number of the computing force nodes in the computing force domain is equal to the preset threshold value F.
Taking a preset threshold F as 100 as an example, assuming that when the computing force domains are divided, 150 computing force nodes are similar in geographic position, the 150 computing force nodes similar in geographic position can be divided into two different computing force domains, and the number of computing force nodes in each computing force domain is smaller than 100 in the two different computing force domains obtained by dividing.
Fig. 7 is a schematic diagram of dividing the computational domain according to an embodiment of the present application.
Illustratively, as shown in fig. 7, the unified quantitative modeling module 22 for computing force resources divides the plurality of computing force nodes into computing force domain 1, computing force domain 2, computing force domain 3a, computing force domain 3b, and other computing force domains, wherein the computing force nodes in computing force domain 3a and computing force domain 3b are close in position in the ground and may be divided into primary computing force domain 3, but the sum of the number of computing force nodes in computing force domain 3a and computing force domain 3b is greater than a preset threshold F, so that primary computing force domain 3 is divided into computing force domain 3a and computing force domain 3b again.
Average geographic location (a) of each computing node in computing domain 1 1 ，b 1 ，c 1 ) Can be seen as the central position of the force domain 1; average geographic location (a) of each computing node in computing domain 2 2 ，b 2 ，c 2 ) Can be seen as the central position of the force field 2; average geographic location (a) of each computing node in computing domain 3a 3 ，b 3 ，c 3 ) Can be seen as the central position of the force field 3 a; average geographic position (a) of each computing node in computing domain 3b 4 ，b 4 ，c 4 ) The central position of the force field 3b can be seen.
Optionally, before S601, the unified quantitative modeling module for computing power resources 22 may further divide one or more terminal devices into a plurality of computing power nodes and determine a geographic location of the computing power nodes.
Fig. 8 is a schematic diagram of an overall architecture of a computing resource scheduling system according to an embodiment of the present application.
In some possible embodiments, as shown in fig. 8, the unified quantitative modeling module 22 for computing power resources may uniformly quantize and divide a mobile phone terminal, a data center, a small data center, a notebook computer, and the like into different computing power nodes, where multiple computing power nodes form a computing power domain. The power resource scheduling module 21 may schedule different power nodes for different AI services according to the requirements of the different AI services to carry the different AI services. For example, AI services may be rendering factories, model training, intelligent security, and speech recognition, among others.
The embodiment of the application also provides a computing power resource scheduling device, which can be used for realizing the computing power resource scheduling method in the previous embodiment. For example, the computing power resource scheduling device may refer to a computing power resource scheduling module in the computing power network mentioned in the above embodiment.
Fig. 9 is a schematic diagram of a component of a computing power resource scheduling device according to an embodiment of the present application.
As shown in fig. 9, the computing power resource scheduling apparatus 90 includes: a receiving unit 91 and a processing unit 92, the receiving unit 91 and the processing unit 92 being connected;
A receiving unit 91, configured to receive service request information of a first service from a user equipment;
a processing unit 92, configured to obtain, according to service request information of the first service, a geographic location where the user equipment is located; the method is also used for sorting all the calculation domains in the calculation network according to the sequence from the near to the far according to the distance between the geographical position of the user equipment and the geographical position of each calculation domain in the calculation network, so as to obtain a first sorting result; wherein each computing domain in the computing network comprises a plurality of computing nodes; the target computing node is used for bearing the first service according to the service request information of the first service and the first sequencing result; the target computing force node is one of at least one computing force node which is included in the target computing force domain and meets the requirement of the first service; the target computational domain is the computational domain of the computational node that first includes the computational node that satisfies the requirements of the first business in the first ordering result.
In one possible design, the processing unit 92 is specifically configured to calculate, by using a first algorithm from a first computing force domain in the first ordering result, an available value of each computing force node in the computing force domain for the first service according to the service request information of the first service; when the computing force node with the largest available value for the first service in the first computing force domain meets the requirement of the first service, determining the computing force node with the largest available value for the first service in the first computing force domain as a target computing force node; when the computing force node with the largest available value in the first computing force domain does not meet the requirement of the first service, determining a target computing force node from the next computing force domain until the target computing force node is obtained.
In another possible design, the processing unit 92 is specifically configured to calculate, by using a first algorithm from a first computing force domain in the first ordering result, an available value of each computing force node in the computing force domain for the first service according to the service request information of the first service; when the first computing domain comprises a plurality of computing nodes meeting the requirement of the first service, sequencing the computing nodes according to the sequence from small to large of the available value of the first service to obtain a second sequencing result; and according to the second sequencing result, determining one of a plurality of computing force nodes with the available value of the first service meeting the requirement of the first service as a target computing force node.
In yet another possible design, the processing unit 92 is specifically configured to determine, from among the plurality of computing nodes in the first computing domain that meet the requirement of the first service, any computing node other than the computing node with the largest value available for the first service as the target computing node.
In yet another possible design, the processing unit 92 is specifically configured to determine, from a plurality of computing nodes in the first computing domain that meet the requirement of the first service, a computing node with a smallest available value for the first service as the target computing node.
In yet another possible design, the processing unit 92 is specifically configured to obtain, according to service request information of the first service, a geographic location where the user equipment is located, a priority of the first service, a calculation type of the first service, and a calculation power requirement of the first service; and calculating the available value of each computing power node in the computing power domain for the first service according to the geographic position of the user equipment, the priority of the first service, the computing type of the first service, the computing power size requirement of the first service, the computing type provided by the preset computing power node and the computing power size provided by the preset computing power node.
In yet another possible design, the processing unit 92 is specifically configured to, for each computing force node in the computing force domain: according to the geographic position of the user equipment and the geographic position of the computing domain of the computing node, calculating the network transmission time delay between the user equipment and the computing node; calculating the ratio of the first weight coefficient to the network transmission delay; calculating the sum of the second weight coefficient and the priority of the first service; determining a first value according to the calculation type of the first service and the calculation type which can be provided by the calculation force node; determining a second value according to the calculation force demand of the first service and the calculation force which can be provided by the calculation force node; and calculating the available value of the computing power node to the first service according to the ratio of the first weight coefficient to the network transmission delay, the sum of the priority of the second weight coefficient and the first service, the first value, the second value and the compensation coefficient.
In yet another possible design, the processing unit 92 is specifically configured to calculate a euclidean distance between a geographic location where the user equipment is located and a geographic location of each computing domain in the computing network; and ordering the computing power domains in the computing power network according to the sequence from the near to the far according to the Euclidean distance between the geographic position of the user equipment and the geographic position of each computing power domain in the computing power network.
In yet another possible design, the processing unit 92 is specifically configured to, before calculating the euclidean distance between the geographical location of the user equipment and the geographical location of each computing domain in the computing network, further configured to, for each computing domain: the geographic location of the computing force domain is determined based on the geographic locations of the respective computing force nodes included in the computing force domain.
In yet another possible design, the processing unit 92 is specifically configured to, for each computing domain: and obtaining the average geographic position of each computing force node according to the geographic position of each computing force node included in the computing force domain, and taking the average geographic position of each computing force node as the geographic position of the computing force domain.
Fig. 10 is a schematic structural diagram of an electronic device according to an embodiment of the present application. As shown in fig. 10, the electronic device includes: a processor 1001 and a memory 1002; memory 1002 stores instructions executable by processor 1001; the processor 1001 is configured to execute the instructions, cause the electronic device to implement the method as described in the previous embodiments.
In an exemplary embodiment, the present application also provides a computer-readable storage medium having stored thereon computer program instructions; the computer program instructions, when executed by an electronic device, cause the electronic device to implement the method as described in the previous embodiments.
The computer readable storage medium may be a non-transitory computer readable storage medium, which may be, for example, ROM, random Access Memory (RAM), CD-ROM, magnetic tape, floppy disk, optical data storage device, etc.
The foregoing is merely a specific embodiment of the present application, but the protection scope of the present application is not limited thereto, and any changes or substitutions within the technical scope of the present disclosure should be covered in the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (10)
1. A method for scheduling computing power resources, the method comprising:
receiving service request information of a first service from user equipment;
acquiring the geographic position of the user equipment according to the service request information of the first service;
according to the distance between the geographical position of the user equipment and the geographical position of each power calculation domain in the power calculation network, sequencing each power calculation domain in the power calculation network according to the sequence from the near to the far to obtain a first sequencing result; wherein each computing domain in the computing network comprises a plurality of computing nodes;
Acquiring the geographic position of the user equipment, the priority of the first service, the calculation type of the first service and the calculation power requirement of the first service according to the service request information of the first service;
starting from a first computational domain in the first ranking result, for each computational node in the computational domain:
according to the geographic position of the user equipment and the geographic position of the computing domain of the computing node, calculating the network transmission time delay between the user equipment and the computing node;
calculating the ratio of the first weight coefficient to the network transmission delay;
calculating the sum of a second weight coefficient and the priority of the first service;
determining a first value according to the calculation type of the first service and the calculation type which can be provided by the computing power node;
determining a second value according to the calculation force requirement of the first service and the calculation force which can be provided by the calculation force node;
calculating an available value of the computing node to the first service according to a ratio of the first weight coefficient to the network transmission delay, a sum of the second weight coefficient and the priority of the first service, the first value, the second value and a compensation coefficient;
When a computing node with the largest available value in a first computing power domain meets the requirement of the first service, determining the computing node with the largest available value in the first computing power domain as a target computing power node; the target computing force node is one of at least one computing force node which is included in a target computing force domain and meets the requirement of the first service; the target computing domain is a computing domain of a computing node of which the first one of the first sequencing results comprises a requirement meeting the first service;
and when the computing force node with the maximum available value of the first service in the first computing force domain does not meet the requirement of the first service, determining the target computing force node from the next computing force domain until the target computing force node is obtained.
2. The method of claim 1, wherein the ordering the computing domains in the computing network in order from near to far according to a distance between a geographic location of the user device and a geographic location of each computing domain in the computing network comprises:
calculating the Euclidean distance between the geographic position of the user equipment and the geographic position of each computing domain in the computing network;
And sequencing all the computing power domains in the computing power network according to the sequence from the near to the far according to the Euclidean distance between the geographic position of the user equipment and the geographic position of each computing power domain in the computing power network.
3. The method of claim 2, wherein prior to calculating the euclidean distance between the geographic location at which the user device is located and the geographic locations of the individual computing domains in the computing network, the method further comprises:
for each of the computational domains:
and determining the geographic position of the computing force domain according to the geographic position of each computing force node included in the computing force domain.
4. A method according to claim 3, wherein for each of the computational domains: determining the geographic location of the computing force domain according to the geographic location of each computing force node included in the computing force domain, including:
for each of the computational domains:
and obtaining the average geographic position of each computing power node according to the geographic position of each computing power node included in the computing power domain, and taking the average geographic position of each computing power node as the geographic position of the computing power domain.
5. An apparatus for scheduling computational resources, the apparatus comprising:
A receiving unit, configured to receive service request information of a first service from a user equipment;
the processing unit is used for acquiring the geographic position of the user equipment according to the service request information of the first service; according to the distance between the geographical position of the user equipment and the geographical position of each power calculation domain in the power calculation network, sequencing each power calculation domain in the power calculation network according to the sequence from the near to the far to obtain a first sequencing result; wherein each computing domain in the computing network comprises a plurality of computing nodes; acquiring the geographic position of the user equipment, the priority of the first service, the calculation type of the first service and the calculation power requirement of the first service according to the service request information of the first service; starting from a first computational domain in the first ranking result, for each computational node in the computational domain: according to the geographic position of the user equipment and the geographic position of the computing domain of the computing node, calculating the network transmission time delay between the user equipment and the computing node; calculating the ratio of the first weight coefficient to the network transmission delay; calculating the sum of a second weight coefficient and the priority of the first service; determining a first value according to the calculation type of the first service and the calculation type which can be provided by the computing power node; determining a second value according to the calculation force requirement of the first service and the calculation force which can be provided by the calculation force node; calculating an available value of the computing node to the first service according to a ratio of the first weight coefficient to the network transmission delay, a sum of the second weight coefficient and the priority of the first service, the first value, the second value and a compensation coefficient; when a computing node with the largest available value in a first computing power domain meets the requirement of the first service, determining the computing node with the largest available value in the first computing power domain as a target computing power node; the target computing force node is one of at least one computing force node which is included in a target computing force domain and meets the requirement of the first service; the target computing domain is a computing domain of a computing node of which the first one of the first sequencing results comprises a requirement meeting the first service; and when the computing force node with the maximum available value of the first service in the first computing force domain does not meet the requirement of the first service, determining the target computing force node from the next computing force domain until the target computing force node is obtained.
6. The apparatus according to claim 5, comprising:
the processing unit is specifically configured to calculate a euclidean distance between a geographic location where the user equipment is located and a geographic location of each computing domain in the computing network; and sequencing all the computing power domains in the computing power network according to the sequence from the near to the far according to the Euclidean distance between the geographic position of the user equipment and the geographic position of each computing power domain in the computing power network.
7. The apparatus according to claim 6, wherein the processing unit is further configured, before calculating the euclidean distance between the geographical location of the user device and the geographical location of each computing domain in the computing network, to: and determining the geographic position of the computing force domain according to the geographic position of each computing force node included in the computing force domain.
8. The apparatus according to claim 7, comprising:
the processing unit is specifically configured to, for each of the computing power domains: and obtaining the average geographic position of each computing power node according to the geographic position of each computing power node included in the computing power domain, and taking the average geographic position of each computing power node as the geographic position of the computing power domain.
9. An electronic device, comprising: a processor and a memory; the memory stores instructions executable by the processor; the processor is configured to, when executing the instructions, cause the electronic device to implement the method of any one of claims 1-4.
10. A computer-readable storage medium, the computer-readable storage medium comprising: computer software instructions; when executed in an electronic device, causes the electronic device to implement the method of any one of claims 1-4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111062337.1A CN113938816B (en) | 2021-09-10 | 2021-09-10 | Computing power resource scheduling method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111062337.1A CN113938816B (en) | 2021-09-10 | 2021-09-10 | Computing power resource scheduling method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113938816A CN113938816A (en) | 2022-01-14 |
| CN113938816B true CN113938816B (en) | 2023-04-28 |
Family
ID=79275433
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111062337.1A Active CN113938816B (en) | 2021-09-10 | 2021-09-10 | Computing power resource scheduling method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113938816B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114968573A (en) * | 2022-05-24 | 2022-08-30 | 中国联合网络通信集团有限公司 | Computing resource scheduling method and device and computer readable storage medium |
| CN114884862A (en) * | 2022-06-06 | 2022-08-09 | 中国联合网络通信集团有限公司 | Terminal service processing method and device |
| CN115086230B (en) * | 2022-06-15 | 2023-06-30 | 中国联合网络通信集团有限公司 | Method, device, equipment and storage medium for determining computing network route |
| CN115277598B (en) * | 2022-06-20 | 2024-04-02 | 中国联合网络通信集团有限公司 | Method and device for scheduling computing power resources and computer readable storage medium |
| CN115361280B (en) * | 2022-08-15 | 2023-11-28 | 中国联合网络通信集团有限公司 | Method, device, equipment and storage medium for invoking calculation power network |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020207264A1 (en) * | 2019-04-08 | 2020-10-15 | 阿里巴巴集团控股有限公司 | Network system, service provision and resource scheduling method, device, and storage medium |
| CN112631758A (en) * | 2020-12-30 | 2021-04-09 | 杭州雾联科技有限公司 | Edge computing resource scheduling method, device and equipment and readable storage medium |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111649724B (en) * | 2020-06-04 | 2022-09-06 | 百度在线网络技术（北京）有限公司 | Visual positioning method and device based on mobile edge calculation |
-
2021
- 2021-09-10 CN CN202111062337.1A patent/CN113938816B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020207264A1 (en) * | 2019-04-08 | 2020-10-15 | 阿里巴巴集团控股有限公司 | Network system, service provision and resource scheduling method, device, and storage medium |
| CN112631758A (en) * | 2020-12-30 | 2021-04-09 | 杭州雾联科技有限公司 | Edge computing resource scheduling method, device and equipment and readable storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 雷波 ; 王江龙 ; 赵倩颖 ; 余勇志 ; 杨明川 ; .基于计算、存储、传送资源融合化的新型网络虚拟化架构.电信科学.2020,(第07期),全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113938816A (en) | 2022-01-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113938816B (en) | Computing power resource scheduling method and device | |
| CN111427679B (en) | Computing task scheduling method, system and device for edge computing | |
| US20150356703A1 (en) | Arranging a transport service based on computed vectors associated with service providers | |
| CN109005211B (en) | Micro-cloud deployment and user task scheduling method in wireless metropolitan area network environment | |
| CN115134371A (en) | Scheduling method, system, equipment and medium containing edge network computing resources | |
| CN106161553B (en) | Community application information pushing method and system | |
| CN112926090A (en) | Service analysis method and device based on differential privacy | |
| Imagane et al. | Evaluation and analysis of system latency of edge computing for multimedia data processing | |
| CN115834586B (en) | Communication device deployment method and device, electronic device and readable storage medium | |
| CN110347973B (en) | Method and device for generating information | |
| CN111737009A (en) | Local end and cloud computing distribution method and device and local server | |
| US6813390B2 (en) | Scalable expandable system and method for optimizing a random system of algorithms for image quality | |
| Fu et al. | Using NearestGraph QoS prediction method for service recommendation in the cloud | |
| CN116902712A (en) | Elevator dispatching method, elevator dispatching device and electronic equipment | |
| CN115309520A (en) | Task migration method and device, electronic equipment and storage medium | |
| CN111836274B (en) | Service processing method and device | |
| CN112866358B (en) | Method, system and device for rescheduling service of Internet of things | |
| CN116107732A (en) | Task decomposition method and system for collaborative task system | |
| CN112258009B (en) | Intelligent government affair request processing method | |
| CN114297041A (en) | Network heterogeneous computing platform testing method and device and computer equipment | |
| CN114254381A (en) | Data evaluation method, device, equipment and medium based on multi-party security calculation | |
| CN113657718A (en) | Multi-robot dynamic alliance task allocation method and related device | |
| WO2002080563A2 (en) | Scalable expandable system and method for optimizing a random system of algorithms for image quality | |
| KR20160150563A (en) | Method of managing resources of public safety network | |
| CN106056231B (en) | Automatic matching recommendation method and device based on multi-party position and time availability |
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 |