CN117955898A

CN117955898A - Load balancing optimization method, device and product for data sharing network

Info

Publication number: CN117955898A
Application number: CN202410353875.3A
Authority: CN
Inventors: 杨国利; 刘艺; 刘昊; 李翔; 郑奇斌; 秦伟; 史殿习; 黄罡
Original assignee: Beijing Big Data Advanced Technology Research Institute
Current assignee: Beijing Big Data Advanced Technology Research Institute
Priority date: 2024-03-27
Filing date: 2024-03-27
Publication date: 2024-04-30
Anticipated expiration: 2044-03-27
Also published as: CN117955898B

Abstract

The application provides a load balancing optimization method, a device and a product for a data sharing network, which relate to the technical field of data sharing and network optimization, and the method comprises the following steps: taking a service application program as a logic network node, and carrying out logic network interconnection to obtain a logic network; the communication equipment is used as a physical network node to carry out physical network interconnection to obtain a physical network; the logic network node is hung on the corresponding physical network node through the binding link, and a data sharing network is obtained; performing robustness design of a main path and a backup path; based on a genetic algorithm, encoding a plurality of load balancing optimization schemes into population individuals to generate an initialized population; designing an objective function and constraint conditions; executing a genetic algorithm to enable the initialized population to carry out population evolution under constraint conditions until an objective function converges, and determining an optimal load balancing optimization scheme; and according to the optimal load balancing optimization scheme, determining a corresponding main path and backup path for the logic link, and distributing load resources.

Description

Load balancing optimization method, device and product for data sharing network

Technical Field

The application relates to the technical field of data sharing and network optimization, in particular to a load balancing optimization method, device and product for a data sharing network.

Background

In the big data age, due to business requirements, shared data is generally required to be exchanged with one or more internal and external organizations, but achieving stable data sharing is not as simple as building a data platform. From data acquisition to data processing and cleaning, the data transmission flow is transferred to management and application for value, and a large amount of basic work is needed. Particularly, under the condition that the load resources provided by the wireless and optical fiber and other physical transmission networks are limited, how to realize reliable, efficient, continuous and stable data sharing among service application programs needs to depend on robust information-physical network support.

However, in the existing information-physical network, in the process of providing real-time network services for each service application program, there are often the problem that part of service application programs have insufficient load resources supplied by the physical transmission network because of larger load requirements, and the problem that part of service application programs have idle and wasted load resources configured by the physical transmission network because of smaller load requirements. Therefore, it is necessary to develop a load balancing optimization method, device and product for a data sharing network, so as to allocate balanced load resources for data sharing among various service applications, so as to realize stable and continuous data sharing.

Disclosure of Invention

In view of the above, embodiments of the present application provide a method, apparatus, and product for load balancing optimization for a data sharing network, so as to overcome or at least partially solve the above problem.

In a first aspect of the embodiment of the present application, a load balancing optimization method for a data sharing network is provided, where the method includes:

taking each service application program as a logic network node, and taking data sharing requirements existing among the service application programs as logic links, and performing logic network interconnection to obtain a logic network;

taking each communication equipment as a physical network node, and carrying out physical network interconnection according to physical links existing among the communication equipment to obtain a physical network;

Each logic network node is hung on a corresponding physical network node through a binding link to obtain a data sharing network, wherein the data sharing network is a logic-physical double-layer network model facing data sharing; the physical links are used for providing load resources for one or more corresponding logic links, and the logic links utilize the provided load resources to realize data communication between two connected logic network nodes;

performing robustness design of a main path and a backup path; the main path is used for providing normal load resources for the corresponding logic links, and the backup path is used for providing load resources for the corresponding logic links when the main path fails;

based on a genetic algorithm, encoding a plurality of load balancing optimization schemes into population individuals to generate an initialized population; the load balancing optimization scheme comprises the following steps: based on the robustness design, a main path and a backup path are selected for each logic link in the logic network, and the main path and the backup path are the size of load resources configured for the logic link;

Designing an objective function and constraint conditions; the objective function is used for calculating the population individual fitness of each population individual, and the constraint condition is used for constraining the load resource size in the load balancing optimization scheme corresponding to the population individual;

Executing the genetic algorithm to enable the initialized population to carry out population evolution under the constraint condition until the objective function converges, and determining an optimal load balancing optimization scheme;

And according to the optimal load balancing optimization scheme, determining a corresponding main path and backup path in the physical link for each logic link in the data sharing network, and distributing load resources of the main path and the backup path for the logic links.

In one possible implementation manner, the robust design of the main path and the backup path includes:

The load requirement of each logic link is defined; the load demand represents the size of load resources required for realizing data sharing between two business application programs corresponding to the logic link;

defining the upper load limit of each physical link; the upper load limit represents the maximum value of the load resources which can be provided by the communication equipment corresponding to the physical link;

Defining a path selection rule for selecting a corresponding main path and backup path for each logic link from the physical links; the path selection rule indicates that the size of load resources configured for the logic link by the selected main path and the backup path is equal and is not smaller than the load requirement of the logic link and is not greater than the load upper limit of the logic link.

In a possible implementation manner, the executing the genetic algorithm, subjecting the initialized population to population evolution under the constraint condition until the objective function converges, and determining an optimal load balancing optimization scheme includes:

under the constraint condition, performing cross mutation operation on the initialized population;

calculating the individual fitness of the optimal population of the population after evolution according to the objective function every time the population finishes evolution;

judging whether the individual fitness of the optimal population converges to a stable state or not;

under the condition that the individual fitness of the optimal population does not converge to a stable state, carrying out the next round of population evolution, and recalculating the individual fitness of the optimal population of the population after the next round of evolution;

And under the condition that the individual fitness of the optimal population converges to a stable state, ending the population evolution to obtain the optimal load balancing optimization scheme.

In one possible embodiment, the constraint includes: a demand constraint and a load constraint;

The requirement constraint condition indicates that the load resource of the main path or the backup path configured for each logic link is not smaller than the load requirement of the logic link and not greater than the load upper limit of the logic link; and the load resource of the main path is equal to the load resource of the backup path in size;

The bearer constraint indicates that, for each physical link, the configured load resources of all logical links that are carried as primary and backup paths are not greater than the upper load limit of that physical link.

In a possible implementation manner, the calculating the optimal population individual fitness of the population after evolution according to the objective function includes:

calculating a main and standby path overlapping degree parameter, a load resource surplus degree parameter and a network load balancing degree parameter of each population individual in the current population;

determining population individual fitness corresponding to population individuals according to the main and standby path overlapping degree parameter, the load resource surplus degree parameter and the network load balancing degree parameter;

determining the optimal population individual fitness according to the sequence from the large population individual fitness to the small population individual fitness;

The main path and backup path overlap degree parameter represents the average overlap degree of the main path and the backup path of the logic link in the load balancing optimization scheme corresponding to the population individuals; the load resource surplus degree parameter represents the average surplus degree of the load resources configured by the logic links in the load balancing optimization scheme corresponding to the population individuals; and the network load balancing degree parameter represents the balancing degree of the load surplus of the logic link in the load balancing optimization scheme corresponding to the population individuals.

In one possible implementation, the primary and backup path overlap degree parameters are calculated according to the following formula:

；

Wherein f ₀ represents the overlapping degree of the active and standby paths, and M represents the total number of the logic links; representing the primary path of the ith logical link,/> Representing a standby path of an ith logical link;

The load resource margin parameter is calculated according to the following formula:

；

Wherein f _r denotes the load resource margin parameter, Load resources configured for the ith logical link; /(I)Representing the load demand of the ith logical link;

The network load balancing degree parameter is calculated according to the following formula:

；

Wherein f _b represents the network load balancing degree parameter;

Determining the population individual fitness corresponding to the population individuals according to the main and standby path overlapping degree parameter, the load resource surplus degree parameter and the network load balancing degree parameter according to the following formula:

；

Wherein f represents the fitness of individuals in the population.

In one possible implementation, the encoding the plurality of load balancing optimization schemes into population individuals based on the genetic algorithm generates an initialized population, including:

for each logic link, determining an reachable path set of the logic link in a physical network according to physical network nodes respectively bound by two connected logic network nodes;

Determining multiple combinations of a main path and a backup path of the logic link from the reachable path set, and determining the main path, the backup path and the load resource configured for the logic link as solving elements of the logic link by each combination as the load resource configured for the logic link;

generating a plurality of load balancing optimization schemes according to the solving elements;

And encoding each load balancing optimization scheme into population individuals to generate an initialized population.

The second aspect of the embodiment of the present application further provides a load balancing optimization device for a data sharing network, where the device includes:

The logic network construction module is used for carrying out logic network interconnection by taking each service application program as a logic network node and taking the data sharing requirement existing among the service application programs as a logic link to obtain a logic network;

The physical network construction module is used for carrying out physical network interconnection by taking each communication device as a physical network node according to physical links existing among the communication devices to obtain a physical network;

the data sharing network construction module is used for hanging each logic network node on a corresponding physical network node through a binding link to obtain a data sharing network, and the data sharing network is a logic-physical double-layer network model facing data sharing; the physical links are used for providing load resources for one or more corresponding logic links, and the logic links utilize the provided load resources to realize data communication between two connected logic network nodes;

The robustness design module is used for carrying out robustness design of the main path and the backup path; the main path is used for providing normal load resources for the corresponding logic links, and the backup path is used for providing load resources for the corresponding logic links when the main path fails;

The coding module is used for coding a plurality of load balancing optimization schemes into population individuals based on a genetic algorithm to generate an initialized population; the load balancing optimization scheme comprises the following steps: based on the robustness design, a main path and a backup path are selected for each logic link in the logic network, and the main path and the backup path are the size of load resources configured for the logic link;

The function and constraint design module is used for designing an objective function and constraint conditions;

the scheme determining module is used for executing the genetic algorithm, so that the initialized population is subjected to population evolution under the constraint condition until the objective function converges, and an optimal load balancing optimization scheme is determined;

And the load balancing optimization module is used for determining a corresponding main path and backup path in the physical link for each logic link in the data sharing network according to the optimal load balancing optimization scheme, and distributing load resources of the main path and the backup path for the logic links.

In one possible implementation, the robustness design module includes:

The load demand determining submodule is used for determining the load demand of each logic link; the load demand represents the size of load resources required for realizing data sharing between two business application programs corresponding to the logic link;

the load upper limit determining submodule is used for determining the load upper limit of each physical link; the upper load limit represents the maximum value of the load resources which can be provided by the communication equipment corresponding to the physical link;

The rule definition submodule is used for defining a path selection rule and selecting a corresponding main path and backup path for each logic link from the physical links; the path selection rule indicates that the size of load resources configured for the logic link by the selected main path and the backup path is equal and is not smaller than the load requirement of the logic link and is not greater than the load upper limit of the logic link.

In one possible implementation manner, the scheme determining module includes:

the cross mutation sub-module is used for executing cross mutation operation on the initialized population under the constraint condition;

The objective function calculation sub-module is used for calculating the optimal population individual fitness of the population after evolution according to the objective function when the population finishes evolution once;

The judging submodule is used for judging whether the individual fitness of the optimal population converges to a stable state or not;

The iteration sub-module is used for carrying out the next round of seed group evolution under the condition that the individual fitness of the optimal population is not converged to a stable state, and recalculating the individual fitness of the optimal population of the population after the next round of evolution;

And the scheme determining submodule is used for ending the population evolution under the condition that the individual fitness of the optimal population converges to a stable state to obtain the optimal load balancing optimization scheme.

In one possible implementation, the objective function calculation sub-module includes:

the parameter calculation unit is used for calculating the main and standby path overlapping degree parameter, the load resource surplus degree parameter and the network load balancing degree parameter of each population individual in the current population;

the individual fitness determining unit is used for determining population individual fitness corresponding to population individuals according to the main and standby path overlapping degree parameter, the load resource surplus degree parameter and the network load balancing degree parameter;

The optimal value determining unit is used for determining the optimal population individual fitness according to the sequence from the large population individual fitness to the small population individual fitness;

In one possible implementation manner, the parameter calculation unit calculates the primary and backup path overlapping degree parameter according to the following formula:

；

The parameter calculation unit calculates the load resource surplus degree parameter according to the following formula:

；

the parameter calculation unit calculates the network load balancing degree parameter according to the following formula:

；

Wherein f _b represents the network load balancing degree parameter;

The individual fitness determining unit determines population individual fitness corresponding to the population individuals according to the main and standby path overlapping degree parameter, the load resource surplus degree parameter and the network load balancing degree parameter according to the following formula:

；

Wherein f represents the fitness of individuals in the population.

In one possible embodiment, the coding module includes:

An reachable path set determining sub-module, configured to determine, for each logical link, a reachable path set of the logical link in a physical network according to physical network nodes to which the two connected logical network nodes are respectively bound;

The solving element determining submodule is used for determining multiple combinations of a main path and a backup path of the logic link from the reachable path set, each combination is configured as a load resource of the logic link, and the main path, the backup path and the load resource configured for the logic link are determined as solving elements of the logic link;

the scheme generation sub-module is used for generating a plurality of load balancing optimization schemes according to the solving elements;

And the population individual coding sub-module is used for coding each load balancing optimization scheme into population individuals to generate an initialized population.

The third aspect of the embodiment of the application also provides an electronic device, which comprises a memory, a processor and a computer program stored on the memory, wherein the processor executes the computer program to realize the steps in the load balancing optimization method for the data sharing network according to the first aspect of the embodiment of the application.

The fourth aspect of the embodiment of the present application further provides a computer readable storage medium, on which a computer program/instruction is stored, where the computer program/instruction implements the steps in the load balancing optimization method for a data sharing network according to the first aspect of the embodiment of the present application when the computer program/instruction is executed by a processor.

The fifth aspect of the embodiment of the present application further provides a computer program product, which when executed on an electronic device, causes a processor to implement the steps in the load balancing optimization method for a data sharing network according to the first aspect of the embodiment of the present application.

According to the load balancing optimization method for the data sharing network, which is provided by the embodiment of the application, each service application program is used as a logic network node, and the data sharing requirement existing among each service application program is used as a logic link, so that logic network interconnection is carried out to obtain a logic network; taking each communication equipment as a physical network node, and carrying out physical network interconnection according to physical links existing among the communication equipment to obtain a physical network; each logic network node is hung on a corresponding physical network node through a binding link to obtain a data sharing network, wherein the data sharing network is a logic-physical double-layer network model facing data sharing; the physical links are used for providing load resources for one or more corresponding logic links, and the logic links utilize the provided load resources to realize data communication between two connected logic network nodes; performing robustness design of a main path and a backup path; the main path is used for providing normal load resources for the corresponding logic links, and the backup path is used for providing load resources for the corresponding logic links when the main path fails; based on a genetic algorithm, encoding a plurality of load balancing optimization schemes into population individuals to generate an initialized population; the load balancing optimization scheme comprises the following steps: based on the robustness design, a main path and a backup path are selected for each logic link in the logic network, and the main path and the backup path are the size of load resources configured for the logic link; designing an objective function and constraint conditions; the objective function is used for calculating the population individual fitness of each population individual, and the constraint condition is used for constraining the load resource size in the load balancing optimization scheme corresponding to the population individual; executing the genetic algorithm to enable the initialized population to carry out population evolution under the constraint condition until the objective function converges, and determining an optimal load balancing optimization scheme; and according to the optimal load balancing optimization scheme, determining a corresponding main path and backup path in the physical link for each logic link in the data sharing network, and distributing load resources of the main path and the backup path for the logic links.

The concrete beneficial effects are that:

On one hand, the embodiment of the application obtains the optimal load balancing optimization scheme by utilizing the genetic algorithm calculation, and allocates proper load resource size for each logic link (corresponding to the data sharing between two business application programs), thereby avoiding insufficient load resources or idle waste and improving the utilization rate of the load resources; on the other hand, the embodiment of the application sets the corresponding main path and backup path in the physical link for each logic link by utilizing the determined optimal load balancing optimization scheme, and in general, two service application programs corresponding to the logic links use the load resources provided by the main path to share data, and when the main path is abnormal and can not provide the load resources, the load resources of the backup path are automatically converted to be used for data sharing, thereby realizing the redundant design of the load resources and further improving the stability of a data sharing network.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the description of the embodiments of the present application 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 step flowchart of a load balancing optimization method for a data sharing network according to an embodiment of the present application;

FIG. 2 is a diagram of a logical-physical dual-layer network topology according to an embodiment of the present application;

FIG. 3 is a schematic diagram of a primary path and a backup path according to an embodiment of the present application;

fig. 4 is a schematic structural diagram of a load balancing and optimizing device according to an embodiment of the present application;

fig. 5 is a schematic diagram of an electronic device according to an embodiment of the present application.

Detailed Description

Exemplary embodiments of the present application will be described in more detail below with reference to the accompanying drawings in the embodiments of the present application. While exemplary embodiments of the present application are shown in the drawings, it should be understood that the present application may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the application to those skilled in the art.

The information-physical system (Cyber-PHYSICAL SYSTEM, CPS) refers to a multidimensional complex system integrating computing, network and physical environment, and realizes real-time sensing, dynamic control and information service of a large complex system through organic fusion and deep cooperation of a 3C (Computation, communication, control) technology. The information-physical network is connected with the information world and the physical world through the data transmission layer, so that efficient data transmission is realized, real-time reliability of physical network number distribution, transmission and reception is ensured, and real-time network service is provided for each business application system.

However, in the existing information-physical network, in the process of providing real-time network services for each service application program, there are often the problem that part of service application programs have insufficient load resources supplied by the physical transmission network because of a large data volume, and the problem that part of service application programs have idle and wasted load resources configured by the physical transmission network because of a small data volume.

In view of the above problems, an embodiment of the present application proposes a server-oriented system for scheduling tasks with multiple resource functions without perceived computation, so as to solve the problems of low resource utilization and the like. The server-oriented non-aware computing multi-resource function task scheduling system provided by the embodiment of the application is described in detail below through some embodiments and application scenes thereof with reference to the accompanying drawings.

The first aspect of the embodiment of the present application provides a load balancing optimization method for a data sharing network, referring to fig. 1, fig. 1 is a step flowchart of the load balancing optimization method for a data sharing network, as shown in fig. 1, where the method includes:

Step S101, using each service application program as a logic network node, using data sharing requirements existing among the service application programs as logic links, and performing logic network interconnection to obtain a logic network.

Step S102, each communication device is used as a physical network node, and physical network interconnection is performed according to physical links existing between the communication devices, so as to obtain a physical network.

Step S103, each logic network node is hung on a corresponding physical network node through a binding link to obtain a data sharing network, wherein the data sharing network is a logic-physical double-layer network model facing data sharing; the physical links are used for providing load resources for one or more corresponding logic links, and the logic links utilize the provided load resources to realize data communication between two connected logic network nodes.

In this embodiment, the implementation of data sharing needs to rely on logical network interconnection and physical network interconnection. Each logical network node in the logical network represents a kind of business application, also called application service, representing different business applications running within different organizations or the same organization, but within different authorization boundaries. In the process of constructing a logic network in a data sharing network, when a data sharing requirement exists between two business application programs, such as file sharing service or application, and data exchange, information service (such as identity verification and log record) providing and the like occur in a session layer, a presentation layer or an application layer, logic network interconnection is performed, and a logic link between two corresponding logic network nodes is generated.

Each physical network node in the physical network represents a physical communication node or communication device, e.g. a switching station. In the process of constructing a physical network in a data sharing network, when any one or more of optical fiber, microwave, wireless, bluetooth and other physical links are actually erected between two communication devices, physical network interconnection is performed, and a physical link between two physical network nodes is generated, wherein the physical link indicates the data accessibility between the two connected physical network nodes.

And hanging each logic network node in the logic network on a corresponding physical network node in the physical network through a binding link to obtain a data sharing network, so that the data sharing network is a logic-physical double-layer network model oriented to data sharing. Referring to fig. 2, fig. 2 shows a topology structure diagram of a logical-physical dual-layer network, as shown in fig. 2, a data sharing network is a logical-physical dual-layer network model oriented to data sharing, an upper layer in fig. 2 is a logical network, a lower layer in fig. 2 is a physical network, nodes in the logical network are logical network nodes, and the logical network nodes are connected with each other through logical links to indicate that a data sharing requirement exists between two logical network nodes; the nodes in the physical network are physical network nodes, the physical network nodes are connected with each other through physical links, the physical links are used for realizing data communication for the two connected physical network nodes, so that transmitted data are sent to corresponding logic network nodes through the physical network nodes, load resources are provided for one or more logic links corresponding to the physical links, and the logic links utilize the provided load resources to realize data communication between the two connected logic network nodes.

As shown in fig. 2, the logical network includes service applications (i.e., logical network nodes a, b, c, d, e, f, g) and their logical links with each other, which characterize the data sharing requirements between two service applications. The physical network comprises individual communication devices (i.e., physical network nodes A, B, C, D, E, F, G, etc.) and their physical links to each other, which characterize the provision of load resources between the communication devices. The realization of data sharing also needs to rely on the interconnection between a logical network and a physical network, and each logical network node is hung on a certain physical network node through a binding link.

The physical network is a bearer of the logical network, the logical network is an application of the physical network, and the physical network and the logical network are mutually coupled and interleaved. If one physical link in the physical network fails, the logical link carried by the physical link also fails correspondingly, so that data sharing cannot be realized between two service applications corresponding to two logical network nodes connected by the logical link. As shown in fig. 2, a logical link exists between the logical network nodes b and d, which means that there is a data sharing requirement between two service applications corresponding to b and d, where b needs to obtain data of d, and/or d needs to obtain data of b. The logical network node B is suspended from the physical network node B and the logical network node D is suspended from the physical network node F via the data sharing network, so that data sharing between B and D can be achieved via the physical transmission paths B-D-F. If the physical link B-D or D-F fails, which would result in the failure of the logical link B-D, no data sharing between the logical network nodes B and D can be achieved.

Step S104, robustness design of a main path and a backup path is carried out; the main path is used for providing normal load resources for the corresponding logic link, and the backup path is used for providing load resources for the corresponding logic link when the main path fails.

Referring to fig. 3, fig. 3 is a schematic diagram of a primary path and a backup path, where as shown in fig. 3, for a logical link logi between logical network nodes b and d, it indicates that there is a data sharing requirement between two service applications corresponding to b and d. The logical network node B is suspended from the physical network node B and the logical network node D is suspended from the physical network node F, so that data sharing between B and D can be achieved through the physical transmission path B-D-F or through the physical transmission path B-C-E-G-F. And taking B-D-F as a main path, and providing load resources for the logic link logi to realize data sharing between B and D. When the physical link B-D or D-F fails, the backup path B-C-E-G-F automatically changes to provide load resources for the logical link logi, so that the failure of the logical link logi caused by the problems of failure of the main path and the like is avoided, the load redundancy is realized, and the stability of data sharing is improved.

In a possible implementation manner, the step S104 performs robust design of the primary path and the backup path, including:

Step S1041, defining the load requirement of each logic link; the load requirement represents the size of load resources required to realize data sharing between two business applications corresponding to the logical link.

Each logical link in the logical network has its corresponding load requirement. The load requirement represents the size of the load required to achieve data sharing between the two business applications to which the logical link corresponds. In order to realize data sharing between service applications, each logic link has a corresponding load requirement, and the load requirement is determined by the loaded service, and a relatively fixed value is maintained. As shown in fig. 3, data of 10Gbps is required to be normally transmitted between the logical network nodes b and D, and then the load requirement D ^b-d = 10Gbps of the corresponding logical link b-D.

Step S1042, defining the upper load limit of each physical link; the upper load limit represents the maximum value of the load resources which can be provided by the communication equipment corresponding to the physical link.

Each physical link in the physical network may provide a certain load resource based on the corresponding communication device to satisfy the corresponding data transmission bearing capacity. The upper load limit of each physical link is determined by the maximum data transmission rate R of the communication device corresponding to the physical link. As shown in fig. 3, the maximum transmission rate between the physical network nodes B and D is 15Gbps, i.e. the upper load limit R _phyB-D =15 Gbps corresponding to the physical link phyB-D.

Step S1043, defining a path selection rule for selecting a corresponding main path and backup path for each logical link from the physical links; the path selection rule indicates that the size of load resources configured for the logic link by the selected main path and the backup path is equal and is not smaller than the load requirement of the logic link and is not greater than the load upper limit of the logic link.

In selecting the primary and backup paths for each logical link, defined path selection rules need to be followed. Specifically, the main path is configured for the corresponding logical link with the load resource sizeLoad resource size/>, configured for the corresponding logical link with the backup pathEqual. And the load resource provided by the selected path is not less than the load requirement of the logic link and is not greater than the load upper limit of the logic link. Illustratively, as shown in fig. 3, the load requirement D ^b-d = 10Gbps of the logical link b-D, and the optional physical transmission path corresponding to the logical link b-D includes: B-D-F, B-D-E-G-F, B-C-E-D-F. Since the physical link D-E exists in the path b→d→e→g→f and the path b→c→e→d→f, and the upper load limit of the physical link D-E is 8Gbps, which is smaller than the load demand D ^b-d (10 Gbps) of the logical link B-D, the path involving the physical link D-E does not satisfy the path selection rule, so the two paths do not satisfy the condition of the data sharing transmission path as the bearer logical link B-D. The feasible transmission paths of the logic links B-D are B-D-F and B-C-E-G-F, and can be used as a transmission main path and a backup path for data sharing between the business applications B and D respectively.

Furthermore, the load resources provided by each physical link cannot exceed its own upper load limit. The following formula is shown:

；

where n represents the number of all logical links carried by physical link phyj, Load resources configured for logical link logi, denoted physical link phyj,/>Indicating the upper load limit of the physical link phyj. Each physical link in the physical network can provide a certain load resource based on the corresponding communication equipment so as to realize the corresponding data transmission bearing capacity. The upper load limit of each physical link is determined by the maximum data transmission rate R of the communication device corresponding to the physical link. When one physical link carries a plurality of logic links simultaneously to execute data sharing service, it is necessary to ensure that the sum of configuration load resources of each logic link is not greater than the upper load limit of the physical link. For example, if the maximum data transmission rate between the physical network nodes B and D is 15Gbps, the upper load limit R _phyB-D = 15Gbps corresponding to the physical link phyB-D. Assuming that the total two logical links that can be carried by the physical link phyB-D are two log1 and log2, the load requirement of the physical link phyB-D provided for the logical link log1 is 10Gbps, and the load requirement of the logical link log2 is 8Gbps, the sum of the load requirements of the two logical links log1 and log2 exceeds the upper load limit R _phyB-D (15 Gbps) of the physical link phyB-D, and the physical link phyB-D cannot be used as the main path or the backup path of the logical link log2 on the basis that the logical link log1 has been configured for the physical link phyB-D.

Step S105, encoding a plurality of load balancing optimization schemes into population individuals based on a genetic algorithm, and generating an initialized population; the load balancing optimization scheme comprises the following steps: and selecting a main path and a backup path for each logic link in the logic network based on the robustness design, and configuring load resource sizes for the logic link by the main path and the backup path.

In this embodiment, after the data sharing network is constructed, load resources of the corresponding physical link are also required to be allocated to the logical link therein. Because the load resources of the physical links are limited, one physical link may carry one or more logical links, and each logical link needs to be provided with load resources with a proper size, so that data sharing is realized between corresponding logical network nodes. Because of the different logical links, the corresponding service applications are different, and the required load resources are different in size. If the load resource provided by the physical link is less than the load resource actually required by the logical link, the service performance is reduced and the data sharing efficiency is lowered easily due to insufficient load resource in the data sharing process; if the load resources provided by the physical links are too many, the problems of idle load resources, waste, reduced load resource utilization rate and the like are easy to exist. In order to realize load balancing and improve data sharing efficiency, a plurality of load balancing optimization schemes are encoded into population individuals through a genetic algorithm, and then an optimal load balancing optimization scheme is determined, so that a corresponding physical link and load resources of the physical link are configured for each logic link according to the optimization scheme. Specifically, in the genetic algorithm, each population of individuals corresponds to one feasible solution of the load balancing optimization problem, namely one load balancing optimization scheme, and the feasible solution is encoded as follows:

；

Wherein, Load resources configured for logical links logi, M represents the total number of logical links in the logical network, and the primary path of logical link logi is/>The backup path is/>. The main path and the backup path of each logical link are determined based on the robust designs of the main path and the backup path in step S104, that is, each main path and each backup path satisfy the defined path selection rule (the size of the load resources configured for the logical link by the selected main path and backup path is equal to and not less than the load requirement of the logical link and not greater than the load upper limit of the main path and backup path).

In a possible implementation manner, the step S105 encodes a plurality of load balancing optimization schemes into population individuals based on a genetic algorithm, and generates an initialized population, including:

Step S1051, for each logical link, determining a set of reachable paths of the logical link in the physical network according to the physical network nodes to which the two connected logical network nodes are respectively bound.

In this embodiment, for each logical link logi, the physical network nodes to which the two logical network nodes connected according to the logical link are respectively boundAnd/>. To enable data sharing of logical links, it is necessary to determine the physical network node/>To/>Is a physical link to the network. When the physical network G _p has a complex structure and more nodes, two physical network nodes/>And/>There are multiple physical links between, and an reachable path set pi ⁱ is available, which can be expressed as pi ⁱ:

；

Wherein, Two physical network nodes/>, representing a physical networkAnd/>Extraction functions of all reachable path sets in between.

Step S1052, determining multiple combinations of the primary path and the backup path of the logical link from the set of reachable paths, and determining the primary path, the backup path, and the load resource configured for the logical link as solution elements of the logical link for each combination as the load resource configured for the logical link.

Specifically, two physical links are selected from the set of reachable paths as the primary path for the logical linkAnd backup Path/>. Main Path/>And backup Path/>All belong to the elements in the reachable path set, and the corresponding numbers are/>, respectivelyAnd/>. When multiple reachable paths exist in the reachable path set, multiple combinations of the main path and the backup path can be generated, and for each combination of the main path and the backup path, the load resources/>, of the main path and the backup path are calculatedThe load resources provided by the primary path are the same size as the load resources provided by the backup path assigned to the logical link logi. For each logical link logi, the corresponding primary path/>Backup Path/>And configured load resources/>A solution element for the logical link load configuration scheme.

And step S1053, generating a plurality of load balancing optimization schemes according to the solving elements.

And S1054, encoding each load balancing optimization scheme into population individuals to generate an initialized population.

Step S106, designing an objective function and constraint conditions; the objective function is used for calculating the population individual fitness of each population individual, and the constraint condition is used for constraining the load resource size in the load balancing optimization scheme corresponding to the population individual.

In this embodiment, the objective function is used to calculate the fitness of each individual population after each population has evolved. Specifically, population individual fitness is designed based on three factors: the method comprises the steps of a primary path overlapping degree parameter, a standby path overlapping degree parameter, a load resource surplus degree parameter and a network load balancing degree parameter. The main path and backup path overlap degree parameter represents the average overlap degree of the main path and the backup path of the logic link in the load balancing optimization scheme corresponding to the population individuals; the load resource surplus degree parameter represents the average surplus degree of the load resources configured by the logic links in the load balancing optimization scheme corresponding to the population individuals; and the network load balancing degree parameter represents the balancing degree of the load surplus of the logic link in the load balancing optimization scheme corresponding to the population individuals.

In this embodiment, the requirement constraint and the load constraint need to be satisfied when each population of individuals is encoded. The requirement constraint indicates that the load resource of the primary path or the backup path configured for each logical link logi is not less than the load requirement of the logical link, that is:

；

Wherein, Load resource configured for logical link logi,/>Representing the load demand of logical link logi. The load requirement of the logical links b-d is, for example, 10Gbps. As shown in fig. 3, the physical transmission paths in the set of reachable paths corresponding to the logical links b-d include: B-D-F, B-D-E-G-F, B-C-E-D-F. Since the physical link D-E exists in the path b→d→e→g→f and the path b→c→e→d→f, and the upper load limit of the physical link D-E is 8Gbps, which is smaller than the load demand D ^b-d (10 Gbps) of the logical link B-D, the demand constraint condition is not satisfied, so the two paths do not satisfy the condition as the data sharing transmission path carrying the logical link B-D.

In this embodiment, the bearer constraint indicates that, for each physical link, the configured load resources of all the logical links that are carried by the primary path and the backup path are not greater than the upper load limit of the physical link, that is:

；

where n represents the number of all logical links carried by physical link phyj, Load resources configured for logical link logi, denoted physical link phyj,/>Indicating the upper load limit of the physical link phyj. For example, if the maximum data transmission rate between the physical network nodes B and D is 15Gbps, the upper load limit R _phyB-D = 15Gbps corresponding to the physical link phyB-D. If the two logical links that can be carried by the physical link phyB-D have two log1 and log2, the load requirement of the logical link log1 is 10Gbps, and the load requirement of the logical link log2 is 8Gbps, the sum of the load requirements of the two logical links log1 and log2 exceeds the upper load limit R _phyB-D (15 Gbps) of the physical link phyB-D, the carrying constraint condition is not satisfied, and the physical link phyB-D cannot be used as the main path or the backup path of the logical link log2 on the basis that the logical link log1 has been configured for the physical link phyB-D.

And S107, executing the genetic algorithm, and enabling the initialized population to carry out population evolution under the constraint condition until the objective function converges, so as to determine an optimal load balancing optimization scheme.

In a possible implementation manner, the step S107 executes the genetic algorithm to make the initialized population undergo population evolution under the constraint condition until the objective function converges, and determines an optimal load balancing optimization scheme, including:

Step S1071, under the constraint condition, performing cross mutation operation on the initialized population. Specifically, the population evolution is performed through the cross mutation operation, so that the population individuals after each population evolution still meet the requirement constraint condition and the bearing constraint condition.

Step S1072, each time the population finishes evolution, calculating the optimal population individual fitness of the population after evolution according to the objective function.

In a possible implementation manner, the step S1072 calculates, according to the objective function, an optimal population individual fitness of the population after evolution, including:

Step S1072-a, calculating the main and standby path overlapping degree parameter, the load resource surplus degree parameter and the network load balancing degree parameter of each population individual in the current population.

；

Wherein f ₀ represents the overlapping degree of the active and standby paths, and M represents the total number of the logic links; representing the primary path of the ith logical link,/> Representing a standby path of an ith logical link; as shown in the formula, the larger the value of the main and standby path overlapping degree parameter is, the higher the overlapping degree of the main path and the standby path in the scheme is, and when a certain section of physical link is abnormal, the higher the possibility that the main path and the standby path are simultaneously invalid, namely the higher the possibility that the corresponding logic link is invalid, the worse the robustness of the data sharing network is.

；

Wherein f _r denotes the load resource margin parameter, Load resources configured for the ith logical link; /(I)Representing the load demand of the ith logical link; as shown in the formula, the larger the value of the load resource surplus degree parameter is, the higher the average resource surplus degree of each logic link in the scheme is, the smaller the probability of insufficient load resource of the logic link is, and the better the transmission performance of the data sharing network is.

；

Wherein f _b represents the network load balancing degree parameter; as shown in the formula, the larger the network load balancing degree parameter value is, the lower the load balancing degree in the data sharing network is, for example, the load resource surplus degree configured for the logic link is too large, the idle waste of load resources is easy to be caused, and the network resource allocation is unreasonable.

Step S1072-b, determining the population individual fitness corresponding to the population individuals according to the primary and standby path overlapping degree parameter, the load resource surplus degree parameter and the network load balancing degree parameter.

In one possible implementation manner, the population individual fitness corresponding to the population individual is determined according to the main and standby path overlapping degree parameter, the load resource surplus degree parameter and the network load balancing degree parameter according to the following formula:

；

Wherein f represents the fitness of the population individuals, f _b represents the network load balancing degree parameter, and f _r represents the load resource surplus degree parameter.

Step S1072-c, determining the optimal population individual fitness according to the sequence from the large population individual fitness to the small population individual fitness. Specifically, after each population is evolved, the population individual fitness of each population individual in the evolved population is calculated, and then the maximum value is determined from the population individual fitness to be used as the optimal population individual fitness after the current population is evolved.

And step S1073, judging whether the fitness of the individuals in the optimal population converges to a stable state.

Step S1074, under the condition that the individual fitness of the optimal population does not converge to a stable state, the next round of population evolution is carried out, and the individual fitness of the optimal population of the population after the next round of evolution is recalculated.

And step S1075, ending the population evolution under the condition that the individual fitness of the optimal population converges to a stable state, and obtaining the optimal load balancing optimization scheme. Specifically, under the condition that the individual fitness of the optimal population converges to a stable state (for example, k generations of continuous evolution of the population, and the value of the individual fitness of the optimal population remains unchanged), the population evolution is ended to obtain an optimal solution, so that an optimal load balancing optimization scheme is determined, and a main path, a backup path and configured load resources of each logic link of the data sharing network are determined.

Step S108, according to the optimal load balancing optimization scheme, determining a corresponding main path and backup path in the physical link for each logic link in the data sharing network, and distributing load resources of the main path and the backup path for the logic link.

The embodiment of the application constructs a logic-physical double-layer network model oriented to data sharing; the main path and the backup path are transmitted for each logic link data sharing in the model, and corresponding load resources are configured, so that when the main path is abnormal, the backup path can automatically switch to provide the load resources for the logic links, and the robustness of network data sharing is maintained through link backup and redundant loads. And based on genetic algorithm, the constraint conditions (demand constraint condition and bearing constraint condition) are utilized to ensure that the load resources on the main path and the backup path are not smaller than the load demands of corresponding logic links and are not larger than the load upper limit of the physical links, population individual fitness is calculated from three aspects (main and backup path overlapping degree parameters, load resource surplus degree parameters and network load balancing degree parameters), and an optimal load balancing optimization scheme is determined according to the population individual fitness, so that the transmission main path and the backup path corresponding to each logic link are relatively independent (the main path and the backup path have no repeated physical links as far as possible), and the surplus of each logic link is balanced as much as possible while the load surplus of each logic link is ensured, thereby realizing efficient, stable and reliable data sharing.

The second aspect of the embodiment of the present application further provides a load balancing optimization device for a data sharing network, referring to fig. 4, fig. 4 shows a schematic structural diagram of the load balancing optimization device, as shown in fig. 4, where the device includes:

In one possible implementation, the robustness design module includes:

In one possible implementation manner, the scheme determining module includes:

；

Wherein f _b represents the network load balancing degree parameter;

；

Wherein f represents the fitness of individuals in the population.

In one possible embodiment, the coding module includes:

The embodiment of the application also provides an electronic device, and referring to fig. 5, fig. 5 is a schematic diagram of the electronic device according to the embodiment of the application. As shown in fig. 5, the electronic device 100 includes: the memory 110 and the processor 120 are connected through bus communication, and a computer program is stored in the memory 110 and can run on the processor 120, so that the steps in the load balancing optimization method for the data sharing network disclosed by the embodiment of the application are realized.

The embodiment of the application also provides a computer readable storage medium, on which a computer program/instruction is stored, which when executed by a processor, implements the steps in the load balancing optimization method for the data sharing network as disclosed in the embodiment of the application.

The embodiment of the application also provides a computer program product, which when being run on the electronic equipment, causes a processor to realize the steps of the load balancing optimization method for the data sharing network, which is disclosed by the embodiment of the application.

In this specification, each embodiment is described in a progressive manner, and each embodiment is mainly described by differences from other embodiments, and identical and similar parts between the embodiments are all enough to be referred to each other.

Embodiments of the present application are described with reference to flowchart illustrations and/or block diagrams of methods, apparatus, electronic devices, and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

While preferred embodiments of the present application have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiment and all such alterations and modifications as fall within the scope of the embodiments of the application.

Finally, it is further noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or terminal that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or terminal. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or terminal device that comprises the element.

The above describes in detail a load balancing optimization method, device and product for a data sharing network, and specific examples are applied to illustrate the principles and embodiments of the present application, and the description of the above examples is only used to help understand the method and core idea of the present application; meanwhile, as those skilled in the art will have variations in the specific embodiments and application scope in accordance with the ideas of the present application, the present description should not be construed as limiting the present application in view of the above.

Claims

1. The load balancing optimization method for the data sharing network is characterized by comprising the following steps of:

2. The method for load balancing optimization of a data sharing network according to claim 1, wherein the performing the robust design of the primary path and the backup path includes:

3. The load balancing optimization method for a data sharing network according to claim 1, wherein the executing the genetic algorithm to make the initialized population perform population evolution under the constraint condition until the objective function converges, and determining an optimal load balancing optimization scheme comprises:

4. The load balancing optimization method for a data sharing network according to claim 1, wherein the constraint condition includes: a demand constraint and a load constraint;

5. The load balancing optimization method for a data sharing network according to claim 3, wherein the calculating the optimal population individual fitness of the evolved population according to the objective function comprises:

6. The load balancing optimization method for the data sharing network according to claim 5, wherein the primary and backup path overlapping degree parameters are calculated according to the following formula:

；

Wherein f _r denotes the load resource margin parameter, Load resources configured for the ith logical link; representing the load demand of the ith logical link;

；

Wherein f _b represents the network load balancing degree parameter;

；

Wherein f represents the fitness of individuals in the population.

7. The method for optimizing load balancing for a data sharing network according to claim 3, wherein the encoding a plurality of load balancing optimization schemes into population individuals based on a genetic algorithm, generating an initialized population, comprises:

8. A load balancing optimization device for a data sharing network, the device comprising:

9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor when executed implementing the steps in the data sharing network oriented load balancing optimization method of any one of claims 1-7.

10. A computer readable storage medium, characterized in that a computer program/instruction is stored which, when executed by a processor, implements the steps of the load balancing optimization method for a data sharing network according to any of the claims 1-7.