Network orchestration method, device, equipment and storage medium

Abstract

This application proposes a network orchestration method, apparatus, electronic device, and computer storage medium. The network orchestration method includes: obtaining an initial slicing particle swarm of the network to be orchestrated; iteratively processing the initial slicing particle swarm; if a locally optimal slicing particle exists in the iteratively processed initial slicing particle swarm, mapping a pre-acquired chaotic variable to the solution space of the initial slicing particle swarm to obtain a mapped slicing particle swarm; and determining the globally optimal slicing particle in the initial slicing particle swarm based on the mapped slicing particle swarm. Because the network orchestration method updates slicing particles by mapping pre-acquired chaotic variables to the solution space of the initial slicing particle swarm, it facilitates obtaining an optimal routing solution while also reducing the complexity of the algorithm.

Description

Technical Field

[0001] The present application relates to network orchestration technology, and in particular to a network orchestration method, apparatus, device, and storage medium. Background Art

[0002] Currently, the Genetic Algorithm-Particle Swarm Optimization (GA-PSO) algorithm can be used to transform network optimization into the orchestration of network slices. By leveraging the genetic principles of hybridization and mutation, multiple initial slice groups are generated based on slice type. These groups are then continuously optimized using genetics and mutation to generate new network slices, and the optimal routing solution is finally determined through iteration. However, updating slice particles through genetics and mutation is relatively complex, and the algorithmic complexity is high. When the network is densely populated and data traffic is high, it is likely difficult to output the optimal routing solution in a timely manner. Summary of the Invention

[0003] This application hopes to provide a network orchestration method, apparatus, device and storage medium.

[0004] In a first aspect, the present application provides a network orchestration method, the method comprising:

[0005] Obtaining an initial slicing particle group of the network to be orchestrated, wherein the initial slicing particle group includes N basic slicing particles obtained by hybridizing the basic network slices;

[0006] Iteratively processing the initial slice particle group;

[0007] In the case where a local optimal slicing particle exists in the initial slicing particle swarm after the iterative process, the pre-obtained chaotic variable is mapped to the solution space of the initial slicing particle swarm to obtain a mapped slicing particle swarm;

[0008] Based on the mapped slicing particle group, a globally optimal slicing particle in the initial slicing particle group is determined.

[0009] In one embodiment, after iteratively processing the initial slice particle group, the method further includes:

[0010] Determining basic slicing particles in the initial slicing particle group after the iterative processing;

[0011] Determine the average particle distance between the basic slice particles after iteration;

[0012] When the average particle distance between the basic slicing particles is less than a preset average particle distance threshold, determining the fitness variance corresponding to the basic slicing particles after iteration;

[0013] When the fitness variance corresponding to the basic slicing particle is less than a preset fitness variance threshold, the basic slicing particle is determined to be a local optimal slicing particle.

[0014] In one embodiment, after iteratively processing the initial slice particle group, the method further includes:

[0015] When there is no locally optimal slicing particle in the initial slicing particle group after the iterative process, a globally optimal slicing particle in the initial slicing particle group is determined based on the initial slicing particle group.

[0016] In one embodiment, mapping the pre-obtained chaotic variables to the solution space of the initial slice particle swarm includes:

[0017] Based on the randomly generated initial chaotic variables, the chaotic variables are generated by a chaotic mapping equation;

[0018] The chaotic variables are mapped to the solution space of the initial slice particle swarm through carrier transformation.

[0019] In one embodiment, mapping the chaotic variable to the solution space of the initial slice particle swarm through carrier transformation includes:

[0020] Obtaining an initial slicing particle to be updated; wherein the initial slicing particle to be updated is a plurality of basic slicing particles determined in the initial slicing particle group with the local optimal slicing particle as the center;

[0021] Slice particles to be updated having chaotic motion characteristics obtained by carrier transformation; the chaotic motion characteristics indicate that the position, velocity, and direction of the particles all change;

[0022] Determining the slice particle to be updated having chaotic motion characteristics as the target slice particle to be updated;

[0023] The to-be-updated initial slice particles are updated based on the target to-be-updated slice particles.

[0024] In one embodiment, the iterative processing of the initial slice particle group includes:

[0025] Obtaining initialization parameters for performing the iterative process, wherein the initialization parameters include at least one of the following: an initialization learning factor, a maximum inertia weight, a minimum inertia weight, a maximum number of iterations, a speed upper limit, a threshold, and an iteration number;

[0026] The initialized slice particle swarm is iteratively processed based on the initialization parameters.

[0027] In one embodiment, the iterative processing of the initialized slice particle swarm based on the initialization parameter includes:

[0028] Based on the initialization parameters, the positions and velocities of the basic slice particles in the initialized slice particle group are updated.

[0029] In a second aspect, the present application further proposes a network orchestration device, comprising: a first obtaining module, an iteration module, a mapping module, and a first determining module, wherein:

[0030] The first obtaining module is used to obtain an initial slicing particle group of the network to be orchestrated, wherein the initial slicing particle group includes N basic slicing particles obtained by hybridizing the basic network slices;

[0031] The iteration module is used to iteratively process the initial slice particle swarm;

[0032] The mapping module is used to map the pre-obtained chaotic variables to the solution space of the initial slicing particle swarm after iterative processing to obtain a mapped slicing particle swarm when there is a local optimal slicing particle in the initial slicing particle swarm after iterative processing;

[0033] The first determining module is configured to determine a globally optimal slicing particle in the initial slicing particle group based on the mapped slicing particle group.

[0034] In one embodiment, the device further includes: a second determining module and a third determining module, wherein:

[0035] The second determining module is used to determine the basic slicing particles in the initial slicing particle group after the iterative processing; and determine the average particle distance between the basic slicing particles after the iteration;

[0036] The third determination module is configured to determine the fitness variance corresponding to the iterated basic slicing particles when the average particle distance between the basic slicing particles is less than a preset average particle distance threshold; and determine that the basic slicing particles are local optimal slicing particles when the fitness variance corresponding to the basic slicing particles is less than a preset fitness variance threshold.

[0037] In one embodiment, the apparatus further includes: a fourth determining module, wherein:

[0038] The fourth determining module is configured to determine a global optimal slicing particle in the initial slicing particle group based on the initial slicing particle group when there is no locally optimal slicing particle in the initial slicing particle group after the iterative processing.

[0039] In one embodiment, the mapping module is configured to generate the chaotic variables through a chaos mapping equation based on randomly generated initial chaotic variables; and map the chaotic variables to the solution space of the initial slice particle swarm through carrier transformation.

[0040] In one embodiment, the mapping module is used to obtain initial slicing particles to be updated; wherein the initial slicing particles to be updated are multiple basic slicing particles determined in the initial slicing particle group with the local optimal slicing particle as the center; the slicing particles to be updated having chaotic motion characteristics are obtained by carrier transformation; the chaotic motion characteristics represent characteristics in which the position, speed, and direction of the particles all change; the slicing particles to be updated having chaotic motion characteristics are determined as target slicing particles to be updated; and the initial slicing particles to be updated are updated based on the target slicing particles to be updated.

[0041] In one embodiment, the iteration module is used to obtain initialization parameters for performing the iterative processing, and the initialization parameters include at least one of the following: an initialization learning factor, a maximum inertia weight, a minimum inertia weight, a maximum number of iterations, a speed upper limit, a threshold, and an iteration number; and the initialized slice particle swarm is iteratively processed based on the initialization parameters.

[0042] In one embodiment, the iteration module is used to update the position and velocity of the basic slicing particles in the initialized slicing particle group based on the initialization parameters.

[0043] In a third aspect, the present application further provides an electronic device, comprising a processor and a memory for storing a computer program that can be run on the processor; wherein,

[0044] When the processor is used to run the computer program, it executes any one of the above-mentioned network orchestration methods.

[0045] In a fourth aspect, the present application also provides a computer storage medium on which a computer program is stored, and when the computer program is executed by a processor, any one of the above-mentioned network orchestration methods is implemented.

[0046] In this application, the network orchestration method updates the slice particles by mapping pre-obtained chaotic variables into the solution space of the initial slicing particle swarm, rather than updating the slice particles through genetics or mutation. This facilitates obtaining the optimal routing solution while also reducing the algorithm's complexity. Furthermore, by determining whether a basic slicing particle is the optimal one based on the average particle distance between the basic slicing particles and the corresponding fitness variance of the basic slicing particles, it can more accurately determine whether the basic slicing particles are trapped in a local optimal solution.

[0047] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The drawings herein are incorporated into and constitute a part of the specification. These drawings illustrate embodiments consistent with the present application and, together with the specification, are used to illustrate the technical solutions of the present application.

[0049] Figure 1 A flowchart of a network orchestration method of the present application;

[0050] Figure 2 A flowchart of another network orchestration method of the present application;

[0051] Figure 3 A flowchart of another network orchestration method of the present application;

[0052] Figure 4 A flowchart of another network orchestration method of the present application;

[0053] Figure 5 A flowchart of another network orchestration method of the present application;

[0054] Figure 6 This is a schematic diagram of the process of implementing network slicing based on the Software Defined Network (SDN) architecture of this application;

[0055] Figure 7 This is a schematic diagram of the structure of the network orchestration device of the present application;

[0056] Figure 8 This is a schematic structural diagram of the electronic device of this application. DETAILED DESCRIPTION

[0057] The present application will be further described in detail below in conjunction with the accompanying drawings and examples. It should be understood that the embodiments provided herein are merely intended to explain the present application and are not intended to limit the present application. In addition, the embodiments provided below are partial embodiments for implementing the present application, rather than providing all embodiments for implementing the present application. In the absence of conflict, the technical solutions described in the embodiments of the present application may be implemented in any combination.

[0058] It should be noted that, in the embodiments of the present application, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a method or apparatus comprising a series of elements includes not only the elements explicitly stated, but also other elements not explicitly listed, or also includes elements inherent to the implementation of the method or apparatus. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the presence of other related elements (such as steps in the method or units in the apparatus, for example, a unit may be a portion of a circuit, a portion of a processor, a portion of a program or software, etc.) in the method or apparatus comprising the element.

[0059] The term "and / or" herein simply describes an association relationship between associated objects, indicating that three relationships can exist. For example, "A and / or B" can represent the existence of three situations: A alone, A and B simultaneously, and B alone. Furthermore, the term "at least one" herein refers to any combination of at least two of any one or more of a plurality of items. For example, "at least one of A, B, and C" can represent any one or more elements selected from the set consisting of A, B, and C.

[0060] In the related art, commonly used network orchestration algorithms include greedy strategy-based network orchestration algorithms and GA-PSO-based network orchestration algorithms. The greedy strategy-based network orchestration algorithm refers to the pre-planning stage, that is, before the actual traffic arrives, based on the existing historical traffic data and the entire network topology information, pre-planning the path through the network orchestration control platform, completing the construction of the network slice, and converting the node reservation information contained in the construction result into queue configuration information and sending it to the underlying network equipment. In this way, the entire network is divided into multiple different virtual networks, each of which is a network slice. Each network slice contains routing information and the resource load status of each node. Then, according to the actual traffic demand, the network demand is divided through the greedy algorithm, and finally the divided network is orchestrated to assign reasonable routing solutions to different users.

[0061] The GA-PSO-based particle swarm optimization network orchestration algorithm first uses the shortest path algorithm. Specifically, this network orchestration algorithm can generate three basic network slices according to the specific application scenarios of the 5G network, namely: low-latency slices, high-bandwidth slices, and high-reliability slices.

[0062] Furthermore, based on the hybridization and mutation concepts of the genetic algorithm, the three basic network slices can be hybridized pairwise. The N particles with the highest fitness are selected based on the fitness function to form an initial network slice group. Each network slice is an M-dimensional matrix representing a feasible solution, or routing plan, where M is the number of nodes included in the routing plan. Selecting the optimal routing plan for the user involves selecting the slice with the highest fitness within the group—in other words, finding the optimal solution within the group. The particle swarm optimization algorithm is used to find the optimal solution. The initial network slice group is the initial slice particle swarm in the particle swarm optimization algorithm, and each network slice in the initial network slice group is a particle in the initial slice particle swarm. After each population iteration, the subgraph represented by the current particle is hybridized with the local optimal subgraph and the global optimal subgraph, respectively. The hybridized subgraphs are then optimized. Finally, based on the iteration termination criteria, the optimal solution, or routing plan, is output.

[0063] Greedy-based network orchestration algorithms use a greedy algorithm to partition pre-defined network slices based on actual traffic demand. This type of network orchestration algorithm is suitable for optimizing network resources in data centers with relatively simple network conditions, but it is inadequate for the complex demands of high bandwidth, low latency, and high reliability in 5G application scenarios. Furthermore, this type of network orchestration algorithm can only optimize for single objectives, such as network resource utilization or Quality of Service (QoS), lacking a holistic approach.

[0064] The GA-PSO-based network orchestration algorithm transforms network optimization into the orchestration of network slices. Specifically, this network orchestration algorithm uses the genetic hybridization and mutation concepts to generate N initial swarms of slice particles based on a basic network slice. It then continuously optimizes the swarm using genetics and mutation to generate new network slice particles, ultimately iterating to find the optimal routing solution. However, updating slice particles through genetics and mutation is complex, leading to high algorithmic complexity. When the network is densely populated and data traffic is high, it can be difficult to output the optimal routing solution in a timely manner.

[0065] In order to solve the above technical problems, the present application embodiment proposes a network orchestration method. Figure 1 A flow chart of a network orchestration method according to an embodiment of the present application is shown in FIG. Figure 1 As shown, the process may include:

[0066] Step 101: obtaining an initial slicing particle group of a network to be orchestrated, wherein the initial slicing particle group includes N basic slicing particles obtained by hybridizing a basic network slice; N is an integer greater than or equal to 2.

[0067] In one embodiment, basic network slices can generate multiple types of basic network slices based on specific application scenarios of the 5G network. For example, three types of basic network slices can be generated, specifically low-latency network slices, high-bandwidth network slices, and high-reliability network slices. The initial slice particle swarm can be a collection of multiple basic particle slices. The basic network slices can be hybridized using the subgraph hybridization and subgraph mutation methods of the genetic algorithm, performing a pairwise hybridization process. Of course, other hybridization processes besides pairwise hybridization can also be performed.

[0068] A basic slicing particle can be an M-dimensional matrix, where M represents the number of network nodes in the network topology to be orchestrated. Each basic slicing particle can be characterized by speed, position, direction, and fitness. Furthermore, a fitness evaluation function can be used to evaluate the quality of each basic slicing particle.

[0069] As for the implementation method of obtaining the initial slice particle group of the network to be orchestrated, for example, it can be achieved by performing pairwise hybridization processing on low-latency network slices, high-bandwidth network slices and high-reliability network slices to obtain N hybridized basic slice particles.

[0070] Step 102: performing iterative processing on the initial slice particle group.

[0071] Here, the GA-PSO algorithm can be used to iterate the initial slicing particle swarm. The purpose of iteratively processing the initial slicing particle swarm is to find the globally optimal slicing particle in the initial slicing particle swarm. In each iteration, the basic slicing particle can update the position, velocity, and direction of the basic slicing particle through the local optimal slicing particle and the global optimal slicing particle. It is understood that in the process of finding the optimal solution, the initial particle slicing swarm uses the position to represent the current solution of the basic slicing particle, and uses the velocity to represent the direction and speed of the slicing particle approaching the optimal solution. Updating the position and velocity of the slicing particle is the process of all swarm particles continuously approaching the optimal solution.

[0072] In one example, iteration can be performed according to the following iterative formula (1) and iterative formula (2):

[0073]

[0074]

[0075] Among them, c1 and c2 are learning factors, which are generally set to 2; rand() represents taking a random number between 0 and 1, and w is the inertia weight, which is used to adjust the ability of global search and local search. Generally, the larger w is, the stronger the global search ability is, and the smaller w is, the stronger the local search ability is; p irepresents the local optimal solution of the current particle; g i represents the global optimal solution, and They represent the speed of the particle at time k and the next moment respectively; and represent the position of the particle at time k and the next moment respectively.

[0076] Step 103: When there is a locally optimal slicing particle in the initial slicing particle swarm after the iterative process, the pre-obtained chaotic variable is mapped to the solution space of the initial slicing particle swarm to obtain a mapped slicing particle swarm.

[0077] Here, the local optimal slicing particle can be expressed as an optimal solution found by the slicing particle itself during the iteration process.

[0078] In one embodiment, before the iterative processing, it can be determined whether the current slicing particle in the initial slicing particle swarm after the iterative processing is the locally optimal slicing particle; when it is determined that the current slicing particle is the locally optimal slicing particle, the chaotic variable is obtained, and the obtained chaotic variable is mapped to the solution space of the slicing particle swarm, the initial slicing particle swarm is updated, and the mapped slicing particle swarm is obtained.

[0079] Step 104: Based on the mapped slicing particle group, determine the globally optimal slicing particle in the initial slicing particle group.

[0080] Here, the global optimal slicing particle can be expressed as the optimal slicing particle that can currently be found in the entire initialized slicing particle group.

[0081] In one embodiment, based on the mapped slicing particle group, determining the global optimal slicing particle in the initial slicing particle group may be performed by iteratively processing the updated initial slicing particle group to determine the global optimal slicing particle in the initial slicing particle group.

[0082] In actual applications, steps 101 to 104 can be implemented using a processor in a business cluster. The processor can be at least one of an application-specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), an FPGA, a central processing unit (CPU), a controller, a microcontroller, and a microprocessor.

[0083] It can be seen that the embodiment of the present disclosure proposes a network orchestration method, which obtains an initial slicing particle group of the network to be orchestrated, wherein the initial slicing particle group includes N basic slicing particles obtained after hybridization processing of the basic network slice; N is an integer greater than or equal to 2; the initial slicing particle group is iteratively processed; when there is a locally optimal slicing particle in the initial slicing particle group after iterative processing, the pre-obtained chaotic variable is mapped to the solution space of the initial slicing particle group to obtain a mapped slicing particle group; based on the mapped slicing particle group, the globally optimal slicing particle in the initial slicing particle group is determined. In the embodiment of the present application, since the network orchestration method updates the slicing particles by mapping the pre-obtained chaotic variables to the solution space of the initial slicing particle group, rather than updating the slicing particles in a genetic and mutation manner, it is conducive to obtaining the optimal routing solution and also reduces the complexity of the algorithm.

[0084] This application embodiment proposes a network orchestration method. Figure 2 This is a flow chart of another network orchestration method according to an embodiment of the present application. Figure 2 As shown, the process may include:

[0085] Step 201: obtaining an initial slicing particle group of a network to be orchestrated, wherein the initial slicing particle group includes N basic slicing particles obtained by hybridizing a basic network slice; N is an integer greater than or equal to 2.

[0086] Step 202: performing iterative processing on the initial slice particle group.

[0087] Step 203: determining the basic slicing particles in the initial slicing particle group after the iterative processing.

[0088] Step 204: Determine the average particle distance between the basic slice particles after iteration.

[0089] Here, the average particle distance between particles can reflect the discrete distribution between individuals from a spatial perspective. The calculation formula for the average particle distance of basic slice particles is as follows:

[0090]

[0091] Among them, L is the maximum diagonal length of the search space, N is the size of the initial slice particle swarm, d is the dimension of the solution space, and p is the maximum diagonal length of the search space. id is the d-dimensional coordinate value of the i-th basic slice particle, p d Represents the mean of the d-th dimension coordinate values ​​of all basic slice particles.

[0092] Step 205: When the average particle distance between the basic slicing particles is less than a preset average particle distance threshold, determine the fitness variance corresponding to the basic slicing particles after iteration.

[0093] Here, the preset average particle distance threshold may be an average particle distance threshold pre-set according to demand; the average particle distance between basic slice particles being less than the preset average particle distance threshold may be achieved by the average particle distance between basic particles being less than the preset average particle distance threshold.

[0094] In one embodiment, the fitness variance reflects the particle distribution from the function value. The fitness variance is calculated as follows:

[0095]

[0096] Where N is the number of individuals in the initial slice particle group, f is the normalized scaling factor, and f i is the fitness of the i-th basic slice particle, f avg is the average fitness of the current initial slice particle swarm.

[0097] Here, the fitness of the basic slice particles can be determined by the fitness evaluation function. The fitness evaluation function can be:

[0098] Fitness(α,β,D,Β)=-αe D +βe B (5);

[0099] Where D is the maximum delay path in a single subgraph after normalization; B is the minimum bandwidth in the subgraph link after normalization; α is the ratio of low-latency slices to all slices; β is the ratio of high-bandwidth demand slices to all slices.

[0100] Here, the formula for parameter normalization can be:

[0101]

[0102] Among them, V nor represents the normalized parameters, v is the performance parameter, u is the mean of the performance parameter, and σ is the variance of the performance parameter.

[0103] In one embodiment, the fitness of the current basic slice particle can be determined using a fitness evaluation function based on a threshold value of actual demand. Here, the threshold value of actual demand can be the maximum latency, the minimum bandwidth, the proportion of low-latency slices in actual demand, and the proportion of high-bandwidth slices. For example, the threshold value of actual demand can be substituted as an input parameter into Formula (5) to calculate the fitness of the target basic slice particle.

[0104] The value of f is determined by the following formula:

[0105]

[0106] Step 206: When the fitness variance corresponding to the basic slicing particle is less than a preset fitness variance threshold, determine that the basic slicing particle is a local optimal slicing particle.

[0107] Here, the preset fitness variance threshold may be a fitness variance threshold preset according to requirements.

[0108] In one embodiment, using only fitness variance to describe the diversity of the initial slice particle swarm is not sufficient, as it is possible for the fitness variance corresponding to the basic slice particles in the initial slice particle swarm to be small, while the average particle spacing is large. Only when both the fitness variance and the average particle spacing corresponding to the basic slice particles are small is the basic slice considered to be prone to falling into a local optimal solution.

[0109] Step 207: When there is a locally optimal slicing particle in the initial slicing particle swarm after the iterative process, the pre-obtained chaotic variable is mapped to the solution space of the initial slicing particle swarm to obtain a mapped slicing particle swarm.

[0110] Step 208: Based on the mapped slicing particle group, determine the globally optimal slicing particle in the initial slicing particle group.

[0111] The network orchestration algorithm proposed in this application can determine whether a basic slicing particle is the optimal slicing particle by the average particle distance between the basic slicing particles and the fitness variance corresponding to the basic slicing particles. Therefore, it can more accurately determine whether the basic slicing particles are trapped in a local optimal solution.

[0112] Example 3

[0113] This application proposes a network orchestration method. Figure 3This is a flow chart of another network orchestration method of the present application, such as Figure 3 As shown, the process may include:

[0114] Step 301: obtaining an initial slicing particle group of a network to be orchestrated, wherein the initial slicing particle group includes N basic slicing particles obtained by hybridizing a basic network slice; N is an integer greater than or equal to 2.

[0115] Step 302: performing iterative processing on the initial slice particle group.

[0116] Step 303: when there is a local optimal slicing particle in the initial slicing particle group after the iterative process, the chaotic variable is generated by a chaotic mapping equation based on the randomly generated initial chaotic variable.

[0117] Here, the initial chaotic variable can be a d-dimensional vector, which can be expressed as Z0. Each dimension of Z0 can be a value in the range of -1 to 1, and the number of the initial chaotic variables is the same as the number of basic slice particles.

[0118] In one example, the chaos mapping equation may be:

[0119] z n+1 =sin(5.56 / z n ),-1<z n <1 (8);

[0120] Among them, Z n Represents the chaotic variable at the current moment, Z n+1 represents the chaotic variable at the next moment, n is an integer greater than or equal to 0 and less than or equal to S*R. S*R d-dimensional chaotic variables can be generated by formula (8), where S is the size of the basic slice particle group and R is the chaos radius, which ranges from 0 to 1.

[0121] Step 304: obtaining an initial slicing particle to be updated; wherein the initial slicing particle to be updated is a plurality of basic slicing particles determined in the initial slicing particle group with the local optimal slicing particle as the center.

[0122] In one example, the initial slicing particles to be updated may be all basic slicing particles in the initial slicing particle group within a chaotic radius R centered around the local optimal slicing particle. Since R ranges from 0 to 1, the initial slicing particles to be updated may be some of the basic slicing particles in the basic slicing particle group.

[0123] Step 305: Obtain slice particles to be updated with chaotic motion characteristics through carrier transformation; the chaotic motion characteristics indicate that the position, speed and direction of the particles all change.

[0124] Here, the carrier transformation formula can be:

[0125] x ij =a j +(b j -a j )z ij (9);

[0126] Among them, i represents the i-th basic slice particle in the population, j represents the dimension in the solution space, and a j and b j Indicates the value range of the current basic slice particle in the solution space, x ij represents the i-th basic slice particle after carrier transformation, Z ij Represents the chaotic variable corresponding to the i-th basic slice particle.

[0127] Step 306: Determine the slice particle to be updated having chaotic motion characteristics as the target slice particle to be updated.

[0128] Step 307: Update the to-be-updated initial slice particles based on the target to-be-updated slice particles.

[0129] In one example, updating the to-be-updated initial slicing particles based on the target to-be-updated slicing particles may be performed by replacing the to-be-updated initial slicing particles with the target to-be-updated slicing particles to obtain an updated initial slicing particle group.

[0130] The network orchestration algorithm proposed in this application can change the chaotic motion characteristics of the initial slicing particles to be updated through carrier transformation based on the obtained chaotic variables when the basic slicing particles fall into a local optimal solution, so that the position, speed and direction of the initial slicing particles to be updated are changed, thereby jumping out of the local optimal solution and facilitating obtaining the global optimal solution. At the same time, the slicing particles are updated by mapping the pre-obtained chaotic variables to the solution space of the initial slicing particle group, rather than updating the slicing particles by inheritance and mutation, which reduces the complexity of the algorithm.

[0131] This application embodiment proposes a network orchestration method. Figure 4 This is a flow chart of another network orchestration method according to an embodiment of the present application. Figure 4 As shown, the process may include:

[0132] Step 401: obtaining an initial slicing particle group of a network to be orchestrated, wherein the initial slicing particle group includes N basic slicing particles obtained by hybridizing a basic network slice.

[0133] Step 402: Obtain initialization parameters for performing the iterative process, wherein the initialization parameters include at least one of the following: an initialization learning factor, a maximum inertia weight, a minimum inertia weight, a maximum number of iterations, a speed upper limit, a threshold, and an iteration number.

[0134] Here, the maximum number of iterations can be used as the termination condition of the iteration.

[0135] Step 403: Based on the initialization parameters, the positions and velocities of the basic slicing particles in the initialized slicing particle group are updated.

[0136] Here, position represents the current solution of a particle, and velocity represents the direction and speed at which the particle approaches the optimal solution. Updating the position and velocity of a particle is the process by which all particles in the population continuously approach the optimal solution.

[0137] Step 404: When there is a locally optimal slicing particle in the initial slicing particle swarm after the iterative process, the pre-obtained chaotic variable is mapped to the solution space of the initial slicing particle swarm to obtain a mapped slicing particle swarm.

[0138] Step 405: Based on the mapped slicing particle group, determine the globally optimal slicing particle in the initial slicing particle group.

[0139] Step 406: when there is no locally optimal slicing particle in the initial slicing particle group after the iterative process, a globally optimal slicing particle in the initial slicing particle group is determined based on the initial slicing particle group.

[0140] The network orchestration method proposed in this application can improve the speed and accuracy of obtaining the global optimal solution by setting reasonable initialization parameters.

[0141] The network slicing problem for 5G networks essentially involves finding a relatively optimal routing solution for user computations, a problem that falls under the network routing allocation problem. Based on the specific application scenarios of 5G networks, three basic network slice types are generated: low-latency slices, high-bandwidth slices, and high-reliability slices. These three network slice types are used as the original basic slice particles. Using the subgraph hybridization and subgraph mutation methods of the genetic algorithm, these three types of network slices are pairwise hybridized to obtain N hybridized basic slice particles, forming the initial slice particle swarm.

[0142] The M hybridized basic slice particles are the initially divided network slices. Each network slice is an M-dimensional matrix, where M represents the number of network nodes contained in the network topology.

[0143] When selecting the optimal routing solution for a user, the fitness evaluation function for the slice particle is determined based on the actual demand thresholds. The actual demand thresholds can be maximum latency, minimum bandwidth, the proportion of low-latency slices in the actual demand, and the proportion of high-bandwidth slices. These thresholds are then incorporated into the fitness evaluation function to calculate the current basic slice particle. The evaluation function's input parameters are these thresholds. By incorporating these thresholds into the fitness evaluation function, the fitness of the current particle can be calculated. The fitness evaluation function is as follows:

[0144] Fitness(α,β,D,Β)=-αe D +βe B (5);

[0145] Where D is the maximum delay path in a single subgraph after normalization; B is the minimum bandwidth in the subgraph link after normalization; α is the ratio of low-latency slices to all slices; β is the ratio of high-bandwidth demand slices to all slices.

[0146] The parameter normalization formula is:

[0147]

[0148] Among them, V nor represents the normalized parameters, v is the performance parameter, u is the mean of the performance parameter, and σ is the variance of the performance parameter.

[0149] The PSO algorithm is used to update the initial slice group. Before each update of the initial slice group, the N-dimensional subgraph represented by the current slice is hybridized with the optimal slice subgraph found locally and the optimal slice subgraph found globally. Then, according to the idea of ​​mutation in the genetic algorithm, unsuitable slices in the population are eliminated. According to the slice fitness function, the local optimal slice and the global optimal slice are determined. Finally, according to the iteration conditions, the optimal routing solution is output.

[0150] During the iterative search for the optimal solution in the initial slice group, slice particles are prone to premature maturation, becoming trapped in a local optimal solution and losing the ability to find the global optimal solution. Based on this, the slice group is abstracted as a population in the PSO algorithm, with each slice abstracted as a basic particle in the population. In the GA-PSO algorithm, through inheritance and mutation, particles are led out of the local optimal solution. This application uses the GA-PSO algorithm based on chaos theory, using chaotic variables to update the particle swarm, replacing the hybridization and mutation update method in the GA-PSO algorithm.

[0151] In one embodiment, before each iteration of the elementary particle slicing group, a determination is made as to whether the current elementary slice particle is experiencing premature maturation. This determination is based on the average interparticle spacing and the fitness variance of each elementary slice particle. The average interparticle spacing reflects the discrete distribution of individual elementary slice particles from a spatial perspective, while the average fitness variance reflects the particle distribution from a function value perspective.

[0152] The calculation formula of the average particle distance of basic slice particles is as follows:

[0153]

[0154] Among them, L is the maximum diagonal length of the search space, N is the size of the initial slice particle swarm, d is the dimension of the solution space, and p is the maximum diagonal length of the search space. id is the d-dimensional coordinate value of the i-th basic slice particle, p d Represents the mean of the d-th dimension coordinate values ​​of all basic slice particles.

[0155] The fitness variance calculation formula of a single slice particle is as follows:

[0156]

[0157] Where N is the number of individuals in the initial slice particle group, f is the normalized scaling factor, and f i is the fitness of the i-th basic slice particle, f avg is the average fitness of the current initial slice particle swarm.

[0158] The value of f is determined by the following formula:

[0159]

[0160] In related technologies, using only fitness variance to describe population diversity is incomplete, as it is possible for the fitness variance of basic slice particles to be small, while the average particle distance of basic slice particles is large. Only when both the fitness variance and the average particle distance are very small is it considered that the algorithm is prone to falling into a local optimum and experiencing premature convergence.

[0161] In one embodiment, when the fitness variance is less than a preset fitness variance threshold C, and when the average particle distance of the basic slicing particles is less than a preset average particle distance threshold D, it is considered that the basic slicing particles are trapped in a local optimal solution; wherein C and D are thresholds preset according to actual needs.

[0162] In one example, if a basic slice particle exhibits premature maturation, a set of chaotic variables is generated using the chaos mapping characteristic equation, where the number of chaotic variables is the same as the number of slices. For example, a d-dimensional vector can be randomly generated as the initial value Z0, where each dimension of Z0 can be a value between -1 and 1. Then, using the chaos mapping equation, N*R d-dimensional chaotic variables are generated, where N is the population size and R is the chaos search radius, ranging from 0 to 1.

[0163] The chaotic variables are transformed into the solution space where the slice particles are located, so that all the slice particles obtain chaotic motion characteristics. For example, carrier transformation can be used to map the chaotic variables into the solution space range, randomly replacing the original N*R basic slice particles.

[0164] The carrier transformation formula can be:

[0165] x ij =a j +(b j -a j )z ij (9);

[0166] Among them, i represents the i-th basic slice particle in the population, j represents the dimension in the solution space, and a j and b j Indicates the value range of the current basic slice particle in the solution space, x ij represents the i-th basic slice particle after carrier transformation, Z ij Represents the chaotic variable corresponding to the i-th basic slice particle.

[0167] Then the mapped slicing particle swarm is directly used to iterate and update the local optimal slicing scheme and the global optimal slicing scheme.

[0168] In one embodiment, the present application also proposes a network orchestration method. Figure 5 A flowchart of another network orchestration method according to an embodiment of the present application is shown in FIG. Figure 5 As shown, the process may include:

[0169] Step 501: Use the shortest path algorithm to generate three basic types of network slices, namely low-latency slices, high-bandwidth slices, and high-reliability slices.

[0170] Step 502: basic slicing particles are hybridized in pairs to obtain an initial slicing particle group.

[0171] Step 503: Evaluate the fitness of each basic slice particle in the initial slice particle group through the fitness evaluation function.

[0172] Step 504: Determine the fitness value of the slice particle with the highest fitness as the initial local optimal solution and the global optimal solution.

[0173] Here, the fitness value of the slice particle with the highest fitness is determined as the local optimal solution, which is also the global optimal solution, and the current local optimal solution is recorded as p best , and denote the current global optimal solution as g best .

[0174] Step 505: Set the initialization parameters of the PSO algorithm.

[0175] Here, setting the initialization parameters may include setting the initialization learning factors c1 and c2, the maximum inertia weight W max , minimum inertia weight W min , maximum number of iterations T max , Speed ​​limit V max and the number of chaotic search iterations T.

[0176] Step 506: Determine whether the current number of iterations is less than the maximum number of iterations.

[0177] Here, if the current number of iterations is less than the maximum number of iterations, go to step 507 ; if not, go to step 513 .

[0178] Step 507: performing iterative processing on the initial slice particle swarm.

[0179] Here, iteration can be performed according to the following iterative formula (1) and iterative formula (2):

[0180]

[0181]

[0182] Among them, c1 and c2 are learning factors, which are generally set to 2; rand() represents taking a random number between 0 and 1, and w is the inertia weight, which is used to adjust the ability of global search and local search. Generally, the larger w is, the stronger the global search ability is, and the smaller w is, the stronger the local search ability is; p i represents the local optimal solution of the current particle; g i represents the global optimal solution, and They represent the speed of the particle at time k and the next moment respectively; and represent the position of the particle at time k and the next moment respectively.

[0183] Step 508: Update the position and velocity of the initial slice particle group.

[0184] Here, when the initial slicing particle swarm is searching for the optimal solution, the position of the slicing particle is used to represent the current solution, and the velocity is used to represent the direction and speed at which the particle is approaching the optimal solution. Updating the position and velocity of the slicing particle is the process by which all swarm particles continuously approach the optimal solution.

[0185] Step 509: Calculate the fitness value of each slice particle and update the local optimal solution and the global optimal solution.

[0186] Step 510: Calculate the average particle distance and fitness variance of each slice particle.

[0187] Step 511: Determine whether the sliced ​​particles have "premature" phenomenon. If so, go to step 512; if not, go to step 508.

[0188] Step 512 : Generate chaotic variables and map the chaotic variables to the solution space, replace the original basic slice particles, and go to step 506 .

[0189] In one example, the chaos mapping equation may be:

[0190] z n+1 =sin(5.56 / z n ),-1<z n <1 (8);

[0191] Among them, Z n Represents the chaotic variable at the current moment, Z n+1 represents the chaotic variable at the next moment, n is an integer greater than or equal to 0 and less than or equal to S*R. S*R d-dimensional chaotic variables can be generated by formula (8), where S is the size of the basic slice particle swarm and R is the chaotic search radius, which ranges from 0 to 1.

[0192] Then, through carrier transformation, the chaotic variables are mapped to the solution space, randomly replacing the original N*R basic slice particles. The carrier transformation formula can be:

[0193] x ij =a i +(b j -a j )z ij (9)

[0194] Among them, i represents the i-th basic slice particle in the population, j represents the dimension in the solution space, and a j and b j Indicates the value range of the current basic slice particle in the solution space, x ij represents the i-th basic slice particle after carrier transformation, Z ij Represents the chaotic variable corresponding to the i-th basic slice particle.

[0195] Step 513: The iterative process ends, and the global optimal solution is returned as the optimal routing solution.

[0196] In the embodiment of the present application, since the network orchestration method determines whether the basic slicing particles are the optimal slicing particles based on the average particle distance between the basic slicing particles and the fitness variance corresponding to the basic slicing particles, rather than determining the optimal slicing particles based solely on the average particle distance between the basic slicing particles or the fitness corresponding to the basic slicing particles, it can more accurately determine whether the basic slicing particles are trapped in a local optimal solution; and the network orchestration algorithm updates the slicing particles by mapping the pre-obtained chaotic variables to the solution space of the initial slicing particle group, rather than updating the slicing particles in a genetic and mutation manner, which is conducive to obtaining the optimal routing solution and also reduces the complexity of the algorithm.

[0197] The prerequisites for network slicing are network function virtualization (VNF) and SDN.

[0198] Figure 6 This is a schematic diagram of the process of implementing network slicing based on the SDN architecture in an embodiment of the present application. Figure 6 As shown in the figure, the main hardware devices included are: network slicing access dedicated equipment (Dedicated Equipment), access network dedicated equipment RAN (Radio Access Network): wireless access network, core network dedicated equipment NF (Network Function): network function.

[0199] Using NFV technology, access network functions are virtualized into an edge cloud, and core network functions are virtualized into a core cloud. The Edge Cloud and Core Cloud are connected and managed using an SDN architecture. Network slicing is the product of virtualizing physical devices, and all slices must be generated based on real physical devices.

[0200] When a terminal accesses, the network orchestration algorithm proposed in this application can be used to obtain the best routing solution for accessing Edge Cloud and Core Cloud, and finally create a slice for the terminal service based on the routing solution. Here, in each layer of network equipment, such as access network equipment, there are multiple physical devices. On these physical devices, certain software needs to be installed. The function of these software is to virtualize the functions of the physical machine and generate network slices; different networks are connected through the intranet or the extranet, such as the RAN slice 1 (RAN Slice1) and the CN slice 1 (CN Slice1) in the access network, which are connected through the intranet, so that a whole connected link can be formed.

[0201] Figure 7 This is a schematic diagram of the composition structure of the network orchestration device according to an embodiment of the present application. Figure 7 As shown, the apparatus may include: a first obtaining module 701, an iterating module 702, a mapping module 703 and a first determining module 704, wherein:

[0202] The first obtaining module 701 is used to obtain an initial slicing particle group of the network to be orchestrated, wherein the initial slicing particle group includes N basic slicing particles obtained by hybridizing the basic network slices;

[0203] The iteration module 702 is used to iteratively process the initial slice particle swarm;

[0204] The mapping module 703 is used to map the pre-obtained chaotic variables to the solution space of the initial slicing particle swarm after the iterative processing to obtain a mapped slicing particle swarm when there is a local optimal slicing particle in the initial slicing particle swarm after the iterative processing;

[0205] The first determining module 704 is configured to determine a globally optimal slicing particle in the initial slicing particle group based on the mapped slicing particle group.

[0206] In one embodiment, the apparatus further includes: a second determining module 705 and a third determining module 706, wherein:

[0207] The second determining module 705 is used to determine the basic slicing particles in the initial slicing particle group after the iterative processing; and determine the average particle distance between the basic slicing particles after the iteration;

[0208] The third determination module 706 is configured to determine the fitness variance corresponding to the iterated basic slicing particles when the average particle distance between the basic slicing particles is less than a preset average particle distance threshold; and determine that the basic slicing particles are local optimal slicing particles when the fitness variance corresponding to the basic slicing particles is less than a preset fitness variance threshold.

[0209] In one embodiment, the apparatus further includes: a fourth determining module 707, wherein:

[0210] The fourth determining module 707 is configured to determine a global optimal slicing particle in the initial slicing particle group based on the initial slicing particle group when there is no locally optimal slicing particle in the initial slicing particle group after the iterative process.

[0211] In one embodiment, the mapping module 703 is configured to generate the chaotic variables through a chaos mapping equation based on the randomly generated initial chaotic variables; and map the chaotic variables to the solution space of the initial slice particle swarm through carrier transformation.

[0212] In one embodiment, the mapping module 703 is used to obtain an initial slicing particle to be updated; wherein the initial slicing particle to be updated is a plurality of basic slicing particles determined in the initial slicing particle group with the local optimal slicing particle as the center; the slicing particle to be updated having chaotic motion characteristics is obtained by carrier transformation; the chaotic motion characteristics represent characteristics in which the position, speed, and direction of the particle all change; the slicing particle to be updated having chaotic motion characteristics is determined as a target slicing particle to be updated; and the initial slicing particle to be updated is updated based on the target slicing particle to be updated.

[0213] In one embodiment, the iteration module 702 is used to obtain initialization parameters for performing the iterative processing, and the initialization parameters include at least one of the following: an initialization learning factor, a maximum inertia weight, a minimum inertia weight, a maximum number of iterations, a speed upper limit, a threshold, and an iteration number; and the initialized slice particle swarm is iteratively processed based on the initialization parameters.

[0214] In one embodiment, the iteration module 702 is used to update the position and velocity of the basic slicing particles in the initialization slicing particle group based on the initialization parameters.

[0215] In actual applications, the first acquisition module 701, the iteration module 702, the mapping module 703, the first determination module 704, the second determination module 705, the third determination module 706 and the fourth determination module 707 can be implemented using a processor in an electronic device. The above-mentioned processor can be at least one of an ASIC, a DSP, a DSPD, a PLD, an FPGA, a CPU, a controller, a microcontroller, and a microprocessor.

[0216] In addition, the functional modules in this embodiment may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated units may be implemented in the form of hardware or software functional modules.

[0217] If the integrated unit is implemented as a software functional module and is not sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this embodiment, or the portion that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) or a processor to execute all or part of the steps of the method described in this embodiment. The aforementioned storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0218] Specifically, the computer program instructions corresponding to a network orchestration method in this embodiment can be stored on a storage medium such as a CD, a hard disk, or a USB flash drive. When the computer program instructions corresponding to a network orchestration method in the storage medium are read or executed by an electronic device, any one of the network orchestration methods in the aforementioned embodiments is implemented.

[0219] Based on the same technical concept as the above embodiment, see Figure 8 , which shows an electronic device provided by an embodiment of the present application, which may include: a memory 801 and a processor 802; wherein,

[0220] The memory 801 is used to store computer programs and data;

[0221] The processor 802 is configured to execute the computer program stored in the memory to implement any one of the network orchestration methods of the foregoing embodiments.

[0222] In practical applications, the memory 801 may be a volatile memory, such as RAM; or a non-volatile memory, such as ROM, flash memory, hard disk drive (HDD) or solid-state drive (SSD); or a combination of the above types of memory, and provide instructions and data to the processor 802.

[0223] The processor 802 may be at least one of an ASIC, a DSP, a DSPD, a PLD, an FPGA, a CPU, a controller, a microcontroller, and a microprocessor. It is understood that for different augmented reality cloud platforms, the electronic device used to implement the above processor functions may also be other, and this embodiment of the application does not specifically limit this.

[0224] In some embodiments, the functions or modules provided by the apparatus of the present application can be used to execute the method described in the above method embodiment. The specific implementation thereof can refer to the description of the above method embodiment. For the sake of brevity, it will not be repeated here.

[0225] The above description of the various embodiments tends to emphasize the differences between the various embodiments. The same or similar aspects can be referenced to each other. For the sake of brevity, this article will not repeat them.

[0226] The methods disclosed in the various method embodiments provided in this application can be arbitrarily combined without conflict to obtain new method embodiments.

[0227] The features disclosed in the various product embodiments provided in this application can be arbitrarily combined without conflict to obtain new product embodiments.

[0228] The features disclosed in the various method or device embodiments provided in this application can be arbitrarily combined without conflict to obtain new method embodiments or device embodiments.

[0229] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0230] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of this application, ordinary technicians in this field can also make many forms without departing from the purpose of this application and the scope of protection of the claims, all of which are protected by this application.

Claims

1. A network orchestration method, characterized in that: The method comprises: Obtaining an initial slicing particle group of the network to be orchestrated, wherein the initial slicing particle group includes N basic slicing particles obtained by hybridizing the basic network slices; N is an integer greater than or equal to 2; Iteratively processing the initial slice particle group; In the case where a local optimal slicing particle exists in the initial slicing particle swarm after the iterative process, the pre-obtained chaotic variable is mapped to the solution space of the initial slicing particle swarm to obtain a mapped slicing particle swarm; Determining a globally optimal slicing particle in the initial slicing particle swarm based on the mapped slicing particle swarm; The step of mapping the pre-obtained chaotic variables to the solution space of the initial slice particle swarm includes: Based on randomly generated initial chaotic variables, the chaotic variables are generated by a chaotic mapping equation; the number of the initial chaotic variables is the same as the number of basic slice particles; Obtaining an initial slicing particle to be updated; wherein the initial slicing particle to be updated is a plurality of basic slicing particles determined in the initial slicing particle group with the local optimal slicing particle as the center; Obtaining slice particles to be updated with chaotic motion characteristics through carrier transformation; the chaotic motion characteristics indicate that the position, speed and direction of the particles all change; Determining the slice particle to be updated having chaotic motion characteristics as the target slice particle to be updated; The to-be-updated initial slice particles are updated based on the target to-be-updated slice particles.

2. The method according to claim 1, characterized in that After iteratively processing the initial slice particle group, the method further includes: Determining basic slicing particles in the initial slicing particle group after the iterative processing; Determine the average particle distance between the basic slice particles after iteration; When the average particle distance between the basic slicing particles is less than a preset average particle distance threshold, determining the fitness variance corresponding to the basic slicing particles after iteration; When the fitness variance corresponding to the basic slicing particle is less than a preset fitness variance threshold, the basic slicing particle is determined to be a local optimal slicing particle.

3. The method according to claim 1, characterized in that After iteratively processing the initial slice particle group, the method further includes: When there is no locally optimal slicing particle in the initial slicing particle group after the iterative process, a globally optimal slicing particle in the initial slicing particle group is determined based on the initial slicing particle group.

4. The method according to claim 1, wherein The iterative processing of the initial slice particle swarm includes: Obtaining initialization parameters for performing the iterative process, wherein the initialization parameters include at least one of the following: an initialization learning factor, a maximum inertia weight, a minimum inertia weight, a maximum number of iterations, a speed upper limit, a threshold, and an iteration number; The initialized slice particle swarm is iteratively processed based on the initialization parameters.

5. The method according to claim 4, characterized in that The iterative processing of the initialized slice particle swarm based on the initialization parameter includes: Based on the initialization parameters, the positions and velocities of the basic slice particles in the initialized slice particle group are updated.

6. A network arrangement device, characterized in that: The device includes: a first obtaining module, an iterative module, a mapping module and a first determining module, wherein: The first obtaining module is used to obtain an initial slicing particle group of the network to be orchestrated, wherein the initial slicing particle group includes N basic slicing particles obtained by hybridizing the basic network slices; N is an integer greater than or equal to 2; The iteration module is used to iteratively process the initial slice particle swarm; The mapping module is used to map the pre-obtained chaotic variables to the solution space of the initial slicing particle swarm after iterative processing to obtain a mapped slicing particle swarm when there is a local optimal slicing particle in the initial slicing particle swarm after iterative processing; The first determining module is configured to determine a globally optimal slicing particle in the initial slicing particle group based on the mapped slicing particle group; The mapping module is further configured to generate the chaotic variables through a chaotic mapping equation based on randomly generated initial chaotic variables; the number of the initial chaotic variables is the same as the number of basic slicing particles; obtain the initial slicing particles to be updated; wherein the initial slicing particles to be updated are multiple basic slicing particles determined with the local optimal slicing particle as the center in the initial slicing particle group; obtain the slicing particles to be updated with chaotic motion characteristics through carrier transformation; the chaotic motion characteristics represent characteristics in which the position, speed, and direction of the particles all change; determine the slicing particles to be updated with chaotic motion characteristics as target slicing particles to be updated; and update the initial slicing particles to be updated based on the target slicing particles to be updated.

7. An electronic device, characterized in that: comprising a processor and a memory for storing a computer program capable of running on the processor; wherein, When the processor is used to run the computer program, it executes the network orchestration method according to any one of claims 1 to 5.

8. A computer storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the network orchestration method according to any one of claims 1 to 5 is implemented.

Images