A load balancing method, device, equipment and readable storage medium

Abstract

The present invention discloses a load balancing method, device, equipment and readable storage medium, which relate to the field of communication technology, to provide a load balancing solution that is simple to implement and saves computing power. The method comprises: receiving business data of a target type; determining the business arrival rate of a target network slice according to the business data of the target type, wherein the target network slice is used to process the business data of the target type, and the number of the network slices is greater than or equal to 2; and distributing the business data of the target type among the target network slices according to the business arrival rate. The embodiments of the present invention are simple to implement and require fewer computing resources.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a load balancing method, device, equipment and readable storage medium. Background Art

[0002] The industry and academia have done a lot of work to address the problem of unbalanced service loads. The industry mainly deploys and installs load balancing devices between servers and external networks to achieve optimal load balancing requirements, while the academia focuses on diverse load balancing strategies to achieve high-quality service transmission effects.

[0003] For existing load balancing technology, the installation of hardware load balancing equipment can improve the overall performance of the system, but the configuration is complex and the cost is high; and the diversified load balancing strategies use the real-time load information of the server to determine the distribution of business, which requires huge computing power.

[0004] It can be seen from the above that the solutions in the prior art have the problem of complex implementation process and high consumption of computing power. Summary of the invention

[0005] Embodiments of the present invention provide a load balancing method, apparatus, device and readable storage medium to provide a load balancing solution that is simple to implement and saves computing power.

[0006] In a first aspect, an embodiment of the present invention provides a load balancing method, which is applied to an SDN controller (Software Defined Network, software defined network), including:

[0007] Receive business data of the target type;

[0008] Determine, according to the service data of the target type, a service arrival rate of a target network slice, wherein the target network slice is used to process the service data of the target type, and the number of the network slices is greater than or equal to 2;

[0009] According to the service arrival rate, the service data of the target type is distributed among the target network slices.

[0010] The determining, according to the service data of the target type, the service arrival rate of the target network slice includes:

[0011] Determine a first utilization transfer probability matrix of the target network slice using a DTMC (discrete-time Markov chain);

[0012] Obtaining a second utilization rate transfer probability matrix according to the first utilization rate transfer probability matrix;

[0013] Obtaining a stable state probability matrix of the target network slice according to the second utilization transition probability matrix;

[0014] According to the steady state probability matrix and the service arrival rate in the SDN-based slice network service queuing processing model, the service arrival rate matrix of the target network slice is obtained.

[0015] The first utilization rate transfer probability matrix is ​​expressed as follows:

[0016]

[0017] Among them, the elements in the matrix

[0018] ρ k represents the utilization rate of the kth network slice, K represents the total number of target network slices, i and j represent the row number and column number respectively, i=1, 2...K, j=1, 2,...K.

[0019] The step of obtaining a second utilization rate transfer probability matrix according to the first utilization rate transfer probability matrix includes:

[0020] The second utilization rate transfer probability matrix is ​​obtained according to the first utilization rate transfer probability matrix in the following manner:

[0021]

[0022] Among them, * represents the second utilization transfer probability matrix, P represents the first utilization transfer probability matrix, α is a constant and 0<α<1, and K represents the total number of target network slices.

[0023] Wherein, obtaining the service arrival rate matrix of the target network slice according to the steady state probability matrix and the service arrival rate in the SDN-based slice network service queuing processing model includes:

[0024] The service arrival rate matrix of the target network slice is obtained according to the steady state probability matrix and the service arrival rate in the SDN-based slice network service queuing processing model in the following manner:

[0025] λ * =λ total π

[0026] Among them, λ * represents the service arrival rate matrix of the target network slice, represents the sum of the service arrival rates of the target network slice, λ kis the service arrival rate of the kth target network slice; π represents the steady-state probability matrix, π = [π1, π2, ..., π K ],π k represents the stable state probability of the kth target network slice.

[0027] Wherein, before determining the service arrival rate of the target network slice according to the service data of the target type, the method further includes:

[0028] Calculating a standard deviation of the utilization of the target network slice;

[0029] The determining, according to the service data of the target type, a service arrival rate of the target network slice includes:

[0030] When the standard deviation is greater than a preset threshold, the service arrival rate of the target network slice is determined based on the service data of the target type.

[0031] The standard deviation of the utilization of the target network slice is calculated as follows:

[0032]

[0033] Where σ represents the standard deviation of the utilization of the target network slice, K represents the total number of target network slices, and ρ k represents the system utilization of the kth network slice, Represents the average utilization of K network slices.

[0034] In a second aspect, an embodiment of the present invention further provides a load balancing device, applied to an SDN controller, comprising:

[0035] A first receiving module, used for receiving business data of a target type;

[0036] A first determination module is used to determine a service arrival rate of a target network slice according to the service data of the target type, wherein the target network slice is used to process the service data of the target type, and the number of the network slices is greater than or equal to 2;

[0037] The first processing module is used to distribute the service data of the target type among the target network slices according to the service arrival rate.

[0038] Wherein, the first determining module includes:

[0039] A first determination submodule, configured to determine a first utilization transfer probability matrix of the target network slice using DTMC;

[0040] A first acquisition submodule, configured to obtain a second utilization rate transfer probability matrix according to the first utilization rate transfer probability matrix;

[0041] A second acquisition submodule, configured to obtain a stable state probability matrix of the target network slice according to the second utilization transfer probability matrix;

[0042] The third acquisition submodule is used to obtain the service arrival rate matrix of the target network slice according to the steady state probability matrix and the service arrival rate in the SDN-based slice network service queuing processing model.

[0043] The first utilization rate transfer probability matrix is ​​expressed as follows:

[0044]

[0045] Among them, the elements in the matrix

[0046] ρ k represents the utilization rate of the kth network slice, K represents the total number of target network slices, i and j represent the row number and column number respectively, i=1, 2...K, j=1, 2,...K.

[0047] The first acquisition submodule is used to obtain a second utilization rate transfer probability matrix according to the first utilization rate transfer probability matrix in the following manner:

[0048]

[0049] Among them, * represents the second utilization transfer probability matrix, P represents the first utilization transfer probability matrix, α is a constant and 0<α<1, and K represents the total number of target network slices.

[0050] The third acquisition submodule is used to obtain the service arrival rate matrix of the target network slice according to the steady state probability matrix and the service arrival rate in the SDN-based slice network service queuing processing model in the following manner:

[0051] λ * =λ total π

[0052] Among them, λ * represents the service arrival rate matrix of the target network slice, represents the sum of the service arrival rates of the target network slice, λ k is the service arrival rate of the kth target network slice; π represents the steady-state probability matrix, π = [π1, π2, ..., π K ],π krepresents the stable state probability of the kth target network slice.

[0053] Wherein, the device further comprises:

[0054] A first acquisition module is used to obtain a standard deviation of the utilization of the target network slice;

[0055] The first determination module is used to determine the service arrival rate of the target network slice based on the service data of the target type when the standard deviation is greater than a preset threshold.

[0056] The first acquisition module is used to calculate the standard deviation of the utilization of the target network slice in the following manner:

[0057]

[0058] Where σ represents the standard deviation of the utilization of the target network slice, K represents the total number of target network slices, and ρ k represents the system utilization of the kth network slice, Represents the average utilization of K network slices.

[0059] In a third aspect, an embodiment of the present invention further provides a load balancing device, applied to an SDN controller, comprising: a processor and a transceiver;

[0060] Wherein, the transceiver is used to receive service data of the target type;

[0061] The processor is used to determine the service arrival rate of the target network slice based on the service data of the target type, wherein the target network slice is used to process the service data of the target type, and the number of network slices is greater than or equal to 2; according to the service arrival rate, the service data of the target type is distributed among the target network slices.

[0062] Wherein, the processor is used for:

[0063] Determine a first utilization transfer probability matrix of the target network slice using DTMC;

[0064] Obtaining a second utilization rate transfer probability matrix according to the first utilization rate transfer probability matrix;

[0065] Obtaining a stable state probability matrix of the target network slice according to the second utilization transition probability matrix;

[0066] According to the steady state probability matrix and the service arrival rate in the SDN-based slice network service queuing processing model, the service arrival rate matrix of the target network slice is obtained.

[0067] The first utilization rate transfer probability matrix is ​​expressed as follows:

[0068]

[0069] Among them, the elements in the matrix

[0070] ρ k represents the utilization rate of the kth network slice, K represents the total number of target network slices, i and j represent the row number and column number respectively, i=1, 2...K, j=1, 2,...K.

[0071] The processor is used to obtain a second utilization rate transfer probability matrix according to the first utilization rate transfer probability matrix in the following manner:

[0072]

[0073] Among them, * represents the second utilization transfer probability matrix, P represents the first utilization transfer probability matrix, α is a constant and 0<α<1, and K represents the total number of target network slices.

[0074] The processor is used to obtain a service arrival rate matrix of the target network slice according to the steady state probability matrix and the service arrival rate in the SDN-based slice network service queuing processing model in the following manner:

[0075] λ * =λ total π

[0076] Among them, λ * represents the service arrival rate matrix of the target network slice, represents the sum of the service arrival rates of the target network slice, λ k is the service arrival rate of the kth target network slice; π represents the steady-state probability matrix, π = [π1, π2, ..., π K ],π k represents the stable state probability of the kth target network slice.

[0077] Wherein, the processor is used for:

[0078] Obtaining a standard deviation of the utilization of the target network slice;

[0079] When the standard deviation is greater than a preset threshold, the service arrival rate of the target network slice is determined based on the service data of the target type.

[0080] The processor is used to calculate the standard deviation of the utilization of the target network slice in the following manner:

[0081]

[0082] Where σ represents the standard deviation of the utilization of the target network slice, K represents the total number of target network slices, and ρ k represents the system utilization of the kth network slice, Represents the average utilization of K network slices.

[0083] In a fourth aspect, an embodiment of the present invention further provides a communication device, comprising: a transceiver, a memory, a processor, and a program stored in the memory and executable on the processor, wherein the processor implements the steps in the load balancing method as described above when executing the program.

[0084] In a fifth aspect, an embodiment of the present invention further provides a readable storage medium, on which a program is stored, and when the program is executed by a processor, the steps in the load balancing method as described above are implemented.

[0085] In the embodiment of the present invention, the service arrival rate of each network slice to carry services is recalculated by the SDN controller, thereby completing the service load balancing of the network slice. Therefore, the solution of the embodiment of the present invention is simple to implement and requires less computing resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] Figure 1 is a schematic diagram of a load balancing system provided by an embodiment of the present invention;

[0087] Figure 2 This is one of the flow charts of the load balancing method provided by an embodiment of the present invention;

[0088] Figure 3 It is the traditional slicing network service processing queue model;

[0089] Figure 4 It is a slicing network service queuing processing model based on SDN provided by an embodiment of the present invention;

[0090] Figure 5(a)-Figure 5(d) It is a schematic diagram of a URLLC slice network service load balancing solution provided by an embodiment of the present invention;

[0091] Figure 6 It is one of the structural diagrams of the load balancing device provided by the embodiment of the present invention;

[0092] Figure 7 This is the second structural diagram of the load balancing device provided in the embodiment of the present invention. DETAILED DESCRIPTION

[0093] In the embodiments of the present invention, the term "and / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent three situations: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the associated objects before and after are in an "or" relationship.

[0094] In the embodiments of the present application, the term "plurality" refers to two or more than two, and other quantifiers are similar.

[0095] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0096] See also Figure 1 , Figure 1 Schematic diagram of a load balancing system provided by an embodiment of the present invention. The system may include: an SDN controller 101, a MEC (Mobile Edge Computing) device 102 and a cloud 103.

[0097] In actual applications, business data from different industries (such as telemedicine, autonomous driving, smart meter reading, smart home, cloud gaming, etc.) is transmitted to MEC through the wireless side. MEC classifies business data with different service requirements and maps different types of business data to different types of network slices (or slices). Different business data needs to be carried by different types of network slices, and then the business data is uploaded to the cloud for processing according to business needs. Among them, the business data in MEC will be adapted to the network slice through the SDN controller, so that the business data can be flexibly scheduled through the SDN controller and the appropriate network slice can be selected to realize dynamic diversion of the business.

[0098] Similarly, when the server in the cloud transmits business data to the end user, the business data with different service requirements is adapted to the corresponding network slice through the SDN controller, and then the subsequent business data is processed and distributed.

[0099] Take URLLC (Ultra-Reliable and Low Latency Communications) slices as an example. Assume that there are 4 URLLC slices in the entire network system, and autonomous driving services need to be carried by a certain URLLC slice. At this time, URLLC slices 1 to 3 have more access to services than URLLC slice 4. Therefore, the SDN controller adapts the autonomous driving services to URLLC slice 4, thus avoiding the problem of unbalanced service carrying when multiple network slices carry services.

[0100] In order to improve the resource utilization of network slices and avoid the imbalance of service load between network slices, based on the following Figure 1 As shown in the schematic diagram of the network slicing system structure, in an embodiment of the present invention, a slicing network service load balancing method based on SDN is proposed to ensure effective load distribution between slices.

[0101] See also Figure 2 , Figure 2 is a flow chart of a load balancing method provided by an embodiment of the present invention, which is applied to an SDN controller, such as Figure 2 As shown, the following steps are included:

[0102] Step 201: Receive business data of a target type.

[0103] The target type can be any type, such as the aforementioned telemedicine, autonomous driving, smart meter reading, smart home, cloud gaming and other business data. The SDN controller can receive business data from the MEC or from the cloud. In both cases, the SDN controller can perform load balancing as needed, and the implementation principle is the same.

[0104] Step 202: Determine the service arrival rate of the target network slice based on the service data of the target type, wherein the target network slice is used to process the service data of the target type, and the number of network slices is greater than or equal to 2.

[0105] This step may include:

[0106] (1) Determine a first utilization transfer probability matrix of the target network slice using DTMC.

[0107] In the embodiment of the present invention, any network slice can be modeled as an M / M / n queuing system with n servers when processing services. Assume that the number of network slices of the same type is K, λ k , μ k 、n kare the service arrival rate, service departure rate, and number of servers of the services carried by the kth slice, respectively. In the embodiment of the present invention, it is assumed that the service processing capability of each network slice is the same, that is, the service departure rate is the same, so there exists μ k =μ. Figure 3 The figure shows the traditional slice network service processing queue model. In this slice network service processing queue model, the utilization rate of the kth slice is ρ k It can be expressed as the service arrival rate of the kth slice divided by the product of the number of servers carrying services on the kth slice and the service departure rate, which can be expressed as formula (1):

[0108]

[0109] In the embodiment of the present invention, since each service flow is controlled by an SDN controller to achieve dynamic diversion of the service flow, in the slice network service queuing processing model, the service arriving at each network slice for service transmission processing can be regarded as the service arriving at an SDN queue, and the SDN plays the role of service flow diversion.

[0110] Figure 4 The figure shows a slicing network service queuing processing model based on SDN in an embodiment of the present invention. Assume that the service arrival rate, service departure rate, and number of servers in the slicing network service queuing processing model based on SDN controller are respectively μ total =μ k , That is, the service arrival rate λ in the SDN-based slice network service queuing processing model total 、Number of servers n total They are respectively the sum of the service arrival rate and the sum of the number of servers in the traditional slice network service processing queue model. The service departure rate of the SDN-based slice network service queuing processing model is the same as the service departure rate in the traditional slice network service processing queue model.

[0111] In the slice network service queuing processing model of the embodiment of the present invention, the service arriving at each network slice for service transmission processing can be regarded as the service arriving at an SDN queue, and the SDN controller plays the role of service flow diversion. Therefore, it can avoid the service processing bottleneck problem caused by the surge in service volume and the limited storage resources and computing power of the VNF (Virtual Network Function) module of a certain network slice.

[0112] In the embodiment of the present invention, DTMC is used to describe the state of the network slice under the control of the SDN controller. A network slice can be regarded as a state in DTMC. In the embodiment of the present invention, it is assumed that there are K network slices of a certain type in the system, and the utilization rate of the kth network slice is k It can be regarded as the probability of a service transferring from one slice to itself in DTMC. Therefore, there is P in the transfer probability matrix in DTMC. ij =ρ k , (if i=j), that is, the probability of business transferring from the kth network slice to itself is the utilization rate of the kth network slice.

[0113] In the embodiment of the present invention, it is assumed by default that the probability of a service being transferred from the kth network slice to other K-1 slices is the same. Therefore, the probability of a service being transferred from a certain network slice to other K-1 network slices is the probability of the service being transferred from one network slice to another K-1 slices, which is equal to K-1 parts after removing the probability of the service being transferred to its own slice. Therefore, the transition probability P in this DTMC ij It can be expressed as:

[0114]

[0115] For intuitive expression, equation (2) can be expressed as the following matrix expression:

[0116]

[0117] Among them, ρ k represents the utilization of the kth network slice, K represents the total number of target network slices, i, j represent the row number and column number respectively, i = 1, 2 ... K, j = 1, 2, ... K. The above matrix P is the first utilization transfer probability matrix.

[0118] (2) Obtaining a second utilization rate transfer probability matrix according to the first utilization rate transfer probability matrix.

[0119] In order to reduce the difference in the amount of business carried by each network slice and achieve load balancing of the business, the transfer probability P in the first utilization transfer probability matrix is ij After the change, the transition probability matrix is ​​α multiplied by the transition probability matrix of the network slice in DTMC plus the K equal parts of the constant term 1-α, that is:

[0120]

[0121] Among them, *represents the second utilization transfer probability matrix, P represents the first utilization transfer probability matrix, α is a constant and 0<α<1 (α can be set as needed), and K represents the total number of target network slices. αP can reduce the difference between each state. At the same time, in order to ensure that the sum of the transition probability of each state is 1, it is necessary to add a constant term on the basis of αP By adjusting α, the traffic carried by each network slice can be dynamically adjusted, so as to make full use of the idle resources of other network slices and improve the efficiency of network resource utilization.

[0122] (3) Obtain a stable state probability matrix of the target network slice based on the second utilization transfer probability matrix.

[0123] According to the second utilization rate transition probability matrix, based on the discrete time Markov chain, the steady state probability matrix of each network slice utilization can be obtained as follows: K ],π k represents the steady-state probability of the kth target network slice. Since in a discrete-time Markov chain, the steady-state probability indicates that the service flow state will converge to a stable value after several transfers, the probability of the service flow being in a certain network slice state can be obtained by the above method.

[0124] (4) According to the steady-state probability matrix and the service arrival rate in the SDN-based slice network service queuing processing model, a service arrival rate matrix of the target network slice is obtained.

[0125] In this step, the service arrival rate matrix of the target network slice is obtained in the following manner:

[0126] λ * =λ total π (5)

[0127] Among them, λ * represents the service arrival rate matrix of the target network slice, represents the sum of the service arrival rates of the target network slice, λ k is the service arrival rate of the kth target network slice; π represents the steady-state probability matrix, π = [π1, π2, ..., π K ],π k represents the stable state probability of the kth target network slice.

[0128] Step 203: Allocate the service data of the target type among the target network slices according to the service arrival rate.

[0129] In the embodiment of the present invention, the service arrival rate of each network slice to carry services is recalculated by the SDN controller, thereby completing the service load balancing of the network slice. Therefore, the solution of the embodiment of the present invention is simple to implement and requires less computing resources.

[0130] On the basis of the above embodiments, in order to trigger the load balancing operation in time, the standard deviation σ of the network slice utilization is used as a measurement indicator for evaluating the difference between the access service loads of different network slices. Therefore, in an embodiment of the present invention, the standard deviation of the utilization of the target network slice is obtained, and when the standard deviation is greater than a preset threshold, the service arrival rate of the target network slice is determined according to the service data of the target type.

[0131] Specifically, the standard deviation of the utilization of the target network slice is calculated according to the following formula (6):

[0132]

[0133] Where σ represents the standard deviation of the utilization of the target network slice, K represents the total number of target network slices, and ρ k represents the system utilization of the kth network slice, Represents the average utilization of K network slices.

[0134] When the index is greater than a certain threshold σ0, the load balancing mechanism is triggered and the load balancing strategy is implemented. After load balancing, since the differences in the services carried by each network slice are relatively small, the standard deviation σ obtained after load balancing is * It should be less than σ before load balancing. In this way, the differences in system load services of the entire slice network are considered from a macro perspective, and the service processing capability of the entire slice network system is improved.

[0135] Figure 5(a)-Figure 5(d) The following are examples of load balancing solutions for URLLC slice network services. Figure 5(a) and Figure 5(b) are examples of services carried by three URLLC slices and examples of services after load balancing the services through the above solution, where α = 0.2. Figure 5(c) and Figure 5(d) are examples of services carried by four URLLC slices and examples of services after load balancing the services through the above solution, where α = 0.2.

[0136] The steady-state probabilities and service arrival rates corresponding to the state transfer matrices shown in FIG5(a) and FIG5(b) are shown in Table 1.

[0137] Table 1

[0138]

[0139] Among them, π URLLC1, π URLLC2 , π URLLC3 are the steady-state probabilities of URLLC1, URLLC2, and URLLC3 slices, respectively, URLLC1 , URLLC2 , URLLC3 They are the service arrival rates of URLLC1, URLLC2, and URLLC3 slices before load balancing. They are the service arrival rates of URLLC1, URLLC2, and URLLC3 slices after load balancing.

[0140] The standard deviation between the slice network utilization shown in Figure 5(a) and Figure 5(b) is shown in Table 2.

[0141] Table 2

[0142] Before load balancing After load balancing σ=0.309 <![CDATA[σ * =0.0618]]>

[0143] Where σ is the standard deviation of the slice network utilization before load balancing; σ * is the standard deviation of the network utilization of slices after load balancing.

[0144] The steady-state probabilities and service arrival rates corresponding to the state transfer matrices shown in FIG5(c) and FIG5(d) are shown in Table 3.

[0145] Table 3

[0146]

[0147] Among them, π URLLC1 , π URLLC2 , π URLLC3 , π URLLC4 are the steady-state probabilities of URLLC1, URLLC2, URLLC3, and URLLC4 slices, respectively, URLLC1 , URLLC2 , URLLC3 , URLLC4 They are the service arrival rates of URLLC1, URLLC2, URLLC3, and URLLC4 slices before load balancing. They are the service arrival rates of URLLC1, URLLC2, URLLC3, and URLLC4 slices after load balancing.

[0148] The standard deviation between the slice network utilization shown in Figure 5(c) and Figure 5(d) is shown in Table 4.

[0149] Table 4

[0150] Before load balancing After load balancing σ=0.3625 <![CDATA[σ * =0.2725]]>

[0151] Where σ is the standard deviation of the slice network utilization before load balancing; σ * is the standard deviation of the network utilization of slices after load balancing.

[0152] It can be seen from the results of Table 1 and Table 2 that before load balancing, the traffic volume of URLLC3 slice network load is much larger than that of URLLC1 and URLLC2, and the traffic loads of the three slices are unbalanced; through the SDN controller, and based on the SDN-based slice network traffic load balancing method of the embodiment of the present invention, the traffic is flexibly scheduled. After load balancing, the traffic rates corresponding to the three slices are relatively small, and the traffic carried between the slices is effectively load balanced. It can be seen from the results of Table 3 and Table 4 that before load balancing, the traffic volume carried between the slices is unbalanced. After the traffic is flexibly scheduled based on the SDN-based slice network traffic load balancing method of the embodiment of the present invention, the traffic volume carried between the slices is relatively small.

[0153] From the above description, it can be seen that in this embodiment of the present invention, DTMC is used to describe the state of the network slice under SDN control, and the service arrival rate of each slice to carry the service is recalculated, thereby completing the load balancing of the slice network service. From the above URLLC slice network service load balancing solution example, it can be seen that this method has a good load balancing effect, is relatively simple and easy to implement, and uses less computing resources.

[0154] The embodiment of the present invention also provides a load balancing device, which is applied to an SDN controller. Figure 6 , Figure 6 1 is a structural diagram of a load balancing device provided in an embodiment of the present invention. Since the principle of solving the problem by the load balancing device is similar to that of the load balancing method in an embodiment of the present invention, the implementation of the load balancing device can refer to the implementation of the method, and the repeated parts will not be repeated.

[0155] like Figure 6 As shown, the load balancing device 600 includes:

[0156] The first receiving module 601 is used to receive service data of a target type; the first determining module 602 is used to determine a service arrival rate of a target network slice based on the service data of the target type, wherein the target network slice is used to process the service data of the target type, and the number of network slices is greater than or equal to 2; the first processing module 603 is used to distribute the service data of the target type among the target network slices based on the service arrival rate.

[0157] Wherein, the first determining module includes:

[0158] A first determination submodule, configured to determine a first utilization transfer probability matrix of the target network slice using DTMC;

[0159] A first acquisition submodule, configured to obtain a second utilization rate transfer probability matrix according to the first utilization rate transfer probability matrix;

[0160] A second acquisition submodule, configured to obtain a stable state probability matrix of the target network slice according to the second utilization transfer probability matrix;

[0161] The third acquisition submodule is used to obtain the service arrival rate matrix of the target network slice according to the steady state probability matrix and the service arrival rate in the SDN-based slice network service queuing processing model.

[0162] The first utilization rate transfer probability matrix is ​​expressed as follows:

[0163]

[0164] Among them, the elements in the matrix

[0165] ρ k represents the utilization rate of the kth network slice, K represents the total number of target network slices, i and j represent the row number and column number respectively, i=1, 2...K, j=1, 2,...K.

[0166] The first acquisition submodule is used to obtain a second utilization rate transfer probability matrix according to the first utilization rate transfer probability matrix in the following manner:

[0167]

[0168] Among them, * represents the second utilization transfer probability matrix, P represents the first utilization transfer probability matrix, α is a constant and 0<α<1, and K represents the total number of target network slices.

[0169] The third acquisition submodule is used to obtain the service arrival rate matrix of the target network slice according to the steady state probability matrix and the service arrival rate in the SDN-based slice network service queuing processing model in the following manner:

[0170] λ * =λ total π

[0171] Among them, λ * represents the service arrival rate matrix of the target network slice, represents the sum of the service arrival rates of the target network slice, λ kis the service arrival rate of the kth target network slice; π represents the steady-state probability matrix, π = [π1, π2, ..., π K ],π k represents the stable state probability of the kth target network slice.

[0172] Wherein, the device further comprises:

[0173] A first acquisition module is used to obtain a standard deviation of the utilization of the target network slice;

[0174] The first determination module is used to determine the service arrival rate of the target network slice based on the service data of the target type when the standard deviation is greater than a preset threshold.

[0175] The first acquisition module is used to calculate the standard deviation of the utilization of the target network slice in the following manner:

[0176]

[0177] Where σ represents the standard deviation of the utilization of the target network slice, K represents the total number of target network slices, and ρ k represents the system utilization of the kth network slice, Represents the average utilization of K network slices.

[0178] The device provided in the embodiment of the present invention can execute the above method embodiment, and its implementation principle and technical effect are similar, so this embodiment will not be repeated here.

[0179] The embodiment of the present invention also provides a load balancing device, which is applied to an SDN controller. Figure 7 , Figure 7 1 is a structural diagram of a load balancing device provided in an embodiment of the present invention. Since the principle of solving the problem by the load balancing device is similar to that of the load balancing method in an embodiment of the present invention, the implementation of the load balancing device can refer to the implementation of the method, and the repeated parts will not be repeated.

[0180] like Figure 7 As shown, the load balancing device 700 includes: a processor 701 and a transceiver 702;

[0181] Wherein, the transceiver 702 is used to receive service data of a target type;

[0182] The processor 701 is used to determine the service arrival rate of the target network slice according to the service data of the target type, wherein the target network slice is used to process the service data of the target type, and the number of network slices is greater than or equal to 2; according to the service arrival rate, the service data of the target type is distributed among the target network slices.

[0183] The processor 701 is used for:

[0184] Determine a first utilization transfer probability matrix of the target network slice using DTMC;

[0185] Obtaining a second utilization rate transfer probability matrix according to the first utilization rate transfer probability matrix;

[0186] Obtaining a stable state probability matrix of the target network slice according to the second utilization transition probability matrix;

[0187] According to the steady state probability matrix and the service arrival rate in the SDN-based slice network service queuing processing model, the service arrival rate matrix of the target network slice is obtained.

[0188] The first utilization rate transfer probability matrix is ​​expressed as follows:

[0189]

[0190] Among them, the elements in the matrix

[0191] ρ k represents the utilization rate of the kth network slice, K represents the total number of target network slices, i and j represent the row number and column number respectively, i=1, 2...K, j=1, 2,...K.

[0192] The processor 701 is configured to obtain a second utilization rate transfer probability matrix according to the first utilization rate transfer probability matrix in the following manner:

[0193]

[0194] Among them, * represents the second utilization transfer probability matrix, P represents the first utilization transfer probability matrix, α is a constant and 0<α<1, and K represents the total number of target network slices.

[0195] The processor is used to obtain a service arrival rate matrix of the target network slice according to the steady state probability matrix and the service arrival rate in the SDN-based slice network service queuing processing model in the following manner:

[0196] λ * =λ total π

[0197] Among them, λ * represents the service arrival rate matrix of the target network slice, represents the sum of the service arrival rates of the target network slice, λk is the service arrival rate of the kth target network slice; π represents the steady-state probability matrix, π = [π1, π2, ..., π K ],π k represents the stable state probability of the kth target network slice.

[0198] The processor 701 is used for:

[0199] Obtaining a standard deviation of the utilization of the target network slice;

[0200] When the standard deviation is greater than a preset threshold, the service arrival rate of the target network slice is determined based on the service data of the target type.

[0201] The processor 701 is used to calculate the standard deviation of the utilization of the target network slice in the following manner:

[0202]

[0203] Where σ represents the standard deviation of the utilization of the target network slice, K represents the total number of target network slices, and ρ k represents the system utilization of the kth network slice, Represents the average utilization of K network slices.

[0204] The device provided in the embodiment of the present invention can execute the above method embodiment, and its implementation principle and technical effect are similar, so this embodiment will not be repeated here.

[0205] An embodiment of the present invention further provides a communication device, comprising: a transceiver, a memory, a processor, and a program stored in the memory and executable on the processor, wherein the processor implements the steps in the load balancing method described above when executing the program.

[0206] The embodiment of the present invention also provides a readable storage medium, on which a program is stored. When the program is executed by a processor, each process of the above-mentioned load balancing method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it is not repeated here. Among them, the readable storage medium can be any available medium or data storage device that can be accessed by the processor, including but not limited to magnetic storage (such as floppy disk, hard disk, magnetic tape, magneto-optical disk (MO), etc.), optical storage (such as CD, DVD, BD, HVD, etc.), and semiconductor storage (such as ROM, EPROM, EEPROM, non-volatile memory (NAND FLASH), solid state drive (SSD)), etc.

[0207] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0208] 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 a 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 such an understanding, the technical solution of the present invention, 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, disk, CD), and includes a number of instructions for a terminal (which can be a mobile phone, computer, server, air conditioner, or network equipment, etc.) to execute the methods described in each embodiment of the present invention.

[0209] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims, all of which are within the protection of the present invention.

Claims

1. A load balancing method, applied to a software defined network (SDN) controller, characterized in that: include: Receive business data of the target type; Determine, according to the service data of the target type, a service arrival rate of a target network slice, wherein the target network slice is used to process the service data of the target type, and the number of the target network slices is greater than or equal to 2; Allocating the service data of the target type among the target network slices according to the service arrival rate; The determining, according to the service data of the target type, the service arrival rate of the target network slice includes: Determine a first utilization transition probability matrix of the target network slice using a discrete time Markov chain DTMC; Obtaining a second utilization rate transfer probability matrix according to the first utilization rate transfer probability matrix; Obtaining a stable state probability matrix of the target network slice according to the second utilization transition probability matrix; According to the steady state probability matrix and the service arrival rate in the SDN-based slice network service queuing processing model, the service arrival rate matrix of the target network slice is obtained.

2. The method according to claim 1, characterized in that The first utilization rate transfer probability matrix is ​​expressed as follows: Among them, the elements in the matrix ρ k represents the utilization rate of the kth network slice, K represents the total number of target network slices, i and j represent the row number and column number respectively, i=1, 2...K, j=1, 2,...K.

3. The method according to claim 1, characterized in that The obtaining a second utilization rate transfer probability matrix according to the first utilization rate transfer probability matrix comprises: The second utilization rate transfer probability matrix is ​​obtained according to the first utilization rate transfer probability matrix in the following manner: Among them, * represents the second utilization transfer probability matrix, P represents the first utilization transfer probability matrix, α is a constant and 0<α<1, and K represents the total number of target network slices.

4. The method according to claim 1, characterized in that The obtaining, according to the steady state probability matrix and the service arrival rate in the SDN-based slice network service queuing processing model, a service arrival rate matrix of the target network slice includes: The service arrival rate matrix of the target network slice is obtained according to the steady state probability matrix and the service arrival rate in the SDN-based slice network service queuing processing model in the following manner: l * =λ total p Among them, λ * represents the service arrival rate matrix of the target network slice, represents the sum of the service arrival rates of the target network slice, λ k is the service arrival rate of the kth target network slice; π represents the steady-state probability matrix, π = [π1, π2, ..., π K ],π k represents the stable state probability of the kth target network slice.

5. The method according to claim 1, characterized in that Before determining the service arrival rate of the target network slice according to the service data of the target type, the method further includes: Obtaining a standard deviation of the utilization of the target network slice; The determining, according to the service data of the target type, a service arrival rate of the target network slice includes: When the standard deviation is greater than a preset threshold, the service arrival rate of the target network slice is determined based on the service data of the target type.

6. The method according to claim 5, characterized in that The standard deviation of the utilization of the target network slice is calculated as follows: Where σ represents the standard deviation of the utilization of the target network slice, K represents the total number of target network slices, and ρ k represents the system utilization of the kth network slice, Represents the average utilization of K network slices.

7. A load balancing device, applied to an SDN controller, characterized in that: include: A first receiving module, used for receiving business data of a target type; A first determination module is used to determine a service arrival rate of a target network slice according to the service data of the target type, wherein the target network slice is used to process the service data of the target type, and the number of the target network slices is greater than or equal to 2; A first processing module, configured to distribute the service data of the target type among the target network slices according to the service arrival rate; Wherein, the first determining module includes: A first determination submodule, configured to determine a first utilization transfer probability matrix of the target network slice using DTMC; A first acquisition submodule, configured to obtain a second utilization rate transfer probability matrix according to the first utilization rate transfer probability matrix; A second acquisition submodule, configured to obtain a stable state probability matrix of the target network slice according to the second utilization transfer probability matrix; The third acquisition submodule is used to obtain the service arrival rate matrix of the target network slice according to the steady state probability matrix and the service arrival rate in the SDN-based slice network service queuing processing model.

8. A load balancing device, applied to an SDN controller, characterized in that: include: processors and transceivers; Wherein, the transceiver is used to receive service data of the target type; The processor is configured to determine a service arrival rate of a target network slice according to the service data of the target type, wherein the target network slice is used to process the service data of the target type, and the number of the target network slices is greater than or equal to 2; and distribute the service data of the target type among the target network slices according to the service arrival rate; Wherein, the processor is further used for: Determine a first utilization transfer probability matrix of the target network slice using DTMC; Obtaining a second utilization rate transfer probability matrix according to the first utilization rate transfer probability matrix; Obtaining a stable state probability matrix of the target network slice according to the second utilization transition probability matrix; According to the steady state probability matrix and the service arrival rate in the SDN-based slice network service queuing processing model, the service arrival rate matrix of the target network slice is obtained.

9. A communication device, comprising: A transceiver, a memory, a processor, and a program stored in the memory and executable on the processor; characterized in that: The processor is used to read the program in the memory to implement the steps of the load balancing method according to any one of claims 1 to 6.

10. A readable storage medium for storing a program, characterized in that: When the program is executed by a processor, the steps in the load balancing method according to any one of claims 1 to 6 are implemented.

Images