A network slice access control method and system based on congestion control

Abstract

The present invention discloses a network slice access control method and system based on congestion control, including the following steps: S1, setting up an initial congestion control coefficient for any network slice, and calculating the slice throughput mean and congestion control threshold; S2, realizing differentiated service priority identification according to the service priority calculation of the network slice; S3, evaluating the severity of the current timeout of each service and realizing dynamic control of service timeout; S4, selecting a slice network to realize slice access of newly arrived services; S5, adjusting the congestion control coefficient by comparing the congestion rate according to the target slice applied for the new service access, realizing congestion window update, and thus realizing dynamic access control according to the slice status. The present invention realizes access control of different services by controlling the slice congestion degree under the premise of certain 5G network slice resources, effectively controlling the congestion rate within the slice network, and solving the collision problem of services within the slice.

Description

Technical Field

[0001] The present invention belongs to the application field of 5G network slicing scenarios, and in particular, relates to a network slicing access control method and system based on congestion control. Background Art

[0002] The overall goal of wireless access of 5G communication technology is to provide high-density traffic, low-cost and low-latency services, and to achieve flexible switching and management of access networks to meet different business needs. In order to achieve the above goals, the 5G system is mainly improved in terms of random access reconstruction, edge enhancement and network slicing. Wireless access will be optimized and enhanced according to the needs of different application scenarios.

[0003] For this reason, Li Xin and others proposed a 5G network slice access control method based on service type (patent number: 202211365786.8), which first determines the service type, and then selects different principles to implement network slice access control based on whether it is a delay-sensitive service. This method is simple, easy to implement, and highly efficient, but it does not consider the issue of slice load too much.

[0004] The applicant proposed a 5G network slice access control method and system based on load evaluation (application number: 202410213625.X), starting from the goal of controlling the total load of the slice, starting from the end, and performing differentiated access according to the attributes of the service. However, these two methods do not intuitively and effectively solve the congestion problem that often occurs in the slice network, and with more and more access, the possibility of service collision has increased significantly. Therefore, how to effectively control the congestion rate within the slice network, select different access control methods, and solve the collision problem of services within the slice has become more and more urgent. Summary of the invention

[0005] The purpose of the present invention is to solve the congestion problem that often occurs in the slice network and effectively control the congestion rate in the slice network; solve the dynamic access problem of differentiated services in 5G scenarios; solve the robustness and security problems of the 5G network slicing system and provide a 5G network slice access control method and system based on congestion control; on the basis of considering the slice load, access control of different services is achieved by controlling the slice congestion degree under the premise of certain 5G network slice resources. Each network slice has an appropriate amount of services, and the priority of the newly arrived services is estimated to seek target slices that meet resource requirements; according to the differentiated services in different scenarios, different access criteria are set, and the services are dynamically adjusted and controlled, so as to ensure the implementation of access to different network slices and meet the index requirements of different services.

[0006] The technical solution adopted by the present invention to achieve its invention purpose is: a network slice access control method based on congestion control, comprising the following steps: S1, establishing an initial congestion control coefficient for any network slice, and calculating the slice throughput mean and congestion control threshold; S2, realizing differentiated service priority identification based on network slice service priority calculation; S3, evaluating the severity of the current timeout of each service and realizing dynamic control of service timeout; S4, selecting a slice network to realize slice access for newly arrived services; S5, adjusting the congestion control coefficient according to the target slice applied for new service access by comparing the congestion rate, realizing congestion window update, and thereby realizing dynamic access control according to the slice status.

[0007] The present invention first establishes an initial congestion control coefficient at the beginning stage, starts with setting a slice throughput threshold for controlling congestion, and assigns a higher priority to all services with throughput less than the threshold; calculates the weights of all service priorities, evaluates the severity of the current timeout of each service, and gives high priority to services that have serious timeouts and may cause congestion; calculates the weighted sum of the service priorities of all slice networks that can characterize the current congestion status, and then selects the slice network with the minimum weighted sum value to complete the access to the newly arrived service; after access, if the congestion rate of the target slice increases, the congestion control coefficient can be reduced by reducing the congestion window value, and the service is dynamically adjusted and controlled according to the slice status; compared with the prior art mentioned in the background technology, the present invention effectively controls the congestion rate in the slice network, solves the congestion and collision problems that often occur in the slice network, ensures that the service is always in a slice environment with a good environment, and provides real-time protection for improving customer perception.

[0008] The present invention not only controls congestion, that is, when network slice congestion occurs, it will automatically shrink the congestion window, thereby greatly reducing the probability of slice congestion, but also takes into account the problem of slice load, thereby ensuring a certain load control effect.

[0009] Preferably, the step S1 comprises the following steps: S101. Establish any network slice Sec i The initial congestion control coefficient is used to solve the mathematical expectation of all service throughputs of any network slice. S102, multiplying the mathematical expectation value obtained by the solution by the initial congestion control coefficient to obtain Sec i Congestion control throughput threshold within a slice.

[0010] The present invention sets an initial congestion control coefficient and starts from setting a slice throughput threshold for controlling congestion. i The throughput and threshold within the slice will give higher priority to all services with throughput less than the threshold.

[0011] Preferably, step S2 comprises the following steps: S201, for any slice Sec i Any business within Obtaining a corresponding service priority by dividing the congestion control throughput threshold by the throughput within the corresponding service; S202, sum up the above service priorities to obtain Sec i The sum of all business priorities; S203: Divide the service priority by the sum of all service priorities to obtain a service priority. The priority weight of .

[0012] The present invention can calculate the transmission priority of each service, can measure the service congestion weight according to the service priority, and realize differentiated service priority identification.

[0013] Preferably, step S3 comprises: for any slice Sec i Any business within The severity of the current timeout of each service is evaluated by comparing the waiting delay and the delay tolerance.

[0014] Preferably, the step S3 specifically comprises the following steps: S301. When the waiting delay is greater than or equal to the delay tolerance, that is, the slice times out, the service timeout coefficient is calculated by dividing the waiting delay by the delay tolerance. S302: When the waiting delay is less than the delay tolerance, the slice is not timed out. The difference between the delay tolerance and the waiting delay is calculated, and the quotient obtained by subtracting the difference divided by the waiting delay from one is used to obtain the service timeout coefficient, which is used to estimate the possibility that the service is about to time out. S303: According to the calculation results of steps S301 and S302, the services that may cause congestion due to serious timeout are given high priority, and the corresponding services are calculated according to the service timeout status. The priority weight value of the business The product of the priority weight and the business timeout coefficient is used to achieve weighted evaluation of the comprehensive degree of timeout.

[0015] The present invention measures the service congestion weight according to the service priority, and calculates the corresponding service weight value according to the service timeout state; the service timeout state is calculated by waiting for the delay and delay tolerance By comparison, when a slice times out, the timeout coefficient is calculated according to the timeout coefficient calculation method when it times out, which represents the severity of the timeout; if a slice does not time out, the possibility of the service being about to time out is estimated and evaluated according to the timeout coefficient calculation method when it does not time out; the priority weights of all services are calculated through the priority weights of the above steps; the severity of the current timeout of each service is evaluated to achieve dynamic control of service timeouts, laying the foundation for the subsequent selection of slice networks to achieve slice access for newly arrived services.

[0016] Preferably, step S4 comprises the following steps: S401, calculate any slice Sec i The sum of all business priority weights in SJpri i , the SJpri i Recorded as the access decision value, it is used to characterize the current congestion status of all slice networks; S402. Select a slice network with a minimum weighted sum value, i.e., a minimum access decision value, to complete access to the newly arrived service.

[0017] Preferably, step S5 comprises the following steps: S501, setting a congestion window decrement value and calculating a congestion window increment value; S502, when new service SRV arrives new Access target slice Sec t Then, the access decision value SJpri of the target slice is determined. t increase or decrease.

[0018] Preferably, step S5 comprises the following steps: S503, if the target slice Sec t SJpri t The value increases, that is, the congestion rate of the target slice increases, then the congestion control coefficient of the target slice is reduced by the congestion window reduction value, and the congestion control coefficient Coef t ′=Coef t *Ccv sub ; If the target slice Sec t SJpri t If the value decreases, the congestion control coefficient of the target slice is increased, and the congestion control coefficient Coef t ′=Coef t *Ccv pls .

[0019] A 5G network slice access control system based on congestion control includes a congestion control throughput threshold module, a differentiated service priority identification module, a service timeout dynamic control module and a new arrival service access control module, wherein each module is connected via a wireless network.

[0020] Preferably, the congestion control throughput threshold module is used to establish an initial congestion control coefficient; calculate the throughput mean of all services in all slices, and calculate the congestion control throughput threshold corresponding to the service according to the initial congestion control coefficient; The differentiated service priority identification module includes calculating the service priority of any service according to the congestion control throughput threshold; in the initial stage, the smaller the throughput, the higher the priority; summing up all priorities in a single slice and calculating the corresponding priority weight of each service.

[0021] Preferably, the service timeout dynamic control module is used to perform a weighted evaluation of the comprehensive degree of timeout, specifically including: when a slice times out, calculating the timeout coefficient under the timeout condition to characterize the severity of the timeout; if it does not time out, estimating the possibility that the service is about to time out, and taking the timeout coefficient evaluation under the condition of no timeout; calculating the weighted value of all service priority weights through the priority weights of the above steps; The newly arrived service access control module is used to dynamically adjust and control the service according to the slice status; specifically, it includes calculating the sum of the weighted values ​​of all the above services in any slice, and representing the decision value of access feasibility; for the newly arrived service, the slice with the smallest decision value is selected for access; if the congestion of the target slice increases after the new service is accessed, the congestion control coefficient of the slice is reduced, otherwise the congestion control coefficient of the slice is increased.

[0022] In summary, the present invention includes the following beneficial technical effects: the present invention calculates the transmission priority of each service through the congestion control throughput threshold calibration, and measures the service congestion weight according to the service priority, and calculates the corresponding service weighted value according to the service timeout status to achieve dynamic control of service timeout, thereby solving the congestion control problem of 5G network slicing. The present invention solves the dynamic access problem of differentiated services in 5G scenarios and the robustness and security problems of 5G network slicing systems by capturing the service weighted sum value that is ultimately used to judge the congestion degree of the slice, and implementing dynamic control based on the impact of the access of newly arrived services; it can ensure that the service is always in a slice environment with a good environment, providing real-time protection for improving customer perception. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 is a flow chart of the 5G network slice access control method in the present invention; Figure 2 This is a specific flow chart of selecting a slice access method according to a newly arrived service type in an embodiment of the present invention; Figure 3 is a comparison diagram of average congestion of service access between the embodiment of the present invention and the comparative technology; Figure 4 This is a comparison chart of the average load of slices between the embodiment of the present invention and the comparative technology. DETAILED DESCRIPTION

[0024] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments.

[0025] It should be noted that if Figure 1 , Figure 2 In the embodiment shown, a network slice access control method based on congestion control includes m network slices SEC = {Sec1, Sec2, ..., Sec m}, Network Slice Sec i , there are already n i Business Sec i Any business within The current waiting delay is marked as The delay tolerance is marked as Throughput is marked as The current newly arrived service is marked as SRV new .

[0026] Specifically, a network slice access control method based on congestion control includes the following steps: S1, congestion control throughput threshold calibration: establish an initial congestion control coefficient of any network slice, and calculate the slice throughput mean and congestion control threshold; Step S1 includes the following steps: s101、Establish any network slice Sec i The initial congestion control coefficient Coef i , solve the mathematical expectation of all service throughputs of any network slice Right now S102, using the mathematical expectation value obtained by the solution and the initial congestion control coefficient Coef i Multiply to get Sec i Congestion control throughput threshold CoThr within a slice i ,Right now The present invention sets an initial congestion control coefficient and starts from setting a slice throughput threshold for controlling congestion. i The throughput and threshold within the slice will give higher priority to all services with throughput less than the threshold.

[0027] When i>m, set slice i=1, service j=1 in slice i and go to step S2.

[0028] When i≤m, slice i=i+1, and return to calculate the mean value of all service throughputs in slice i.

[0029] S2. Differentiated service priority identification is realized based on the service priority calculation of the network slice; Step S2 includes the following steps: S201, for any slice Sec i Any business within The congestion control throughput threshold CoThr i The throughput of the corresponding business Divide to get the corresponding business priority Right now S202: Prioritize the above services Sum to get Sec i The sum of all business priorities Spri i ;Right now S203. Business Priority Divide by the sum of all business priorities Spri i Get business Priority weight Right now

[0030] Similarly, after the above steps are calculated, it is necessary to determine whether j>n. If the condition is not met, it is necessary to return to calculate the priority of service j; if both j>n and i>m are met, then go to step S3.

[0031] S3. Evaluate the severity of current timeouts of each service and implement dynamic control of service timeouts; Step S3 includes: for any slice Sec i Any business within By comparing the waiting delay and delay tolerance Evaluate the severity of the current timeout of each service.

[0032] The specific steps include: S301: When the waiting delay is greater than or equal to the delay tolerance, that is, When the slice times out, the service timeout coefficient is calculated by dividing the waiting delay by the delay tolerance, that is, S302: When the waiting delay is less than the delay tolerance, When , the slice has not timed out, the difference between the delay tolerance and the waiting delay is calculated, and the quotient obtained by subtracting the difference divided by the waiting delay from one is used to obtain its service timeout coefficient, that is, And used to estimate the possibility that the business will time out; S303: According to the calculation results of steps S301 and S302, the services that may cause congestion due to serious timeout are given high priority, and the corresponding services are calculated according to the service timeout status. The priority weight value of the business The product of the priority weight and the service timeout coefficient, that is, To achieve a weighted assessment of the comprehensive degree of overtime.

[0033] When j>n is not satisfied, then j=j+1, and the process reenters step S3 to determine whether service j has timed out; When j>n is satisfied, but i>m is not satisfied, i=i+1, and service j=1 is set in slice i, and step S3 is re-entered to determine whether service j has timed out; When j>n and i>m are satisfied at the same time, the process goes to step S4 to select access to the slice network, and at this time, slice i=i+1.

[0034] The present invention measures the service congestion weight according to the service priority, and calculates the corresponding service weight value according to the service timeout state; the service timeout state is calculated by waiting for the delay and delay tolerance By comparison, when a slice times out, the timeout coefficient is calculated according to the timeout coefficient calculation method when it times out, which represents the severity of the timeout; if a slice does not time out, the possibility of the service being about to time out is estimated and evaluated according to the timeout coefficient calculation method when it does not time out; the priority weights of all services are calculated through the priority weights of the above steps; the severity of the current timeout of each service is evaluated to achieve dynamic control of service timeouts, laying the foundation for the subsequent selection of slice networks to achieve slice access for newly arrived services.

[0035] S4. Select a slice network to implement slice access for the newly arrived service. Step S4 includes the following steps: S401, calculate the access decision value of slice i, that is, calculate any slice Sec i The sum of all business priority weights in is as follows: The SJpri i Recorded as the access decision value, it is used to characterize the current congestion status of all slice networks; S402. When i>m is satisfied, select the slice network t with the minimum weighted sum value, that is, the minimum access decision value, to complete the access to the newly arrived service.

[0036] S5. According to the target slice t applied for the new service access, the congestion control coefficient is adjusted by comparing the congestion rate to update the congestion window, thereby realizing dynamic access control according to the slice status.

[0037] The congestion window is a state variable maintained by the sender, which changes dynamically according to the degree of network congestion. The essence of congestion control is to use an algorithm to control the congestion window to avoid injecting too much data into the network.

[0038] Specifically, step S5 includes the following steps: S501. Set the congestion window reduction value Ccv sub And calculate the congestion window core value Ccv pls =1 / Ccv sub ; S502, when new service SRV arrives new Access target slice Sec t After that, the access decision value SJpri of the target slice is determined t increase or decrease; S503, if the target slice Sec t SJpri t The value increases, that is, the congestion of the target slice t increases. At this time, the congestion control coefficient of the target slice t is reduced by the congestion window reduction value. The congestion control coefficient Coef t ′=Coef t *Ccv sub ; If the target slice Sec t SJpri t The value decreases, then the congestion control coefficient of the target slice t is increased, and the congestion control coefficient Coef t ′=Coef t *Ccv pls .

[0039] A 5G network slice access control system based on congestion control includes a congestion control throughput threshold module, a differentiated service priority identification module, a service timeout dynamic control module and a new arrival service access control module, wherein each module can be connected via a wireless network.

[0040] The congestion control throughput threshold module is used to set up the initial congestion control coefficient; calculate the average throughput of all services in all slices, and calculate the congestion control throughput threshold corresponding to the service based on the initial congestion control coefficient; The differentiated service priority identification module includes calculating the service priority of any service based on the congestion control throughput threshold; in the initial stage, the smaller the throughput, the higher the priority; summing up all priorities within a single slice and calculating the corresponding priority weight of each service.

[0041] The service timeout dynamic control module is used to perform a weighted evaluation of the comprehensive degree of timeout, including: when a slice times out, the timeout coefficient is calculated to characterize the severity of the timeout; if it does not time out, the possibility of the service being about to time out is estimated, and the timeout coefficient evaluation is adopted when there is no timeout; the priority weights of all services are calculated through the priority weights of the above steps; The newly arrived service access control module is used to dynamically adjust and control the service according to the slice status; specifically, it includes calculating the sum of the weighted values ​​of all the above services in any slice, and representing the decision value of access feasibility; for the newly arrived service, the slice with the smallest decision value is selected for access; if the congestion of the target slice increases after the new service is accessed, the congestion control coefficient of the slice is reduced, otherwise the congestion control coefficient of the slice is increased.

[0042] The present invention is described in detail below by taking m=3 as an example. The network slices of 5G are shown in Table 1: Table 1 Services available in each 5G network slice The basic data are shown in Table 2: project data Operating frequency (GHz) 2.6 Operating bandwidth (MHz) 100 Subcarrier bandwidth index 1 <![CDATA[Congestion window core write-down value Ccv sub > 0.9 Table 2 Basic data A network slice access control method based on congestion control described in this embodiment includes the following steps: congestion control throughput threshold calibration, differentiated service priority identification, service timeout dynamic control, new arrival service access control, etc.; the specific calculation process is as follows: Step 1: Congestion control throughput threshold calibration: Step 1-1: Calculate Sec i The mathematical expectation of all business throughput is as follows: Step 1-2: Calculate Sec i The congestion control throughput threshold within a slice is as follows: Step 2: Identify differentiated business priorities: Step 2-1: For any slice Sec i Any business within The corresponding business priorities are calculated as follows: Step 2-2: Calculate Sec i The sum of all business priorities Step 2-3: Calculate business The priority weights are as follows: Step 3: Dynamic control of service timeout; Step 3-1: For any slice Sec i Any business within When the conditions are met When , it means the slice has timed out, and its service timeout coefficient is calculated. When the conditions are met When , the slice has not timed out, and its service timeout coefficient is calculated. Used to estimate the possibility that the service will time out.

[0043] From the above, we can obtain the timeout coefficients of any service as {{0.8, 0.93, 0.6, 0.9}, {0.8, 0.83}, {0.8, 0.53, 1.3}}.

[0044] Step 3-2: Prioritize services that have serious timeouts and may cause congestion, and calculate the service timeout status based on the service timeout status. The priority weights are as follows: Step 4: Access control of newly arrived services; Step 4 specifically includes calculating task slice Sec i The sum of all business priority weights in It is called the access decision value and is used to characterize the current congestion situation of all slice networks. For the current new arrival service SRV new , select the one with the minimum weighted sum value SJpri i =The network slice Sec2 with a judgment value of 0.81 is accessed to complete the access of the newly arrived service.

[0045] Step 5: According to the target slice applied for the new service access, the congestion control coefficient is adjusted by comparing the congestion rate to update the congestion window, thereby realizing dynamic access control according to the slice status.

[0046] Step 5-1: Set the congestion window reduction value Ccv sub And calculate the congestion window core value Ccv pls =1 / Ccv sub ; Step 5-2: When the new business SRV new After accessing the target slice Sec2, determining whether the access decision value Sec2 of the target slice increases or decreases; If Sec t＝2 SJpri t The value increases, that is, the congestion of the target slice increases, and the congestion rate increases. At this time, the congestion control coefficient of the slice is reduced, and the congestion control coefficient of the slice Coef′2=Coef t *Ccv sub =0.63; If Sec t＝2 SJpri t If the value decreases, the congestion control coefficient of the target slice is increased, and the congestion control coefficient Coef′ of the target slice t =Coef t *Ccv pls =0.78.

[0047] The present invention designs a network slice access control method based on congestion control. In the initial stage, an initial congestion control coefficient is established. Starting from setting a slice throughput threshold for controlling congestion, all services with throughput less than the threshold are given a higher priority; the weights of all service priorities are calculated, the severity of the current timeout of each service is evaluated, and high priority is weighted for services that cause congestion due to severe timeouts; the weighted sum of service priorities that can characterize the current congestion status of all slice networks is calculated, and then the slice network with the minimum weighted sum value is selected to complete the access to the newly arrived service; after access, if the congestion rate of the target slice increases, the congestion control coefficient can be reduced by reducing the congestion window value, and vice versa; through the above method, dynamic adjustment and control of the service can be implemented according to the slice status.

[0048] The following are the results of the simulation experiment: A network slice access control method based on congestion control of the present invention is simulated on the MATLAB platform together with a slice access control method based on load judgment with application number 202410213625.X and a 5G network slice access control method based on service type with application number 202211365786.8.

[0049] According to the above table, the network and service configuration are performed, and the average access congestion rate and average slice load of the three slice dismissal control algorithms are obtained as shown in the attached figure. Figure 3 and Figure 4 shown.

[0050] For the convenience of description, the slice access control algorithm of the present invention is abbreviated as CC-NSAC; the load evaluation-based slice access control method with application number 202410213625.X is abbreviated as LJ-NSAC.

[0051] like Figure 3 As shown in the figure, the slice access control method of the present invention and the other two slice access control algorithms have similar congestion rates when accessed at the beginning of the simulation. The reason is that in the early stage of access, the number of access services is small and the slice load is low, so the probability of network congestion is not high. However, as the number of access services increases, the slice link becomes more and more congested, and the congestion rates of the above three slice control methods all show an increasing trend. If not controlled, the robustness of the system will be sharply reduced, resulting in service collisions.

[0052] However, since the CC-NSAC of the present invention has an automatic adjustment mechanism for the congestion window through the steps of congestion control throughput threshold calibration, differentiated service priority identification, service timeout dynamic control, and new service access control, when network slice congestion occurs, the congestion window will be automatically reduced, thereby greatly reducing the probability of slice congestion. The other two slice access control methods do not have an automatic adjustment mechanism for the congestion window. Figure 3 As shown in the curve, with the continuous increase of access traffic, the other two slice access control methods have much higher final average congestion rates than the CC-NSAC of the present invention.

[0053] The difference is that the LJ-NSAC slice control method controls the final load of the slice to a certain extent and makes a certain contribution to controlling slice congestion. However, the CC-NSAC slice control access method of the present invention directly controls congestion without artificially limiting the slice load upper limit. Even when the load in the network is very high, it can maintain and ensure that the service is always in a slice environment with a good environment.

[0054] like Figure 4 As shown, it can be seen that there is a certain relationship between slice load and congestion, but it is not a linear relationship. When the slice access control method of the present invention and the other two slice access control algorithms are accessed at the beginning of the simulation, the average slice load is almost the same; as the number of access services increases, the average slice loads of the three are different. Specifically, the slice access control method based on service type has the largest average slice load, and LJ-NSAC has the least average slice load and the best control.

[0055] LJ-NSAC controls the slice load, which is more effective in load control. However, the slice access control method based on service type does not restrict load and congestion. It controls the delay and non-delay indicators of the service more, which greatly increases the probability of service collision, resulting in an increase in slice load in the later stage of service access, thereby reducing the robustness of the slice network.

[0056] Although the load control effect of CC-NSAC in the present invention is not as obvious as that of LJ-NSAC, the present invention not only controls congestion, that is, when network slice congestion occurs, it will automatically reduce the congestion window, thereby greatly reducing the probability of slice congestion, but also takes the problem of slice load into consideration, thus ensuring a certain load control effect. The present invention calculates the transmission priority of each service through the congestion control throughput threshold calibration, and measures the service congestion weight according to the service priority. At the same time, the corresponding service weighted value is calculated according to the service timeout status to achieve dynamic control of service timeout, thereby solving the congestion control problem of 5G network slicing. The present invention solves the dynamic access problem of differentiated services in 5G scenarios and the robustness and security problems of 5G network slicing systems by capturing the service weighted sum value that is ultimately used to judge the congestion degree of the slice, and implementing dynamic control based on the impact of the access of newly arrived services; it can ensure that the service is always in a slice environment with a good environment, providing real-time protection for improving customer perception.

[0057] The load-judgment-based slice access control method with application number 202410213625.X evaluates and limits the slice load, estimates the load of newly arrived services, seeks target slices that meet resource requirements, and sets different access methods according to differentiated services in different scenarios. This can ensure the implementation of access to different network slices and meet the indicator requirements of different services and the robustness and security of communications.

[0058] The application number is 202211365786.8. The 5G network slicing access control method based on service type minimizes the delay of delay-sensitive services; achieves relative fairness in the transmission delay size of each service; and can minimize the transmission waiting time of non-delay-sensitive services to improve user perception.

[0059] Compared with the prior art, the present invention can calculate the throughput threshold within a slice according to the congestion control coefficient through the above simulation experiments; It can calculate the transmission priority of each service; it can measure the service congestion weight according to the service priority, and calculate the corresponding service weighted value according to the service timeout status; it can capture the service weighted sum value ultimately used to judge the congestion of the slice, and implement dynamic control according to the impact of the access of newly arrived services; it can ensure that the service is always in a slice environment with a good environment, providing real-time guarantee for improving customer perception.

[0060] The application scenarios of the present invention include 5G large-scale deployment of networks and typical application scenarios of 5G network slicing, which will promote the scenario deployment and application of 5G slicing.

[0061] The above embodiments are further elaborations and illustrations of the present invention for ease of understanding, and are not limitations of the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A network slice access control method based on congestion control, characterized in that: The following steps are involved: S1. Set up the initial congestion control coefficient of any network slice and calculate the slice throughput mean and congestion control threshold; S2. Differentiated service priority identification is realized based on the service priority calculation of the network slice; S3. Evaluate the severity of current timeouts of each service and implement dynamic control of service timeouts; S4. Select a slice network to implement slice access for the newly arrived service. S5. According to the target slice applied for the new service access, the congestion control coefficient is adjusted by comparing the congestion rate to update the congestion window, thereby realizing dynamic access control according to the slice status.

2. A network slice access control method based on congestion control according to claim 1, characterized in that: The step S1 comprises the following steps: S101. Establish any network slice Sec i The initial congestion control coefficient is used to solve the mathematical expectation of all service throughputs of any network slice. S102, multiplying the mathematical expectation value obtained by the solution by the initial congestion control coefficient to obtain Sec i Congestion control throughput threshold within a slice.

3. A network slice access control method based on congestion control according to claim 1 or 2, characterized in that: The step S2 comprises the following steps: S201, for any slice Sec i Any business within Obtaining a corresponding service priority by dividing the congestion control throughput threshold by the throughput within the corresponding service; S202, sum up the above service priorities to obtain Sec i The sum of all business priorities; S203: Divide the service priority by the sum of all service priorities to obtain a service priority. The priority weight of .

4. A network slice access control method based on congestion control according to claim 3, characterized in that: The step S3 comprises: for any slice Sec i Any business within The severity of the current timeout of each service is evaluated by comparing the waiting delay and the delay tolerance.

5. A network slice access control method based on congestion control according to claim 4, characterized in that: The step S3 specifically comprises the following steps: S301. When the waiting delay is greater than or equal to the delay tolerance, that is, the slice times out, the service timeout coefficient is calculated by dividing the waiting delay by the delay tolerance. S302: When the waiting delay is less than the delay tolerance, the slice is not timed out. The difference between the delay tolerance and the waiting delay is calculated, and the quotient obtained by subtracting the difference divided by the waiting delay from one is used to obtain the service timeout coefficient, which is used to estimate the possibility that the service is about to time out. S303: According to the calculation results of steps S301 and S302, the services that may cause congestion due to serious timeout are given high priority, and the corresponding services are calculated according to the service timeout status. The priority weight value of the business The product of the priority weight and the business timeout coefficient is used to achieve weighted evaluation of the comprehensive degree of timeout.

6. A network slice access control method based on congestion control according to claim 1 or 5, characterized in that: The step S4 comprises the following steps: S401, calculate any slice Sec i The sum of all business priority weights in SJpri i , the SJpri i Recorded as the access decision value, it is used to characterize the current congestion status of all slice networks; S402. Select a slice network with a minimum weighted sum value, i.e., a minimum access decision value, to complete access to the newly arrived service.

7. A network slice access control method based on congestion control according to claim 6, characterized in that: The step S5 comprises the following steps: S501, setting a congestion window decrement value and calculating a congestion window increment value; S502, when new service SRV arrives new Access target slice Sec t After that, the access decision value SJpri of the target slice is determined t increase or decrease.

8. A network slice access control method based on congestion control according to claim 7, characterized in that: The step S5 comprises the following steps: S503, if the target slice Sec t SJpri t The value increases, that is, the congestion rate of the target slice increases, then the congestion control coefficient of the target slice is reduced by the congestion window reduction value, and the congestion control coefficient coef t ′=Coef t *Ccv sub ; If the target slice Sec t SJpri t If the value decreases, the congestion control coefficient of the target slice is increased, and the congestion control coefficient Coef t ′=Coef t *Ccv pls .

9. A network slice access control system based on congestion control, characterized in that: A network slice access control method based on congestion control as described in any one of claims 1-8, comprising a congestion control throughput threshold module, a differentiated service priority identification module, a service timeout dynamic control module and a newly arrived service access control module, wherein each module is connected via a wireless network.

10. A network slice access control system based on congestion control according to claim 8, characterized in that: The congestion control throughput threshold module is used to set up an initial congestion control coefficient; calculate the throughput mean of all services in all slices, and calculate the congestion control throughput threshold corresponding to the service according to the initial congestion control coefficient; The differentiated service priority identification module includes calculating the service priority of any service according to the congestion control throughput threshold; In the initial stage, the smaller the throughput, the higher the priority. All priorities in a single slice are summed up, and the corresponding priority weight of each service is calculated. The service timeout dynamic control module is used to perform a weighted evaluation of the comprehensive degree of timeout, specifically including: when a slice times out, calculating the timeout coefficient under the timeout condition to characterize the severity of the timeout; If it has not timed out, estimate the possibility that the service will time out, and use the timeout coefficient evaluation when it has not timed out; calculate the weighted value of all service priority weights through the priority weights in the above steps; The newly arrived service access control module is used to dynamically adjust and control the service according to the slice status; specifically, it includes calculating the sum of the weighted values ​​of all the above services in any slice, and representing the decision value of access feasibility; for the newly arrived service, the slice with the smallest decision value is selected for access; if the congestion of the target slice increases after the new service is accessed, the congestion control coefficient of the slice is reduced, otherwise the congestion control coefficient of the slice is increased.

Images