A network slice access control method based on adaptive evaluation of connection density

Abstract

The application provides a network slice access control method based on connection density adaptive evaluation, and relates to the technical field of 5G network slices. The method generates a congestion control value by dynamically comparing a slice congestion rate with a threshold, combines the connection density high and low thresholds, calculates the high and low density components by using a sine / cosine function, and fuses to generate a connection density adaptive value; all slice adaptive values are accumulated to obtain a global sum, and a single slice density slope is calculated, multiplied by a new service specified target density value to output a connection density evaluation value; finally, the slice signal-to-noise ratio is converted into a decimal value, a comprehensive evaluation value is generated by weighting according to a configurable weight and the density evaluation value, the maximum value slice accesses the new service and triggers a closed-loop dynamic adjustment. The beneficial effects are that the congestion response speed is improved, the time delay is reduced by 22%, the resource utilization rate is improved, the signal-to-noise ratio and density dual-dimensional decision optimization make the collision times decrease by 20%, and the closed-loop control ensures that the service continuously stays in the optimal slice environment.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of access control technology of 5G system, and particularly relates to a network slice access control method based on adaptive evaluation of connection density. BACKGROUND

[0002] 5G and later communication networks are no longer limited to solving the communication problem between people, but are aimed at realizing the interconnection of all things. Therefore, the Internet of Things (IoT) business will become the main driving force for the development of future mobile communication networks. Machine-to-Machine (M2M) or Machine-Type Communication (MTC) as the main form of the Internet of Things will have a wide development prospect. With the explosive growth of M2M / MTC demand for wide-area connectivity, traditional wireless access technology has been difficult to adapt to its needs. Massive Machine-Type Communication (mMTC) has emerged and is defined as one of the three general scenarios of 5G. In the mMTC scenario, due to the massive M2M business terminals sending access requests to the base station, it is easy to cause access network overload and signaling congestion, and thus cause the business access delay to increase, the packet data to be severely lost, and even the service to be interrupted.

[0003] In view of the above phenomenon, the skilled in the art has proposed a wireless access control strategy with congestion control function in the M2M business access process. This way can alleviate the access congestion problem by effectively controlling the number of terminals initiating access requests to the base station, and can reduce the access delay through a reward function. But the precondition for implementing this method is that each MTC device must have sufficient bandwidth, and the data aggregation function of the aggregation node must be implemented only after the connection is successfully established, which causes the loss of network utilization. With more and more access, the load of 5G network (including slices) is getting larger and larger, the possibility of business collision is rising sharply, and the decline in network performance is also increasing sharply. Therefore, how to consider the quality of the slice network and select different access control methods to solve the collision problem of the business in the slice has become more and more urgent.

[0004] Chinese patent document CN118555612A discloses a "5G network slice access control method based on connection balance degree", which starts from analyzing sub-slice coverage area, calculates the number of connections in each sub-slice area, and thus calculates the high and low threshold values of service connections; according to the drop rate of each service, the average drop rate of the slice where the service is located is evaluated; according to the connection density, the connection balance degree of the new service before and after connecting each slice is calculated; finally, according to the connection quantity applied by the new service access, the distribution interval is calculated, and dynamic access control is implemented differently; it can ensure that the service is always in a good slice environment, and provides real-time protection for improving customer perception. The above technical solution is innovative in sub-slice coverage area, connection high and low threshold, and connection balance degree calculation, but does not involve dynamic weighting processing of connection density high and low threshold and signal-to-noise ratio weight using trigonometric function; at the same time, the method does not mention dynamic feedback adjustment of closed-loop control mechanism. SUMMARY

[0005] The present application proposes a network slice access control method based on adaptive evaluation of connection density, which is aimed at 5G network slice scenarios and aims to achieve optimal access control of new services through dynamic evaluation of slice status; the core innovation lies in establishing a three-level evaluation system of congestion rate-> connection density-> channel quality, and completing data aggregation and decision generation through a double-loop structure. The present application solves the problems of low resource utilization and congestion response lag caused by static threshold mechanism, realizes dynamic optimization of network slice access, and has significant advantages in real-time performance, resource utilization and anti-congestion capability.

[0006] The purpose of the present application also includes: under the premise of a certain 5G network slice resource, more intuitively realizing the access control of different services through differentiated strategies. Each network slice has an appropriate amount of services, and the newly arrived services are estimated to seek the target slice that meets the resource demand. According to the differentiated services in different scenarios, different access criteria are set to implement dynamic adjustment control of services, guarantee the implementation of different network slice access, and meet the index requirements of different services.

[0007] To achieve the above purpose, the present application proposes a network slice access control method based on adaptive evaluation of connection density, comprising the following steps:

[0008] Calculate the sum of the congestion rates of all services in each network slice, and determine the slice congestion control value in combination with the preset congestion threshold;

[0009] Set the high and low thresholds of connection density, calculate the high and low density values of the slice through trigonometric function processing in combination with the slice congestion control value, and sum to obtain the adaptive value of connection density;

[0010] Calculate the sum of all slice connection density adaptive values, and determine the connection density slope and connection density evaluation value of each slice accordingly;

[0011] The slice SNR is converted into a decimal value, and a preset SNR weight and a connection density evaluation value are combined to calculate an access evaluation comprehensive value, and a slice with the largest evaluation value is selected to access new services;

[0012] After the new service access is completed, a new round of slice state monitoring is triggered to form a closed loop control.

[0013] As preferred, the specific steps of summing to obtain the connection density adaptive value are:

[0014] The slice low density value is calculated based on the low threshold of the connection density, and the slice high density value is calculated based on the high threshold.

[0015] The low density value and the high density value are added to obtain the connection density adaptive value of the slice.

[0016] As preferred, the specific steps of calculating the slice high and low density values are: after the congestion control value is multiplied by two divided by pi, the high and low thresholds of the connection density are respectively processed by a trigonometric function to calculate, the slice low density value is obtained by processing by a cosine function and multiplying the low threshold, and the slice high density value is obtained by processing by a sine function and multiplying the high threshold.

[0017] As preferred, the calculation of the connection density evaluation value specifically includes:

[0018] The connection density adaptive values of all slices are accumulated as a sum;

[0019] The connection density slope of the current slice is determined according to the ratio of the slice connection density adaptive value to the sum, and the connection density evaluation value of the current slice is calculated in combination with the target connection density value.

[0020] As preferred, the target connection density value is the slice connection density value expected to be accessed by the new service at the current time, which is dynamically transmitted when the new service is requested.

[0021] As preferred, the specific steps of determining the slice congestion control value include: for any slice, the sum of the congestion rates of all services thereof is counted, and the slice congestion rate is calculated; if the slice congestion rate is equal to a preset congestion threshold, the congestion control value is set to zero; if it is less than the threshold, the congestion control value is calculated by a difference.

[0022] As preferred, the difference calculation is that the difference between the congestion rate and the congestion threshold is divided by the absolute difference value, and the result is taken as the congestion control value.

[0023] As preferred, the specific steps of calculating the access evaluation comprehensive value include:

[0024] Each slice SNR is converted into a decimal value by a power function.

[0025] The signal-to-noise ratio weight is set, the decimal signal-to-noise ratio and the connection density evaluation value are weighted and calculated by using the preset signal-to-noise ratio weight, and then the access evaluation comprehensive value is obtained.

[0026] The connection density evaluation value of each slice is calculated in combination with the connection density value expected by the new service, the signal-to-noise ratio index is introduced, the access priority of each slice is evaluated by weighting, and finally the optimal slice is selected for service access.

[0027] Preferably, the preset signal-to-noise ratio weight is a configurable parameter, the value range is 0~1, and the proportion of the signal-to-noise ratio and the connection density in decision-making is dynamically adjusted.

[0028] Preferably, the new round of slice state monitoring includes the following closed-loop control process:

[0029] The slice congestion rate, connection density and signal-to-noise ratio data are collected in real time;

[0030] The congestion control value, connection density adaptive value and connection density evaluation value are dynamically updated;

[0031] When the change of the congestion rate of any slice exceeds 10% or the change of the connection density exceeds 15%, the access evaluation comprehensive value is recalculated immediately;

[0032] Based on the recalculated result, the service access dynamic migration is performed, and the services in the over-limit slice are migrated to the slice with a higher comprehensive value according to the priority.

[0033] Compared with the prior art, the present application has the following beneficial effects:

[0034] 1) Congestion response speed-up: the threshold difference triggers the dynamic adjustment of the control value, and the time delay is reduced by 22% (compared with the traditional static threshold);

[0035] 2) Multi-dimensional decision optimization: signal-to-noise ratio (reliability) and connection density (resource efficiency) are weighted and fused, and the collision times are reduced by 20%; the present application effectively avoids service shock and improves resource utilization by using a trigonometric function to smoothly migrate the high and low density thresholds; at the same time, it has closed-loop adaptability, and can update the slice state in real time after service access, significantly reducing the blocking rate in the burst traffic scenario.

[0036] In summary, the application can calculate the congestion rate of the slice where the service is located according to the congestion rate of the service; can calculate the congestion control value and the adaptive value of the connection density of all slices; can calculate the connection density slope and the connection density evaluation value of the slice; can calculate the comprehensive evaluation value of the slice in combination with the signal-to-noise ratio, and on this basis, implement differentiated access to new services; can ensure that the service is always in a good slice environment, and provide real-time protection for improving customer perception. The application solves the three major pain points of poor adaptability of static threshold, incomplete single-dimensional evaluation and lagging response to sudden congestion in the traditional method, and provides a dynamic access framework for mathematical guarantee of high-density 5G scenarios. BRIEF DESCRIPTION OF DRAWINGS

[0037] Figure 1 is a network slice access control method flowchart based on adaptive evaluation of connection density provided by the embodiment 1 of the application.

[0038] Figure 2 is a service access average delay comparison chart of the embodiment 3 of the application and other algorithms.

[0039] Figure 3 is a service collision times comparison chart of the embodiment 3 of the application and other algorithms. DETAILED DESCRIPTION

[0040] In order to make the purpose, technical scheme and advantages of the application more clear and explicit, the technical scheme of the application is further described in detail below through embodiments, and combined with the drawings. It should be understood that the specific implementation scheme described here is only one of the best embodiments of the application, and is only used to explain the technical scheme of the application, and does not limit the protection scope of the application. All other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.

[0041] First, the professional terms involved in the application are explained as follows:

[0042] Link Density: The number of connections or link quality (such as signal strength, interference level) between terminal devices and base stations in a network slice within a unit of time or area, reflecting the real-time load state of the slice. High-density connection may cause channel congestion and increased latency, while low-density may result in insufficient resource utilization.

[0043] Adaptive Evaluation: Real-time collection of connection density data of each slice (such as through base station reporting or terminal measurement); automatic adjustment of access control threshold according to slice type (such as mMTC needs to support massive connections, uRLLC needs low latency) and current network state; optimization of evaluation model based on historical data and service QoS feedback to improve long-term adaptability.

[0044] Network Slice: A new network architecture, a networking-on-demand approach, provides multiple logical networks on the same shared network infrastructure, each serving a specific business type or industry user. Each network slice can flexibly define its own logical topology, SLA requirements, reliability and security levels to meet the differentiated needs of different businesses, industries or users.

[0045] Admission Control: The core goal is to dynamically decide whether to allow new service requests to access the slice while ensuring the QoS (Quality of Service) of the accessed services, avoiding network overload.

[0046] NSAC (Network Slice Admission Control): Network Slice Admission Control, a key mechanism in 5G / 6G network slicing technology, is used to dynamically manage service requests for access, ensuring reasonable allocation of slice resources and guaranteeing service quality (QoS).

[0047] With the explosive growth of M2M / MTC demand for wide-area connectivity, traditional wireless access technology has been difficult to meet its needs. Massive Machine Communication (mMTC) has emerged and is defined as one of the three general scenarios for 5G. In the mMTC scenario, due to the massive M2M service terminals sending access requests to the base station, it is easy to cause access network overload and signaling congestion, which in turn leads to increased service access latency, severe packet data loss, and even service interruption.

[0048] Therefore, some people in the field have proposed a wireless access control strategy with congestion control function in the M2M service access process (from He Duoyi, 5G Scenario Wireless Access Control Strategy for M2M Massive Connection, Guangdong Communications Technology [J], 2022 (1): 22-29), hereinafter referred to as Paper Hedy. This approach effectively controls the number of terminals initiating access requests to the base station to alleviate access congestion, and uses a reward function to reduce access latency. However, the prerequisite for implementing this method is that each MTC device must have sufficient bandwidth, and the data aggregation function of the aggregation node is only implemented after a successful connection is established, resulting in a loss of network utilization.

[0049] As more and more access, the load of 5G network (including slices) becomes larger, the possibility of service collision increases significantly, and the magnitude of network performance degradation also increases dramatically. Therefore, how to consider the quality of the slice network and choose different access control methods to solve the collision problem of services within the slice has become increasingly urgent.

[0050] To this end, the application designs a 5G network slice access control method LAE-NSAC (Link density Adaptive Evaluation based 5G Network Slice Admission Control algorithm) based on link density adaptive evaluation. First, the slice congestion control value is calculated by analyzing and calculating the slice congestion rate, the link density adaptive value of each slice is calculated according to the target threshold requirement of the slice link density, and the corresponding link density slope is calculated on this basis. According to the expected value of the link density of the newly arrived service, the link density evaluation value of any slice is calculated. The signal-to-noise ratio and the link density are weighted in combination with the signal-to-noise ratio index of each slice, and the final access evaluation value is calculated. Then, the dynamic adjustment control of the new service access is implemented according to the maximum value principle.

[0051] Embodiment 1:

[0052] The application provides a network slice access control method based on link density adaptive evaluation, which specifically comprises the following steps:

[0053] S1: calculate the sum of the congestion rates of all services in each network slice, and determine the slice congestion control value in combination with the preset congestion threshold;

[0054] S2: set the high and low thresholds of the link density, calculate the high and low density values of the slice by triangular function processing in combination with the slice congestion control value, and sum to obtain the link density adaptive value;

[0055] S3: calculate the sum of all slice link density adaptive values, and determine the link density slope and the link density evaluation value of each slice accordingly;

[0056] S4: convert the slice signal-to-noise ratio to a decimal value, weight the access evaluation comprehensive value by combining the preset signal-to-noise ratio weight and the link density evaluation value, and select the slice with the maximum evaluation value to access the new service;

[0057] S5: trigger a new round of slice state monitoring after the new service access is completed, and form a closed loop control.

[0058] The above step S1 mainly summarizes the calculation of the slice congestion control value. The specific steps for determining the slice congestion control value include: for any slice, the sum of the congestion rates of all services is counted, and the slice congestion rate is calculated; if the slice congestion rate is equal to the preset congestion threshold, the congestion control value is set to zero; if it is less than the threshold, the congestion control value is calculated by difference.

[0059] The difference is calculated as the difference between the congestion rate and the congestion threshold divided by the absolute difference, and the result is taken as the congestion control value.

[0060] The step S2 mainly summarizes the calculation of the slice connection density adaptive value, and the specific steps of the summation to obtain the connection density adaptive value are as follows:

[0061] The slice low density value is calculated based on the low threshold of the connection density, and the slice high density value is calculated based on the high threshold.

[0062] The low density value and the high density value are added to obtain the connection density adaptive value of the slice.

[0063] The specific steps of calculating the slice high and low density values are as follows: after the congestion control value is multiplied by two-thirds of the circumference, the connection density high and low thresholds are respectively processed by a trigonometric function to calculate the slice low density value by cosine function processing and multiply the low threshold, and the slice high density value is obtained by sine function processing and multiply the high threshold.

[0064] The step S3 mainly summarizes the calculation of the slice connection density evaluation value, and the calculation of the connection density evaluation value specifically includes the following steps:

[0065] The connection density adaptive values of all slices are accumulated as a sum;

[0066] The connection density slope of the current slice is determined according to the ratio of the slice connection density adaptive value to the sum, and the connection density evaluation value of the current slice is calculated in combination with the target connection density value.

[0067] The target connection density value is the connection density value expected to be accessed by the new service at the current time, which is dynamically transmitted when the new service is requested.

[0068] The step S4 mainly summarizes the calculation of the slice access evaluation comprehensive value, and the specific steps of the access evaluation comprehensive value calculation include the following steps:

[0069] The signal-to-noise ratio of each slice is converted to a decimal value by a power function;

[0070] The signal-to-noise ratio weight is set, and the decimal signal-to-noise ratio and the connection density evaluation value are weighted and calculated by using the preset signal-to-noise ratio weight, so as to obtain the access evaluation comprehensive value.

[0071] In combination with the expected connection density value of the new service, the connection density evaluation value of each slice is calculated; at the same time, the signal-to-noise ratio index is introduced, the access priority of each slice is evaluated by weighting, and the optimal slice is finally selected for service access.

[0072] The preset signal-to-noise ratio weight is a configurable parameter, and the value range is 0-1, which is used to dynamically adjust the proportion of signal-to-noise ratio and connection density in decision-making.

[0073] The step S5 mainly summarizes the new round of closed-loop control mechanism of the network slice access control method, and the new round of slice state monitoring includes the following closed-loop control process:

[0074] Real-time acquisition of slice congestion rate, connection density and signal-to-noise ratio data;

[0075] Dynamic updating of congestion control value, connection density adaptive value and connection density evaluation value;

[0076] When the change of congestion rate of any slice exceeds 10% or the change of connection density exceeds 15%, the recalculation of the access evaluation comprehensive value is triggered immediately;

[0077] Based on the recalculation result, the dynamic migration of business access is performed, and the business in the over-limit slice is migrated to the slice with higher comprehensive value according to the priority.

[0078] Reference Figure 1 The flowchart systematically describes the network slice access control method based on connection density adaptive evaluation proposed by the application, and the whole flow contains 6 sequentially executed stages: system initialization stage, slice internal business congestion index aggregation stage (inner loop), slice level congestion control value generation stage, connection density adaptive value calculation and global accumulation stage (outer loop), connection density evaluation and channel fusion stage, and access decision execution stage. The complete closed loop from single business index collection to global access decision is covered, and each stage will be described in detail below.

[0079] System initialization stage,

[0080] After starting the control engine, reset all historical state variables, clear the temporary storage area, and prepare to receive real-time network data; initialize the slice traversal pointer, set the initial value of the slice counter to 1, and identify the first network slice (Sec 1) for processing. Each execution is an independent decision cycle to avoid residual data interference and ensure the timeliness of the evaluation results.

[0081] Business level congestion index aggregation (inner loop) stage,

[0082] The first step is to initialize the business pointer, set the initial value of the internal business counter of the current slice Sec i to 1 (pointing to the first business Srv (i, 1)), and then the second step is to perform the congestion rate accumulation operation, read the real-time congestion rate value of the business Srv (i, j), and accumulate it to the congestion rate total register of the slice; the third step checks whether the processing of all businesses in the current slice has been completed, if not, the business counter is incremented by 1, and the next business is returned to the second step for processing, if it has been completed, the slice level calculation stage is entered. The congestion state of each business is collected independently to avoid the overall average covering up local exceptions; the processing time is only positively related to the number of businesses, which is suitable for hardware parallel acceleration.

[0083] Slice level congestion control value generation stage,

[0084] The first step is to calculate the slice congestion rate, which divides the accumulated total congestion rate by the total number of services in the current slice to obtain the average congestion rate of the slice; the second step is to perform dynamic decision of congestion threshold (core adaptive mechanism), which compares the slice congestion rate with the preset congestion threshold: equal to the threshold value, immediately set the congestion control value to zero, trigger preventive control to prevent congestion deterioration trend; lower than the threshold value, calculate the proportion of the difference between the congestion rate and the threshold to the absolute difference, generate a negative control value according to the difference degree, and take this proportion as the congestion control value to provide an adjustment parameter for subsequent density calculation.

[0085] Connection density adaptive value calculation and accumulation (outer loop),

[0086] The first step is to perform high-low threshold fusion calculation, and the congestion control value is used as an input parameter to make the connection density evaluation reflect the network congestion degree in real time; the stepless switching of the low threshold and the high threshold is realized through the sine / cosine function to avoid service shock caused by threshold jump. Among them, the low density component of the slice is generated by using the product of pi and the congestion control value, which is transformed by the cosine function and then multiplied by the connection density low threshold value; the high density component of the slice is generated by using the product of pi and the congestion control value, which is transformed by the sine function and then multiplied by the connection density high threshold value. Based on this, the low density component and the high density component are added to obtain the connection density adaptive value of the current slice;

[0087] The second step is global accumulation update, which adds the connection density adaptive value of the current slice to the global connection density adaptive value accumulator; the third step is to traverse and judge all network slices to check whether all processing is completed: if not, the slice counter is incremented by 1, and the next slice is returned to the service level congestion index aggregation (inner loop) stage processing; if it is completed, enter the global evaluation stage.

[0088] Connection density evaluation and channel fusion stage,

[0089] The first step is to calculate the connection density evaluation value, which divides the connection density adaptive value of the above single slice by the global accumulation value to obtain a normalized density influence coefficient; multiplying this coefficient by the target connection density value of the new service can output the slice connection density evaluation value; the second step is to perform channel quality fusion, which first converts the decibel value of the slice signal-to-noise ratio into a decimal value (note: the conversion rule is 10 power operation); then perform two-dimensional weighted synthesis: weight the linear signal-to-noise ratio according to the preset signal-to-noise ratio weight, and weight the connection density evaluation value according to the complementary weight, and add them to generate the comprehensive evaluation value of the slice.

[0090] The comprehensive evaluation system is designed for scientific evaluation: the normalization eliminates the scale deviation, the density influence coefficient ensures the comparability of different scale slices, the signal-to-noise ratio conversion conforms to the 3GPP signal processing specification, and the weight parameter allows the operator to flexibly configure the strategy between transmission reliability (signal-to-noise ratio) and resource utilization (connection density).

[0091] The access decision execution stage,

[0092] The first step is to traverse all the comprehensive evaluation values of the slices, select the slice with the maximum value as the target access position; the second step is to direct the new service request to the selected slice, update the service set and state parameters of the slice; the third step is to record the log and generate the access operation report when the process terminates, and wait for the next service to trigger the process start.

[0093] The advantages of the decision mechanism are: the traversal comparison avoids the local optimal trap, ensures the global optimality, and guarantees the real-time closed-loop control.

[0094] The technical advantages of the method of the application are embodied in the flowchart as follows:

[0095] 1) Dynamic congestion prevention: when the congestion rate of a slice reaches the threshold (such as 3%), the congestion control value of the slice is set to zero immediately; this operation causes the following connection density calculation:

[0096] Low density component = cosine function result x low threshold → since the control value is zero, the cosine output is 1, and the component is equal to the low threshold value;

[0097] High density component = sine function result x high threshold → the sine output is zero, and the component is zero;

[0098] The final effect is that the adaptive connection density value is suppressed to the low threshold level, limiting the access of new services and preventing congestion from worsening.

[0099] 2) Resource utilization optimization: assuming that the congestion rate of slice A is 2.5% (threshold 3%), the control value is a negative proportion; the congestion rate of slice B is 1% (threshold 3%), and the control value is a larger negative proportion. Then in the connection density calculation:

[0100] Slice A: control value is a small negative value → cosine output is close to 1, and sine output is close to 0 → adaptive value is slightly higher than the low threshold;

[0101] Slice B: control value is a large negative value → cosine output decreases and sine output increases → adaptive value is close to the high threshold;

[0102] The final effect is that the low-congestion slice obtains a higher connection density capacity, guiding the service to shunt to the idle area, and improving the overall resource utilization.

[0103] 3) Multi-objective collaborative decision-making, such as slice X: signal-to-noise ratio -1 dB (linear value 0.79), connection density evaluation value 80; slice Y: signal-to-noise ratio 2 dB (linear value 1.58), connection density evaluation value 60; assuming that the signal-to-noise ratio weight is 40%, then:

[0104] Slice X comprehensive value = 40% x 0.79 + 60% x 80 = 48.316;

[0105] Slice Y comprehensive value = 40% x 1.58 + 60% x 60 = 36.632;

[0106] According to the principle of higher density evaluation value, the decision result is finally to select slice X access, which reflects the congestion control priority principle.

[0107] The above process also has a guarantee mechanism to achieve high efficiency: business-level congestion rate accumulation supports distributed parallel processing, shortening the processing time delay by 60%; the results of sine / cosine functions are pre-stored in a lookup table to avoid real-time calculation overhead. When there is no service in the slice, automatically skip the congestion calculation and assign a default density value; when the congestion rate exceeds the threshold, forcibly start the independent service migration process. The newly added slice automatically expands the loop processing scale, and resets the global accumulation benchmark when the slice is removed; the signal-to-noise ratio weight supports dynamic online adjustment according to the network load state.

[0108] This process realizes four breakthroughs in 5G network slice access control through three core designs of business-slice double-loop, congestion-density coupled control, and multi-objective weighted evaluation: dynamic congestion threshold decision mechanism improves congestion suppression speed, connection density adaptive algorithm improves slice resource utilization, two-dimensional evaluation improves the probability of high-value business access to high-quality slices, and weight parameter adjustment reduces the frequency of manual intervention. Figure 1 The process design fully meets the differentiated needs of ITU-defined 5G three scenarios (eMBB, uRLLC, mMTC), providing an evolutionary technical foundation for 6G network intelligent slice architecture.

[0109] Embodiment 2:

[0110] This embodiment 2 takes the case where the number of network slices is equal to 3 as an example to specifically illustrate the method of the present application. The 5G network slice conditions are shown in Table 1, and the basic data are shown in Table 2:

[0111] Table 1: Services stored in each 5G network slice

[0112]

[0113] In Table 1, Sec i represents the network slice number, i∈[1,3], and there are services Srv (i,j) in slice Sec i, j∈[1,n]. In Table 1, the number of network slices is 3, which are Sec 1, Sec 2 and Sec 3, wherein there are services Srv (1,1)~Srv (1,4) in slice Sec 1, services Srv (2,1) and Srv (2,2) in slice Sec 2, and services Srv (3,1)~Srv (3,3) in slice Sec 3.

[0114] The 5G network slice access control method based on adaptive evaluation of connection density described in Example 2 includes the following steps: slice congestion index calculation, connection density adaptive calibration, connection density evaluation value determination, and new service implementation access:

[0115] The slice congestion index calculation step calculates the congestion rate and value of all services Srv (i,j) in any slice Sec i, obtaining {10.5%, 5.9%, 9.8%}. On this basis, the congestion rate of slice Sec i is calculated as {2.75%, 2.95%, 3.27%}. All slices Sec i do not meet the preset congestion threshold condition (3%) in Table 2 below, and the congestion control value at this time is calculated as {1, 1, 0};

[0116] The connection density adaptive calibration step calculates the low density value of any slice Sec i as {0, 0, 80} (ten thousand / km 2 ) and the high density value as {150, 150, 0} (ten thousand / km 2 ). The sum of the high / low density values of slice Sec i is calculated to obtain the connection density adaptive value as {150, 150, 80} (ten thousand / km 2 );

[0117] The connection density evaluation value determination step calculates the sum of the connection density adaptive values of all slices, obtaining 380 (ten thousand / km 2 ). The connection density slope of slice Sec i is calculated as {0.39, 0.39, 0.21}, and on this basis, the connection density evaluation value of slice Sec i is calculated as {47.37, 47.37, 25.26} (ten thousand / km 2 );

[0118] The new service implementation access step converts the signal-to-noise ratio of each slice Sec i to a decimal value {0.63, 1, 1.26}. The access evaluation comprehensive value of slice Sec i is calculated as {28.67, 28.82, 15.66} (ten thousand / km 2 ). Finally, the slice Sec 2 with the largest access evaluation comprehensive value is selected for access.

[0119] Example 3:

[0120] This example 3 describes a comparison experiment between the 5G network slice access control method based on connection density adaptive evaluation proposed by the application and the access control algorithm based on congestion control used in the paper Hedy, which is simulated on the MATLAB platform. In the 5G network slice access control method of the application, the connection adaptive evaluation control algorithm of LAE-NSAC is used. The comparison experiment configures the network and service according to the basic data in table 2, and the specific data of table 2 is as follows:

[0121] Table 2 basic data

[0122]

[0123] In order to compare the superior performance of the algorithm proposed in this paper, the comparison of the average access delay and the average collision times of MTC devices is used to illustrate the feasibility of the algorithm in this paper. The comparison of the average access delay and the collision times of the service is shown in Figure 2 and Figure 3 .

[0124] As shown in Figure 2 , with the continuous increase of the number of devices, the average delay also increases. Compared with the access control algorithm based on congestion control used in the paper Hedy, the LAE-NSAC algorithm of the application is always lower than the paper Hedy in access delay, and its average delay growth rate is smaller, and the average delay is reduced by about 20ms. The mechanism is that LAE-NSAC will adaptively match the comprehensive connection density evaluation value according to the personalized congestion threshold, combined with the different congestion degree of slice, and select the most suitable target slice to implement access; in the same connection density, it will select the slice with the best signal-to-noise environment for access, avoiding the problem of limiting the number of M2M access devices in the paper Hedy. Furthermore, the matching of connection density of the LAE-NSAC algorithm used in the application is linear and adaptive, and there is basically no jitter on the index evaluation, which is more flexible in the curve and can save system resources. The same point of the two is that with the increase of the number of access services, the average access delay will continue to rise.

[0125] The collision times reflect the congestion degree of the base station, which reflects the difficulty of successful access of the base station. Business collision will cause network congestion, and in severe cases, it will cause packet loss and even service interruption, which will greatly affect customer perception. As Figure 3As shown, with the increasing number of device accesses, the collision times are also increasing, compared with the access control algorithm based on congestion control adopted in the paper Hedy, the collision times of the LAE-NSAC algorithm of the application are always lower than those of the paper Hedy, and the average collision times are smaller. The reason is that the LAE-NSAC algorithm of the application directly analyzes the congestion result, and then adaptively evaluates in combination with the threshold value of the connection density, according to the signal-to-noise environment of the slice, and comprehensively calculates the target slice with the minimum collision probability; while the collision avoidance mechanism of the paper Hedy is relatively simple, that is, to reduce the number of accessed services. Therefore, when the access service increases sharply, the collision suppression ability will be weakened and lag behind; reflected in the curve, the collision times of LAE-NSAC are always lower than those of the paper Hedy when the same number of access services.

[0126] In summary, the application proposes a network slice access control method (LAE-NSAC) based on adaptive evaluation of connection density, which relates to the field of 5G network slicing technology. The LAE-NSAC algorithm introduces connection density as a key evaluation index, optimizes slice resource utilization and service experience by adaptively adjusting access strategy. The method compares the slice congestion rate with the threshold dynamically to generate a congestion control value (set to zero when the congestion rate is equal to the threshold, and calculate according to the difference ratio when the congestion rate is less than the threshold), combines with the high and low thresholds of connection density, and uses sine / cosine function to calculate high and low density components and fuse to generate adaptive connection density value; accumulate all slice adaptive values to get global sum, and then calculate single slice density slope, multiply by new service specified target density value to output connection density evaluation value; finally, convert the slice signal-to-noise ratio to decimal value, weight the comprehensive evaluation value according to the configurable weight and density evaluation value, select the maximum value slice to access new service and trigger closed-loop dynamic adjustment. Its beneficial effects are: the congestion response speed is improved, the time delay is reduced by 22%, the resource utilization is improved, the signal-to-noise ratio and density dual-dimensional decision optimization make the collision times decrease by 20%, and the closed-loop control ensures that the service is continuously in the optimal slice environment.

[0127] The above is only the preferred embodiment of the application, but the protection scope of the application is not limited thereto, it should be pointed out that for ordinary skilled in the art, any easily thought change or replacement without departing from the technical principles of the application should be covered in the protection scope of the application.

Claims

1. A network slice access control method based on adaptive evaluation of connection density, characterized in that, The method comprises the following steps: calculating the total congestion rate of all services in each network slice, and determining the slice congestion control value in combination with a preset congestion threshold; setting a high and low connection density threshold, and calculating the high and low density values of the slice through trigonometric function processing in combination with the slice congestion control value, and summing to obtain the connection density adaptive value; calculating the total sum of the connection density adaptive values of all slices, and determining the connection density slope and the connection density evaluation value of each slice accordingly; converting the slice signal-to-noise ratio into a decimal value, and calculating the access evaluation comprehensive value by weighting the preset signal-to-noise ratio weight and the connection density evaluation value, and selecting the slice with the maximum evaluation value to access the new service; triggering a new round of slice state monitoring after the new service access is completed to form a closed-loop control. 2.The network slice access control method based on adaptive evaluation of connection density according to claim 1, characterized in that, The specific steps for obtaining the connection density adaptive value are as follows: calculating the slice low density value based on the low connection density threshold, and calculating the slice high density value based on the high threshold; adding the low density value and the high density value to obtain the connection density adaptive value of the slice. 3.The network slice access control method based on adaptive evaluation of connection density according to claim 1 or 2, characterized in that, The specific steps for calculating the high and low density values of the slice are as follows: after multiplying the congestion control value by two-thirds of the circumference, the product of two-thirds of the circumference multiplied by the congestion control value is processed through trigonometric function processing, the slice low density value is obtained through cosine function processing multiplied by the low threshold, and the slice high density value is obtained through sine function processing multiplied by the high threshold.

4. The network slice access control method based on adaptive evaluation of connection density according to claim 1, characterized in that, The calculation of the connection density evaluation value specifically includes: accumulating the connection density adaptive values of all slices as the total sum; determining the connection density slope of the current slice according to the ratio of the connection density adaptive value of the slice to the total sum, and then calculating the connection density evaluation value of the current slice in combination with the target connection density value.

5. The network slice access control method based on adaptive evaluation of connection density according to claim 4, characterized in that, The target connection density value is the connection density value of the slice expected to be accessed by the new service at the current time, which is dynamically transmitted when the new service is requested.

6. The network slice access control method based on adaptive evaluation of connection density according to claim 1, characterized in that, The specific steps for determining the slice congestion control value include: for any slice, the sum of the congestion rates of all its services is counted, and the slice congestion rate is calculated; if the slice congestion rate is equal to the preset congestion threshold, the congestion control value is set to zero; if it is less than the threshold, the congestion control value is calculated by the difference.

7. The network slice access control method based on connection density adaptive evaluation according to claim 1, characterized in that, The specific steps for calculating the access evaluation comprehensive value include: convert the signal-to-noise ratio of each slice into a decimal value through a power function; set the signal-to-noise ratio weight, and calculate the access evaluation comprehensive value by weighting the decimal signal-to-noise ratio and the connection density evaluation value using the preset signal-to-noise ratio weight. 8.The network slice access control method based on adaptive evaluation of connection density according to claim 7, characterized in that, The preset signal-to-noise ratio weight is a configurable parameter, with a value range of 0~1, and is used to dynamically adjust the proportion of signal-to-noise ratio and connection density in decision-making. 9.The network slice access control method based on adaptive evaluation of connection density according to claim 1, characterized in that, The new round of slice state monitoring includes the following closed-loop control process: real-time acquisition of slice congestion rate, connection density and signal-to-noise ratio data; dynamic updating of congestion control value, connection density adaptive value and connection density evaluation value; when the change of the congestion rate of any slice exceeds 10% or the change of the connection density exceeds 15%, the access evaluation comprehensive value is recalculated immediately; based on the recalculated result, perform dynamic migration of service access, and migrate the services in the out-of-limit slice to the slice with a higher comprehensive value according to the priority.

Images