Method for allocating resources among 5G network slices based on extensible adjustment

Abstract

The invention discloses a 5G network inter-slice resource allocation method based on extensible adjustment. The problems that existing slice resource allocation is complex, the safety is low and the management efficiency is low are solved. The method comprises the steps of determining slice high-quality and non-high-quality channel services based on a signal-to-noise ratio; calculating a signal-to-noise ratio coefficient for the high-quality channel service according to a signal-to-noise ratio, calculating a synthesis coefficient for the non-high-quality channel service according to throughput and time delay, and calculating a slice service performance coefficient; determining a slice physical resource occupancy rate, and allocating a corresponding resource occupancy coefficient as a slice resource coefficient according to the physical resource occupancy rate; calculating a slice service quality coefficient according to the congestion rate; and determining a slice allocation priority according to the slice service performance coefficient, the resource coefficient and the service quality coefficient, and allocating the unallocated resources. According to the method, the distribution priority is evaluated from multiple aspects of service performance, resources and quality, the service is ensured to be in a good slice environment, and a real-time guarantee is provided for improving customer perception.

Description

Technical Field

[0001] The present invention relates to the field of 5G communication technologies, and particularly to a method for resource allocation between 5G network slices based on scalable adjustment. Background Art

[0002] Network slicing refers to dividing a piece of physical resources into multiple logically independent slices, and each slice can be configured according to different requirements and performance requirements to provide personalized network services. The resource allocation strategy between network slices can be based on application requirements, network status, user experience, and historical data prediction. However, due to reasons such as the complexity, security, and management efficiency of slice resource allocation, up to now, how to implement a simple and efficient method for resource allocation between slices remains an unsolved problem.

[0003] For example, in the invention application with the patent application number 202411435611.9 and the patent name "A Method for Resource Allocation between Network Slices Based on the Idle Coefficient", by using the number of idle resources and the total resources of each slice, and combining the congestion status of each service within the slice, the idle coefficient of each slice is calculated, and by combining the load priority and the signal priority, a comprehensive priority that can evaluate the resource allocation of each slice is obtained. This patent can effectively reduce the congestion rate of slices, but there are also some problems: firstly, the slice marking type is fixed and cannot account for mixed-type services; secondly, the signal quality level index is not refined according to the specific service type and level value, resulting in inaccurate calculation of the final comprehensive priority and affecting resource allocation. Summary of the Invention

[0004] The present invention mainly solves the problems of complex existing slice resource allocation, low security, and low management efficiency, and provides a method for resource allocation between 5G network slices based on scalable adjustment.

[0005] The present invention also solves the problem that the existing resource allocation method cannot account for mixed-type services, the signal quality level signal is not refined according to the specific service type and level value, and there is inaccurate priority calculation, which affects resource allocation, and provides a method for resource allocation between 5G network slices based on scalable adjustment.

[0006] The above technical problems of the present invention are mainly solved by the following technical solutions: A method for resource allocation between 5G network slices based on scalable adjustment, including the following steps: Determine high-quality and non-high-quality channel services for slices based on the signal-to-noise ratio; Calculate the signal-to-noise ratio coefficient for high-quality channel services according to the signal-to-noise ratio, calculate the composite coefficient for non-high-quality channel services according to the throughput and delay, and calculate the slice service performance coefficient based on the signal-to-noise ratio coefficient and the composite coefficient; Determine the physical resource occupancy rate of the slice, and allocate the corresponding resource occupancy coefficient according to the size of the physical resource occupancy rate, which is the slice resource coefficient; Calculate the slice service quality coefficient according to the congestion rate; Determine the slice allocation priority according to the slice service performance coefficient, resource coefficient, and service quality coefficient; Allocate the unallocated resources according to the allocation priority.

[0007] The present invention starts from screening high-quality channel services, calculates the signal-to-noise ratio coefficient of the calculator, and simultaneously calculates the throughput coefficient and delay coefficient of non-high-quality channel services. On this basis, the performance coefficients of all slices are measured. Based on the physical resource occupancy status, the resource coefficients of all slices are calculated. Based on the service congestion status, the service quality coefficients of all slices are calculated. A synthetic and scalable adjustable comprehensive allocation priority is synthesized, and the slices that were not allocated in the previous round are dynamically adjusted. Through differentiated strategies, the resource allocation between slices can be more intuitively realized. The present invention can ensure that the service is always in a slice environment with good conditions, providing real-time guarantee for improving customer perception.

[0008] As a preferred solution, set a signal-to-noise ratio threshold, obtain the services with a service signal-to-noise ratio greater than the signal-to-noise ratio threshold as high-quality channel services, and obtain the services with a service signal-to-noise ratio not greater than the signal-to-noise ratio threshold as non-high-quality channel services.

[0009] This solution screens high-quality channel services, sets a signal-to-noise ratio threshold, and divides the slice services into high-quality channel services and non-high-quality channel services according to the service signal-to-noise ratio with the signal-to-noise ratio threshold as the boundary. The signal-to-noise ratio threshold divides the service signal-to-noise ratio to determine high-quality channel services and non-high-quality channel services.

[0010] As a preferred solution, calculate the signal-to-noise ratio coefficient of each high-quality channel service according to the service signal-to-noise ratio, and calculate the sum of all signal-to-noise ratio coefficients as the slice signal-to-noise ratio coefficient.

[0011] In this solution, for each high-quality channel service, calculate the signal-to-noise ratio coefficient according to its signal-to-noise ratio, specifically the reciprocal of the value obtained by adding 1 to the exponential function of the negative decimal signal-to-noise ratio of the high-quality channel service. Add up the signal-to-noise ratio coefficients of all high-quality channel services, and the obtained sum is the slice service signal-to-noise ratio coefficient. The decimal signal-to-noise ratio is the result of performing decimal conversion calculation on the signal-to-noise ratio.

[0012] As a preferred solution, calculate the throughput coefficient of each non-high-quality channel service according to the service throughput, calculate the delay coefficient of each non-high-quality channel service according to the delay, calculate the synthetic coefficient according to the throughput coefficient and the delay coefficient, and calculate the sum of all synthetic coefficients as the slice synthetic coefficient.

[0013] For each non-high-quality channel service in this solution, the throughput coefficient is calculated based on the throughput, and the delay coefficient is calculated based on the delay. The throughput here is the average throughput, and the delay is the average delay. Specifically, the throughput coefficient is the ratio of the throughput to the ideal throughput, and the delay coefficient is the exponential function of the negative of the ratio of the delay to the delay tolerance. Further, the composite coefficient is calculated based on the throughput coefficient and the delay coefficient. The composite coefficient of each non-high-quality channel service is specifically the product of the throughput coefficient and the delay coefficient. The sum of the composite coefficients of all non-high-quality channel services is obtained as the slice service composite coefficient.

[0014] As a preferred solution, the service performance coefficient is calculated based on the signal-to-noise ratio coefficient and the composite coefficient, including: The ratio of the sum of the slice service signal-to-noise ratio and the slice composite coefficient to the number of slice services.

[0015] For any slice, the service performance coefficient of the slice is calculated as the quotient of the sum of the slice service signal-to-noise ratio and the slice composite coefficient divided by the number of slice services.

[0016] As a preferred solution, a physical resource occupancy threshold is set, and several physical resource occupancy ranges are divided; Corresponding resource occupancy coefficients are set for each physical resource occupancy range respectively; Judge the range to which the slice physical resource occupancy belongs, and obtain the corresponding resource occupancy coefficient as the slice resource coefficient.

[0017] According to the physical resource occupancy threshold and the set points, several physical resource occupancy ranges are divided, and corresponding resource occupancy coefficients are set for each physical resource occupancy range. The slice physical resource occupancy is the quotient of the total number of physical resource blocks occupied by all services in the slice and the total number of physical resource blocks allocated to the slice, that is, the ratio of the total number of physical resource blocks occupied by services in the slice to the total number of physical resource blocks already available in the slice. Judge the physical resource block occupancy range in which the slice physical resource occupancy falls, and use the resource occupancy coefficient corresponding to this range as the slice resource coefficient of this slice.

[0018] Set a physical resource occupancy threshold and a point. The physical resource occupancy range is preferably divided into four ranges. At the same time, high, medium, and low resource occupancy coefficients and their specific values are set, and then combined with 1, and paired with each physical resource occupancy range respectively.

[0019] As a preferred solution, a service congestion threshold is set, and congested services and non-congested services are determined based on the service congestion rate; Determine the service quality coefficient of the congested services according to the service congestion threshold and the congestion rate, and set the service quality coefficient of the non-congested services; Determine the slice service quality coefficient according to the service quality coefficients of all services in the slice.

[0020] By setting the service congestion threshold, services with a service congestion rate not less than the service congestion threshold are obtained as congested services, and services with a service congestion rate less than the service congestion threshold are obtained as non-congested services. The service quality coefficient of a congested service is the ratio of the service congestion threshold to the service congestion rate, and the service quality coefficient of a non-congested service is set, preferably set to 1. The service quality coefficient is jointly determined according to the service quality coefficients of all services of the slice.

[0021] As a preferred solution, it is the ratio of the sum of the service quality coefficients of all services of the slice to the number of slice services.

[0022] For any slice, add the service quality coefficients of each of its services, and then divide by the number of all services of the slice. The value obtained is the slice service quality coefficient.

[0023] As a preferred solution, the slice allocation priority is the product of the slice service performance coefficient, the slice resource coefficient, and the slice service quality coefficient; Set an adjustment coefficient, and dynamically adjust the slice priority that was not resource-allocated in the previous round to the final allocation priority. The final allocation priority is the product of the adjustment coefficient and the slice allocation priority.

[0024] The allocation priority of all slices is the product of the slice service performance coefficient, the resource coefficient, and the service quality coefficient. This allocation priority is extensible, and the corresponding adjustment coefficient can be increased as appropriate to make the measurement of the priority more reasonable. Set an adjustment coefficient to dynamically adjust the allocation priority of the slices that were not resource-allocated in the previous round, and use the product of the adjustment coefficient and the allocation priority as the final allocation priority for the current resource allocation.

[0025] The set adjustment coefficient range is (1, 2], and the adjustment coefficient is increased to dynamically adjust the priority of the slices that were not allocated in the previous round.

[0026] As a preferred solution, according to the allocation priority, use the PF algorithm to allocate the unallocated resources and update the total number of physical resource blocks of all slices.

[0027] Therefore, the advantages of the present invention are as follows: starting from screening high-quality channel services, calculating the signal-to-noise ratio coefficient of the calculator, and at the same time calculating the throughput coefficient and delay coefficient of non-high-quality channel services. On this basis, measure the performance coefficients of all slices, calculate the resource coefficients of all slices based on the physical resource occupancy status, calculate the service quality coefficients of all slices based on the service congestion status, synthesize an extensible and adjustable comprehensive allocation priority, dynamically adjust the slices that were not allocated in the previous round, and can more intuitively achieve resource allocation between slices through a differentiated strategy. The present invention can ensure that services are always in a slice environment with good conditions, providing real-time guarantee for improving customer perception. Description of the Drawings

[0028] Figure 1 is a flowchart of a process of the present invention.

[0029] Figure 2 is a comparison chart of RB allocation of random TTI between the algorithm of the present invention and other algorithms.

[0030] Figure 3 is a comparison chart of average RB allocation between the algorithm of the present invention and other algorithms. Detailed Implementation Manner

[0031] The technical solution of the present invention will be further specifically described below through embodiments in conjunction with the drawings.

[0032] Embodiment 1:

[0033] A resource allocation method between 5G network slices based on scalable adjustment in this embodiment is as Figure 1 shown, and includes the following steps: S1. Determine slice high-quality and non-high-quality channel services based on signal-to-noise ratio; Calculate the slice signal-to-noise ratio coefficient for high-quality channel services according to the signal-to-noise ratio, calculate the slice synthesis coefficient for non-high-quality channel services according to throughput and delay, and calculate the slice service performance coefficient based on the service signal-to-noise ratio coefficient and synthesis coefficient.

[0034] This step is used to calculate the performance coefficient of slice services by determining high-quality channel services and according to the signal-to-noise ratio, throughput, and delay of the services.

[0035] S11. Set a signal-to-noise ratio threshold, obtain services with a service signal-to-noise ratio greater than the signal-to-noise ratio threshold as high-quality channel services, and obtain services with a service signal-to-noise ratio not greater than the signal-to-noise ratio threshold as non-high-quality channel services. By setting the signal-to-noise ratio threshold, the signal-to-noise ratio threshold divides the service signal-to-noise ratio to determine high-quality channel services and non-high-quality channel services. Specifically, the slice services are divided into high-quality channel services and non-high-quality channel services according to the service signal-to-noise ratio with the signal-to-noise ratio threshold as the boundary.

[0036] S12. For high-quality channel services, convert the signal-to-noise ratio of all services into a decimal signal-to-noise ratio, calculate the service signal-to-noise ratio coefficient according to the decimal signal-to-noise ratio, and calculate the sum of the signal-to-noise ratio coefficients of all high-quality channel services as the service signal-to-noise ratio coefficient.

[0037] Specifically, the signal-to-noise ratio coefficient is the reciprocal of the value obtained by adding 1 to the exponential function of the negative decimal signal-to-noise ratio of high-quality channel services. The decimal signal-to-noise ratio conversion calculation is the power of 10 to the ratio of the service signal-to-noise ratio to 10, expressed as power(10, signal-to-noise ratio / 10) in a power function.

[0038] S13. For non-high-quality channel services, calculate the throughput coefficient for each non-high-quality channel service according to the service throughput, calculate the delay coefficient for each non-high-quality channel service according to the delay, calculate the composite coefficient according to the throughput coefficient and the delay coefficient, and calculate the sum of all composite coefficients as the slice composite coefficient.

[0039] Specifically, the throughput coefficient is the ratio of the service throughput to the ideal throughput, and the throughput is the average throughput. The delay coefficient is the exponential function of the negative of the ratio of the delay to the delay tolerance, and the delay is the average delay.

[0040] The composite coefficient of each non-high-quality channel service is the product of the throughput coefficient and the delay coefficient. Add all the composite coefficients, and the obtained sum value is the slice service composite coefficient.

[0041] S14. For any slice, calculate the slice service performance coefficient based on the slice signal-to-noise ratio coefficient and the composite coefficient. Specifically, it is the ratio of the sum of the slice service signal-to-noise ratio and the slice composite coefficient to the number of services of this slice.

[0042] S2. Determine the physical resource occupancy rate of the slice, and allocate the corresponding resource occupancy coefficient according to the size of the physical resource occupancy rate as the slice resource coefficient.

[0043] S21. Determine the physical resource occupancy rate of the slice. Specifically, it is the quotient of the total number of physical resource blocks occupied by all services of the slice and the total number of physical resource blocks already allocated to the slice, that is, the proportion of the total number of physical resource blocks occupied by services in the slice to the total number of physical resource blocks already available in the slice.

[0044] S22. Set the physical resource occupancy rate threshold and divide several physical resource occupancy rate ranges.

[0045] Divide the physical resource occupancy rate range by setting the physical resource occupancy rate threshold and points. The physical resource occupancy rate range is preferably divided into four ranges.

[0046] S23. Set the corresponding resource occupancy coefficients for each physical resource occupancy rate range respectively. Simultaneously set the high resource occupancy coefficient, the medium resource occupancy coefficient, the low resource occupancy coefficient, and coefficient 1 for the physical resource occupancy rate range, and pair each coefficient with each physical resource occupancy rate range respectively.

[0047] S23. Determine the range to which the physical resource occupancy rate of the slice belongs, and obtain the corresponding resource occupancy coefficient as the slice resource coefficient.

[0048] For each slice, determine which physical resource occupancy rate range its slice physical resource rate falls into, obtain the physical resource occupancy rate range where the slice physical resource rate is located, and use the resource occupancy coefficient corresponding to this range as the slice resource coefficient.

[0049] S3. Calculate the slice service quality coefficient based on the congestion rate.

[0050] S31. Set the service congestion threshold, and determine the congested services and non-congested services based on the service congestion rate.

[0051] Specifically, set the service congestion threshold, obtain the services with a service congestion rate not less than the service congestion threshold as congested services, and obtain the services with a service congestion rate less than the service congestion threshold as non-congested services.

[0052] S32. Determine the service quality coefficient of the congested services based on the service congestion threshold and the congestion rate, and set the service quality coefficient of the non-congested services.

[0053] Among them, the service quality coefficient of the congested services is the ratio of the service congestion threshold to the service congestion rate. The service quality coefficient of the non-congested services is set, preferably set to 1.

[0054] S33. Determine the slice service quality coefficient according to the service quality coefficients of all services in the slice.

[0055] Specifically, add up the service quality coefficients of all services in the slice, and the ratio of the added value to the number of slice services is used as the slice service quality coefficient.

[0056] S4. Determine the slice allocation priority according to the slice service performance coefficient, resource coefficient, and service quality coefficient.

[0057] The slice allocation priority is the product of the slice service performance coefficient, the slice resource coefficient, and the slice service quality coefficient; Set an adjustment coefficient, and dynamically adjust the priority of the slices for which resource allocation was not implemented in the previous round to the final allocation priority. The final allocation priority is the product of the adjustment coefficient and the slice allocation priority.

[0058] The allocation priority is scalable, and by increasing the corresponding adjustment coefficient, the priority calculation can be made more reasonable. Set an adjustment coefficient, with a range of (1, 2]. For the slices for which allocation was not implemented in the previous round, the product of the adjustment coefficient and the allocation priority is used as their final slice allocation priority.

[0059] S5. Allocate the unallocated resources according to the allocation priority.

[0060] According to the allocation priority, use the PF algorithm to allocate the unallocated resources, and update the total number of physical resource blocks of all slices.

[0061] The present invention starts from screening high-quality channel services, calculates the signal-to-noise ratio coefficient of the calculator, and simultaneously calculates the throughput coefficient and delay coefficient of non-high-quality channel services. On this basis, the performance coefficients of all slices are measured. Based on the physical resource occupancy status, the resource coefficients of all slices are calculated. Based on the service congestion status, the service quality coefficients of all slices are calculated, and a synthetically scalable and adjustable comprehensive allocation priority is synthesized. The slices that were not allocated in the previous round are dynamically adjusted, and through differential strategies, the resource allocation between slices can be more intuitively achieved. The present invention can ensure that services are always in a slice environment with good conditions, providing real-time guarantee for improving customer perception.

[0062] Embodiment 2:

[0063] In this embodiment, a method for resource allocation between 5G network slices based on scalable adjustment is described in combination with specific formulas.

[0064] It includes m network slice sets SEC = {Sec 1 , Sec 2 , …, Sec m}, corresponding to any network slice Sec i , i ∈ [1, m].

[0065] The total number of allocated physical resource blocks of network slice Sec i is NRB i . There are ni services SRV i stored in slice Sec i = {Srv i1 , Srv i2 , …, Srv ini}, corresponding to any service Srv ij , j ∈ [1, ni]. The signal-to-noise ratio of the service at the current moment is Snr ij (dB), the number of occupied physical resource blocks is Prb ij , and the service congestion rate is Coj ij . The average throughput Thr ij (Mbps) generated by service Srv ij in two time slots at the current moment, the average delay Dly ij (ms), the ideal throughput Ttr ij (Mbps), and the delay tolerance Dty ij (ms).

[0066] The method includes the following steps: S1. Determine high-quality and non-high-quality channel services of the slice based on the signal-to-noise ratio.

[0067] Calculate the signal-to-noise ratio coefficient for high-quality channel services based on the signal-to-noise ratio, and calculate the composite coefficient for non-high-quality channel services based on throughput and delay. Calculate the slice service performance coefficient based on the signal-to-noise ratio coefficient and the composite coefficient.

[0068] This step is used to calculate the performance coefficient of slice services by determining high-quality channel services and based on the signal-to-noise ratio, throughput, and delay of the services.

[0069] S11. Set the signal-to-noise ratio threshold SNR th (dB), and obtain Snr ij >SNR th All services Srv ij are high-quality channel services and are included in the high-quality channel service set Sot i , i ∈[1, m], obtain Snr ij ≤SNR th The services are non-high-quality channel services and are included in the non-high-quality channel service set Sot i '.

[0070] S12. For the high-quality channel service set Sot i , all services Srv in i ∈[1, m] ij , convert the signal-to-noise ratio of all services to the decimal signal-to-noise ratio Sdr ij , Sdr ij =power(10, Snr ij / 10).

[0071] Calculate the service signal-to-noise ratio coefficient Cof_snr based on the decimal signal-to-noise ratio Sdr ij , Cof_snr ij =1 / (1+exp((-1)* Sdr ij ))), where exp() represents the exponential function with the natural base. ij ))), where exp() represents the exponential function with the natural base.

[0072] Calculate the sum of the signal-to-noise ratio coefficients Cof_snr of all high-quality channel services ij as the slice signal-to-noise ratio coefficient COF_snr i , COF_snr i =∑ N Cof_snr ij , N is Sdr ij ∈Sot i .

[0073] S13. For the non-high-quality channel service Sot i ', calculate the throughput coefficient Cof_thr of each non-high-quality channel service according to the service throughput ij =Thrij / Ttr ij Calculate the delay coefficient Cof_dly for each non-high-quality channel service according to the delay ij =exp((-1)*Dly ij / Dty ij ).

[0074] According to the throughput coefficient Cof_thr ij and the delay coefficient Cof_dly ij calculate the composite coefficient Cof_ctd ij , Cof_ctd ij =Cof_thr ij *Cof_dly ij .

[0075] Calculate the sum of all composite coefficients as the slice composite coefficient COF_ctd i .

[0076] COF_ctd i =∑ M Cof_ctd ij , where M is Sdr ij ∈Sot i '.

[0077] S14. For any slice Sec i , i ∈ [1, m], based on the slice signal-to-noise ratio coefficient COF_snr i and the composite coefficient COF_ctd i calculate the slice service performance coefficient COF_spf i , The slice service performance coefficient is the ratio of the sum of the slice service signal-to-noise ratio and the slice composite coefficient to the number of slice services COF_spf i =(COF_snr i +COF_ctd i ) / n i .

[0078] S2. Determine the slice physical resource occupancy rate, and allocate the corresponding resource occupancy coefficient according to the size of the physical resource occupancy rate, which is the slice resource coefficient

[0079] S21. Determine the slice physical resource occupancy rate

[0080] Calculate the total number of physical resources Pod i occupied by any slice Sec i =∑ j=1 ni Prbij , the physical resource occupancy rate Roc of the slice i = Pod i / NRB i .

[0081] S22. Set the physical resource occupancy threshold ROC th ∈ (0.5, 1), and divide several physical resource occupancy ranges. According to the physical resource occupancy threshold ROC th and the points 0.5, 1, it is preferably divided into four ranges, namely (-∞, 0.5), [0.5, ROC th , (ROC th , ROC th + (1 - ROC th ) / 2], (ROC th + (1 - ROC th ) / 2, 1).

[0082] S23. Set the corresponding resource occupancy coefficients for each physical resource occupancy range, which are the high resource occupancy coefficient cof min ∈ (0, 0.5], the medium resource occupancy coefficient cof med ∈ (0.5, 1], the low resource occupancy coefficient cof max ∈ (1, 1.5), and the coefficient 1. Pair each coefficient with each physical resource occupancy range respectively.

[0083] S23. Determine the range to which the physical resource occupancy rate of the slice belongs, and obtain the corresponding resource occupancy coefficient as the slice resource coefficient.

[0084] When Roc i < 0.5, calculate the slice resource coefficient COF_src i = cof max ; When 0.5 ≤ Roc i ≤ ROC th , calculate the slice resource coefficient COF_src i = 1; When ROC th < Roc i ≤ ROC th + (1 - ROC th ) / 2, calculate the slice resource coefficient COF_src i = cof med ; When ROC th + (1 - ROC th ) / 2 < Roc iWhen it is < 1, calculate the slice resource coefficient COF_src i =cof min 。

[0085] S3. Calculate the slice service quality coefficient according to the congestion rate.

[0086] S31. Set the service congestion threshold Coj th , and determine the congested services and non-congested services based on the service congestion rate. Specifically, obtain the service congestion rate Coj ij ≥Coj th The services are congested services and are included in the congestion set Sct i , i ∈ [1, m]. Obtain the service congestion rate Coj ij <Coj th The services are non-congested services and are included in the non-congestion set Sct i '.

[0087] S32. Determine the service quality coefficient of the congested services according to the service congestion threshold and the congestion rate, Cof_sql ij =Coj th / Coj ij ; Set the service quality coefficient Cof_sql of the non-congested services ij = 1.

[0088] S33. Determine the slice service quality coefficient according to the service quality coefficients of all services of the slice.

[0089] Add up the service quality coefficients of all services of the slice, and the ratio of the added value to the number of slice services is used as the slice service quality coefficient, which is expressed as follows: COF_ sql i =∑ j=1 ni Cof_sql ij / ni.

[0090] S4. Determine the slice allocation priority according to the slice service performance coefficient, resource coefficient, and service quality coefficient.

[0091] Specifically, the slice allocation priority is the product of the slice service performance coefficient, slice resource coefficient, and slice service quality coefficient, which is expressed as follows: Pri i = COF_spf i *COF_src i * COF_ sql i 。

[0092] Set the adjustment coefficient σ ∈ (1, 2], and the current final slice allocation priority for the slices where resource allocation was not implemented in the previous round is as follows: Pri i ' = σ * Pri i 。

[0093] S5. Allocate the unallocated resources according to the allocation priority.

[0094] According to the final slice allocation priority, use the PF algorithm to allocate the unallocated resources and update the total number of physical resource blocks of all slices.

[0095] The following uses a practical example to specifically illustrate the present invention. Assuming m = 3 as an example, the situations of each 5G network slice are shown in Table 1-3: Table 1 Services stored in each 5G network slice 1

[0096] Table 2 Services stored in each 5G network slice 2

[0097] Table 3 Services stored in each 5G network slice 3

[0098] The basic data is shown in Table 4: Table 4 Basic data

[0099] A method for resource allocation between 5G network slices based on scalable adjustment in this embodiment includes the following steps: S1. Determine the high-quality and non-high-quality channel services of the slices based on the signal-to-noise ratio.

[0100] S11. Set the signal-to-noise ratio threshold SNR th (dB), and obtain all services Srv ij > SNR th as high-quality channel services and include them in the high-quality channel service set Sot ij , i ∈ [1, m], i Sot Sot i = {{Srv(1,3), Srv(1,4)}, {Srv(2,2)}, {Srv(3,1), Srv(3,2)}}.

[0101] Obtain the services with Snr ij ≤ SNR th as non-high-quality channel services and include them in the non-high-quality channel service set Sot i', Sot i '={{Srv(1,1), Srv(1,2)}, {Srv(2,1)},{Srv(3,3)}}。

[0102] S12. For the set of high-quality channel services Sot i , for all services Srv in i ∈ [1, m] ij , convert the signal-to-noise ratio of all services to the decimal signal-to-noise ratio Sdr ij , Sdr ij =power(10, Snr ij / 10) ={{0.63,1,2,1.26},{0.79,2.51},{2,2051,0.63}}。

[0103] Calculate the signal-to-noise ratio coefficient Cof_snr ij =1 / (1+exp((-1)* Sdr ij )) = {{0.88,0.78},{0.92},{0.88,0.92}}。

[0104] Calculate the sum of the signal-to-noise ratio coefficients of all high-quality channel services as the slice signal-to-noise ratio coefficient COF_snr i =∑ N Cof_snr ij ={1.66,0.92,1.81}, N is Sdr ij ∈Sot i 。

[0105] S13. For the non-high-quality channel service Sot i ', calculate the throughput coefficient of each non-high-quality channel service according to the service throughput Cof_thr ij =Thr ij / Ttr ij ={{0.9,0.8},{0.3},{0.25}}, Calculate the delay coefficient of each non-high-quality channel service according to the delay Cof_dly ij =exp((-1)*Dly ij / Dty ij )= {{0.72,0.67},{0.67},{0.45}}。

[0106] According to the throughput coefficient Cof_thr ij and the delay coefficient Cof_dly ij calculate the composite coefficient Cof_ctd ij = Cof_thr ij * Cof_dly ij ={{0.64, 0.54}, {0.2}, {0.11}}.

[0107] Calculate the sum of all composite coefficients as the slice composite coefficient COF_ctd i = ∑ M Cof_ctd ij ={1.18, 0.2, 0.11}, where M is Sdr ij ∈ Sot i '.

[0108] S14. For any slice Sec i , i ∈ [1, m], based on the slice signal-to-noise ratio coefficient COF_snr i and the composite coefficient COF_ctd i calculate the slice service performance coefficient COF_spf i =(COF_snr i + COF_ctd i ) / n i ={0.71, 0.56, 0.64}.

[0109] S2. Determine the physical resource occupancy rate of the slice, and allocate the corresponding resource occupancy coefficient according to the size of the physical resource occupancy rate as the slice resource coefficient.

[0110] S21. Determine the physical resource occupancy rate of the slice.

[0111] Calculate the total number of physical resources occupied by any slice Sec i , i ∈ [1, m] Pod i = ∑ j=1 ni Prb ij ={30, 19, 14}; Calculate the physical resource occupancy rate of the slice Roc i = Pod i / NRB i ={0.75, 0.63, 0.47}.

[0112] S22. Set the physical resource occupancy rate threshold ROC th= 0.7, divide into several ranges of physical resource occupancy rates.

[0113] S23. Set corresponding resource occupancy coefficients for each range of physical resource occupancy rates, which are high resource occupancy coefficient cof min = 0.3, medium resource occupancy coefficient cof med = 0.8, low resource occupancy coefficient cof max = 1.2, and coefficient 1. Pair each coefficient with each range of physical resource occupancy rates respectively.

[0114] S23. Determine the range to which the slice physical resource occupancy rate belongs, and obtain the corresponding resource occupancy coefficient as the slice resource coefficient.

[0115] When Roc i <0.5, calculate the slice resource coefficient COF_src i = cof max ; When 0.5 ≤ Roc i ≤ ROC th , calculate the slice resource coefficient COF_src i = 1; When ROC th <Roc i ≤ ROC th +(1 - ROC th ) / 2, calculate the slice resource coefficient COF_src i = cof med ; When ROC th +(1 - ROC th ) / 2 < Roc i <1, calculate the slice resource coefficient COF_src i = cof min .

[0116] Obtain the slice resource coefficient COF_src i ={0.8, 1, 1.2}.

[0117] S3. Calculate the slice service quality coefficient according to the congestion rate.

[0118] S31. Set the service congestion threshold Coj th = 0.4%, and obtain the service congestion rate Coj ij ≥ Coj th The services are congested services and are included in the congestion set Sct i , i ∈ [1, m], Sct i={{Srv(1,1), Srv(1,3)}, {Srv(2,1), Srv(2,2)}, {Srv(3,1), Srv(3,2), Srv(3,3)}}。

[0119] Obtain the service congestion rate Coj ij <Coj th The services are non-congested services and are included in the non-congested set Sct i ', Sct i '={{Srv(1,2), Srv(1,4)}, {}, {}}。

[0120] S32. Determine the service quality coefficient of congested services according to the service congestion threshold and congestion rate Cof_sql ij =Coj th / Coj ij ={{0.8,0.5},{0.8,0.36},{0.8,1,0.67}}; Calculate the service quality coefficient of non-congested services Coefficient Cof_sql ij ={{1,1},{},{}}。

[0121] S33. Determine the slice service quality coefficient according to the service quality coefficients of all services in the slice COF_ sql i =∑ j=1 ni Cof_sql ij / ni={0.83,0.58,0.82}。

[0122] S4. Determine the slice allocation priority according to the slice service performance coefficient, resource coefficient, and service quality coefficient.

[0123] Calculate the slice allocation priority Pri i = COF_spf i *COF_src i * COF_ sql i ={0.47,0.33,0.63}; The slices for which resource allocation was not implemented in the previous round are {Sec 1 ,Sec 3}, and adjust the allocation priority Pri of the slices for which real-time resource allocation was not performed i '=σ* Pri i, the final slice allocation priority is {0.52, 0.33, 0.69}. The proportional fair PF algorithm is adopted to allocate the unallocated resources, and the total number of physical resource blocks of all slices is updated.

[0124] Simulation experiment: The resource allocation method between 5G network slices based on scalable adjustment of the present invention (hereinafter referred to as SA-NSRA) is simulated on the MATLAB platform with other algorithms based on the idle coefficient (hereinafter referred to as IC-NSRA). The network and service configurations are carried out according to the above table. The RB allocation in the obtained random TTI and the average RB allocation are respectively as Figure 2 and Figure 3 shown.

[0125] As Figure 2 shown, in the process of RB allocation in random TTI, the RB resources allocated by the IC-NSRA algorithm are generally less than those of the SA-NSRA algorithm of the present invention. The reason is that IC-NSRA only considers the amount of idle resources, while SA-NSRA can comprehensively adjust many factors including signal-to-noise ratio, throughput, delay, resource occupancy rate, and service congestion status. Moreover, this adjustment is scalable and can be dynamically increased according to the actual needs of the service. In terms of congestion rate control, the effects of the two algorithms are comparable; in addition, the fundamental difference between SA-NSRA and IC-NSRA is that the former will preferentially schedule services with good channel quality, but in order to prevent services with poor channel quality from starving, it can also compulsorily increase their priorities according to their starvation degree, so as to make each service receive more fair treatment, which can significantly improve the efficiency of system resource allocation; whether it is eMBB, uRLLC or mMTC slices, the resources allocated by SA-NSRA are higher than those of IC-NSRA.

[0126] As Figure 3 shown, the RB allocations of the two schemes are not much different after averaging. Similarly, both schemes allocate more RBs to the eMBB slice. However, since the synthesis of the service performance coefficients of the SA-NSRA algorithm of the present invention treats throughput and service delay equally, the performance performances of the uRLLC and uRLLC slices are close, and thus the final obtained RB quantities are not much different. And mMTC mainly considers service connections and is relatively less emphasized.

[0127] The specific embodiments described in this article are only examples to illustrate the spirit of the present invention. Those skilled in the technical field to which the present invention belongs can make various modifications or supplements to the described specific embodiments or use similar ways to replace them, but will not deviate from the spirit of the present invention or exceed the scope defined by the appended claims.

[0128] Although terms such as high-quality channel service, signal-to-noise ratio coefficient, throughput coefficient, synthesis coefficient, slice resource coefficient, and service quality coefficient are used more frequently in this article, the possibility of using other terms is not excluded. The use of these terms is only for more convenient description and explanation of the essence of the present invention; interpreting them as any additional restrictions is contrary to the spirit of the present invention.

Claims

1. A 5G network slice resource allocation method based on scalable adjustment, characterized in that: The following steps are involved: Determine slice premium and non-premium channel services based on signal-to-noise ratio; For high-quality channel services, the signal-to-noise ratio coefficient is calculated based on the signal-to-noise ratio. For non-high-quality channel services, the synthesis coefficient is calculated based on the throughput and delay. The slice service performance coefficient is calculated based on the signal-to-noise ratio coefficient and the synthesis coefficient. Determine the physical resource occupancy rate of the slice, and allocate the corresponding resource occupancy coefficient according to the size of the physical resource occupancy rate, which is the slice resource coefficient; Calculate the slice service quality coefficient according to the congestion rate; Determine the slice allocation priority based on the slice service performance coefficient, resource coefficient, and service quality coefficient; Unassigned resources are allocated according to their allocation priorities.

2. According to claim 1, a 5G network slice resource allocation method based on scalable adjustment is characterized in that: A signal-to-noise ratio threshold is set, and services whose signal-to-noise ratio is greater than the signal-to-noise ratio threshold are high-quality channel services, and services whose signal-to-noise ratio is not greater than the signal-to-noise ratio threshold are non-high-quality channel services.

3. According to a scalable and adjustable 5G network slice resource allocation method according to claim 1, it is characterized by: The signal-to-noise ratio coefficient of each high-quality channel service is calculated according to the service signal-to-noise ratio, and the sum of all signal-to-noise ratio coefficients is calculated as the slice signal-to-noise ratio coefficient.

4. According to claim 3, a 5G network slice resource allocation method based on scalable adjustment is characterized in that: The throughput coefficient of each non-premium channel service is calculated according to the service throughput, the delay coefficient of each non-premium channel service is calculated according to the delay, the synthesis coefficient is calculated according to the throughput coefficient and the delay coefficient, and the sum of all synthesis coefficients is calculated as the slice synthesis coefficient.

5. According to claim 4, a 5G network slice resource allocation method based on scalable adjustment is characterized in that: The service performance coefficient is calculated based on the signal-to-noise ratio coefficient and the synthesis coefficient, including: The ratio of the sum of the slice service signal-to-noise ratio and the slice synthesis coefficient to the number of slice services.

6. The method for allocating resources between 5G network slices based on scalable adjustment according to claim 1 is characterized in that: The resource occupancy coefficient is set according to the physical resource occupancy rate, which is the slice resource coefficient, including: Set physical resource occupancy rate thresholds and divide physical resource occupancy rate ranges; Set the corresponding resource occupancy coefficient for each physical resource occupancy range; Determine the range to which the physical resource occupancy rate of the slice belongs, and obtain the corresponding resource occupancy coefficient as the slice resource coefficient.

7. The method for allocating resources between 5G network slices based on scalable adjustment according to claim 1 is characterized in that: The service quality coefficient is calculated based on the congestion rate, including: Set service congestion thresholds and determine congested services and non-congested services based on service congestion rates; Determine the service quality coefficient of the congested service according to the service congestion threshold and congestion rate, and set the service quality coefficient of the non-congested service; The service quality coefficient of the slice is determined according to the service quality coefficients of all services of the slice.

8. The method for allocating resources between 5G network slices based on scalable adjustment according to claim 7 is characterized in that: The slice service quality coefficient is determined based on the service quality coefficients of all services in the slice, including: The ratio of the sum of the service quality coefficients of all services in a slice to the number of services in the slice.

9. A 5G network slice resource allocation method based on scalable adjustment according to any one of claims 1 to 8, characterized in that: The slice allocation priority is the product of the slice service performance coefficient, the slice resource coefficient and the slice service quality coefficient; Set the adjustment coefficient and dynamically adjust the slice priority for which resource allocation was not implemented in the previous round to the final allocation priority. The final allocation priority is the product of the adjustment coefficient and the slice allocation priority.

10. The method for allocating resources between 5G network slices based on scalable adjustment according to claim 9 is characterized in that: According to the allocation priority, the PF algorithm is used to allocate the unallocated resources and update the total number of physical resource blocks of all slices.

Images