Method for allocating resources among network slices based on idle coefficient

Abstract

The invention discloses a method for allocating resources among network slices based on idle coefficients, which relates to the technical field of resource allocation, aims to solve the problem of collision of services in the network slices, and comprises the following steps of: S1, determining mark types of the network slices, and respectively measuring and calculating the load of each type of network slices; s2, calculating a load priority and a signal priority of any slice at the current moment according to a set reference signal receiving power threshold value; s3, calculating to obtain a mathematical expected value according to the congestion rate of all the services, and then calculating idle coefficients of all the slices; and S4, calculating the composite priority and the final allocation priority of any slice, and carrying out differentiated resource allocation according to the final allocation priority. Dynamic resource control is performed on the slice according to the load state of the slice and the service quality, so that the service can be ensured to be always in a good slice environment, and real-time guarantee is provided for improving customer perception.

Description

Technical Field

[0001] The present invention relates to the technical field of resource allocation, and in particular to a method for allocating resources between network slices based on an idle coefficient. Background Art

[0002] 5G realizes the interconnection of everything and diversified business needs. Network slicing can logically isolate the network. The business differences between different slices are very large, and the resource requirements are also different. In order to provide better services for the business, this puts forward a higher requirement for resource allocation between 5G network slices. How to achieve a simple and efficient resource allocation method between slices has become an important research direction.

[0003] Wang Wenqi proposed an improved slice resource allocation method based on PF proportional fairness in the Journal of Optical Communication Research (QoS-based 5G inter-slice resource allocation, Optical Communication Research [J], 2018 (3): 59-63). This method first determines the slice type, then dynamically calculates the priority based on the number of users whose services in the slice do not reach the limit, and then allocates resources between slices based on the PF (Proportional Fair) algorithm. This method can operate each slice independently, which is simple and easy; but it does not consider the slice load problem too much, let alone the quality of the network slice. As the scale of business increases, the slice load becomes larger and larger, the possibility of business collision increases significantly, and the efficiency of the resource allocation algorithm will naturally decrease sharply. Therefore, how to comprehensively consider the quality of the slice network, select different inter-slice resource allocation methods, and solve the problem of collision of services within the slice has become increasingly urgent. Summary of the invention

[0004] In order to solve the collision problem of services within a network slice, the present invention proposes a resource allocation method between network slices based on an idle coefficient. Dynamic resource control is implemented on the slice according to the load status of the slice and the service quality. This can ensure that the service is always in a slice environment with a good environment, providing real-time protection for improving customer perception.

[0005] In order to achieve the above object, the present invention adopts the following technical solution: a method for allocating resources between network slices based on idle coefficient, comprising the following steps: S1, determine the marking type of the network slice and calculate the load of each type of network slice respectively; S2, calculate the load priority and signal priority of any slice at the current moment according to the set reference signal receiving power threshold; S3, calculates the mathematical expectation value according to the congestion rate of all services, and then calculates the idle coefficient of all slices; S4, calculates the composite priority and final allocation priority of any slice, and performs differentiated resource allocation according to the final allocation priority.

[0006] The present invention provides a 5G network slice resource allocation method based on idle coefficient. First, starting from the service type in the subdivided slice, the slice load caused by different services in the slice is analyzed differentially, and on this basis, it is converted into a load priority; the signal priority is calculated according to the reference signal reception level value of the slice; the idle resource number and the total resource number of each slice are used, combined with the congestion status of each service in the slice, the idle coefficient of each slice is calculated, and the load priority and the signal priority are weighted by the idle coefficient, and finally a comprehensive priority for evaluating the resource allocation of each slice is obtained, and the PF algorithm is adopted to allocate resources between slices; in this way, dynamic resource control can be implemented on the slice according to the load status and service quality of the slice.

[0007] The present invention is further configured as follows: Step S1 comprises: S11, if the network slice is marked as an eMBB slice, the sum of the throughput of all services and the sum of the ideal throughput are calculated; if the network slice is marked as an uRLLC slice, the sum of the waiting delays and the sum of the delay tolerance of all services are calculated; if the network slice is marked as an mMTC slice, the sum of the connection density of all services and the sum of the maximum allowed connection density are calculated; S12, calculate the current load of the eMBB slice, uRLLC slice and mMTC slice respectively.

[0008] In this technical solution, the load of different types of network slices is calculated differentially according to the service type of the network slice, preparing for the subsequent slice resource allocation priority evaluation.

[0009] The present invention is further configured as follows: Step S2 comprises: S21, calculate the load priority of any slice at the current moment, where the load priority is the current load of the slice divided by the average load of the two time slots before the current moment; S22, setting a reference signal receiving power threshold, and calculating the signal priority of any slice at the current moment.

[0010] In this technical solution, the load priority is converted according to the above-mentioned slice load, and the signal priority is calculated to complete the priority evaluation of network slice resource allocation.

[0011] The present invention is further configured as follows: Step S4 comprises: S41, setting a transmission time interval window and calculating the composite priority of any slice; S42, calculate the final allocation priority of the slice, and the system adopts the PF algorithm to allocate unallocated resources according to the final allocation priority, and updates the total number of physical resource blocks of all slices.

[0012] In this technical solution, the comprehensive allocation priority is calculated through steps S41 and S42, and dynamic allocation control is implemented on resources between slices.

[0013] The present invention is further configured as follows: the final allocation priority of the slice is equal to the product of the composite priority and the idle coefficients of all slices, and the composite priority is the weighted sum of the load priority of the slice and the signal priority of the slice.

[0014] In this technical solution, differentiated resource allocation is ultimately performed based on the final allocation priority.

[0015] The present invention is further configured as follows: Step S3 comprises: S31, for any slice, calculate the mathematical expectation value of all service congestion rates, which is expressed as: in, is the traffic congestion rate, n i is the number of businesses; S32, calculate the idle coefficients of all slices, and the idle coefficients are expressed as: Idc i =NIRB i / NRB i *(1-ECoj i ) NIRB i is the number of idle physical resource blocks at the current moment, NRB i Network SliceSec i , i∈[1,m] is the total number of allocated physical resource blocks.

[0016] In this technical solution, the idle coefficient of each slice is calculated by using the number of idle resources and the total number of resources of each network slice and combining the congestion status of each service in the slice.

[0017] The present invention is further configured as follows: the composite priority is expressed as: Cpri i =(1-1 / T ti )*Pdi i +1 / T ti *Psi i Among them, T ti To set the transmission time interval window, Pdi i Any slice Sec at the current momenti Load priority, Psi i Any slice Sec at the current moment i The signal priority of

[0018] In the technical solution, the composite priority comprehensively considers the load priority and the signal priority, and uses the set transmission time interval window to finally calculate the composite priority.

[0019] The present invention is further configured as follows: Step S12 comprises: The current load when the slice is marked as an eMBB slice is: LOD i =SThr i / SThg i ; The current load when the slice is a uRLLC slice is: LOD i =SDly i / SDlt i ; SliceSec i The current load when marked as mMTC slice is: LOD i =SLkd i / SLmd i ; Among them, SThr i and SThg i are the sum of throughput of all services and the sum of ideal throughput, SDly i and SDlt i are the sum of waiting delays and delay tolerance of all services, SLkd i and SLmd i They are the sum of the connection density of all services and the sum of the maximum allowed connection density, respectively.

[0020] In this technical solution, slice loads are calculated separately according to different marking types of network slices.

[0021] The present invention is further configured as follows: the signal priority is expressed as: Psi i =(RSRP i -RSRP th ) / abs(RSRP th ) Among them, RSRP i is the reference signal received power, RSRP th To set the reference signal receiving power threshold, abs(·) is the absolute value function.

[0022] In the technical solution, the signal priority is calculated by setting the reference signal received power threshold and the reference signal received power.

[0023] The present invention is further configured as follows: the sum of the throughput of all services is expressed as The ideal sum of throughput is expressed as: in, is the current throughput, For ideal throughput.

[0024] In this technical solution, the current load of the eMBB slice is calculated by calculating the throughput and ideal throughput of all services.

[0025] The present invention can bring the following beneficial effects: The present invention relates to a 5G network inter-slice resource allocation method based on idle coefficient. Starting from the service type in the subdivided slice, the slice load caused by different services in the slice is differentially analyzed, and on this basis, it is converted into a load priority; the signal priority is calculated according to the reference signal reception level value of the slice; the idle resource number and the total resource number of each slice are used, combined with the congestion status of each service in the slice, the idle coefficient of each slice is calculated, and the load priority and the signal priority are weighted by the idle coefficient, and finally a comprehensive priority for evaluating the resource allocation of each slice is obtained, and the PF algorithm is adopted to allocate resources between slices; in this way, dynamic resource control can be implemented on the slice according to the load status and service quality of the slice. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a flowchart of a method for allocating resources between 5G network slices based on idle coefficients in the present application.

[0027] Figure 2 This is a comparison chart of RB allocation of random TTI of a 5G network slice resource allocation method based on idle coefficient in the present application and other algorithms.

[0028] Figure 3 This is a comparison chart of the average RB allocation between a 5G network slice resource allocation method based on idle coefficient and other algorithms in this application. DETAILED DESCRIPTION

[0029] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific implementation method described herein is only an optimal embodiment of the present invention, which is only used to explain the present invention and does not limit the scope of protection of the present invention. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0030] Example 1 According to the technical problems in the background technology, it is becoming more and more urgent to comprehensively consider the quality of the slice network, select different inter-slice resource allocation methods, and solve the collision problem of services within the slice. Therefore, this embodiment proposes a 5G network inter-slice resource allocation method based on idle coefficient to solve the above problem. Figure 1 , which includes the following steps.

[0031] Step S1, first determine the marking type of the network slice, and calculate the load of each type of network slice according to the marking type. In this technical solution, starting from the service type in the subdivided network slice, the slice load caused by different services in the network slice is differentially analyzed.

[0032] Step S2, calculate the load priority of any slice at the current moment, and further calculate the signal priority of any slice at the current moment according to the set reference signal receiving power threshold. In this technical solution, based on step S1, the slice load is converted into load priority, and the signal priority is calculated according to the reference signal receiving level value of the slice.

[0033] Step S3, calculate the mathematical expectation value according to the congestion rate of all services, and then calculate the idle coefficient of all slices; in this technical solution, the idle coefficient of each slice is calculated by using the number of idle resources and the total number of resources of each slice, combined with the congestion status of each service in the slice.

[0034] Step S4, calculate the composite priority and final allocation priority of any slice, and perform differentiated resource allocation according to the final allocation priority. In this technical solution, the load priority and signal priority are weighted by the idle coefficient, and finally a comprehensive priority that can evaluate the resource allocation of each slice is obtained, and the PF algorithm is used to allocate resources between slices.

[0035] The purpose of the present invention is to more intuitively realize resource allocation between slices through differentiated strategies under the premise of certain 5G network slice resources. Each network slice has an appropriate amount of business, the slice load is evaluated, the service quality status is analyzed, the idle coefficient is calculated according to the number of idle resources, and finally the comprehensive allocation priority is calculated, and dynamic allocation control of resources between slices is implemented, thereby ensuring the resource requirements of different network slices to meet the index requirements of different services.

[0036] The present invention discloses an idle coefficient based 5G Network Slice Resource Allocation algorithm (IC-NSRA), which first starts from the service types in the subdivided slices, analyzes the slice loads caused by different services in the slices in a differentiated manner, and converts them into load priorities on this basis; calculates the signal priority according to the reference signal reception level value of the slice; calculates the idle coefficient of each slice by using the number of idle resources and the total number of resources of each slice, combined with the congestion status of each service in the slice, and uses the idle coefficient to weight the load priority and the signal priority, and finally obtains a comprehensive priority for evaluating the resource allocation of each slice, and adopts the PF algorithm to allocate resources between slices; in this way, dynamic resource control can be implemented on the slices according to the load status and service quality of the slices.

[0037] Example 2 On the basis of Example 1, this embodiment proposes a method for allocating resources between network slices based on an idle coefficient. More specifically, it mainly includes m network slices SEC = {Sec1, Sec2, ..., Sec m}, marking type is eMBB, uRLLC and mMTC slice; network slice Sec i , i∈[1,m] the total number of allocated physical resource blocks NRB i , the number of idle physical resource blocks NIRB at the current moment i , reference signal received power RSRP i (dB); the average load LBD of the two time slots before the current time i Sec i There are already n i Business Corresponding to any business Current throughput and ideal throughput Waiting delay Time delay tolerance Connection density and the maximum allowed connection density Service congestion rate

[0038] As for step S1, it mainly includes step S11 and step S12.

[0039] Step S11, if the network slice is marked as an eMBB slice, then the sum of the throughput of all its services and the sum of the ideal throughput are calculated. If the network slice is marked as an uRLLC slice, then the sum of the waiting delays and the sum of the delay tolerance of all its services are calculated; if the network slice is marked as an mMTC slice, then the sum of the connection density of all its services and the sum of the maximum allowed connection density are calculated.

[0040] Step S12: upon completion of step S11, current loads of the eMBB slice, the uRLLC slice, and the mMTC slice are calculated respectively.

[0041] In the above technical solution, the loads of different types of network slices are calculated differentially according to the service types of the network slices, in preparation for the subsequent slice resource allocation priority evaluation.

[0042] For any slice Sec i , i∈[1,m], if its type is marked as eMBB slice, then all services in it are calculated The sum of the throughputs is specifically expressed as: The ideal sum of throughput is expressed as:

[0043] If its type is marked as uRLLC slice, all services in it are calculated The sum of the waiting delays is expressed as: The delay tolerance is expressed as:

[0044] If its type is marked as mMTC slice, all services in it are calculated The sum of the connection density is specifically expressed as: The sum of the maximum allowed connection densities is expressed as:

[0045] After completing the above calculations, calculate the load of various types of network slices.

[0046] If the network slice Sec i Mark it as eMBB slice and calculate its current load, specifically: LOD i =SThr i / SThg i If the network slice Sec i Mark the slice as uRLLC and calculate its current load, which is: LOD i =SDly i / SDlt i If the network slice Sec i Mark it as mMTC slice and calculate its current load, specifically: LOD i =SLkd i / SLmd i Among them, SThr i and SThg i are the sum of throughput of all services and the sum of ideal throughput, SDly i and SDlt i are the sum of waiting delays and delay tolerance of all services, SLkd i and SLmd i They are the sum of the connection density of all services and the sum of the maximum allowed connection density, respectively.

[0047] As for step S2, it mainly includes step S21 and step S22.

[0048] Step S21, first calculate the load priority of any slice at the current moment, the load priority is equal to the quotient of the current load of the slice and the average load of the two time slots before the current moment; specifically, any slice Sec i The load priority is expressed as: Pd i =LOD i / LBD i .

[0049] Step S22, then set the reference signal received power threshold, and then calculate the signal priority of any slice at the current moment. Specifically, set the reference signal received power threshold RSRP th (dB); Any slice Sec at the current moment i The signal priority is expressed as: Psi i =(RSRP i -RSRP th ) / abs(RSRP th ) Among them, RSRP iis the reference signal received power, RSRP th To set the reference signal receiving power threshold, abs(·) is the absolute value function.

[0050] In the above technical solution, the load priority is calculated according to the above slice load, and the signal priority is calculated to complete the priority evaluation of network slice resource allocation.

[0051] After completing the process of step S2, proceed to step S3, which mainly includes step S31 and step S32.

[0052] Step S31, calculate the mathematical expectation value of all service congestion rates of any slicer, for any slice Sec i , i∈[1,m], calculate all the services The mathematical expectation value of the congestion rate is expressed as: in, is the traffic congestion rate, n i The number of businesses.

[0053] Step S32, calculate all slices Sec i The idle coefficient is expressed as: Idc i =NIRB i / NRB i *(1-ECoj i ) Among them, NIRB i Represented as the number of idle physical resource blocks at the current moment, NRB i Represented as network slice Sec i , i∈[1,m] is the total number of allocated physical resource blocks.

[0054] In the above technical solution, the idle coefficient of each slice is calculated by using the number of idle resources and the total number of resources of each network slice and combining the congestion status of each service in the slice.

[0055] Step S4 mainly includes step S41 and step S42.

[0056] Step S41, by setting the corresponding transmission time interval window, the composite priority of any slice is calculated.

[0057] Specifically, set the transmission time interval window T ti (ms); calculate any slice Sec i The composite priority is expressed as: Cpri i =(1-1 / T ti)*Pdi i +1 / T ti *Psi i Among them, T ti To set the transmission time interval window, Pdi i Any slice Sec at the current moment i Load priority, Psi i Any slice Sec at the current moment i The signal priority of

[0058] Step S42, calculate the final allocation priority of the slice, and the system adopts the PF algorithm to allocate unallocated resources according to the final allocation priority, and updates the total number of physical resource blocks of all slices.

[0059] The final allocation priority of the slice is equal to the product of the composite priority and the idle coefficients of all slices, and the composite priority is the weighted sum of the load priority of the slice and the signal priority of the slice.

[0060] Specifically, slice Sec i The final allocation priority is expressed as: Fpri i =Cpri i *Idc i The system then uses the proportional fairness PF algorithm to allocate unallocated resources based on this priority and updates the total number of physical resource blocks of all slices.

[0061] The present invention has the following beneficial effects: it can perform detailed analysis on the service types within a slice, and calculate the differentiated load states of different slices, and on this basis calculate the load priority; it can calculate the signal priority of the slice according to the signal state of the slice; it can combine the idle resources and total resources within the slice, and use the service congestion state to calculate the idle coefficient of the slice; it can use the idle coefficient to comprehensively weight all priority indicators; it can implement dynamic resource allocation between slices according to the comprehensive priority; it can ensure that the service is always in a slice environment with a good environment, and provide real-time guarantee for improving customer perception.

[0062] Example 3 Step S1 mainly includes step S11 and step S12.

[0063] Step S11, if the network slice is marked as an eMBB slice, then the sum of the throughput of all its services and the sum of the ideal throughput are calculated. If the network slice is marked as an uRLLC slice, then the sum of the waiting delays and the sum of the delay tolerance of all its services are calculated; if the network slice is marked as an mMTC slice, then the sum of the connection density of all its services and the sum of the maximum allowed connection density are calculated.

[0064] Step S12: upon completion of step S11, current loads of the eMBB slice, the uRLLC slice, and the mMTC slice are calculated respectively.

[0065] In the above technical solution, the loads of different types of network slices are calculated differentially according to the service types of the network slices, in preparation for the subsequent slice resource allocation priority evaluation.

[0066] For any slice Sec i , i∈[1,m], if its type is marked as eMBB slice, then all services in it are calculated The sum of the throughputs is specifically expressed as: The ideal sum of throughput is expressed as:

[0067] If its type is marked as uRLLC slice, all services in it are calculated The sum of the waiting delays is expressed as: The delay tolerance is expressed as:

[0068] If its type is marked as mMTC slice, all services in it are calculated The sum of the connection density is specifically expressed as: The sum of the maximum allowed connection densities is expressed as:

[0069] After completing the above calculations, calculate the load of various types of network slices.

[0070] If the network slice Sec i Mark it as eMBB slice and calculate its current load, specifically: LOD i =SThr i / SThg i If the network slice Seci Mark the slice as uRLLC and calculate its current load, which is: LOD i =SDly i / SDlt i If the network slice Sec i Mark it as mMTC slice and calculate its current load, specifically: LOD i =SLkd i / SLmd i Among them, SThr i and SThg i are the sum of throughput of all services and the sum of ideal throughput, SDly i and SDlt i are the sum of waiting delays and delay tolerance of all services, SLkd i and SLmd i They are the sum of the connection density of all services and the sum of the maximum allowed connection density, respectively.

[0071] As for step S2, it mainly includes step S21 and step S22.

[0072] Step S21, first calculate the load priority of any slice at the current moment, the load priority is equal to the quotient of the current load of the slice and the average load of the two time slots before the current moment; specifically, any slice SeC at the current moment i The load priority is expressed as: Pd i =LOD i / LBD i .

[0073] Step S22, then set the reference signal received power threshold, and then calculate the signal priority of any slice at the current moment. Specifically, set the reference signal received power threshold RSRP th (dB); Any slice Sec at the current moment i The signal priority is expressed as: Psi i =(RSRP i -RSRP th ) / abs(RSRP th ) Among them, RSRP i is the reference signal received power, RSRP th To set the reference signal receiving power threshold, abs(·) is the absolute value function.

[0074] In the above technical solution, the load priority is calculated according to the above slice load, and the signal priority is calculated to complete the priority evaluation of network slice resource allocation.

[0075] After completing the process of step S2, proceed to step S3, which mainly includes step S31 and step S32.

[0076] Step S31, calculate the mathematical expectation value of all service congestion rates of any slicer, for any slice Sec i , i∈[1,m], calculate all the services The mathematical expectation value of the congestion rate is expressed as: in, is the traffic congestion rate, n i The number of businesses.

[0077] Step S32, calculate all slices Sec i The idle coefficient is expressed as: Idc i =NIRB i / NRB i *(1-ECoj i ) Among them, NIRB i Represented as the number of idle physical resource blocks at the current moment, NRB i Represented as network slice Sec i , i∈[1,m] is the total number of allocated physical resource blocks.

[0078] In the above technical solution, the idle coefficient of each slice is calculated by using the number of idle resources and the total number of resources of each network slice and combining the congestion status of each service in the slice.

[0079] Step S4 mainly includes step S41 and step S42.

[0080] Step S41, by setting the corresponding transmission time interval window, the composite priority of any slice is calculated.

[0081] Specifically, set the transmission time interval window T ti (ms); calculate any slice Sec i The composite priority is expressed as: Cpri i =(1-1 / T ti )*Pdi i +1 / T ti *Psi i Among them, T tiTo set the transmission time interval window, Pdi i Any slice Sec at the current moment i Load priority, Psi i Any slice Sec at the current moment i The signal priority of

[0082] Step S42, calculate the final allocation priority of the slice, and the system adopts the PF algorithm to allocate unallocated resources according to the final allocation priority, and updates the total number of physical resource blocks of all slices.

[0083] The final allocation priority of the slice is equal to the product of the composite priority and the idle coefficients of all slices, and the composite priority is the weighted sum of the load priority of the slice and the signal priority of the slice.

[0084] Specifically, slice Sec i The final allocation priority is expressed as: Fpri i =Cpri i *Idc i The system then uses the proportional fairness PF algorithm to allocate unallocated resources based on this priority and updates the total number of physical resource blocks of all slices.

[0085] This embodiment proposes a specific implementation method, and takes m equal to 3 as an example to specifically illustrate the present invention, wherein the conditions of each 5G network slice are shown in Tables 1 to 3, and the basic data are shown in Table 4.

[0086] Table 1 Services available in each 5G network slice 1 Table 2 Services available in each 5G network slice 2 Table 3 Services available in each 5G network slice 3 Table 4 Basic data project data Operating frequency (GHz) 2.6 Operating bandwidth (MHz) 100 Subcarrier bandwidth index 1 <![CDATA[Reference Signal Received Power Threshold RSRP th (dB)]]> -90 <![CDATA[Transmission time interval window T ti (ms)]]> 3 .

[0087] A method for allocating resources between network slices based on an idle coefficient in this embodiment includes the following steps: slice load measurement, slice resource allocation priority evaluation, idle coefficient generation, and differentiated resource allocation.

[0088] Step 1: Slice load calculation.

[0089] Step 1-1: According to Tables 1 to 3, slices Sec1, Sec2, and Sec3 are marked as eMBB, uRLLC, and mMTC, respectively; then calculate all services in slice Sec1 The sum of the throughputs is: and the ideal throughput, specifically: Compute all services in slice Sec2 The sum of the waiting delays is: The sum of the delay tolerance is: Compute all services in slice Sec3 The sum of the connection densities is: and the sum of the maximum allowed connection density, specifically:

[0090] Step 1-2: Calculate the current load LOD of all slices according to the tag type of all slices i ={0.74, 0.4, 0.69}.

[0091] Step 2: Slice resource allocation priority evaluation.

[0092] Step 2-1: Calculate any slice Sec at the current moment i The load priority is: Pdi i =LOD i / LBD i ={1.06, 0.89, 1.05}.

[0093] Step 2-2: Calculate any slice Sec at the current moment i The signal priority is expressed as: Psi i =(RSRP i -RSRP th ) / abs(RSRP th )={0.22, 0.11, 0.33}.

[0094] Step 3: Idle coefficient generation.

[0095] Step 3-1: For any slice Sec i , i∈[1,m], calculate all the services The mathematical expectation of the congestion rate is:

[0096] Step 3-2: Calculate all slices Sec i The idle coefficient is expressed as: Idc i =NIRB i / NRB i *(1-ECoj i )={0.25, 0.44, 0.46}.

[0097] Step 4: Differentiate resource allocation.

[0098] Step 4-1: Calculate any slice Sec i The composite priority is expressed as follows: Cpri i =(1-1 / T ti )*Pdi i +1 / T ti *Psi i ={0.78, 0.63, 0.81}.

[0099] Step 4-2: Calculate slice Sec i The final allocation priority is expressed as follows: Fpri i =Cpri i *Idc i ={0.19, 0.28, 0.37}; The system uses the PF algorithm to allocate unallocated resources according to this priority and updates the total number of physical resource blocks of all slices.

[0100] A resource allocation method between network slices based on an idle coefficient in this embodiment is simulated on a MATLAB platform with a PF algorithm and a PF algorithm based on QoS (comparative document in the background technology). The network and service configuration are performed according to the above table. The obtained RB allocation in the random TTI and the RB average allocation are respectively shown in the attached Figures 2 to 3 shown.

[0101] refer to Figure 2In the RB allocation process in random TTI, PF resource scheduling allocates similar number of RBs to each slice, focusing only on fairness; the QoS-based PF algorithm allocates the most RBs to eMBB slices, the least RBs to mMTC slices, and the RBs allocated to uRLLC slices are similar to those allocated to mMTC slices. The IC-NSRA algorithm of this embodiment will make a comprehensive trade-off analysis based on the load and signal level of each slice, and with the support of the idle coefficient, it can prioritize RB allocation to the slice with the lowest congestion rate, which is reflected in the allocation diagram and has no direct relationship with the slice type.

[0102] refer to Figure 3 , the RB allocation of each scheme is not much different after averaging. Although all three schemes allocate more RBs to eMBB slices, the PF and QoS-based PF algorithms are more due to the algorithm itself, that is, the amount of transmitted data in the eMBB slice is large, so it has a higher priority; while IC-NSRA is directly due to the lower congestion rate of the eMBB slice, and the signal level and load weighting are more balanced.

[0103] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention is described in detail with reference to the embodiments, it should be understood by those skilled in the art that any modification or equivalent replacement of the technical solutions of the present invention does not depart from the spirit and scope of the technical solutions of the present invention and should be included in the scope of the claims of the present invention.

Claims

1. A method for allocating resources between network slices based on idle coefficient, characterized in that: The following steps are involved: S1, determine the marking type of the network slice and calculate the load of each type of network slice; S2, calculate the load priority and signal priority of any slice at the current moment according to the set reference signal receiving power threshold; S3, calculates the mathematical expectation value according to the congestion rate of all services, and then calculates the idle coefficient of all slices; S4, calculates the composite priority and final allocation priority of any slice, and performs differentiated resource allocation according to the final allocation priority.

2. The method for allocating resources between network slices based on idle coefficient according to claim 1, characterized in that: The step S1 comprises: S11, if the network slice is marked as an eMBB slice, the sum of the throughput of all services and the sum of the ideal throughput are calculated; if the network slice is marked as an uRLLC slice, the sum of the waiting delays and the sum of the delay tolerance of all services are calculated; if the network slice is marked as an mMTC slice, the sum of the connection density of all services and the sum of the maximum allowed connection density are calculated; S12, calculate the current loads of the eMBB slice, uRLLC slice and mMTC slice respectively.

3. A method for allocating resources between network slices based on idle coefficient according to claim 1 or 2, characterized in that: The step S2 comprises: S21, calculate the load priority of any slice at the current moment, where the load priority is the current load of the slice divided by the average load of the two time slots before the current moment; S22, setting a reference signal receiving power threshold, and calculating the signal priority of any slice at the current moment.

4. The method for allocating resources between network slices based on idle coefficient according to claim 3, characterized in that: The step S4 comprises: S41, setting a transmission time interval window and calculating the composite priority of any slice; S42, calculate the final allocation priority of the slice, and the system adopts the PF algorithm to allocate unallocated resources according to the final allocation priority, and updates the total number of physical resource blocks of all slices.

5. The method for allocating resources between network slices based on idle coefficient according to claim 4, characterized in that: The final allocation priority of the slice is equal to the product of the composite priority and the idle coefficients of all slices, and the composite priority is the weighted sum of the load priority of the slice and the signal priority of the slice.

6. The method for allocating resources between network slices based on idle coefficient according to claim 3, characterized in that: The step S3 comprises: S31, for any slice, calculate the mathematical expectation value of all service congestion rates, which is expressed as: in, is the traffic congestion rate, n i is the number of businesses; S32, calculate the idle coefficients of all slices, and the idle coefficients are expressed as: Idc i =NIRB i / NRB i *(1-ECoj i ) Among them, NIRB i is the number of idle physical resource blocks at the current moment, NRB i Network SliceSec i , i∈[1,m] is the total number of allocated physical resource blocks.

7. The method for allocating resources between network slices based on idle coefficient according to claim 5, characterized in that: The composite priority is expressed as: Cpri i =(1-1 / T ti )*Pdi i +1 / T ti *Psi i Among them, T ti To set the transmission time interval window, Pdi i Any slice Sec at the current moment i Load priority, Psi i Any slice Sec at the current moment i The signal priority of 8. The method for allocating resources between network slices based on idle coefficient according to claim 2, characterized in that: The step S12 comprises: The current load when the slice is marked as an eMBB slice is: LOD i =SThr i / SThg i ; The current load when the slice is a uRLLC slice is: LOD i =SDly i / SDlt i ; SliceSec i The current load when marked as mMTC slice is: LOD i =SLkd i / SLmd i ; Among them, SThr i and SThg i are the sum of throughput of all services and the sum of ideal throughput, SDly i and SDlt i are the sum of the waiting delays and the delay tolerance of all services, SLkd i and SLmd i They are the sum of the connection density of all services and the sum of the maximum allowed connection density, respectively.

9. The method for allocating resources between network slices based on idle coefficient according to claim 1, 2 or 7, characterized in that: The signal priority is expressed as: Psi i =(RSRP i -RSRP th ) / abs(RSRP th ) Among them, RSRP i is the reference signal received power, RSRP th To set the reference signal receiving power threshold, abs(·) is the absolute value function.

10. The method for allocating resources between network slices based on idle coefficient according to claim 2, characterized in that: The sum of the throughput of all the services is expressed as The ideal sum of throughput is expressed as: in, is the current throughput, For ideal throughput.

Images