CN114630441B

CN114630441B - Resource scheduling method and device

Info

Publication number: CN114630441B
Application number: CN202210526211.3A
Authority: CN
Inventors: 李兰兰; 张铖; 黄永明; 尤肖虎
Original assignee: Network Communication and Security Zijinshan Laboratory
Current assignee: Network Communication and Security Zijinshan Laboratory
Priority date: 2022-05-16
Filing date: 2022-05-16
Publication date: 2022-08-02
Anticipated expiration: 2042-05-16
Also published as: CN114630441A

Abstract

The embodiment of the application provides a method and a device for scheduling resources, wherein the method comprises the following steps: selecting a target subset of slices from all the subsets of slices based on a priority order of each subset of slices within the corresponding set of slices; selecting a first slice from the target subset of slices based on a priority order of each slice in the target subset of slices; performing resource scheduling within the first slice. The resource scheduling method provided by the embodiment of the application determines a scheduling target of a resource by taking a slice subgroup in a slice group as a granularity, selects a target slice as a finer-granularity target of resource scheduling according to a priority order of the slices in the slice subgroup, and executes related scheduling in the target slice, so that a resource scheduling process is more flexible, and the resource utilization rate is more sufficient.

Description

Resource scheduling method and device

Technical Field

The present application relates to the field of communications network technologies, and in particular, to a method and an apparatus for resource scheduling.

Background

The 5th Generation mobile communication (5G) system has The advantages of high speed, high reliability, high capacity and low latency, and The 3rd Generation Partnership Project (3 GPP) defines three application scenarios of 5G: enhanced mobile broadband, large-scale machine communication, high reliability and low time delay communication. The requirements of three major application scenes of the 5G network are different, different devices in different application fields are connected to the network in a large quantity, and the service with high network requirements can be preferentially ensured and the service with low priority can be considered under the condition that the 5G network resources are certain through network slicing. Network slices in the 5G system allow an operator to configure a Network and define specific Functions for the slices, and the slices can be flexibly and dynamically created and revoked through Software Defined Networking (SDN) and Network Function Virtualization (NFV) according to policies of the operator. The network slicing technology is to divide a single physical network into a plurality of virtual end-to-end networks, each virtual network (including devices, access, transmission and core networks in the network) is independent, and the failure of any one divided virtual network does not affect the normal operation of other virtual networks. The 5G Radio access network supports Resource isolation between slices, which can be customized specifically for different clients, and can be implemented by Radio Resource Management (RRM) policy and protection mechanisms that should prevent the shortage of shared resources in one slice from damaging the service level agreement of another slice.

In the prior art, network slices are isolated from each other, and the congestion, overload and configuration adjustment of any one network slice do not affect other network slices. A large amount of signaling interaction generated in the scheduling process of all slices can be maintained inside the base station, so that signaling congestion in the network is caused. Such a scheduling of slices may also result in uneven use of radio resources. The exclusive occupation and use of radio resources by a certain slice results in resource congestion in part of the network slices and insufficient scheduling when part of the network slices are idle. When no relevant service data occurs, the wireless resources are idle and wasted; and other slices may have radio resource shortage due to more service data. Therefore, how to solve the uneven use of wireless resources caused by network slicing in the prior art has become a technical problem to be solved in the field.

Disclosure of Invention

To solve the problems in the prior art, embodiments of the present application provide a method and an apparatus for resource scheduling.

In a first aspect, an embodiment of the present application provides a method for resource scheduling, including:

selecting a target subset of slices from all the subsets of slices based on a priority order of each subset of slices within the corresponding set of slices;

selecting a first slice from the target subset of slices based on a priority order of each slice in the target subset of slices;

performing resource scheduling within the first slice;

wherein the subset of slices is determined within the set of slices based on a K-means clustering algorithm.

Optionally, the performing resource scheduling in the first slice includes:

and comparing the total demand of the users in the first slice on the resources with the size of allocable resources of a cell for providing service for the users in the first slice, and determining whether to execute resource scheduling according to the total demand of the users in the first slice on the resources.

Optionally, the comparing the total demand of the users in the first slice for the resources with the size of the allocable resources of the cell serving the users in the first slice, and determining whether to perform resource scheduling according to the total demand of the users in the first slice for the resources includes:

performing cross-slice scheduling in the target slice subgroup if allocable resources of a cell serving the user in the first slice do not meet the total demand of the user in the first slice for the resources;

and if the allocable resources of the cell providing service for the users in the first slice meet the total demand of the users in the first slice on the resources, performing resource scheduling according to the total demand of the users in the first slice on the resources.

Optionally, the performing cross-slice scheduling within a slice subgroup in the target slice subgroup comprises:

selecting a second slice from the target subset of slices based on a priority order of each slice in the target subset of slices, the first slice having a higher priority than the second slice;

determining whether a first cell serving the user in the first slice and a second cell serving the user in the second slice are the same cell;

and if the first cell and the second cell are the same cell, executing cross-slice scheduling in the same cell.

Optionally, if the first cell and the second cell are the same cell, performing cross-slice scheduling in the same cell, including:

determining whether allocable resources of a cell serving users within the first slice meet a total demand for resources by users within the second slice;

if so, executing resource scheduling according to the total requirement of the user in the second slice on the resource;

and if not, executing the cross-slice scheduling in the same cell in the target slice subgroup.

Optionally, the method further comprises:

if the first cell and the second cell are different cells, performing cross-cell and cross-slice scheduling;

the performing cross-cell cross-slice scheduling is accomplished by sending cross-slice scheduling information from the first cell to the second cell, the cross-slice scheduling information including an identification of the target subset of slices and the first slice overload indication information.

Optionally, the slice subgroup in the slice group is pre-generated, and the specific generation method includes:

determining a cluster to which each slice in the slice group belongs based on a K-means clustering algorithm and a first number, wherein the first number is the number of the clusters;

and selecting one or more slices from different clusters to form the slice subgroup, wherein the QoS difference value corresponding to any two slices in the slice subgroup is within a preset range.

Optionally, the determining a cluster to which each slice in the slice group belongs based on the K-means clustering algorithm and the first number includes:

initially dividing all of the slices into the first number of clusters based on the throughputs of all of the slices within the slice group and the first number;

sequentially determining the difference between the throughput of each of said slices and a randomly selected centroid in each of said clusters, one of said centroids included in each of said clusters; updating the cluster to which the slice belongs by taking the minimum difference value as a target; and repeating the steps until the number of the slices in each cluster is not changed any more, and determining the cluster to which each slice in the slice group belongs.

Optionally, the priority order of each slice subgroup within the corresponding slice group is determined based on a slice subgroup priority determination rule; the slice subgroup priority rule comprises any one of:

determining the priority of the slice subgroup by adopting a weighted average algorithm based on the priorities of all the slices in the slice subgroup;

determining a priority of the subset of slices based on a priority of a lowest priority or a highest priority slice within the subset of slices.

Optionally, if the allocable resources of the cell providing services for the users in the first slice meet the total demand of the users in the first slice for the resources, performing resource scheduling according to the total demand of the users in the first slice for the resources, including:

based on a first scheduling algorithm, performing resource scheduling according to the requirement of each user in the first slice on resources;

the first scheduling algorithm comprises a polling algorithm, a maximum carrier-to-interference ratio algorithm or a proportional fairness algorithm.

In a second aspect, an embodiment of the present application further provides an electronic device, including a memory, a transceiver, and a processor;

a memory for storing a computer program; a transceiver for transceiving data under control of the processor; a processor for reading the computer program in the memory and implementing the method of resource scheduling according to the first aspect as described above.

In a third aspect, an embodiment of the present application further provides an apparatus for resource scheduling, including:

a slice subgroup determination module for selecting a target slice subgroup from all slice subgroups based on the priority order of each slice subgroup within the corresponding slice group;

a slice determination module to select a first slice from the target subset of slices based on a priority order of each slice in the target subset of slices;

a scheduling module for performing resource scheduling within the first slice;

In a fourth aspect, the present application further provides a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the method for resource scheduling according to the first aspect described above.

In a fifth aspect, the present embodiments also provide a processor-readable storage medium, which stores a computer program for causing a processor to execute the method for resource scheduling according to the first aspect.

In a sixth aspect, an embodiment of the present application further provides a communication device readable storage medium, where the communication device readable storage medium stores a computer program, and the computer program is configured to enable a communication device to execute the method for resource scheduling according to the first aspect described above.

In a seventh aspect, an embodiment of the present application further provides a chip product readable storage medium, where the chip product readable storage medium stores a computer program, and the computer program is configured to enable a chip product to execute the method for resource scheduling according to the first aspect.

In an eighth aspect, this application further provides a computer program product, including a computer program, which when executed by a processor implements the method for resource scheduling according to the first aspect described above.

According to the method and the device for scheduling the resources, the slice subgroup in the slice group is used as the granularity, the scheduling target of the resources is determined, then the target slice is selected as the target with finer granularity for scheduling the resources in the slice subgroup according to the priority sequence of the slices, and the related scheduling is executed in the target slice, so that the resource scheduling process is more flexible, and the resource utilization rate is more sufficient.

Drawings

In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings needed for the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

FIG. 1 is a schematic diagram of a conventional 5G network architecture;

fig. 2a is a schematic diagram of a conventional 5G control plane protocol stack architecture;

FIG. 2b is a schematic diagram of a conventional 5G user plane protocol stack architecture;

FIG. 3a is a diagram illustrating a buffer reporting procedure in the prior art;

FIG. 3b is a schematic diagram of a scheduling process in the prior art;

FIG. 4 is a schematic diagram of a distribution of network slices in the prior art;

fig. 5 is a flowchart illustrating a method for resource scheduling according to an embodiment of the present application;

fig. 6a is a schematic diagram of a context establishing/modifying flow of an F1 interface terminal according to an embodiment of the present application;

fig. 6b is a schematic diagram of a NG interface terminal context establishing/modifying flow provided in the embodiment of the present application;

fig. 7 is a diagram illustrating an RRC connection establishment procedure provided in an embodiment of the present application;

fig. 8 is a diagram illustrating an RRC connection reconfiguration procedure according to an embodiment of the present application;

fig. 9 is a schematic diagram of an Xn interface RRC resource status reporting procedure according to an embodiment of the present application;

fig. 10 is a schematic diagram of cross-cell cross-slice scheduling in the method for resource scheduling according to the embodiment of the present application;

FIG. 11 is a schematic diagram of a K-means clustering algorithm for forming slice subgroups according to an embodiment of the present application;

fig. 12a is a schematic view of a resource scheduling hierarchy corresponding to a MAC unit according to an embodiment of the present application;

fig. 12b is a second schematic diagram of a resource scheduling hierarchy corresponding to the MAC unit according to the embodiment of the present application;

fig. 13 is a schematic overall flowchart of a method for resource scheduling according to an embodiment of the present application;

fig. 14 is a schematic structural diagram of an electronic device according to an embodiment of the present application;

fig. 15 is a schematic structural diagram of an apparatus for resource scheduling according to an embodiment of the present application.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

To facilitate a clearer understanding of the embodiments of the present application, some relevant background information is first presented below.

Based on the existing 5G Network architecture (fig. 1) and the existing protocol stack architecture (fig. 2a and 2 b), the Service and System Aspects (SA 2) group in the 3GPP standards organization proposes the concept of Network slice grouping, and also proposes a scheme for cell reselection based on Network slice grouping in the Radio Access Network (RAN) group. But scheduling schemes based on network slices and slice subgroups have not been proposed in the standard organization. In the existing radio resource scheduling process, a terminal (User Equipment, UE) informs a 5G base station (gNodeB) through a buffer status reporting process how much data currently exists in an uplink buffer to be sent, so as to provide uplink and downlink scheduling information for the gNodeB, where a buffer reporting process is as shown in fig. 3a, and a scheduling process is as shown in fig. 3 b.

The traditional wireless resource scheduling algorithm mainly comprises a polling algorithm, a maximum carrier-to-interference ratio algorithm, a proportional fairness algorithm and a related improvement algorithm, system throughput and user fairness are important performance indexes for measuring the quality of the scheduling algorithm, the proportional fairness algorithm is a scheduling algorithm which gives consideration to both the system throughput and the fairness among users at the same time at present, the fairness among the users refers to whether all users requesting resources can obtain scheduled opportunities in a scheduling period or not, and the users can be allocated to the same time-frequency resources. In a RAN slice scene, by improving a traditional scheduling algorithm, aiming at different requirements of each slice on mobility, reliability, speed, time delay and the like, when a user orders a slice service, the priorities of different slices are divided into high and low priorities, the user scheduling priority of a high-priority slice is higher than that of a low-priority slice, in each scheduling period, the user belonging to the high-priority slice preferentially obtains the opportunity of resource scheduling, and the fairness of the user in the RAN slice scene is measured by a scheduler by allocating resources according to needs so as to ensure the service quality of the user. Scheduling with different RAN slice granularities is also disadvantageous because the UE may access multiple slices, the network may have multiple slices, the slices in the network may be located in different cells, and there may be multiple location relationships between the slices, as shown in fig. 4 as 1-1, 1-2, 1-3, and 1-4.

In the prior art, network slices are isolated from each other, and the congestion, overload and configuration adjustment of any one network slice do not influence other network slices; in the scheduling process shown in fig. 3b, the SCH determines a radio resource management policy and a protection mechanism according to a mapping relationship between a logical channel and a network slice group identifier. A unit capable of implementing Media Access Control (MAC) related functions exists in each slice, so that more MAC units are maintained in the base station, and a large amount of signaling interaction is generated in the scheduling process, which causes signaling congestion in the network.

Such a slice distribution manner may also cause uneven use of radio resources, and the use of radio resources exclusively occupied by a certain slice causes resource congestion in a part of network slices, and resource vacancy in a part of network slices does not have sufficient scheduling. When no relevant service data occurs, the RAN resources are idle and wasted; while other slices may have RAN resource shortage due to more traffic data.

Therefore, the method provides a concept of a slice subgroup facing wireless network resource management, a wireless resource management algorithm taking the slice subgroup as granularity, a wireless resource management algorithm introducing a K-means clustering algorithm in the field of artificial intelligence into wireless slice subgroup determination and a flexible wireless resource management algorithm in the slice group, and resource scheduling takes the slice subgroup as granularity and further manages according to wireless resources in the slice-level group, so that the problem of wireless resource shortage in the 5G network slice group is solved.

Fig. 5 is a flowchart illustrating a method for resource scheduling according to an embodiment of the present application; as shown in fig. 5, the method for resource scheduling includes:

step 501, selecting a target slice subgroup from all slice subgroups based on the priority order of each slice subgroup in a corresponding slice group;

step 502, selecting a first slice from the target subset of slices based on a priority order of each slice in the target subset of slices;

step 503, performing resource scheduling in the first slice;

Specifically, network slicing is implemented based on Network Function Virtualization (NFV) and Software Defined Networking (SDN). A physical network may be divided into slices, each of which may be considered a virtual network specifically tailored to a certain service or scenario. Slices and slice groups are a form of partitioning of radio resources. The slice groups can be pre-configured on a network management system of the base station, or the slices are classified according to a corresponding algorithm and are used as one slice group, each slice group comprises one or more slices, and each slice group comprises identification information of the slice group, all slice information in the group and priority information. Each slice information includes identification information of the slice, priority information, and information of the base station and the cell where the slice is located.

When resource scheduling is carried out, firstly, determining a slice subgroup included in each slice group in all slice groups according to corresponding rules, determining the priority of each slice subgroup, and selecting the slice subgroup with the highest priority as a target slice subgroup for pre-allocating resources; then, the first slice of the pre-allocated resource is selected according to the priority of each slice in the target slice subgroup and the high-low order of the priority, specifically, the first slice of the pre-allocated resource may be selected according to the order of the priority from high to low, or according to the order of the priority from low to high. There are also multiple users in each slice, and all users in a slice share these allocated resources. Similarly, the network resources allocated to the first slice may be scheduled among multiple users in order of priority based on the priority requested by the users. When the resource scheduling is carried out, the total resource in each time slot is divided into a plurality of RB groups for carrying out the resource scheduling in a radio bearer RB mode. The corresponding rule is a K-means clustering algorithm.

The resource scheduling method provided by the embodiment of the application determines a scheduling target of a resource by taking a slice subgroup in a slice group as a granularity, selects a target slice as a finer-granularity target of resource scheduling according to a priority order of the slices in the slice subgroup, and executes related scheduling in the target slice, so that a resource scheduling process is more flexible, and the resource utilization rate is more sufficient.

Optionally, the performing resource scheduling in the first slice includes:

Specifically, when resource scheduling is performed in the first slice, the total demand of all users for resources in the first slice is obtained by summing based on the demand of each user for resources in the first slice, and the total demand of all users for resources in the first slice is compared with the size of the cell allocable resources for providing service for the users of the first slice, if the total demand of all users for resources in the first slice is greater than the size of the cell allocable resources, cross-slice scheduling in a target slice subgroup to which the first slice belongs is performed. Since there may be other slices in the target slice subgroup where the user's demand for resources is less than or equal to the size of the cell allocable resources, the corresponding resources are allocated to this slice.

And if the total demand of all the users on the resources in the first slice is less than or equal to the size of the allocable resources of the service cell, scheduling the resources according to the total demand of all the users on the resources in the first slice, so that the demand of the users on the resources can be met, and related network resources are not wasted.

The total resource demand of the user in the first slice may be obtained by summing up the resource demands of the terminal included in a Buffer Status Report (BSR) reported by each terminal. As shown in fig. 3a, the terminal UE provides the network device with information of the amount of data to be transmitted in its Uplink (UL) buffer via the BSR.

The method improves the air interface communication flow and signaling on the network side, and comprises a BSR flow in the Resource management scheduling process, a Radio Resource Control (RRC) connection establishment and reconfiguration flow of Radio Resource management, signaling improvement of an Xn interface between cells, and an up-down file management flow of an F1 interface. The F1 interface UE context setup/modification procedure flow is shown in fig. 6 a. The base station obtains the QoS information of the UE (including the maximum bit rate that the UE can reach in the serving slice group) from the core network through the UE context establishment/modification process of the NG interface, and the flow is shown in fig. 6 b. In the radio resource management scheduling process, a Central Unit (CU) of a 5G base station implements the UE context establishing/modifying process through a Distributed Unit (DU) and an F1 interface message of the CU, and notifies the DU of a network slice associated with each data radio bearer, whereas a Buffer Status Report (BSR) in the current protocol reports no information of a network slice group, and the base station cannot execute a radio resource management policy in the slice group. The resource scheduling method proposed in the present application further relates to an improvement in the context establishment/modification process of the UE at the F1 interface, and adds network slice group information in the corresponding signaling, specifically, the UE context establishment request message or the UE context modification request message carries network slice group identification information, the priority of the network slice group, and QoS information of the UE (including the maximum bit rate that the UE can reach in the service slice group).

Specifically, the implementation is mainly implemented for supplementing the logical channel configuration information at the base station side in the protocol and improving the air interface signaling of the above process, and the implementation is specifically as follows: the mapping relation between the logical channel group LCG and the slice group is established by adding configuration information related to slices and slice groups, including but not limited to network slice group identification information, in a field LogicalChannelConfiguration information related to logical channel configuration in a radio resource control reconfiguration RRCReconfiguration instruction. The related procedures specifically involved include RRC connection establishment procedure and RRC connection reconfiguration procedure.

Fig. 7 is a schematic diagram of an RRC connection establishment procedure provided in an embodiment of the present application, as shown in fig. 7:

1. a terminal (User Equipment, UE) sends an RRC connection establishment request message to a gbb (5G base station), where the message carries an RRC establishment cause and a UE identity.

2. The gNB establishes a context for the UE to perform admission and resource allocation of Signaling Radio Bearer 1 (SRB 1) resources. And replying an RRC establishment message to the UE, wherein the message carries the detailed information of the resource configuration of the SRB 1.

3. The UE performs radio resource configuration according to the SRB1 resource information indicated by the RRC connection establishment message, and then sends an RRC connection establishment completion message to the gNB, wherein the message carries network slice group information including but not limited to slice group identification information and the priority of the network slice group, and the RRC connection establishment is completed.

Fig. 8 is a schematic diagram of an RRC connection reconfiguration procedure provided in an embodiment of the present application, as shown in fig. 8:

1. the gNB issues an RRC connection reconfiguration message to the UE, and indicates to establish a Signaling Radio Bearer 2 (SRB 2) and a Data Radio Bearer (DRB).

2. After receiving the RRC connection reconfiguration message, the UE establishes SRB2 and DRB. After the SRB2 and the DRB are successfully established, the UE replies an RRC connection reconfiguration complete message to the gNB, where the message carries a logical channel configuration cell, where the cell includes information related to a network slice group, including but not limited to network slice identification information and priority of the network slice group, and the RRC connection reconfiguration is complete.

Specifically, if the total demand for resources by all users within a first slice is greater than the size of allocable resources of the serving cell, cross-slice scheduling within a slice subgroup is performed in a target slice subgroup to which the first slice belongs.

According to the priority level of each slice in the target slice subgroup, selecting the slice with the highest priority level except the first slice from the target slice subgroup as a second slice, determining a second cell for providing service for the second slice, comparing whether the first cell for providing service for the first slice and the second cell are the same cell, and if the first cell and the second cell are the same cell, performing cross-slice scheduling in the same cell, namely preparing to schedule the allocable resource of the first cell to the second slice.

Specifically, when preparing to schedule the allocable resources of the first cell to the second slice, it needs to first determine whether the allocable resources of the first cell meet the total demand of the users on the resources in the second slice.

And if the allocable resources of the first cell are more than or equal to the total demand of the users on the resources in the second slice, scheduling the allocable resources to the second slice.

If the allocable resources of the first cell are less than the total demand of the users on the resources in the second slice, continuing to search other slices with lower priorities in the first cell, and if the total demand of the users on the resources in a certain slice is less than or equal to the allocable resources of the first cell, scheduling the allocable resources to the slice. If no such other slice exists in the first cell, switching to cross-cell cross-slice scheduling.

Optionally, the method further comprises:

Specifically, when the first cell and the second cell are different cells, cross-cell and cross-slice scheduling needs to be performed, which is implemented by sending cross-slice scheduling information from the first cell to the second cell, where the cross-slice scheduling information includes an identifier of the target slice subgroup and indication information of overload of the first slice. The resource shortage of other cells where another slice in the slice subgroup is located may be notified through the Xn interface message to trigger cross-cell and cross-slice scheduling, and specifically, the resource status update message of the Xn interface shown in fig. 9 carries the network slice identifier, the slice group identifier, and the resource shortage indication information. The Xn interface RRC resource state includes the flow as shown in fig. 9. The second cell receives the cross-slice scheduling information, and may determine which slice subgroup the slice belongs to, and the reason why the slice performs the cross-slice scheduling, or may bring the total resource demand of the user in the slice, which is convenient for the subsequent comparison, and perform the corresponding resource scheduling when it is determined that there is an allocable resource of a proper size. As shown in fig. 10, a schematic diagram of cross-cell cross-slice scheduling. Indicating that when the radio Resource in slice 1 cannot meet the traffic demand, the radio Resource of another slice (slice 2) within the same slice subgroup can be scheduled, which includes Resource Block Group1 (RBG 1), RBG2, etc.

selecting one or more slices from different clusters to form the slice subgroup, where a Quality of Service (QoS) difference value corresponding to any two slices in the slice subgroup is within a preset range.

Specifically, the slice group may be preconfigured in the network management system of the base station, and includes an ID of the slice group, a priority of the slice group, and slice information; all slices are classified within each slice group according to a corresponding algorithm, classified as a subset of slices. When the slices are classified according to the corresponding algorithm, a K-means clustering algorithm can be adopted.

The method comprises the steps of firstly determining that the K-means clustering algorithm needs to be divided into a plurality of clusters or clusters, clustering all slices into N clusters, and then selecting a plurality of slices with QoS (quality of service) difference values corresponding to the slices within a preset range from N different clusters to form a slice subgroup.

TABLE 1

For example, table 1 lists the slices included in each cluster and the QoS value corresponding to each slice, and in the case that the preset range of the QoS difference is 0.06, the corresponding slice subset may be determined as:

wherein, the slice 1, the slice 4 and the slice 8 are used as a slice subgroup, and the difference value meeting the QoS is less than 0.06;

slice 2, slice 6 and slice 7 are taken as a slice subgroup, and the difference value meeting the QoS is less than 0.06;

slice 3, slice 5, and slice 9, as a subset of slices, satisfy QoS with a difference of less than 0.06.

TABLE 2

For example, the slices included in each cluster and the QoS value corresponding to each slice are listed in table 2, and in the case that the preset range of the QoS difference is 0.06, the corresponding slice subset may be determined as:

the slice 6 and the slice 7 are used as a slice subgroup, and the difference value meeting the QoS is less than 0.06;

slice 3, slice 5 and slice 9 are taken as a slice subgroup, and the difference value meeting the QoS is less than 0.06;

slice 2 serves as a subset of slices.

Fig. 11 is a schematic diagram of slice groups formed by the K-means clustering algorithm provided in the embodiment of the present application. As shown in fig. 11, the slice group1 includes a slice 1, a slice 2, a slice 3, and a slice 4; slice group 2 includes slice 5, slice 6 and slice 7; the cluster 1 comprises a slice 1, a slice 2, a slice 4 and a slice 6; the cluster 2 comprises a slice 3, a slice 5 and a slice 7; all the slices are subjected to clustering analysis in the slice group1 based on a K-means clustering algorithm, the obtained slice subgroup 1 comprises slices 3 and slices 2, the slice subgroup 2 comprises the slices 1, the slice subgroup 3 comprises the slices 4, the slice subgroup 4 comprises the slices 5 and the slices 6, and the slice subgroup 5 comprises the slices 7.

Specifically, based on the K-means clustering algorithm and the first number, the cluster to which each slice in the slice group belongs is determined, that is, all slices in the slice group are determined to be divided into several clusters.

According to the throughput of each slice and the number N of the slices to be classified, initially dividing all the slices into N clusters; randomly determining the center of mass of each cluster, namely determining a certain central slice in each cluster, taking the central slice as a reference, then determining the throughput of each slice to be clustered and comparing the throughput of the determined central slice of each cluster, taking the cluster to which the central slice belongs as the cluster to which the slice to be clustered belongs when the difference of the throughput from the central slice is minimum, and sequentially circulating until the number of slices contained in each cluster does not change any more, determining the cluster to which each slice to be clustered finally belongs, and determining the clustering results of all the slices.

Specifically, according to the K-means clustering algorithm, after determining the slice subgroup in the slice group, the priority of the slice subgroup needs to be further determined, and the specific determination method may determine the priority of the slice subgroup by using a weighted average algorithm on the priorities of all slices included in the slice subgroup, where a weight corresponding to each slice may be set according to a requirement, or the weights of each slice are the same. Or, the priority of the slice with the lowest priority or the highest priority in the slice subgroup is taken as the priority of the slice subgroup.

Specifically, when the allocable resources of the first cell served by the users in the first slice meet the total demand of the users in the first slice for the resources, a polling algorithm may be adopted to perform resource scheduling for each user in the first slice, or according to a proportional fair algorithm, resource scheduling may be performed for each user in the first slice.

The resource scheduling method provided by the embodiment of the application determines a scheduling target of a resource by taking a slice subgroup in a slice group as a granularity, selects a target slice as a finer-granularity target of resource scheduling according to the priority order of the slices in the slice subgroup, and determines whether to execute related scheduling or not based on the actual available resources of a cell and the requirements of users in the slices on the resources, so that the resource scheduling process is more flexible, and the resource utilization rate is more sufficient.

The following describes a method for scheduling resources according to an embodiment of the present application with a specific example.

Configuring slice information and slice group information of a base station on a network management system of the base station, wherein the slice group information comprises a slice group identifier, slice information, the base station where the slice is located and cell information; the slice information comprises identification information and priority information of the slice; after determining a slice clustering result and a corresponding slice subgroup according to a K-means clustering algorithm, when the wireless resource of a certain slice in the slice subgroup in a cell is in short supply, if the wireless resource of another slice in other slice subgroups in the same cell is available, the MAC unit calls the wireless resource of another slice; if the slice in the slice subgroup is located in different cells, it can be notified through an Xn interface message that resources of other cells in which another slice is located are in short supply and cross-cell and cross-slice scheduling is triggered. Specifically, a three-level scheduling mode is adopted: the method comprises the steps that firstly, a network side configures a slice group, a corresponding slice subgroup is determined in the slice group according to a K-means algorithm, the priority of the slice subgroup is further determined according to the priority of slices in the slice subgroup, an MAC unit can determine the slice subgroup with higher priority to preferentially execute slice subgroup level scheduling according to the priority of the slice subgroup, slice level scheduling is executed in the slice subgroup, then user level scheduling is executed, and slice level scheduling is executed when a slice is in the slice subgroup. Fig. 12a and fig. 12b show schematic diagrams of resource scheduling hierarchies corresponding to the MAC units.

In the 5G technology, the unit of physical resources may be an RBG, and the number of RBs in an RB group is determined according to a slice characteristic and the number of users in a scheduling process. As shown in fig. 13, an overall flow diagram of the method for resource scheduling provided in the embodiment of the present application is as follows:

step 1301: configuring a slice group including an ID, a priority, slice information included in the slice group, and the like in an Operation Administration and Maintenance (OAM) network element of a base station.

Step 1302: and at each scheduling moment, dividing the total resources in the time slot into a plurality of RB groups according to the number of users requesting scheduling in the time slot and the corresponding slice. The slices in step 1301 form a slice subgroup in a certain way, and the slice subgroup is determined by adopting a K-means clustering algorithm in the application.

Step 1303: determining the priority of each slice subgroup by adopting a weighted average algorithm according to the priority of the slices in each slice subgroup; carrying out resource scheduling by taking the slice groups as strength;

step 1304: sequentially selecting the slice subgroups to schedule according to the priority level of each slice subgroup; and in the slice subgroup, the slices are sequentially selected for scheduling according to the priority levels of the slices by taking the slices as granularity, if the wireless resources of each slice in a cell are sufficient in the scheduling process, the user scheduling in the slice is executed according to the step 1307, otherwise, the cross-slice scheduling in the slice subgroup is executed.

Step 1305: judging whether different slices are in the same cell or not in the process of executing cross-slice scheduling of the slice subgroup;

step 1306: if the cells are the same, go to step 1306a to execute cross-slice scheduling in the same cell; otherwise go to step 1306b to perform cross-cell cross-slice scheduling. In the process of performing cross-cell and cross-slice scheduling, the source cell is required to send cross-slice scheduling information to a cell in which another slice in the slice subgroup is located through an Xn interface message, where the interface message includes, but is not limited to, a slice subgroup identifier and slice overload indication information.

Step 1307: the scheduling of users in the slice can adopt a polling and proportional fair scheduling algorithm to schedule wireless resources in the slice.

Determining a slice subgroup in step 1302, where slices with close QoS characteristics can be used as the slice subgroup according to the QoS characteristics of the slices, the present application proposes a method for slice subgroup generation:

firstly, adopting a k-means clustering algorithm:

step 1: firstly, determining the number of clusters to be N (planning to divide data into N classes) in each slice group according to the throughput of a user of each slice;

step 2: randomly determining N initial points as centroids (randomly selected within the boundary range of the throughput size);

and step 3: sequentially calculating the distances from the data instances in each slice to the N centroids, selecting the centroid with the minimum distance, distributing the centroid to the cluster corresponding to the centroid until all data in the data set are distributed to the N clusters, and updating the centroids of the N clusters to be the average value of all points of the cluster;

and 4, step 4: and (4) circulating the step 3, and reallocating each slice to a new centroid until the allocation results of all the slices are not changed any more.

And then selecting a plurality of slices with close QoS characteristics in different clusters to form a slice subgroup, wherein a part of slices in the slice subgroup occupy more wireless resources and a part of slices occupy less wireless resources, so that the uniform use of the wireless resources can be realized through cross-slice scheduling among the slices with different throughputs in the cross-slice scheduling process.

Fig. 14 is a schematic structural diagram of an electronic device according to an embodiment of the present application; as shown in fig. 14, the electronic device includes a memory 1420, a transceiver 1410, and a processor 1400; the processor 1400 and the memory 1420 may also be physically separated.

A memory 1420 for storing a computer program; a transceiver 1410 for transceiving data under the control of the processor 1400.

In particular, the transceiver 1410 is used for receiving and transmitting data under the control of the processor 1400.

Where in fig. 14 the bus architecture may include any number of interconnected buses and bridges, in particular one or more processors, represented by the processor 1400, and various circuits of memory, represented by the memory 1420, linked together. The bus architecture may also link together various other circuits such as peripherals, voltage regulators, power management circuits, and the like, which are well known in the art, and therefore, will not be described any further herein. The bus interface provides an interface. The transceiver 1410 may be a number of elements including a transmitter and a receiver that provide a means for communicating with various other apparatus over a transmission medium including wireless channels, wired channels, fiber optic cables, and the like.

The processor 1400 is responsible for managing the bus architecture and general processing, and the memory 1420 may store data used by the processor 1400 in performing operations.

The processor 1400 may be a Central Processing Unit (CPU), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA), or a Complex Programmable Logic Device (CPLD), and may also have a multi-core architecture.

The processor 1400 is used for executing any of the methods provided by the embodiments of the present application according to the obtained executable instructions by calling the computer program stored in the memory 1420, for example:

performing resource scheduling within the first slice;

It should be noted that, the electronic device provided in the embodiment of the present application can implement all the method steps implemented by the method embodiment and achieve the same technical effect, and detailed descriptions of the same parts and beneficial effects as the method embodiment in this embodiment are omitted here.

Fig. 15 is a schematic structural diagram of an apparatus for resource scheduling according to an embodiment of the present application, and as shown in fig. 15, an apparatus for resource scheduling according to an embodiment of the present application includes a slice subgroup determining module 1501, a slice determining module 1502, and a scheduling module 1503, where:

a slice subgroup determining module 1501, configured to select a target slice subgroup from all slice subgroups based on a priority order of each slice subgroup within a corresponding slice group;

a slice determination module 1502 for selecting a first slice from the target subset of slices based on a priority order of each slice in the target subset of slices;

a scheduling module 1503 for performing resource scheduling within the first slice;

Specifically, the apparatus for scheduling resources provided in this embodiment of the present application can implement all the method steps implemented by the method embodiment and achieve the same technical effect, and details of the same parts and beneficial effects as those of the method embodiment in this embodiment are not repeated herein.

In another aspect, the present application further provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, where the computer program includes program instructions, and when the program instructions are executed by a computer, the computer is capable of executing the method for resource scheduling provided by the foregoing embodiments.

On the other hand, the embodiment of the present application further provides a processor-readable storage medium, where the processor-readable storage medium stores a computer program, and the computer program is configured to enable the processor to execute the method for resource scheduling provided in the foregoing embodiments.

The processor-readable storage medium can be any available medium or data storage device that can be accessed by a processor, including, but not limited to, magnetic memory (e.g., floppy disks, hard disks, magnetic tape, magneto-optical disks (MOs), etc.), optical memory (e.g., CDs, DVDs, BDs, HVDs, etc.), and semiconductor memory (e.g., ROMs, EPROMs, EEPROMs, non-volatile memory (NAND FLASH), Solid State Disks (SSDs)), etc.

The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.

Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions in the embodiments of the present application.

Claims

1. A method for resource scheduling, comprising:

performing resource scheduling within the first slice;

wherein the subset of slices is determined by classifying all slices in the set of slices based on a K-means clustering algorithm within the set of slices;

the performing resource scheduling within the first slice comprises:

comparing the total demand of the users in the first slice for resources with the size of allocable resources of a cell serving the users in the first slice, and determining to execute cross-slice scheduling in the slice subgroup in the target slice subgroup to which the first slice belongs, or to execute resource scheduling in the first slice in the target slice subgroup according to the total demand of the users in the first slice for resources.

2. The method of claim 1, wherein the comparing the total demand for resources by the users in the first slice with the size of allocable resources of the cell serving the users in the first slice, and the determining whether to perform cross-slice scheduling in the target slice subgroup to which the first slice belongs or perform resource scheduling in the first slice in the target slice subgroup according to the total demand for resources by the users in the first slice comprises:

3. The method of claim 2, wherein performing cross-slice scheduling within the slice subgroup in the target slice subgroup comprises:

4. The method of claim 3, wherein if the first cell and the second cell are the same cell, performing intra-cell cross-slice scheduling comprises:

5. The method of resource scheduling according to claim 3, further comprising:

6. The method according to claim 1, wherein the subset of slices in the slice group is pre-generated, and the specific generation method comprises:

and selecting one or more slices from different clusters to form the slice subgroup, wherein the difference value of the QoS (quality of service) corresponding to any two slices in the slice subgroup is within a preset range.

7. The method according to claim 6, wherein the determining the cluster to which each slice in the slice group belongs based on the K-means clustering algorithm and the first number comprises:

8. The method according to claim 1, wherein the priority order of each slice subgroup within the corresponding slice group is determined based on a slice subgroup priority determination rule; the slice subgroup priority rule comprises any one of:

9. The method according to claim 2, wherein if the allocable resources of the cell serving the users in the first slice satisfy the total demand for resources by the users in the first slice, performing resource scheduling according to the total demand for resources by the users in the first slice includes:

10. An electronic device comprising a memory, a transceiver, a processor;

a memory for storing a computer program; a transceiver for transceiving data under control of the processor; a processor for reading the computer program in the memory and performing the following operations:

performing resource scheduling within the first slice;

the performing resource scheduling within the first slice comprises:

11. The electronic device of claim 10, wherein the comparing the total demand for resources by users within a first slice with the size of allocable resources of a cell serving users within the first slice, the determining to perform cross-slice scheduling within the target slice subset to which the first slice belongs, or performing resource scheduling within the first slice of the target slice subset in accordance with the total demand for resources by users within the first slice comprises:

12. The electronic device of claim 11, wherein performing intra-slice subgroup cross-slice scheduling in the target slice subgroup comprises:

13. The electronic device of claim 12, wherein if the first cell and the second cell are the same cell, performing intra-cell cross-slice scheduling comprises:

14. The electronic device of claim 12, wherein the operations further comprise:

15. The electronic device of claim 10, wherein the subset of slices within the set of slices is pre-generated, and wherein a particular generation method comprises:

16. The electronic device of claim 15, wherein determining the cluster to which each slice in the slice group belongs based on the K-means clustering algorithm and the first number comprises:

17. The electronic device of claim 10, wherein the priority order of each slice subgroup within the corresponding slice group is determined based on a slice subgroup priority determination rule; the slice subgroup priority rule comprises any one of:

18. The electronic device of claim 11, wherein if allocable resources of a cell serving users in a first slice satisfy the total demand for resources by users in the first slice, performing resource scheduling according to the total demand for resources by users in the first slice comprises:

19. An apparatus for resource scheduling, comprising:

a scheduling module for performing resource scheduling within the first slice;

the scheduling module is specifically configured to, in the process of executing resource scheduling in the first slice:

20. A computer-readable storage medium, characterized in that it stores a computer program for causing a computer to execute the method of resource scheduling of any one of claims 1 to 9.