A network slice resource dynamic allocation apparatus and method

Abstract

The application relates to the technical field of transmission resource allocation, in particular to a network slice resource dynamic allocation device and method, the method comprising the following steps: collecting network slice resource data, and preprocessing the data; calculating the transmission resistance of a virtual logical network through the transmission rate and the transmission queue length of the network slice resource; calculating the congestion increase of the network slice resource through the transmission queue length of the network slice resource; calculating the transmission load supplement of the virtual logical network through the transmission resistance of the virtual logical network and the congestion increase of all the network slice resources in the virtual logical network; allocating the reserved resource quantity to obtain the additional resource allocation quantity of each virtual logical network. The application aims to improve the accuracy of network slice resource allocation and ensure that the resource allocation is more in line with business requirements.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of transmission resource allocation, and particularly relates to a network slice resource dynamic allocation device and method. BACKGROUND

[0002] Network slicing is a technology of dividing a physical network into multiple virtual logical networks based on software defined network (SDN) and network function virtualization (NFV) technology, each virtual logical network is called a "slice", and can independently meet the needs of different users or services. The virtual logical networks after network slicing are relatively independent, and different transmission protocols and transmission modes are used to meet the transmission needs of different services. For example, for services with high bandwidth requirements, more spectrum resources can be allocated; for low-latency services, the network path can be optimized to reduce the latency of data transmission. However, the number of transmission nodes in the virtual logical network determines the performance of the virtual logical network, and the more nodes, the higher the performance. Therefore, it is necessary to reasonably allocate network slice resources to ensure that various services can run efficiently.

[0003] Traditional technologies generally allocate network slice resources by the proportion of the information throughput of each network slice resource in the total information throughput of the network. However, the physical structures of the nodes in the network slice are not necessarily the same, so the load capacity for the same information transmission process is also different. The traditional technology does not fully consider the transmission load capacity of the transmission nodes, resulting in low efficiency of the traditional allocation method and affecting the transmission of various service information. SUMMARY

[0004] In view of the above, it is necessary to provide a network slice resource dynamic allocation device and method to solve the above problems.

[0005] The first aspect of the present application provides a network slice resource dynamic allocation method, the method comprising:

[0006] When the virtual logical network is first allocated with network slice resources, a reserved resource amount is obtained; the byte transmission rate, the transmission queue length and the excess bandwidth of each network slice resource at each time are obtained;

[0007] determine a transmission efficiency quantity and a delay hindering quantity of each network slice resource based on distribution characteristics of byte transmission rates of each network slice resource at all time instants in a preset time period and distribution characteristics of transmission queue lengths, and compose a two-dimensional array of each network slice resource by normalizing values of the two quantities; determine a congestion quantity in each virtual logical network based on position characteristics of the two-dimensional array of all network slice resources in the virtual logical network in a coordinate system; determine a transmission hindering degree of each virtual logical network by analyzing a quantity proportion of the congestion quantity in each virtual network and combining overall distribution of the transmission efficiency quantity;

[0008] analyze numerical characteristics of excess bandwidth of each network slice resource at all time instants in a preset time period, filter transmission queue lengths at all time instants, and compose a congestion length sequence of each network slice resource; perform linear fitting on the congestion length sequence of each network slice resource, and obtain a congestion increasing property of each network slice resource in combination with a change trend of element values in the congestion length sequence;

[0009] obtain a transmission load supplement quantity of each virtual logical network in combination with the transmission hindering degree based on overall distribution of the congestion increasing property of all network slice resources in each virtual logical network.

[0010] allocate a reserved resource quantity based on numerical characteristics of the transmission load supplement quantity of each virtual logical network, and obtain an additional resource allocation quantity of each virtual logical network.

[0011] The transmission efficiency quantity of each network slice resource is specifically a mean value of byte transmission rates of each network slice resource at all time instants in a preset time period.

[0012] The delay hindering quantity is specifically a mean value of transmission queue lengths of each network slice resource at all time instants in a preset time period.

[0013] The congestion quantity in each virtual logical network is determined by specifically:

[0014] map the two-dimensional array of all network slice resources in each virtual logical network to a coordinate system, count a quantity of the two-dimensional array above a y=x straight line in the coordinate system, and obtain the congestion quantity.

[0015] The transmission hindering degree of each virtual logical network is determined by specifically:

[0016] perform positive fusion on a negative correlation mapping result of the quantity proportion of the congestion quantity of each logical virtual network and a mean value of the transmission efficiency quantity of all network slice resources in each virtual logical network, and obtain the transmission hindering degree of each virtual logical network.

[0017] The congestion length sequence of each network slice resource is composed of the congestion length of each network slice resource in the preset time period.

[0018] The time point at which the excess bandwidth of each network slice resource in the preset time period is equal to 0 is obtained as a congestion time point; and the transmission queue length of each network slice resource at each congestion time point in the preset time period is used to compose the congestion length sequence of each network slice resource.

[0019] The congestion increase of each network slice resource is obtained by the following steps.

[0020] The minimum element of the congestion length sequence of each network slice resource is obtained, the mean difference between all elements in the congestion length sequence and the minimum element is calculated, and the normalized result of the slope of the fitting straight line of the congestion length sequence is positively fused to obtain the congestion increase of each network slice resource.

[0021] The transmission load supplement of each virtual logical network is obtained by the following steps.

[0022] The accumulation sum of the congestion increase of all network slice resources in each virtual logical network is calculated, and the transmission load supplement of each virtual logical network is positively fused with the transmission hindering degree of each virtual logical network to obtain the transmission load supplement of each virtual logical network.

[0023] The additional resource allocation of each virtual logical network is the product of the proportion of the transmission load supplement of each virtual logical network in the transmission load supplement of all virtual logical networks and the reserved resource amount.

[0024] In a second aspect, the embodiments of the present application further provide a network slice resource dynamic allocation device, which comprises a memory, a processor, and a computer program stored in the memory and running on the processor, and the processor implements the steps of the method in any of the above aspects when executing the computer program.

[0025] The present application has at least the following beneficial effects:

[0026] The present application accurately evaluates the actual load capacity of the virtual logical network by collecting detailed data of the network slice resource and combining the transmission hindering degree and the congestion increase. Compared with the traditional technology, the present application avoids the problem of uneven allocation caused by not considering the node load capacity, significantly improves the accuracy of resource allocation, and ensures that the resource allocation is more in line with business needs.

[0027] Based on the accurate resource allocation, the present application can dynamically adjust the network slice resource allocation strategy: by calculating the transmission hindering degree and the congestion increase, the network bottleneck is found in time and the resource is allocated to relieve congestion, which effectively reduces the network congestion and transmission delay, significantly improves the overall transmission efficiency of the network, and guarantees the efficient operation of high-bandwidth and low-latency services.

[0028] The present application not only considers the current load, but also predicts future load changes through congestion increase. By dynamically allocating the reserved resource amount, the service demand fluctuation can be responded in time, and the network can still run efficiently when the traffic or service grows, which greatly enhances the stability and scalability of the network, and provides reliable support for complex network environment and service growth. BRIEF DESCRIPTION OF DRAWINGS

[0029] Figure 1 A step flowchart of a network slice resource dynamic allocation method provided by an embodiment of the present application is provided.

[0030] Figure 2 A flowchart of obtaining an additional resource allocation amount provided by an embodiment of the present application is provided. DETAILED DESCRIPTION

[0031] In the description of the embodiments of the present application, the words "exemplary", "or", "for example" are used to mean as an example, instance, or illustration. Any embodiment or design solution described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or having more advantages than other embodiments or design solutions. Rather, the use of "exemplary", "or", "for example" is intended to present the relevant concept in a specific manner.

[0032] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs. The terms used in the specification of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application.

[0033] In addition, it should be noted that the terms "first", "second" in the present application and the drawings are used to distinguish similar objects, and are not intended to describe a specific order or sequence. The method disclosed in the embodiments of the present application or the method shown in the flowchart includes one or more steps for implementing the method, and the execution order of the steps can be interchanged with each other without departing from the scope of the present application, and some steps can also be deleted.

[0034] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs.

[0035] The specific scheme of the network slice resource dynamic allocation device and method provided by the present application will be specifically described below in combination with the drawings.

[0036] Please refer to Figure 1 which shows a step flowchart of a network slice resource dynamic allocation method provided by an embodiment of the present application, and the method includes the following steps:

[0037] The first step: when the network slice resources of the virtual logical network are allocated for the first time, the amount of reserved resources is obtained; the byte transmission rate, the transmission queue length and the excess bandwidth of each network slice resource at each time are obtained.

[0038] When the network slice resources of the virtual logical network are allocated for the first time, the amount of reserved resources is set, and the proportion of the amount of reserved resources in the total amount of resources is generally defaulted to be 1% to 30%, and in this embodiment, the value is 20%. The remaining network slice resources are evenly divided and then allocated to each service, and in this embodiment, the value of the service type is 3, which can be determined by the specific implementer.

[0039] The transmission information of each network slice resource in each virtual logical network is collected by a protocol analyzer, including the byte transmission rate per second, the queue length per second, and the excess bandwidth per second. The length of time for each collection of the present application is 10 minutes, that is, dynamic allocation is performed every 10 minutes. Each kind of data collected for each network slice resource is removed by the LOF algorithm (Local Outlier Factor) for abnormal data, and the removed data is completed by the linear interpolation method for removed data to form a byte transmission rate sequence, a transmission queue length sequence, and an excess bandwidth sequence of each network slice resource. The calculation of the LOF algorithm and the linear interpolation method is a known technology, and the specific calculation steps are not described here.

[0040] The second step: based on the distribution characteristics of the byte transmission rate and the distribution characteristics of the transmission queue length of each network slice resource at all times within a preset time period, the transmission efficiency amount and the delay hindering amount of each network slice resource are determined, and the normalized values of the two form a two-dimensional array of each network slice resource; based on the position characteristics of the two-dimensional array of all network slice resources in the coordinate system in each virtual logical network, the congestion quantity in each virtual logical network is determined; the proportion of the congestion quantity in each virtual network is analyzed, and the overall distribution of the transmission efficiency amount is combined to determine the transmission hindering degree of each virtual logical network.

[0041] For the network slice resources in the same virtual network, the faster the node transmission rate and the shorter the delay time, the better the transmission efficiency of the node. When the transmission rate of the nodes in the virtual logical network is greatly reduced and the delay time is increased, it indicates that the number of network slice resources in the virtual logical network cannot meet the demand of the service information transmission of the virtual logical network, which affects the overall transmission efficiency. Therefore, the network slice resources in the virtual network need to be increased to ensure efficient transmission of service information.

[0042] The greater the transmission rate of service information is, the faster the network node transmits the service information. Therefore, the mean of the byte transmission rate sequence of each network slice resource is taken as a transmission efficiency quantity of the network slice resource, for representing the transmission efficiency of the network slice resource in the data collection. Similarly, the mean of the transmission queue length sequence of each network slice resource is taken as a delay hindering quantity, for representing the delay efficiency of the network slice resource in the data collection. In order to eliminate the influence of the data dimension, the transmission efficiency quantity and the delay hindering quantity of all network slice resources in the same virtual logical network are normalized respectively, and the maximum normalization method is adopted in the embodiment.

[0043] When the transmission efficiency quantity and the delay hindering quantity of the network slice resources in the virtual network exist differences, it indicates that the network slice resources appear to be not smooth in transmission. The more the network slice resources not smooth in transmission are, the higher the congestion of the virtual logical network is. The transmission efficiency quantity and the delay hindering quantity of the network slice resources form a two-dimensional array. Since the two-dimensional array of the network slice resources smooth in transmission is distributed near (1, 0), and the two-dimensional array of the network slice resources congested is distributed near (0, 1), the line y = x is taken as a dividing line, the number of the two-dimensional arrays on the upper side of the line is taken as a congestion number, and the number of the two-dimensional arrays on the lower side of the line is taken as a smooth number.

[0044] Therefore, the transmission hindering degree of each virtual logical network is calculated, for representing the hindering degree of the virtual logical network in the transmission process. Specifically, the number proportion of the congestion number of each virtual logical network is calculated, and the positive fusion of the negative correlation mapping result of the mean of the transmission efficiency quantity of all network slice resources in each virtual logical network is performed, to obtain the transmission hindering degree of each virtual logical network.

[0045] In the embodiment, the transmission hindering degree of the i th virtual logical network is denoted as , and the formula form is: ; in the formula, , respectively represent the congestion number and the smooth number of the i th virtual logical network; represents the number proportion of the congestion number of the i th virtual logical network; represents the mean of the transmission efficiency quantity of all network slice resources in the i th virtual logical network; represents a preset zero adjustment factor, for preventing the denominator to be 0, and the value is 1 in the embodiment.

[0046] It should be understood that when the queue length of the network slice resource is larger, the amount of data to be transmitted is larger, the possibility of delay of the network slice resource is higher, the data transmission rate can be reduced due to congestion, the ratio of the congestion amount of the network slice resource to the total amount of the network slice resource is larger, and the average of the transmission efficiency amount of all network slice resources in the virtual logical network is smaller, that is, the transmission hindering degree of the virtual logical network is larger. In this case, the larger the load of the network slice resource is, the more network slice resources are needed for transmitting service information to ensure smooth operation of the service.

[0047] The third step is to analyze the numerical characteristics of the excess bandwidth of each network slice resource at all time points in a preset time period, filter the transmission queue lengths at all time points, and form a congestion length sequence of each network slice resource; and linear fitting is performed on the congestion length sequence of each network slice resource, and the trend of the element values in the congestion length sequence is combined to obtain the congestion increase of each network slice resource.

[0048] In the process of data transmission of the network slice resource, the transmission rate of the network slice resource is inversely proportional to the excess bandwidth. When the excess bandwidth is 0, it means that the transmission rate of the network slice resource reaches the maximum. At this time, if the queue length of the network slice resource is still increasing, it means that the possibility of congestion of the network slice resource is increasing, and more resources need to be allocated to the virtual logical network for transmission of service information.

[0049] Taking a virtual logical network as an example, the index positions of all elements being 0 in the excess bandwidth sequence of each network slice resource are obtained, the corresponding time points are recorded as congestion time points, the corresponding elements of all congestion time points in the transmission queue length sequence are obtained, the elements in the obtained transmission queue length sequence are arranged in time sequence, and a congestion length sequence of each network slice resource is formed. The larger the element in the congestion length sequence is, the greater the possibility of congestion of the network slice resource is, and when the element in the congestion length sequence is an increasing sequence, the congestion of the network slice resource is increasing. Therefore, the congestion length sequence is taken as the input of the linear least square method, and the output is the slope of the fitted straight line of the congestion length sequence, wherein the calculation of the linear least square method is a known technology, and the specific calculation steps are not described herein.

[0050] Therefore, the congestion increase of the network slice resource is calculated to represent the congestion of the network slice resource in the next stage. Specifically, the minimum element of the congestion length sequence of each network slice resource is obtained, the difference average between all elements in the congestion length sequence and the minimum element is calculated, and the normalized result of the slope of the fitted straight line of the congestion length sequence is positively fused to obtain the congestion increase of each network slice resource.

[0051] In this embodiment, the congestion increase of each network slice resource in the virtual logical network is denoted as , and the formula is ; wherein represents the slope of the fitting straight line of the congestion length sequence corresponding to the jth network slice resource in the virtual logical network. represents the average value of the difference between each element and the minimum element in the congestion length sequence corresponding to the jth network slice resource in the virtual logical network; and sig() represents the sigmoid function.

[0052] It should be noted that when there is no element in the congestion length sequence corresponding to the network slice resource, it means that the network slice resource is in a smooth transmission state, and the congestion increase of the network slice resource is 0.

[0053] It should be understood that when the congestion of the network slice resource is increasing, the length of the accumulated service information to be transmitted in the queue of the network slice resource is also increasing, so that the slope of the fitting straight line of the congestion length sequence is larger, and at the same time, the difference between each element and the minimum element in the congestion length sequence is also increasing, so that the congestion increase of the network slice resource is larger. In order to alleviate the network congestion of the network slice resource, it is necessary to timely increase the number of allocated network slice resources to improve the transmission efficiency of service information, ensure the smoothness and stability of network operation, and meet the growing business demand.

[0054] The fourth step is to obtain the transmission load supplement of each virtual logical network according to the overall distribution of the congestion increase of all network slice resources in each virtual logical network, in combination with the transmission hindering degree.

[0055] The load of the network slice resource includes the current load and the future load, i.e., the current transmission hindering ability and the congestion increase. Therefore, the stronger the hindering ability and the greater the congestion increase of the network slice resource at the current time, the more service information needs to be transmitted in the virtual logical network, which exceeds the load capacity of the virtual logical network and cannot meet the requirements of service information transmission. Therefore, the transmission load supplement of the virtual logical network is calculated, specifically: the cumulative sum of the congestion increase of all network slice resources in each virtual logical network is calculated, and is positively fused with the transmission hindering degree of each virtual logical network to obtain the transmission load supplement of each virtual logical network. In this embodiment, the calculation method of multiplication is adopted for the positive fusion of multiple variables.

[0056] It should be understood that as the congestion increases, the transmission of service information of each network slice resource in the current virtual network is hindered, that is, the transmission hindrance of the virtual logical network increases. At the same time, this will make the congestion of the network slice resource increase, resulting in a significant increase in the sum of the congestion increase of all network slice resources in the virtual logical network. In this case, more network slice resources are needed to transmit service information at the next moment to relieve the congestion pressure, improve the transmission efficiency, ensure the smooth operation of the network, and meet the increase of service demand.

[0057] Fifth step: according to the numerical characteristics of the transmission load supplement of each virtual logical network, the reserved resource amount is allocated to obtain the additional resource allocation amount of each virtual logical network.

[0058] The proportion of the transmission load supplement of each virtual logical network in the transmission load supplement of all virtual logical networks is multiplied by the reserved resource amount to obtain the additional resource allocation amount of each virtual logical network. The network slice allocator is used to allocate the reserved resource amount through the additional resource allocation amount of the virtual logical network, so that the reserved resource amount is connected with the virtual network, and it is guaranteed that the virtual logical network can use the reserved resource amount to transmit service logic information.

[0059] The flow chart of obtaining the additional resource allocation amount is shown in Figure 2

[0060] Based on the same inventive concept as the above method, the embodiments of the present application also provide a network slice resource dynamic allocation device, which comprises a memory, a processor, and a computer program stored in the memory and running on the processor, and the processor implements the steps of the method described in any one of the above embodiments when executing the computer program.

[0061] ​The computer program product of the present application can be a computer program implemented on one or more computers. The program instructions can be stored on a computer-readable medium, such as a floppy disk, CD-ROM, and the like. The computer program product can also include computer programs that are transmitted over a network via, for example, telephone line, LAN, wireless instrument, or others. Accordingly, the computer program product of the present application can be an article of manufacture including a computer usable medium having computer readable program code means distributed therein. The computer readable program code means is means for causing a computer to operate in a specific and predefined manner. The present application can also be embodied in a computer readable medium including transitory signals. Accordingly, the present application can be a product, an article of manufacture, and / or a machine.

[0062] It is apparent that the present application is not limited to the details of the foregoing exemplary embodiments, and thus modifications and / or improvements can be made thereto without departing from the scope of the present application. Accordingly, no matter from which point of view, the foregoing embodiments of the present application should be considered as exemplary and non-limiting; any modification of the technical solutions described in the foregoing embodiments, or equivalent replacement of some of the technical features, without changing the nature of the corresponding technical solutions, should be included in the scope of protection of the present application.

Claims

1. A method for dynamic allocation of network slice resources, characterized in that, The method includes the following steps: When allocating network slice resources for the virtual logical network for the first time, reserve the amount of resources; obtain the byte transmission rate, transmission queue length and excess bandwidth of each network slice resource at each time point; The average byte transmission rate of each network slice resource at all times within a preset time period is used as the transmission efficiency of each network slice resource; the average transmission queue length of each network slice resource at all times within the preset time period is used as the latency barrier of each network slice resource, and the normalized values ​​of the two are combined to form a two-dimensional array of each network slice resource; based on the position characteristics of the two-dimensional array of all network slice resources in each virtual logical network in the coordinate system, the number of congestions in each virtual logical network is determined; the proportion of congestion in each virtual logical network is calculated, and the negative correlation mapping result with the average transmission efficiency of all network slice resources in each virtual logical network is positively fused to obtain the transmission barrier of each virtual logical network; The numerical characteristics of excess bandwidth of each network slice resource at all times within a preset time period are analyzed, and the transmission queue length at all times is filtered to form a congestion length sequence for each network slice resource. The congestion length sequence of each network slice resource is fitted with a straight line, and the congestion increase of each network slice resource is obtained by combining the changing trend of the element values ​​in the congestion length sequence. Based on the overall distribution of congestion increase of all network slice resources in each virtual logical network, and combined with the transmission impedance, the transmission load replenishment amount of each virtual logical network is obtained. Based on the numerical characteristics of the transmission load supplement of each virtual logical network, the reserved resources are allocated to obtain the additional resource allocation for each virtual logical network. The congestion increase of each network slice resource is obtained as follows: Obtain the minimum element of each network slice resource in the congestion length sequence, calculate the mean difference between all elements in the congestion length sequence and the minimum element, and positively fuse it with the normalized result of the slope of the fitted line of the congestion length sequence to obtain the congestion increase of each network slice resource.

2. The method for dynamic allocation of network slice resources as described in claim 1, characterized in that, The specific process for determining the number of congestions in each virtual logical network is as follows: Map the two-dimensional arrays of all network slice resources in each virtual logical network to a coordinate system, and count the number of two-dimensional arrays above the y=x line in the coordinate system to obtain the number of congested areas; the two-dimensional arrays of network slice resources with smooth network access are distributed around (1,0), and the two-dimensional arrays of network slice resources with congested network access are distributed around (0,1).

3. The method for dynamic allocation of network slice resources as described in claim 1, characterized in that, The congestion length sequence that makes up each network slice resource is specifically as follows: The moment when the excess bandwidth of each network slice resource is equal to 0 within a preset time period is taken as the congestion moment; the transmission queue length of each network slice resource at each congestion moment within the preset time period is used to form the congestion length sequence of each network slice resource.

4. The method for dynamic allocation of network slice resources as described in claim 1, characterized in that, The specific method for obtaining the transmission load supplement for each virtual logical network is as follows: Calculate the cumulative sum of the congestion increase of all network slice resources in each virtual logical network, and positively fuse it with the transmission impediment of each virtual logical network to obtain the transmission load supplement of each virtual logical network.

5. The method for dynamic allocation of network slice resources as described in claim 1, characterized in that, The additional resource allocation for each virtual logical network is specifically the product of the proportion of the transmission load supplement of each virtual logical network to the transmission load supplement of all virtual logical networks and the reserved resource amount.

6. A network slice resource dynamic allocation device, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1-5.

Images