Periodic processing method and system for deterministic service

Abstract

The embodiment of the invention provides a periodical processing method and system for a deterministic service, and the method comprises the steps: obtaining service parameters of the deterministic service, the service parameters comprise a target probability, a service type and a message length parameter, determining a target message length according to a preset probability distribution model, the message length parameter and the target probability, and transmitting the target message length to the deterministic service, and determining the period interval of the deterministic service according to the service type and the target message length. According to the embodiment of the invention, the problem of large bandwidth waste caused by arrangement of variable-length services in related technologies can be solved, the reasonable arrangement message length is estimated in combination with the probability distribution model, and periodic processing is carried out on the services according to the estimated message length, so that the complexity of message arrangement can be reduced, and bandwidth resources can be saved.

Description

Technical Field

[0001] This application relates to the field of communications, and more particularly, to a method and system for periodic processing of deterministic services. Background Art

[0002] Deterministic network technology is currently a key area of research in the industry and is one of the four enabling technologies for future 6G communications. Currently, standardization-related work is being carried out in multiple standardization organizations. In the field of packet data communication, in order to meet the requirements of low latency and low jitter services, there are multiple technical routes, each corresponding to different application scenarios. Among them, in Wide Area Network (WAN) and local Area Network (LAN), a time-period-based scheduling method is used.

[0003] In the prior art, the time period is generally of a fixed size, or when there are multiple time periods, the size of each period is also fixed. For example: 1T, 2T, 4T, 8T (here T represents a time size, for example, T can be set to 10 us, note that us = microsecond). For each period, there is a series of queues to support it, and the length of each type of queue is different. For example, according to the port rate R = 10 Gbps and T = 10 us, the queue length corresponding to period 1T is 100,000 bit, the queue length corresponding to period 2T is 200,000 bit, and the queue length corresponding to period 4T is 400,000 bit.

[0004] Figure 1 It is a network diagram of deterministic services in the related art. As Figure 1 shown, in deterministic forwarding, usually a control unit arranges the path according to parameters such as bandwidth and latency. Usually, the traffic data of multiple Provider Edge (PE) nodes will converge at an operator node (such as P1 node). There will be a problem of packet overload in a certain queue of P1. Therefore, it is necessary to schedule and arrange services according to the period.

[0005] In order to improve bandwidth utilization and the number of service accesses, the industry is exploring a technology for precise periodic orchestration based on the control plane. This technology calculates the path and reserves resources based on the remaining bit amount of the periodic queue. Since periodic scheduling is a cyclic scheduling, similar to periodic operation, when performing orchestration, it is assumed that the service is a periodic service. A periodic service means that this service packet will arrive at regular intervals, which can be one packet or multiple packets.

[0006] Figure 2 It is a schematic diagram of periodic service orchestration in the related art. AsFigure 2 As shown in Figure 2 , there are two scheduling cycles: one is a cycle of 80 us, with 8 queues of 10 us in each cycle; the other is a cycle of 320 us, with 32 queues of 10 us in each cycle. Now there is a service that sends a message every 125 us. It can be seen that for the 80 us cycle scheduling, when arranging services, only one queue can be arranged in one cycle to place this service message. For the 320 us cycle scheduling, at least 3 queues need to be arranged in one cycle to place this service message. However, there is a prerequisite for this arrangement method, that is, the service message is a periodic message and the message length is fixed.

[0007] It should be noted that even for periodic services, when the message length is variable, there will also be a problem of how to determine a message length for the scheduling module to perform scheduling to avoid overloading the messages in the periodic queue due to aggregation conflicts. In the related technologies, either a large amount of bandwidth is reserved to avoid aggregation conflicts, or scheduling is performed according to the maximum message length. Both of these two solutions will cause relatively large bandwidth waste.

[0008] In summary, for variable-length services, there is no good solution in the industry on how to perform periodic processing and select an appropriate message length for scheduling. Summary of the Invention

[0009] Embodiments of the present application provide a method and system for periodic processing of deterministic services, so as to at least solve the problem of relatively large bandwidth waste caused by scheduling variable-length services in the related technologies.

[0010] According to an embodiment of the present application, a method for periodic processing of deterministic services is provided. The method includes: obtaining service parameters of a deterministic service, where the service parameters include a target probability, a service type, and a message length parameter; determining a target message length according to a preset probability distribution model, the message length parameter, and the target probability; and determining a period interval of the deterministic service according to the service type and the target message length.

[0011] According to another embodiment of the present application, a system for periodic processing of deterministic services is provided. The system is characterized in that the system includes: a service management unit, configured to send service parameters of a deterministic service to a periodic processing unit, where the service parameters include a target probability, a service type, and a message length parameter; a periodic processing unit, configured to perform periodic processing on the service parameters according to the steps in any one of the above method embodiments, and send the periodically processed service parameters to a periodic scheduling unit; and a periodic scheduling unit, configured to perform service scheduling according to the periodically processed service parameters.

[0012] According to another embodiment of the present application, there is also provided a computer-readable storage medium storing a computer program, wherein when the computer program is run by a processor, the steps in any one of the above method embodiments are executed.

[0013] According to another embodiment of the present application, there is also provided an electronic device including a memory and a processor, where the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0014] Through the embodiments of the present application, a probability distribution model is adopted to implement the periodic processing of variable-length services, which can solve the problem that the scheduling of variable-length services in the related art causes a large amount of bandwidth waste. By combining the probability distribution model, a reasonable scheduling message length is estimated, and the services are periodically processed according to the estimated message length, which can not only reduce the complexity of message scheduling but also save bandwidth resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a schematic diagram of the networking of deterministic services in the related art;

[0016] Figure 2 is a schematic diagram of periodic service scheduling in the related art;

[0017] Figure 3 is a flowchart of the method for periodic processing of deterministic services according to an embodiment of the present application;

[0018] Figure 4 is a distribution schematic diagram of a skewed distribution in an embodiment of the present application;

[0019] Figure 5 is a structural block diagram of the system for periodic processing of deterministic services according to an embodiment of the present application;

[0020] Figure 6 is a schematic flowchart of the periodic processing of deterministic services in an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0021] In the following, the embodiments of the present application will be described in detail with reference to the drawings and in conjunction with the embodiments.

[0022] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above drawings are used to distinguish similar objects and do not necessarily need to describe a specific order or sequence.

[0023] The method embodiments provided in the embodiments of the present application can be executed in network nodes such as computer terminals, routers, gateways, etc.

[0024] In an embodiment of the present application, a method for periodic processing of deterministic services is provided. Figure 3 It is a flowchart of the method for periodic processing of deterministic services according to an embodiment of the present application. As Figure 3 shown, the process includes the following steps:

[0025] Step S302, obtain the service parameters of the deterministic service, where the service parameters include the target probability, service type, and packet length parameter;

[0026] Step S304, determine the target packet length according to the preset probability distribution model, the packet length parameter, and the target probability;

[0027] Step S306, determine the periodic interval of the deterministic service according to the service type and the target packet length.

[0028] In this embodiment, the target probability is used to indicate how many probabilities of packets the target packet length in step S304 should satisfy. The target probability can be set in advance, or by default, it is calculated as the probability 99.73% corresponding to 3σ (standard deviation) in the standard normal distribution to calculate the packet length.

[0029] In the embodiment of the present application, through the above steps S302 to S306, a reasonable arranged packet length can be estimated by combining the probability distribution model, and the service is processed periodically according to the estimated packet length to determine the periodic interval. It can not only reduce the complexity of packet arrangement by uniformly arranging the packet length, but also avoid wasting bandwidth resources by reserving a large bandwidth or arranging according to the maximum packet length in the traditional method, thus solving the problem of large bandwidth waste caused by the arrangement of variable-length services in the related art.

[0030] In some embodiments, the service parameters further include at least one of the following: service identifier, periodic interval, bandwidth parameter, minimum packet length, maximum packet length, mode, and burst number.

[0031] In this embodiment, the service identifier (serviceID) is used to identify the service flow, which can be a parameter or composed of multiple parameters jointly to identify a service flow; the periodic interval (interval) is only valid when the service type is the periodic type; the bandwidth parameter (reqBW) is an optional parameter; the minimum packet length (minPktLen) is less than or equal to the maximum packet length (maxPktLen); the mode is an optional parameter, and when there is no mode, it can be calculated according to the minimum packet length and the maximum packet length; the burst number (brustNum) is an optional parameter used to indicate the number of packets arriving simultaneously, and when there is no burst number, it is defaulted to 1.

[0032] In some embodiments, the probability distribution model includes a normal distribution model or a non - normal distribution model (such as a skewed distribution).

[0033] In some embodiments, step S304 of determining the target packet length according to a preset probability distribution model, the packet length parameter, and the target probability may include the following steps:

[0034] Step S3042: Determine the median and standard deviation of the packet length according to the packet length parameter;

[0035] Step S3044: Determine the target standard deviation coefficient corresponding to the target probability according to the probability distribution model;

[0036] Step S3046: Determine the target packet length according to the median, the standard deviation, and the target standard deviation coefficient.

[0037] In some embodiments, when the packet length parameter only includes the maximum packet length and the minimum packet length, step S3042 of determining the median and standard deviation of the packet length according to the packet length parameter includes: determining half of the sum of the maximum packet length and the minimum packet length as the median; dividing the difference between the median and the minimum packet length into multiple distribution intervals according to the preset number of distribution intervals to obtain the standard deviation, where each distribution interval corresponds to one standard deviation.

[0038] In this embodiment, the number of distribution intervals (i.e., the number of divisions of the standard variance) can be set according to the accuracy requirement. The larger the number of distribution intervals, the higher the accuracy. Exemplarily, the maximum number of distribution intervals only needs to be set to 5 to meet a relatively high accuracy requirement, but the present application is not limited thereto, and the number of distribution intervals can also be greater than 5 or less than 5.

[0039] In some embodiments, if the median of the packet length is not specified in the packet length parameter, the median of the packet length can be calculated and used to replace the median. Exemplarily, the median can be calculated in the following way:

[0040] mode=(minPktLen + maxPktLen) / 2;

[0041] where mode is the median, minPktLen is the minimum packet length, and maxPktLen is the maximum packet length.

[0042] In some other embodiments, the median parameter can be directly obtained from the packet length parameter. Specifying the median usually occurs in scenarios where the packet length shows a skewed normal distribution, including positive skew and negative skew.

[0043] In some embodiments, when the packet length parameter includes the maximum packet length, the minimum packet length, and the mode, step S3046 determines the mode and standard deviation of the packet length according to the packet length parameter, including: directly obtaining the mode from the packet length parameter; determining the median of the packet length as half of the sum of the maximum packet length and the minimum packet length; when the median is greater than or equal to the mode, dividing the difference between the mode and the minimum packet length into multiple distribution intervals according to the preset number of distribution intervals to obtain the standard deviation, where each distribution interval corresponds to one standard deviation; when the median is less than the mode, dividing the difference between the maximum packet length and the mode into multiple distribution intervals according to the number of distribution intervals to obtain the standard deviation, where each distribution interval corresponds to one standard deviation.

[0044] In some embodiments, the median can be calculated in the following manner:

[0045] medianNum = (minPktLen + maxPktLen) / 2;

[0046] where medianNum is the median, minPktLen is the minimum packet length, and maxPktLen is the maximum packet length.

[0047] In some embodiments, taking the preset number of distribution intervals as 5 as an example, the standard deviation can be calculated in the following manner:

[0048]

[0049] where σ is the standard deviation, mode is the mode, medianNum is the median, minPktLen is the minimum packet length, and maxPktLen is the maximum packet length.

[0050] In some embodiments, the probability distribution model includes the mapping relationship between the standard deviation coefficient and the distribution probability value.

[0051] In an exemplary embodiment, the probability distribution model is a normal distribution model, and its mapping relationship between the standard deviation coefficient and the distribution probability value is shown in Table 1 below:

[0052] Table 1

[0053] Coefficient of standard deviation (N) Distribution probability value 1 68.2689492％ 1.28 80％ 1.64 90％ 2 95.4499736％ 3 99.7300204％ 4 99.993666％ 5 99.999943％

[0054] In this embodiment, the accuracy of the distribution probability value and the number of table entries can be adjusted according to requirements, and the present application does not limit this.

[0055] In this embodiment, step S3044 determines the target standard deviation coefficient corresponding to the target probability according to the probability distribution model, including: traversing the probability distribution model in ascending order of the distribution probability values, and determining the first distribution probability value greater than or equal to the target probability as the target distribution probability value; determining the standard deviation coefficient corresponding to the target distribution probability value as the target standard deviation coefficient according to the mapping relationship.

[0056] In an exemplary embodiment, as shown in Table 1, if the target probability is 99%, traversing Table 1 shows that the first target distribution probability value greater than or equal to 99% is 99.7300204%, and the corresponding standard deviation coefficient N = 3, so the target standard deviation coefficient is 3.

[0057] In some embodiments, step S3046 determines the target message length according to the median, the standard deviation, and the target standard deviation coefficient, including: determining the product of the standard deviation and the target standard deviation coefficient as the target standard deviation; determining the sum of the median and the target standard deviation as the target message length.

[0058] In some embodiments, the target message length can be determined in the following manner:

[0059] aptPktLen = mode + N * σ;

[0060] where aptPktLen is the target message length, mode is the median, N is the target standard deviation coefficient, and σ is the standard deviation.

[0061] Furthermore, since the message length is an integer, the calculated target message length needs to be rounded.

[0062] In some embodiments, the service types include: variable-length periodic service, variable-length bandwidth service, or variable-length burst service.

[0063] In some embodiments, step S306, determining the period interval of the deterministic service according to the service type and the target message length, may include the following methods:

[0064] Method 1, when the service type is the variable-length periodic service, obtaining the preset period interval from the service parameters;

[0065] Method 2, when the service type is the variable-length bandwidth service, obtaining the preset bandwidth parameter from the service parameters, and determining the ratio of the target message length to the bandwidth parameter as the period interval;

[0066] In Method 3, when the service type is the variable-length burst service, obtain the preset bandwidth parameter and the number of bursts from the service parameters, and determine the cycle interval as the ratio of the product of the target packet length and the number of bursts to the bandwidth parameter.

[0067] Through the embodiments of the present application, a suitable arranged packet length can be selected in combination with the probability distribution model. After obtaining the suitable arranged packet length, calculate the cycle interval of the service packet to implement the periodic processing of the service packet, and then send the processing result to the cycle arrangement unit for packet arrangement, which solves the problem of large bandwidth waste caused by the arrangement of variable-length services in the related art, can not only reduce the complexity of packet arrangement, but also save bandwidth resources.

[0068] Figure 4 is a distribution schematic diagram of the skewed distribution in an embodiment of the present application, as Figure 4 shown, the packet distribution can be skewed, including two types: positive skewed distribution and negative skewed distribution.

[0069] In this embodiment, when the service packet length is skewed, the median is usually specified in the packet length parameter.

[0070] In this embodiment, the standard deviation in the skewed distribution can be calculated by the following method:

[0071]

[0072] where σ is the standard deviation, mode is the median, medianNum is the median, minPktLen is the minimum packet length, and maxPktLen is the maximum packet length.

[0073] Through the embodiments of the present application, regardless of whether the service packet length is normally distributed, the standard deviation can be calculated according to the method in the above embodiments, and the application scenario and scope of application are more extensive.

[0074] The embodiments of the present application also provide a periodic processing system for deterministic services.

[0075] Figure 5 is a structural block diagram of the periodic processing system for deterministic services according to the embodiments of the present application, as Figure 5 shown, the system includes the following structure:

[0076] A service management unit 52, configured to send the service parameters of the deterministic service to the periodic processing unit, where the service parameters include target probability, service type, and packet length parameter;

[0077] The periodic processing unit 54 is configured to perform periodic processing on the service parameters according to the steps in any of the above method embodiments, and send the periodically processed service parameters to the periodic scheduling unit;

[0078] The periodic scheduling unit 56 is configured to perform service scheduling according to the periodically processed service parameters.

[0079] In some embodiments, the service parameters passed from the service management unit 52 to the periodic processing unit 54 may further include any one of the following parameters or a combination of multiple parameters: service identifier (serviceID), period interval (interval), bandwidth parameter (reqBW), minimum packet length (minPktLen), maximum packet length (maxPktLen), mode, and burst number (brustNum).

[0080] In some embodiments, the periodic processing unit 54 sends the periodically processed service parameters to the periodic scheduling unit to facilitate the periodic queue-related scheduling of services by the periodic scheduling unit. The periodically processed service parameters may include: service identifier, period interval, target packet length (aptPktLen), and packet number (PktNum). The packet number is an optional parameter with a default value of 1.

[0081] Through the embodiments of the present application, the periodic processing unit can obtain specified parameters from the service management unit, implement the periodic processing of variable-length packets, and then send the periodically processed service parameters to the periodic scheduling unit for the periodic queue scheduling of services, which can reduce the complexity of packet scheduling and save bandwidth resources on the premise of avoiding aggregation conflicts.

[0082] Figure 6 is a schematic flow diagram of the periodic processing of deterministic services in an embodiment of the present application, as Figure 6 shown, the process includes the following steps:

[0083] Step S601, obtain service parameters from the service management unit;

[0084] Step S602, calculate the standard deviation according to Formula 1;

[0085] Step S603, obtain the standard deviation coefficient based on Algorithm 1;

[0086] Step S604, determine the target packet length according to Formula 2;

[0087] Step S605, classify according to the service type;

[0088] Step S606: Send the calculated relevant parameters to the periodic scheduling unit, including: service identifier, period interval, target packet length, number of bursts.

[0089] Step S607: Calculate the appropriate period interval INVL_i according to Formula 3, and send the relevant parameters to the periodic scheduling unit, including: service identifier, INVL_i, target packet length, 1 (number of bursts = 1).

[0090] Step S608: Calculate the appropriate period interval INVL_Bi according to Formula 4, and send the relevant parameters to the periodic scheduling unit, including: service identifier, INVL_Bi, target packet length, number of bursts.

[0091] In this embodiment, the service parameters obtained in Step S601 include: service identifier (serviceID), service type (servicetype), period interval (interval), bandwidth parameter (reqBW), minimum packet length (minPktLen), maximum packet length (maxPktLen), median number (mode), number of bursts (brustNum), target probability (probability).

[0092] In this embodiment, in Step S602, first use the two parameters minPktLen and maxPktLen to obtain the median value of the packet length medianNum = (minPktLen + maxPktLen) / 2, and then use the median number mode (when there is no such parameter, it is equal to the median), and calculate the standard deviation value σ according to the following Formula 1.

[0093] Formula 1:

[0094]

[0095] In this embodiment, in Step S603, according to the target probability probablity (when there is no such parameter, it is default to 99.73% corresponding to 3σ), obtain the standard deviation coefficient N according to the following Algorithm 1.

[0096] In an exemplary embodiment, the pseudocode of Algorithm 1 is as follows:

[0097]

[0098] In this embodiment, in Step S604, according to the previously calculated standard deviation value σ, standard deviation coefficient N, and median number mode, calculate the appropriate target packet length aptPktLen according to the following Formula 2.

[0099] Formula 2:

[0100] aptPktLen = mode + N * σ

[0101] In this embodiment, in step S605, the subsequent processing flow is determined according to the service type service-type. If it is a variable-length cycle type service, it goes to step S606. If it is a variable-length bandwidth type service, it goes to step S607. If it is a variable-length burst type service, it goes to step S608.

[0102] In this embodiment, in step S606, for the variable-length cycle type service (non-variable-length services can also be processed in the same way), the target packet length aptPktLen calculated previously is used as the packet length, and combined with the service identifier serviceID, interval period interval, and burst number brustNum obtained from the service management unit, and transmitted to the cycle scheduling unit according to the interface parameter requirements and order of the cycle scheduling unit. The transmitted parameters are: (serviceID, interval, aptPktLen, brustNum), and it ends.

[0103] In this embodiment, in step S607, for the variable-length bandwidth type service, the aptPktLen calculated previously is used as the packet length, and combined with the bandwidth parameter reqBW (Byte / s) obtained from the service management unit. According to the following formula three, the cycle interval INVL_i of the bandwidth type service is obtained, and it is transmitted to the cycle scheduling unit according to the interface parameter requirements and order of the cycle scheduling unit. The input parameters are: (serviceID, INVL_i, aptPktLen, 1).

[0104] Formula three:

[0105] INVL_i = aptPktLen / reqBW

[0106] In this embodiment, in step S608, for the variable-length burst type service, the aptPktLen calculated previously is used as the packet length, and combined with the bandwidth parameter reqBW and burst number brustNum obtained from the service management unit. According to the following formula four, the cycle interval INVL_Bi of the burst type service is obtained, and it is transmitted to the cycle scheduling unit according to the interface parameter requirements and order of the cycle scheduling unit. The input parameters are: (serviceID, INVL_Bi, aptPktLen, brustNum).

[0107] Formula four:

[0108] INVL_Bi = (aptPktLen * brustNum) / reqBW

[0109] Through the embodiments of the present application, reasonable message scheduling for variable-length services of any type can be achieved, reducing the complexity of message scheduling, reducing bandwidth resources, and thus solving the problem of large bandwidth waste caused by the scheduling of variable-length services in the related art.

[0110] The variable-length periodic service, variable-length burst service, and variable-length bandwidth service will be specifically described below in combination with specific service parameters.

[0111] In an exemplary embodiment, assume there is a variable-length periodic service with the following service characteristics:

[0112] Service identifier (serviceID): Source IP: 168.1.1.1, destination IP: 192.1.1.1;

[0113] Minimum packet length (minPktLen): 128 bytes;

[0114] Maximum packet length (maxPktLen): 1500 bytes;

[0115] Period interval (interval): 125 us;

[0116] Target probability (probablity): 95%;

[0117] Number of packet bursts (brustNum): 1;

[0118] Assume the variable-length periodic service is defined as: 1;

[0119] The bandwidth parameter reqBW being 0 means there is no configured bandwidth requirement;

[0120] The median Mode being 0 means there is no configured median;

[0121] The service management unit (or application software) passes the relevant parameters to the periodic processing unit, and the content of the parameters sent down is as follows: (168.1.1.1, 192.1.1.1, 1, 125, 0, 128, 1500, 0, 1, 0.95), as specifically shown in Table 2 below:

[0122] Table 2

[0123] Parameter name Parameter value Description Service identifier serviceID 168.1.1.1,192.1.1.1 Two parameters jointly constitute the service identifier Service type service-type 1 1 indicates periodic service and can be customized Period interval interval 125 Unit is μs Bandwidth parameter reqBW 0 0 indicates that bandwidth configuration is not required Minimum packet length minPktLen 128 Maximum packet length maxPktLen 1500 Median mode 0 0 indicates no configuration Number of packet bursts brustNum 1 Target probability probablity 0.95 Indicates a coverage probability of 95%

[0124] In this embodiment, the processing flow of the periodic processing unit includes the following steps:

[0125] 1. Receive the service parameters shown in Table 2;

[0126] 2. First, using the two parameters minPktLen and maxPktLen, the median of the packet length is obtained as medianNum = (128 + 1500) / 2 = 814. Since mode is not configured, mode = medianNum = 814. According to Formula 1, the standard deviation value σ = (814 - 128) / 5 = 137.2;

[0127] 3. According to the incoming probablity = 0.95, the standard deviation coefficient N = 2 is obtained according to Algorithm 1;

[0128] 4. Based on the previously calculated standard deviation value σ = 137.2, the standard deviation multiple value N = 2, and the median mode = 814, according to Formula 2, the appropriate target packet length aptPktLen = 814 + 2 * 137.2 = 1088.4. After rounding, aptPktLen = 1088;

[0129] 5. The service type service - type = 1 represents a variable - length periodic service;

[0130] 6. The service is a variable - length periodic service. Taking the aptPktLen calculated above as the packet length, combined with the serviceID, interval, and brustNum passed in by the service management unit; it is passed to the periodic scheduling unit according to the interface parameter requirements and order of the periodic scheduling unit. The parameters sent are as follows: (168.1.1.1, 192.1.1.1, 125, 1088, 1), and the specific description is shown in Table 3 below:

[0131] Table 3

[0132] Parameter name Parameter value Description Service identifier serviceID 168.1.1.1,192.1.1.1 Two parameters jointly constitute the service identifier Period interval interval 125 Unit is μs Packet length pktLen 1088 Unit is byte Number of packet bursts brustNum 1

[0133] Through the embodiments of the present application, it is possible to perform periodic processing on the service parameters of variable - length periodic services, determine reasonable packet scheduling lengths and period intervals, avoid aggregation conflicts, save bandwidth resources, and solve the problem of large bandwidth waste caused by the scheduling of variable - length services in related technologies.

[0134] In an exemplary embodiment, assume there is a variable - length bandwidth service with the following service characteristics:

[0135] Service identification (serviceID): flowid (traffic identifier) = 100;

[0136] Minimum packet length (minPktLen): 64 bytes;

[0137] Maximum packet length (maxPktLen): 9000 bytes;

[0138] Bandwidth parameter (reqBW): 1 Gbps;

[0139] Target probability: 99%;

[0140] Mode: 1500;

[0141] Number of message bursts (brustNum): 1;

[0142] Assume that the variable-length bandwidth service is defined as: 2;

[0143] The cycle interval interval being 0 means that there is no configured service cycle;

[0144] The service management unit passes the relevant parameters to the periodic processing unit, and the parameters sent down are: (100, 2, 0, 1000, 64, 9000, 1500, 1, 0.99), and the specific description is shown in Table 4 below:

[0145] Table 4

[0146]

[0147] In this embodiment, the processing flow of the periodic processing unit includes the following steps:

[0148] 1. Receive the service characteristic parameters sent down by the service management unit, etc., as shown in Table 4 above;

[0149] 2. First, using the two parameters minPktLen and maxPktLen, obtain the median medianNum of the message length = (64 + 9000) / 2 = 4532, mode = 1500. According to Formula 1, medianNum > mode, and the standard deviation value σ = (1500 - 64) / 5 = 287.2;

[0150] 3. According to the incoming probablity = 0.99, obtain the standard deviation multiple value N = 3 according to Algorithm 1; / 4. According to the previously calculated standard deviation value σ = 287.2, the standard deviation multiple value N = 3, and the mode = 1500, calculate the appropriate message length aptPktLen = 1500 + 3 * 287.2 = 2361.6 according to Formula 2, and round it up, aptPktLen = 2361;

[0151] 4. The service type service-type = 2 is a variable-length bandwidth service

[0152] 5. The service is a variable-length bandwidth service. Substitute aptPktLen = 2361 obtained from the previous calculation and reqBW = 1000 passed in by the service management unit into Formula 3 to get INVL_i = 2361 * 8 / 1000 = 18.888 us (note that aptPktLen is multiplied by 8 here to convert bytes (BYTE) to bits (bit) to keep the unit consistent with the bandwidth). Then, according to the requirements and order of the interface parameters of the period scheduling unit, pass them to the period scheduling unit. The parameters sent down are: (100, 18.888, 2361, 1), and the specific description is shown in Table 5 below:

[0153] Table 5

[0154] Parameter name Parameter value Description Service identifier serviceID 100 Flowid in the packet, and the parsing rule needs to be agreed upon Period interval interval 18.888 Unit is μs Packet length pktLen 2361 Unit is byte Number of packet bursts brustNum 1

[0155] Through the embodiments of the present application, it is possible to perform periodic processing on the service parameters of variable-length bandwidth services, determine reasonable message scheduling lengths and period intervals, avoid aggregation conflicts, save bandwidth resources, and solve the problem of large bandwidth waste caused by the scheduling of variable-length services in the related art.

[0156] In an exemplary embodiment, assume there is a variable-length burst service, and its service characteristics are as follows:

[0157] Service identification (serviceID): Index index = 100;

[0158] Minimum message length (minPktLen): 256 bytes;

[0159] Maximum message length (maxPktLen): 2048 bytes;

[0160] Bandwidth requirement (reqBW): 500 Mbps;

[0161] Probability: 99%;

[0162] Mode: 1500;

[0163] Number of message bursts (brustNum): 3;

[0164] Service type: 3. Assume that the variable-length burst service is defined as 3;

[0165] The period interval interval being 0 means that the service period is not configured;

[0166] The service management unit passes the relevant parameters to the periodization processing unit. The parameters sent down are as follows: (100, 3, 0, 500, 256, 2048, 1500, 3, 0.99), and the specific description is shown in Table 6 below:

[0167] Table 6

[0168]

[0169] In this embodiment, the processing flow of the periodic processing unit includes the following steps:

[0170] 1. Receive the service characteristic parameters issued by the service management unit, etc., as shown in Table 6 above;

[0171] 2. First, use the two parameters minPktLen and maxPktLen to obtain the median of the packet length medianNum = (256 + 2048) / 2 = 1152, mode = 1500. According to Formula 1, medianNum < mode, and the standard deviation value σ = (2048 - 1500) / 5 = 109.6;

[0172] 3. According to the incoming probablity = 0.99, obtain the standard deviation multiple value N = 3 according to Algorithm 1;

[0173] 4. Based on the previously calculated standard deviation value σ = 109.6, standard deviation multiple value N = 3, and mode value mode = 1500, calculate the appropriate packet length aptPktLen = 1500 + 3 * 109.6 = 1828.8 according to Formula 2. After rounding, aptPktLen = 1828;

[0174] 5. The service type service-type = 3 is a variable-length burst service

[0175] 6. Since the service is a variable-length burst service, substitute the previously calculated aptPktLen = 1828, the reqBW = 500 and brustNum = 3 passed in by the service management unit into Formula 4 to get INVL_Bi = 1828 * 8 * 3 / 500 = 87.744us (note that aptPktLen is multiplied by 8 here to convert bytes BYTE to bits bit to be consistent with the unit of bandwidth), and pass it to the periodic scheduling unit according to the interface parameter requirements and order of the periodic scheduling unit. The issued parameters are: (100, 87.744, 1828, 3), and the specific description is shown in Table 7 below:

[0176] Table 7

[0177]

[0178] Through the embodiments of the present application, it is possible to perform periodic processing on the service parameters of variable-length burst services, determine a reasonable message arrangement length and period interval, avoid aggregation conflicts, save bandwidth resources, and solve the problem of large bandwidth waste caused by the arrangement of variable-length services in the related art.

[0179] The embodiments of the present application also provide a computer-readable storage medium, in which a computer program is stored. When the computer program is run by a processor, it executes the steps in any one of the above method embodiments.

[0180] In an exemplary embodiment, the above computer-readable storage medium may include, but is not limited to: various media such as USB flash drives, read-only memories (ROM for short), random access memories (RAM for short), mobile hard disks, magnetic disks, or optical discs that can store computer programs.

[0181] The embodiments of the present application also provide an electronic device, including a memory and a processor. A computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0182] In an exemplary embodiment, the above electronic device may further include a transmission device and an input / output device, where the transmission device is connected to the above processor, and the input / output device is connected to the above processor.

[0183] Specific examples in this embodiment may refer to the examples described in the above embodiments and exemplary embodiments, and will not be repeated here.

[0184] Obviously, those skilled in the art should understand that the above modules or steps of the present application can be implemented by a general-purpose computing device. They can be concentrated on a single computing device or distributed on a network composed of multiple computing devices. They can be implemented by program code executable by the computing device. Thus, they can be stored in a storage device and executed by the computing device. And in some cases, the steps shown or described can be executed in a different order than here, or they can be separately fabricated into individual integrated circuit modules, or multiple modules or steps among them can be fabricated into a single integrated circuit module to implement. In this way, the present application is not limited to any specific combination of hardware and software.

[0185] The above are only the preferred embodiments of the present application and are not used to limit the present application. For those skilled in the art, the present application can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for periodic processing of deterministic services, characterized in that, The method includes: Obtaining service parameters of deterministic services, where the service parameters include a target probability, a service type, and a message length parameter; Determining a target message length according to a preset probability distribution model, the message length parameter, and the target probability; Determining a periodic interval of the deterministic service according to the service type and the target message length.

2. The method according to claim 1, characterized in that, Determining a target message length according to a preset probability distribution model, the message length parameter, and the target probability, includes: Determining the mode and standard deviation of the message length according to the message length parameter; Determining a target standard deviation coefficient corresponding to the target probability according to the probability distribution model; Determining the target message length according to the mode, the standard deviation, and the target standard deviation coefficient.

3. The method according to claim 2, characterized in that When the message length parameter only includes a maximum message length and a minimum message length, determining the mode and standard deviation of the message length according to the message length parameter, includes: Determining half of the sum of the maximum message length and the minimum message length as the mode; Dividing the difference between the mode and the minimum message length into multiple distribution intervals on average according to a preset number of distribution intervals to obtain the standard deviation, where each distribution interval corresponds to one standard deviation.

4. The method according to claim 2, wherein When the message length parameter includes a maximum message length, a minimum message length, and the mode, determining the mode and standard deviation of the message length according to the message length parameter, includes: Directly obtaining the mode from the message length parameter; Determining half of the sum of the maximum message length and the minimum message length as the median of the message length; When the median is greater than or equal to the mode, dividing the difference between the mode and the minimum message length into multiple distribution intervals on average according to a preset number of distribution intervals to obtain the standard deviation, where each distribution interval corresponds to one standard deviation; When the median is less than the mode, dividing the difference between the maximum message length and the mode into multiple distribution intervals on average according to the number of distribution intervals to obtain the standard deviation, where each distribution interval corresponds to one standard deviation.

5. The method according to claim 2, wherein The probability distribution model contains a mapping relationship between a standard deviation coefficient and a distribution probability value. Determining a target standard deviation coefficient corresponding to the target probability according to the probability distribution model, includes: Traversing the probability distribution model in ascending order of the distribution probability value, and determining the first distribution probability value greater than or equal to the target probability as the target distribution probability value; Determining the standard deviation coefficient corresponding to the target distribution probability value as the target standard deviation coefficient according to the mapping relationship.

6. The method according to claim 2, wherein Determining the target message length according to the mode, the standard deviation, and the target standard deviation coefficient, includes: Determining the product of the standard deviation and the target standard deviation coefficient as the target standard deviation; Determining the sum of the mode and the target standard deviation as the target message length.

7. The method according to claim 1, wherein The service type includes: variable-length periodic service, variable-length bandwidth service, or variable-length burst service.

8. The method according to claim 7, wherein Determining the periodic interval of the deterministic service according to the service type and the target message length includes: When the service type is the variable-length periodic service, obtaining a preset periodic interval from the service parameters; When the service type is the variable-length bandwidth service, obtaining a preset bandwidth parameter from the service parameters, and determining the ratio of the target message length to the bandwidth parameter as the periodic interval; When the service type is the variable-length burst service, obtaining a preset bandwidth parameter and the number of bursts from the service parameters, and determining the ratio of the product of the target message length and the number of bursts to the bandwidth parameter as the periodic interval.

9. The method according to claim 1, wherein The service parameters further include at least one of the following: service identifier, periodic interval, bandwidth parameter, minimum message length, maximum message length, median, number of bursts.

10. The method according to claim 1, wherein The probability distribution model includes a normal distribution model or a non-normal distribution model.

11. A periodic processing system for deterministic services, characterized in that The system includes: A service management unit, configured to send the service parameters of the deterministic service to the periodic processing unit, where the service parameters include a target probability, a service type, and a message length parameter; A periodic processing unit, configured to perform periodic processing on the service parameters according to the method described in any one of claims 1 to 10 above, and send the periodically processed service parameters to the periodic scheduling unit; A periodic scheduling unit, configured to perform service scheduling according to the periodically processed service parameters.

12. A computer-readable storage medium, characterized in that, A computer program is stored in the storage medium, where the computer program, when run by a processor, executes the method described in any one of claims 1 to 10 above.

13. An electronic device, comprising a memory and a processor, characterized in that, A computer program is stored in the memory, and the processor is configured to run the computer program to execute the method described in any one of claims 1 to 10 above.

Images