CN115842766B - Flow simulation method and device - Google Patents

Flow simulation method and device Download PDF

Info

Publication number
CN115842766B
CN115842766B CN202111098758.XA CN202111098758A CN115842766B CN 115842766 B CN115842766 B CN 115842766B CN 202111098758 A CN202111098758 A CN 202111098758A CN 115842766 B CN115842766 B CN 115842766B
Authority
CN
China
Prior art keywords
srv
message
link
flow
length
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202111098758.XA
Other languages
Chinese (zh)
Other versions
CN115842766A (en
Inventor
吴艳芹
张乐
吕田田
章军
刘霖筠
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
China Telecom Corp Ltd
Original Assignee
China Telecom Corp Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by China Telecom Corp Ltd filed Critical China Telecom Corp Ltd
Priority to CN202111098758.XA priority Critical patent/CN115842766B/en
Publication of CN115842766A publication Critical patent/CN115842766A/en
Application granted granted Critical
Publication of CN115842766B publication Critical patent/CN115842766B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Landscapes

  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The disclosure provides a flow simulation method and a flow simulation device, and relates to the technical field of IP networks and big data. The present disclosure proposes a SRv link flow simulation scheme suitable for an IPv6 transmission device and SRv transmission device hybrid networking scenario, which can simulate SRv link flow corresponding to SRv new service based on IPv6 link data, predict SRv link flow corresponding to historical service based on SRv link data, accurately simulate future SRv link flow according to SRv new service and SRv link flow corresponding to historical service, and further provide more effective support for routing decisions of an SDN controller or an IP network controller.

Description

Flow simulation method and device
Technical Field
The present disclosure relates to the field of IP (Internet Protocol ) networks and big data technologies, and in particular, to a flow simulation method and a flow simulation apparatus.
Background
In the Segment Routing (SR) IPv6 application scenario, SRv messages are transmitted in the network. However, there are a large number of devices that do not support SR in the existing network, and these devices that support IPv6 messaging and SRv devices that support messaging are deployed in a mixed manner in the network, and will coexist for a long time.
A software defined network (Software Defined Network, SDN) controller needs to be routed according to future traffic conditions of the optional new link. However, in the scenario of mixed networking of the IPv6 transmission device and the SRv transmission device, since the SRv message is inconsistent with the header information of the IPv6 message, the SRv message has a segment routing header (Segment Routing Header, SRH) added to the IPv6 message, which makes it very difficult to simulate the traffic on the SRv6 link.
Disclosure of Invention
The present disclosure proposes a SRv link flow simulation scheme suitable for an IPv6 transmission device and SRv transmission device hybrid networking scenario, which can simulate SRv link flow corresponding to SRv new service based on IPv6 link data, predict SRv link flow corresponding to historical service based on SRv link data, accurately simulate future SRv link flow according to SRv new service and SRv link flow corresponding to historical service, and further provide more effective support for routing decisions of an SDN controller or an IP network controller.
Some embodiments of the present disclosure provide a flow simulation method, including:
Acquiring SRv link information and IPv6 link information related to simulation, wherein the services on the SRv link comprise historical services originally deployed on the SRv link and SRv new services planned to migrate from the IPv6 link to the SRv link;
Predicting SRv a first message number and a first message length of new service in a future time unit according to first data related to flow simulation of an IPv6 message in a past time unit on an IPv6 link, calculating a second message length according to the first message length and a SRv message extension header length, and calculating SRv flow of new service in the future time unit on a SRv link according to the first message number and the second message length;
predicting a second message number and a third message length of the history service in a future time unit according to second data related to flow simulation of SRv messages in a past time unit on a SRv link, and calculating the flow of the history service in the future time unit on a SRv link according to the second message number and the third message length;
The traffic on SRv link at the future time unit is calculated based on the traffic on SRv link at the future time unit for SRv new traffic and the traffic on SRv link at the future time unit for historical traffic.
In some embodiments, the prediction is performed using LightGBM models.
In some embodiments, the first data includes the number and length of IPv6 messages in past time units on the IPv6 link, traffic, node information and port information of the routing device, and collected time characteristic information; or the second data comprises the number and the length of SRv messages in past time units, the flow, the node information and the port information of the routing equipment and the acquired time characteristic information on SRv links; the time characteristic information comprises at least one of busy time or idle time characteristic information and working day or non-working day characteristic information when data are collected.
In some embodiments, the first message length is an IPv6 message Wen Changdu, the second message length is a SRv message Wen Changdu, the third message length is a SRv message length, and calculating the second message length according to the first message length and the SRv message extension header length includes: and taking the sum of the first message length and the SRv message extension header length as the second message length.
In some embodiments, the product of the first number of messages and the second message length is taken as the flow of SRv new traffic over SRv links in future time units; or taking the product of the second message number and the third message length as the flow of historical service on SRv links in future time units.
In some embodiments, the LightGBM model is trained using data relating to the message and the flow simulation, the data relating to the message and the flow simulation including: the number and the length of the messages, the flow, the node information and the port information of the routing equipment and the acquired time characteristic information.
In some embodiments, parameters to which the LightGBM model is trained include learning rate, tree model depth, number of leaf nodes, where the number of leaf nodes is less than a power value based on 2, and the tree model depth is an exponent.
In some embodiments, a first LightGBM model is used to predict SRv a first number of messages and a first length of messages for a new service in a future time unit, where the first LightGBM model is obtained by using third data training related to flow simulation of an offline IPv6 message on an IPv6 link; and predicting a second message number and a third message length of the history service in a future time unit by using a second LightGBM model, wherein the second LightGBM model is obtained by using fourth data training of SRv messages offline on a SRv link and related to flow simulation.
In some embodiments, the predicted traffic includes one or more of incoming traffic, predicted with received messages, and outgoing traffic, predicted with sent messages.
In some embodiments, the time units include time intervals of one or more different time granularities.
Some embodiments of the present disclosure provide a flow simulation device, including: a memory; and a processor coupled to the memory, the processor configured to execute the flow simulation method based on instructions stored in the memory.
Some embodiments of the present disclosure provide a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the flow simulation method.
Drawings
The drawings that are required for use in the description of the embodiments or the related art will be briefly described below. The present disclosure will be more clearly understood from the following detailed description with reference to the accompanying drawings.
It will be apparent to those of ordinary skill in the art that the drawings in the following description are merely examples of the disclosure and that other drawings may be derived from them without inventive faculty.
Fig. 1 illustrates a flow diagram of a flow simulation method of some embodiments of the present disclosure.
Fig. 2 is a schematic diagram illustrating fields and their meanings in a SRv message extension header according to some embodiments of the present disclosure.
FIG. 3 illustrates a schematic diagram of a flow simulation system of some embodiments of the present disclosure.
Fig. 4 illustrates a schematic structural diagram of a flow simulation device of some embodiments of the present disclosure.
Detailed Description
The technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the drawings in the embodiments of the present disclosure.
Unless specifically stated otherwise, the descriptions of "first," "second," and the like in this disclosure are used for distinguishing between different objects and are not used for indicating a meaning of size or timing, etc.
Fig. 1 illustrates a flow diagram of a flow simulation method of some embodiments of the present disclosure.
As shown in fig. 1, the flow simulation method of this embodiment includes: steps 110-140.
Step 110, obtaining SRv link information and IPv6 link information related to the emulation, wherein the traffic on SRv link includes historical traffic originally deployed on SRv link and SRv new traffic scheduled to migrate from IPv6 link to SRv link.
The emulation-related SRv link information and IPv6 link information are obtained from the emulation requirements, and traffic on the SRv link according to the emulation requirements includes historical traffic originally deployed on the SRv link (either SRv traffic or IPv6 traffic or both) and SRv new traffic intended to migrate from the IPv6 link to the SRv link in order to emulate traffic on the SRv6 link. The emulation requirements may be obtained from an SDN controller or other external system.
Wherein SRv link information includes, for example, a routing list through which the SRv link passes. The IPv6 link information includes, for example, a routing list through which the IPv6 link passes.
At step 120, the traffic for SRv new traffic on SRv link at future time units is calculated.
Wherein the time units comprise one or more time intervals of different time granularity. When the time unit includes time intervals of a plurality of different time granularities, the present disclosure is able to predict traffic on SRv links of different time granularities.
In some embodiments, step 120 specifically includes steps 121-124. The steps are specifically described below.
Step 121, collecting data related to flow simulation of the IPv6 message in the past time unit on the IPv6 link, namely, first data.
The first data includes the number and length of IPv6 messages in past time units on the IPv6 link, traffic, node information and port information of the routing device, and collected time characteristic information.
The message length is an information item newly added in combination with the service scenario of the present disclosure, so as to solve the problem that SRv message and IPv6 message header information are inconsistent in the mixed networking scenario of IPv6 transmission equipment and SRv transmission equipment, which is not beneficial to SRv link flow simulation. Message length = flow/number of messages.
The node information and the port information of the routing device on the IPv6 link, for example, include information such as a local end device node model, an opposite end device node model, a local end port type, and an opposite end port type, which are newly added information items, so as to improve accuracy of flow simulation.
The time feature information includes at least one of busy or idle feature information and working day or non-working day feature information when data are collected, and is also a newly added information item, so as to improve accuracy of flow simulation. Wherein, the characteristic information of busy hour or idle hour is classified into idle hour and busy hour, for example, with the granularity of hour, the idle hour is 0, and the busy hour is 1. The holiday time is classified into time, day 0 and day 1.
Step 122, predicting SRv a first number of messages and a first message length of the new service in a future time unit according to the first data.
The predictions may be made using LightGBM models or other models.
The LightGBM model (or algorithm) is a machine learning algorithm that implements the idea of GBDT (Gradient Boosting Decision Tree, gradient-lifted decision tree). The predictive algorithm formula for LightGBM model is:
fm(x)=fm-1(x)+T(x;θm) (1)
Wherein x is a predicted sample, T (x; θ m) represents a decision tree, θ m represents a decision tree parameter, m is the number of trees, and f m (x) is a sample predicted value.
The trained loss function is expressed as:
where y i is the true value of the i-th sample, f m(xi) is the predicted value of the i-th sample x i.
The minimization of the loss function according to equation (3) yields the parameter θ m:
Where M is the total number of prediction samples x i.
The LightGBM model is trained by using data (historical data) related to the message and the flow simulation, wherein the data related to the message and the flow simulation comprises: the number and the length of the messages, the flow, the node information and the port information of the routing equipment and the acquired time characteristic information. The training process of LightGBM model is described in detail below:
Step a: initializing f 0(xi);
step b: input training samples include, for example: the information such as the number of the transmitted messages, the number of the received messages, the input flow, the output flow, the length of the transmitted messages, the length of the received messages, the type of the local equipment node, the type of the local equipment port, the type of the opposite equipment node, the type of the opposite equipment port, the acquisition time, the busy hour in idle time, the working day or not and the like.
Step c: fitting a residual tree T (x; θ m);
step d: updating the formula (1) based on the formulas (2) (3);
And e, repeating the steps b, c and d continuously by analogy until the model converges.
The trained LightGBM model inputs the data related to the message and the flow simulation in the previous time period, such as the number and the length of the message, the flow, the node information and the port information of the routing equipment, the acquired time characteristic information and the like, so that the number and the length of the message in the next time period can be predicted and output.
Parameters to which the LightGBM model is trained include learning rate, tree model depth, number of leaf nodes, and parameters of the LightGBM model can be adjusted based on error analysis of the prediction data. Wherein the number of leaf nodes affects the complexity of the numerical model, and to prevent overfitting, the number of leaf nodes is less than a power value based on 2, and based on the depth of the tree model. For a detailed description of other parameters see the following table.
In some embodiments, the first LightGBM model may be utilized to predict SRv a first number of messages and a first message length for the new service in future time units. The first LightGBM model is trained by using third data (historical data) related to flow simulation of an offline IPv6 message on an IPv6 link. The training effect of the model is further improved, and the accuracy of model prediction is improved.
Step 123, calculating the second message length according to the first message length and SRv message extension header length.
The first message length is an IPv6 message Wen Changdu, and the second message length is a SRv message length. The calculating the second message length according to the first message length and SRv message extension header length includes: and taking the sum of the first message length and the SRv message extension header length as the second message length.
The length of the SRv message extension header may be obtained by adding the lengths of each field of the SRv message extension header. The calculation formula is as follows:
len (SRv message expansion head)
=len(NextHeader)+len(HdrExtLen)+len(RoutingType)+len(SegmentsLeft)+len(LastEntry)+len(Flags)+len(Tag)+len(SegmentList[n])+len(OptionalTLV)
Wherein, each field and meaning in SRv message extension header refer to fig. 2.
And step 124, calculating the flow of SRv new service on SRv link in future time unit according to the first message number and the second message length.
And taking the product of the first message number and the second message length as the flow of SRv new service on SRv6 links in future time units.
At step 130, the flow of historical traffic over SRv links at future time units is calculated.
In some embodiments, step 130 specifically includes steps 131-133. The steps are specifically described below.
Step 131, collect SRv data related to flow simulation, namely second data, of SRv message over link at past time unit.
The second data includes SRv message number and length in past time units, flow, node information and port information of the routing device, and collected time characteristic information on SRv links.
The message length is an information item newly added in combination with the service scenario of the present disclosure, so as to solve the problem that SRv message and IPv6 message header information are inconsistent in the mixed networking scenario of IPv6 transmission equipment and SRv transmission equipment, which is not beneficial to SRv link flow simulation. Message length = flow/number of messages.
The node information and the port information of the routing device on the SRv link, for example, include information such as a local end device node model, an opposite end device node model, a local end port type, and an opposite end port type, which are newly added information items, so as to improve accuracy of flow simulation.
The time feature information includes at least one of busy or idle feature information and working day or non-working day feature information when data are collected, and is also a newly added information item, so as to improve accuracy of flow simulation. Wherein, the characteristic information of busy hour or idle hour is classified into idle hour and busy hour, for example, with the granularity of hour, the idle hour is 0, and the busy hour is 1. The holiday time is classified into time, day 0 and day 1.
Step 132, predicting the second message number and the third message length of the history service in the future time unit according to the second data.
The predictions may be made using LightGBM models. The LightGBM model is trained by using data (historical data) related to the message and the flow simulation, wherein the data related to the message and the flow simulation comprises: the number and the length of the messages, the flow, the node information and the port information of the routing equipment and the acquired time characteristic information. Parameters to which the LightGBM model is trained include learning rate, tree model depth, number of leaf nodes. Wherein the number of leaf nodes is less than a power value based on 2 and based on the depth of the tree model as an exponent. The relevant description of LightGBM model and specific training method are referred to above and will not be repeated here.
In some embodiments, the second number of messages and the third message length of the history traffic at the future time unit are predicted using a second LightGBM model. The second LightGBM model is trained by using fourth data (historical data) related to flow simulation of an offline SRv message on the SRv6 link. Therefore, the training effect of the model is further improved, and the accuracy of model prediction is improved.
The third message length is SRv message length.
And step 133, calculating SRv the flow of the historical service on the link in the future time unit according to the second message number and the third message length.
And taking the product of the second message number and the third message length as the flow of the historical service on the SRv link in the future time unit.
In step 140, the flow over SRv link at the future time unit is calculated.
The traffic on SRv link at the future time unit is calculated based on the traffic on SRv link at the future time unit for SRv new traffic and the traffic on SRv link at the future time unit for historical traffic. For example, the total traffic of SRv new traffic and historical traffic over the SRv link in future time units is taken as traffic over the SRv link in future time units.
It should be noted that, the flow mentioned in the embodiments of the present disclosure includes one or more of an input flow and an output flow. That is, the input flow rate may be predicted, the output flow rate may be predicted, and both the input flow rate and the output flow rate may be predicted. And, as will be appreciated by those skilled in the art, incoming traffic is predicted using received messages and outgoing traffic is predicted using sent messages.
For example, SRv link ingress traffic = SRv link SRv historical traffic + SRv link SRv new traffic; SRv6 link output traffic =
SRv6 link SRv6 historical traffic + SRv link SRv new traffic.
The above embodiment realizes a SRv link flow simulation scheme suitable for mixed networking scenario of an IPv6 transmission device and a SRv transmission device, which can simulate SRv link flow corresponding to SRv new service based on IPv6 link data, predict SRv link flow corresponding to historical service based on SRv link data, accurately simulate future SRv link flow according to SRv link flow corresponding to SRv new service and historical service, and further provide more effective support for routing decision of an SDN controller or an IP network controller.
FIG. 3 illustrates a schematic diagram of a flow simulation system of some embodiments of the present disclosure.
As shown in FIG. 3, the flow simulation system includes modules 304-313, and may also include devices 301-303, etc., as described in more detail below.
Client device 301: client-side edge network devices.
Router 302: IP network router devices, including common IPv6 routers and SRv routers.
Customer service center/operator service center 303: a cloud server cluster where a customer or operator provides a service.
Simulation demand acquisition module 304: SRv6 emulation requirements are received from an SDN controller or other external system including emulating a related list of routes traversed by an original IPv6 link (i.e., original link), a new SRv link (i.e., new link).
The data acquisition module 305: and collecting data of the number of messages sent, the number of messages received, the input flow, the output flow, the equipment node type, the port type, the opposite-end equipment node type and the opposite-end port type of each network router port through which the IPv6 link and the SRv link pass.
Data storage module 306: the large data is used for efficiently and quickly storing mass collected flow data and equipment data; and storing the preprocessed data.
A data preprocessing module 307: processing the collected flow data, including calculating the flow peak value of day, hour and minute, counting the number of messages received and transmitted by day, hour and minute, discretizing the collection time interval, and the like, and adding the length characteristic and the time characteristic of the messages.
Message number prediction module 308: the number of the data messages is predicted and calculated, and the number of the messages sent and received by the ports of the original IPv6 link and the new SRv link in the next time interval in the future with different time units of day, hour, minute and the like can be predicted.
Data packet length prediction module 309: the length of the received and transmitted data message is predicted and calculated, the length of the header of each IPv6 message is fixed, the length of the data unit in the payload is not fixed, the length of the received and transmitted message can be obtained according to the input and output flow and the quantity of the received and transmitted messages in unit time, and the length of the received and transmitted message of the original link and the new link passing through the port in the next time interval of different time units such as day, hour, minute and the like can be predicted. The payload length field in the IPv6 header is 16bits, and the maximum length is 65535 bytes. The new link currently transmits SRv messages, but the data unit length in the payload still needs to be predicted.
SRv6 message length generation module 310: generating SRv message length, generating an extended segment header SRH according to a new link route list required by simulation, calculating the length of the SRH, and adding the length of the SRH and an original link port receiving and transmitting data message prediction result of the data message length prediction module 309 to obtain SRv message length. The new link has transmitted SRv messages without performing the generation operation of module 310.
Port SRv6 input/output traffic simulation module 311: and multiplying the result of generating the length of the message received by SRv by the port of the original link by the result of predicting the number of messages to obtain SRv input/output flow of the original link, multiplying the result of predicting the length of the message in the new link by the result of predicting the number of messages to obtain the simulation result of the input/output flow of the existing service flow of the new link, and superposing the result of simulating the input/output flow of the new link through the port.
Link bi-directional SRv traffic simulation module 312: and predicting the bidirectional flow of the link, comparing the input and output flow simulation results of the ports SRv at the two ends of the link, and taking the maximum value of the input and output flow simulation to obtain the bidirectional flow simulation result of the link.
Simulation result transmitting module 313: and sending simulation results or bandwidth or link adjustment strategy commands to the SDN controller or an external system which puts forth simulation requirements.
Newly adding message length characteristics in the original characteristic group, and considering configuration characteristics of the ports and the nodes of the added equipment and time-related characteristics; predicting the length and the number of the SRv new service messages by using LightGBM model, and calculating the future input and output flow of SRv port based on the predicted result; and finally, superposing the predicted SRv new business input/output flow into the historical business input/output flow to realize the link bidirectional flow simulation.
The scheme of the disclosure can be applied to SDN controllers or IP network controllers of a communication network, and the simulation network adjusts the change brought to network flow distribution, finds out the flow of a link and loss which possibly generate congestion, and provides clear path suggestions for the SDN controllers or the IP network controllers to perform path optimization; the scheme disclosed by the invention can also be applied to network planning, provides flow twin simulation and situation change in the network, and determines the feasibility of the network planning scheme.
The scheme of the present disclosure can provide an intelligent and automatic means for the maintenance and guarantee of the IP network, solves the problems of long operation time, complex operation, difficult multiparty coordination and the like of the existing network, avoids the influence on the customer service, greatly improves the operation efficiency, reduces the operation and maintenance cost and improves the customer perception; and an intelligent means can be provided for IP network planning, so that the effectiveness and scientificity of a network planning scheme are ensured, and unnecessary investment waste is avoided.
Fig. 4 illustrates a schematic structural diagram of a flow simulation device of some embodiments of the present disclosure.
As shown in fig. 4, the flow simulation apparatus 400 of this embodiment includes: a memory 410 and a processor 420 coupled to the memory 410, the processor 420 being configured to perform the flow simulation method of any of the embodiments described above based on instructions stored in the memory 410.
For example, obtaining emulation-related SRv link information and IPv6 link information, traffic on SRv links including historical traffic originally deployed on SRv links and SRv new traffic intended to migrate from IPv6 links to SRv links; predicting SRv a first message number and a first message length of new service in a future time unit according to first data related to flow simulation of an IPv6 message in a past time unit on an IPv6 link, calculating a second message length according to the first message length and a SRv message extension header length, and calculating SRv flow of new service in the future time unit on a SRv link according to the first message number and the second message length; predicting a second message number and a third message length of the history service in a future time unit according to second data related to flow simulation of SRv messages in a past time unit on a SRv link, and calculating the flow of the history service in the future time unit on a SRv link according to the second message number and the third message length; the traffic on SRv link at the future time unit is calculated based on the traffic on SRv link at the future time unit for SRv new traffic and the traffic on SRv link at the future time unit for historical traffic.
The memory 410 may include, for example, system memory, fixed nonvolatile storage media, and the like. The system memory stores, for example, an operating system, application programs, boot Loader (Boot Loader), and other programs.
The flow simulation device 400 may also include an input-output interface 430, a network interface 440, a storage interface 450, and the like. These interfaces 430, 440, 450 and the memory 410 and the processor 420 may be connected, for example, by a bus 460. The input/output interface 430 provides a connection interface for input/output devices such as a display, a mouse, a keyboard, a touch screen, and the like. Network interface 440 provides a connection interface for various networking devices. Storage interface 450 provides a connection interface for external storage devices such as SD cards, U-discs, and the like.
It will be appreciated by those skilled in the art that embodiments of the present disclosure may be provided as a method, system, or computer program product. Accordingly, the present disclosure may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present disclosure may take the form of a computer program product embodied on one or more non-transitory computer-readable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer program code embodied therein.
The present disclosure is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the disclosure. It will be understood that each flowchart and/or block of the flowchart illustrations and/or block diagrams, and combinations of flowcharts and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The foregoing description of the preferred embodiments of the present disclosure is not intended to limit the disclosure, but rather to enable any modification, equivalent replacement, improvement or the like, which fall within the spirit and principles of the present disclosure.

Claims (12)

1. A flow simulation method, comprising:
Acquiring SRv link information and IPv6 link information related to simulation, wherein the services on the SRv link comprise historical services originally deployed on the SRv link and SRv new services planned to migrate from the IPv6 link to the SRv link;
Predicting SRv a first message number and a first message length of new service in a future time unit according to first data related to flow simulation of an IPv6 message in a past time unit on an IPv6 link, calculating a second message length according to the first message length and a SRv message extension header length, and calculating SRv flow of new service in the future time unit on a SRv link according to the first message number and the second message length;
predicting a second message number and a third message length of the history service in a future time unit according to second data related to flow simulation of SRv messages in a past time unit on a SRv link, and calculating the flow of the history service in the future time unit on a SRv link according to the second message number and the third message length;
The traffic on SRv link at the future time unit is calculated based on the traffic on SRv link at the future time unit for SRv new traffic and the traffic on SRv link at the future time unit for historical traffic.
2. The method of claim 1, wherein the predicting is performed using a LightGBM model.
3. The method of claim 1, wherein the step of determining the position of the substrate comprises,
The first data comprises the number and the length of IPv6 messages in past time units on an IPv6 link, flow, node information and port information of routing equipment and collected time characteristic information; or alternatively
The second data comprises the number and the length of SRv messages in past time units, the flow, the node information and the port information of the routing equipment and the acquired time characteristic information on SRv links;
The time characteristic information comprises at least one of busy time or idle time characteristic information and working day or non-working day characteristic information when data are collected.
4. The method of claim 1, wherein the first message is an IPv6 message Wen Changdu, the second message is a SRv message Wen Changdu, the third message is a SRv message,
The calculating the second message length according to the first message length and SRv message extension header length includes:
And taking the sum of the first message length and the SRv message extension header length as the second message length.
5. The method of claim 1, wherein the step of determining the position of the substrate comprises,
Taking the product of the first message number and the second message length as the flow of SRv new service on SRv link in future time unit; or alternatively
And taking the product of the second message number and the third message length as the flow of the historical service on the SRv link in the future time unit.
6. The method of claim 2, wherein the step of determining the position of the substrate comprises,
The LightGBM model is obtained by training data related to the message and the flow simulation, wherein the data related to the message and the flow simulation comprise: the number and the length of the messages, the flow, the node information and the port information of the routing equipment and the acquired time characteristic information.
7. The method of claim 2, wherein the step of determining the position of the substrate comprises,
Parameters to which the LightGBM model is trained include learning rate, tree model depth, number of leaf nodes, where the number of leaf nodes is less than a power value based on 2, and the tree model depth is an exponent.
8. The method of claim 1, wherein the step of determining the position of the substrate comprises,
Predicting SRv a first message number and a first message length of a new service in a future time unit by using a first LightGBM model, wherein the first LightGBM model is obtained by using third data training related to flow simulation of an offline IPv6 message on an IPv6 link;
And predicting a second message number and a third message length of the history service in a future time unit by using a second LightGBM model, wherein the second LightGBM model is obtained by using fourth data training of SRv messages offline on a SRv link and related to flow simulation.
9. The method of claim 1, wherein the step of determining the position of the substrate comprises,
The predicted flow rate includes one or more of an input flow rate and an output flow rate,
The input flow is predicted by using the received message, and the output flow is predicted by using the transmitted message.
10. The method of claim 1, wherein the step of determining the position of the substrate comprises,
The time units include one or more time intervals of different time granularity.
11. A flow simulation apparatus comprising:
a memory; and
A processor coupled to the memory, the processor configured to perform the flow simulation method of any of claims 1-10 based on instructions stored in the memory.
12. A non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the flow simulation method of any of claims 1-10.
CN202111098758.XA 2021-09-18 2021-09-18 Flow simulation method and device Active CN115842766B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111098758.XA CN115842766B (en) 2021-09-18 2021-09-18 Flow simulation method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111098758.XA CN115842766B (en) 2021-09-18 2021-09-18 Flow simulation method and device

Publications (2)

Publication Number Publication Date
CN115842766A CN115842766A (en) 2023-03-24
CN115842766B true CN115842766B (en) 2024-04-26

Family

ID=85575160

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111098758.XA Active CN115842766B (en) 2021-09-18 2021-09-18 Flow simulation method and device

Country Status (1)

Country Link
CN (1) CN115842766B (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109922004A (en) * 2019-04-24 2019-06-21 清华大学 The traffic engineering method and device of IPv6 network based on partial deployment Segment routing
CN113194037A (en) * 2021-03-31 2021-07-30 新华三大数据技术有限公司 Traffic scheduling method and device
CN113381933A (en) * 2021-06-04 2021-09-10 烽火通信科技股份有限公司 SRv6 bidirectional forwarding detection method and system in network

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10635480B2 (en) * 2018-04-04 2020-04-28 Cisco Technology, Inc. Zero-loss workload mobility with segment routing for virtual machines
US10887799B2 (en) * 2019-01-10 2021-01-05 Cisco Technology, Inc. SRv6 user-plane-based triggering methods and apparatus for session or flow migration in mobile networks

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109922004A (en) * 2019-04-24 2019-06-21 清华大学 The traffic engineering method and device of IPv6 network based on partial deployment Segment routing
CN113194037A (en) * 2021-03-31 2021-07-30 新华三大数据技术有限公司 Traffic scheduling method and device
CN113381933A (en) * 2021-06-04 2021-09-10 烽火通信科技股份有限公司 SRv6 bidirectional forwarding detection method and system in network

Also Published As

Publication number Publication date
CN115842766A (en) 2023-03-24

Similar Documents

Publication Publication Date Title
US10749757B2 (en) Method and apparatus for generating network control policy, and network controller
CN112953778B (en) Intention-driven-based service arrangement system and method in intelligent fusion identification network
CN109995583B (en) Delay-guaranteed NFV cloud platform dynamic capacity expansion and contraction method and system
US11057294B2 (en) Route control method and route setting device
CN112532409B (en) Network parameter configuration method, device, computer equipment and storage medium
US20030061017A1 (en) Method and a system for simulating the behavior of a network and providing on-demand dimensioning
CN105763570A (en) Virtualization-technology-based distributed real-time network simulation system
Huang et al. Intelligent traffic control for QoS optimization in hybrid SDNs
US20230145097A1 (en) Autonomous traffic (self-driving) network with traffic classes and passive and active learning
Mallick et al. Dynamic graph neural network for traffic forecasting in wide area networks
WO2023274304A1 (en) Distributed routing determining method, electronic device, and storage medium
CN114760669A (en) Routing decision method and system based on flow prediction
Lyu et al. Cooperative computing anytime, anywhere: Ubiquitous fog services
CN116599904A (en) Parallel transmission load balancing device and method
CN110601916A (en) Flow sampling and application sensing system based on machine learning
CN115484173B (en) Flow simulation method of digital twin network and digital twin network
CN114448899A (en) Method for balancing network load of data center
CN115842766B (en) Flow simulation method and device
Morais Machine learning in multi-layer optical networks: why and how
Tang et al. Digital Twin Assisted VNF Mapping and Scheduling in SDN/NFV-Enabled Industrial IoT
CN112769620A (en) Network deployment method, equipment and computer readable storage medium
CN116055377A (en) Time-sensitive network traffic scheduling method for power service
CN105959167B (en) A kind of global optimization SDN measurement method based on greedy algorithm
CN114205300B (en) Flow scheduling method capable of guaranteeing coflow transmission deadline under condition of incomplete flow information
Reticcioli et al. Learning SDN traffic flow accurate models to enable queue bandwidth dynamic optimization

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant