Traffic aggregation method, device, electronic device and storage medium

Abstract

The present disclosure provides a traffic aggregation method, device, electronic device and storage medium, the method comprising: obtaining historical service traffic data on the IP side, and inputting a pre-built service traffic prediction model, obtaining the output service traffic prediction result of the IP side in the target time period; pre-arranging the service traffic prediction result to obtain the pre-arrangement result; pre-loading the pre-arrangement result to obtain the pre-loaded network resource status result, and adjusting the service traffic prediction model according to the pre-loaded network resource status result to obtain the adjusted service traffic prediction model, and correcting the pre-arrangement result through the adjusted service traffic prediction model to obtain the arrangement result; according to the arrangement result, in the target time period, the service traffic on the IP side is aggregated to the optical network side. The present disclosure can reduce the waste of optical network pipeline resources and traffic congestion, and improve the bandwidth utilization of the optical network pipeline.

Description

Technical Field

[0001] The present disclosure relates to the field of electronic information technology, and in particular to a traffic aggregation method, device, electronic device and storage medium. Background Art

[0002] This section is intended to provide a background or context to the embodiments of the present disclosure as recited in the claims. No description herein is admitted to be related art by inclusion in this section.

[0003] In related technologies, the bearer network is divided into two layers: IP / MPLS bearer network (IP layer) and optical transport network (OTN optical layer). The optical layer provides statically configured physical links for the IP layer. The rapid growth of network traffic has brought about the need to upgrade the bearer network and expand equipment capacity. With the development of optical network transmission capabilities, the integration of IP and optical transport networks has become the direction of future network architecture evolution.

[0004] The safe and stable operation of related IP networks has higher requirements in terms of reliability, real-time performance, and business volume. The IP+optical network scenario poses greater challenges to the design of the optical layer traffic transmission mechanism. How to effectively achieve the smooth connection between the IP service network and the optical layer transmission technology represented by OTN to meet the IP+optical network's bearing quality and transmission reliability for various services has become one of the key technical issues in the IP+optical network scenario. Summary of the invention

[0005] In view of this, the purpose of the present disclosure is to provide a traffic aggregation method, device, electronic device and storage medium.

[0006] Based on the above purpose, the exemplary embodiment of the present disclosure provides a method for aggregating traffic, including:

[0007] Acquire historical service traffic data on the IP side, and input the historical service traffic data into a pre-built service traffic prediction model to obtain a service traffic prediction result of the IP side in a target time period output by the service traffic prediction model;

[0008] Pre-arranging the service flow prediction result to obtain a pre-arrangement result;

[0009] Pre-loading the pre-scheduling result to obtain a pre-loaded network resource status result, and adjusting the service flow prediction model according to the pre-loaded network resource status result to obtain an adjusted service flow prediction model, and correcting the pre-scheduling result by using the adjusted service flow prediction model to obtain a scheduling result;

[0010] According to the scheduling result, in the target time period, the service traffic on the IP side is aggregated to the optical network side.

[0011] In some exemplary embodiments, the traffic flow prediction model includes:

[0012] Gated recurrent neural network model.

[0013] In some exemplary embodiments, the method further comprises:

[0014] Construct a sample set including several samples; wherein the samples include: sample data and label data; the sample data includes historical service flow data of the training IP side; the label data includes service flow prediction results of the training IP side in the target time period corresponding to the historical service flow data of the training IP side;

[0015] According to the sample set, the business traffic prediction model is constructed and trained through a gated recurrent neural network algorithm.

[0016] In some exemplary embodiments, the pre-arranging the service flow prediction result to obtain the pre-arrangement result includes:

[0017] With the goal of fully loading the granular pipe on the optical network side, the service flow prediction result is pre-arranged to obtain the pre-arranged result.

[0018] In some exemplary embodiments, while pre-arranging the service flow prediction result to obtain the pre-arrangement result, the method further includes:

[0019] The primary and backup services on the IP side are respectively carried on physically separated optical network links to form a service multi-path carrying model.

[0020] In some exemplary embodiments, the step of aggregating the service traffic on the IP side to the optical network side in the target time period according to the scheduling result includes:

[0021] Allocate services of different granularities to corresponding optical network layer link channels.

[0022] In some exemplary embodiments, after aggregating the service traffic on the IP side to the optical network side in the target time period according to the scheduling result, the method further includes:

[0023] Update path resource information.

[0024] Based on the same inventive concept, the exemplary embodiment of the present disclosure further provides a flow aggregation device, including:

[0025] A service flow prediction module is configured to obtain historical service flow data of the IP side, input the historical service flow data into a pre-built service flow prediction model, and obtain a service flow prediction result of the IP side in a target time period output by the service flow prediction model;

[0026] A pre-arrangement module is configured to pre-arrange the service flow prediction result to obtain a pre-arrangement result;

[0027] an orchestration result correction module, configured to preload the pre-orchestrated result to obtain a pre-loaded network resource status result, and adjust the service flow prediction model according to the pre-loaded network resource status result to obtain an adjusted service flow prediction model, and correct the pre-orchestrated result by using the adjusted service flow prediction model to obtain an orchestration result;

[0028] The service traffic scheduling module is configured to converge the service traffic on the IP side to the optical network side in the target time period according to the scheduling result.

[0029] Based on the same inventive concept, an exemplary embodiment of the present disclosure further provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, any of the methods described above is implemented.

[0030] Based on the same inventive concept, an exemplary embodiment of the present disclosure further provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute any of the above methods.

[0031] From the above, it can be seen that the traffic aggregation method, device, electronic device and storage medium provided by the embodiment of the present disclosure include: obtaining historical service traffic data on the IP side, and inputting a pre-built service traffic prediction model to obtain the service traffic prediction result of the IP side in the target time period; pre-arranging the service traffic prediction result to obtain the pre-arrangement result; pre-loading the pre-arrangement result to obtain the pre-loaded network resource status result, and adjusting the service traffic prediction model according to the pre-loaded network resource status result to obtain the adjusted service traffic prediction model, and correcting the pre-arrangement result through the adjusted service traffic prediction model to obtain the arrangement result; according to the arrangement result, in the target time period, the service traffic on the IP side is aggregated to the optical network side. The present disclosure can reduce the waste of optical network pipeline resources and traffic congestion, and improve the bandwidth utilization of optical network pipelines. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the technical solutions in the present disclosure or related technologies, the drawings required for use in the embodiments or related technical descriptions are briefly introduced below. Obviously, the drawings described below are only embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0033] Figure 1 A schematic diagram of an IP+optical network scenario provided according to an exemplary embodiment of the present disclosure;

[0034] Figure 2 A schematic diagram of a structure of an optical transmission unit frame provided according to an exemplary embodiment of the present disclosure;

[0035] Figure 3 A schematic diagram of a flow chart of a method for aggregating traffic provided by an exemplary embodiment of the present disclosure;

[0036] Figure 4 A schematic diagram of a structure of a gated recurrent neural network model provided according to an exemplary embodiment of the present disclosure;

[0037] Figure 5 Another structural schematic diagram of a gated recurrent neural network model provided according to an exemplary embodiment of the present disclosure;

[0038] Figure 6 A schematic diagram of a pre-arrangement provided according to an exemplary embodiment of the present disclosure;

[0039] Figure 7 A schematic diagram of a structure of a flow convergence device provided for an exemplary embodiment of the present disclosure;

[0040] Figure 8 A schematic structural diagram of an electronic device provided by an exemplary embodiment of the present disclosure. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical solutions and advantages of the present disclosure more clear, the principles and spirit of the present disclosure will be described with reference to several exemplary embodiments. It should be understood that these embodiments are provided only to enable those skilled in the art to better understand and implement the present disclosure, and are not intended to limit the scope of the present disclosure in any way. On the contrary, these embodiments are provided to make the present disclosure more thorough and complete, and to fully convey the scope of the present disclosure to those skilled in the art.

[0042] According to an embodiment of the present disclosure, a traffic aggregation method, device, electronic device and storage medium are proposed.

[0043] It should be understood herein that any number of elements in the drawings is for illustration rather than limitation, and any naming is only for distinction rather than having any limiting meaning.

[0044] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the embodiments of the present disclosure should be understood by people with ordinary skills in the field to which the present disclosure belongs. The "first", "second" and similar words used in the embodiments of the present disclosure do not indicate any order, quantity or importance, but are only used to distinguish different components. "Including" or "comprising" and similar words mean that the elements or objects appearing before the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0045] The principle and spirit of the present disclosure are explained in detail below with reference to several representative embodiments of the present disclosure.

[0046] refer to Figure 1 , which is a schematic diagram of an IP+optical network scenario provided according to an exemplary embodiment of the present disclosure.

[0047] Among them, the IP+optical network scenario includes the IP side, namely the IP router, and the optical network side (OTN, optical transport network). The IP router includes several nodes (such as node A, node B and node C), and the OTN includes several OXCs (optical cross-connects) or ROADMs (Reconfigurable Optical Add-Drop Multiplexer).

[0048] In related technologies, the bearer network is divided into two levels: IP / MPLS bearer network (IP layer) and optical transport network (OTN optical layer). The optical layer provides statically configured physical links for the IP layer. The rapid growth of network traffic has brought about the need to upgrade the bearer network and expand equipment capacity. With the development of optical network transmission capabilities, the integration of IP and optical transport networks has become the direction of future network architecture evolution. The safe and stable operation of related IP networks has higher requirements in terms of reliability, real-time performance, and business volume. The IP+optical network scenario poses greater challenges to the design of the optical layer traffic transmission mechanism. How to effectively achieve the smooth connection between the IP service network and the optical layer transmission technology represented by OTN to meet the IP+optical network's bearing quality and transmission reliability for various services has become one of the key technical issues in the IP+optical network scenario.

[0049] In related technologies, traffic intelligent scheduling on the IP service side or the optical transmission side is often considered independently, such as route reconstruction for IP services, traffic grooming in optical transmission networks, etc. However, in the case of IP services converging traffic to optical transmission pipelines in IP+optical network scenarios, when IP services with relatively small granularity converge to large-capacity optical transmission pipelines, service conflicts or resource redundancy may occur, leading to user service congestion and reduced service quality. Currently, there is no better solution for IP+optical network traffic convergence and adaptability to service type transmission.

[0050] The inventors of the present disclosure have found that the causes of the above-mentioned problems in the related art are:

[0051] On the one hand, the optical layer links of the relevant IP+optical network need to be compatible with the small-granularity traffic on the IP side and complete the collaborative carrying of the two-layer traffic. In the IP+optical networking scenario, in order to ensure the effective use of the optical physical layer transmission resources, small-granular IP services are often aggregated in the optical transmission pipeline for transmission. However, the burst nature of IP services makes them often disordered during the aggregation process, presenting a first-come-first-served form. This carrying method is prone to unreasonable waste of pipeline resources. For example, if a pipeline is not fully loaded, the subsequent small-granular services that are adjacent cannot be aggregated on the pipeline due to inappropriate queue sorting. Therefore, how to flexibly coordinate the smooth aggregation of different IP service flows to the optical layer physical pipeline is one of the difficulties in the IP+optical network scenario. On the other hand, the IP+optical network has no targeted design for the protection of key services with high reliability requirements during the aggregation process, and cannot cope with the collective interruption of services caused by the disconnection of optical fiber links. In order to ensure that key services can be stably transmitted end-to-end, how to design a reasonable resource scheduling strategy so that the primary and backup services can be isolated and transmitted in the physical topology and provide reliable N-1 and N-2 protection is another important issue in IP+optical converged networking.

[0052] In order to solve the above problems, the present disclosure provides a traffic aggregation solution: for the resource scheduling problem when small-grained IP services converge to large-capacity optical transmission pipelines in IP+optical network scenarios, consider how to apply ODUk technology on the OTN side to carry small-grained services, and realize integrated traffic aggregation based on machine learning in IP+optical network scenarios. The core lies in the comprehensive use of the iterative data learning characteristics of reinforcement learning and the precise fine-tuning effect of transfer learning on the training results, and the coordinated service characteristics of the IP plane and the pipeline characteristics of the optical plane. More reasonable pipeline resource allocation and aggregation are performed for IP small-grained service traffic with high reliability requirements, ensuring that various types of services can make full use of ODUk pipeline resources as much as possible and improve network resource utilization. At the same time, it can capture the service type according to the real-time network status, dynamically adjust the main and standby services to the links separated by physical locations for transmission, and improve network reliability. In summary, the present disclosure is expected to minimize the redundancy of pipeline resources for small-grained services at the aggregation node, avoid the congestion of backbone optical links, make full use of network resources in the protection and recovery of key services, and realize the stable operation of IP+optical networks.

[0053] The following will introduce the basic concepts involved in this disclosure:

[0054] IP+optical network: With the rapid development of emerging services such as 4K video, VR and the Internet of Things, the backbone network traffic is growing rapidly. The backbone network adopts a cascaded "IP+optical" collaborative architecture to serve the rapidly growing user needs. In the cascade mode, the IP layer adopts a hierarchical networking architecture to aggregate bandwidth, and the optical layer adopts a flat architecture to solve the problem of large capacity and long-distance transmission. At the upper network layer, with SDN technology as the core and virtualization architecture as the means, through pipeline resource pooling and network virtualization, the network can automatically adapt to business changes on demand with the help of advantageous algorithms to achieve network resource and business adaptation. At the IP+optical network exchange, each physical port can be used by IP connections in multiple directions, and the controller can schedule the port resources occupied by IP connections in each direction, thereby forming a router port resource pool. Sub-wavelength-level service scheduling is achieved through OTN, and the controller schedules the wavelength resources occupied in each direction, treating the entire optical network as a resource pool to meet the connection requirements of the IP layer on demand. At the same time, the controller can obtain resource information of the entire network, flexibly schedule the ports and optical network resources of the router, and realize the pooling of pipeline network resources.

[0055] Optical channel frame structure: The optical channel frame structure consists of three parts: optical transmission unit (OTUk) frame structure, optical data unit (ODUk) frame structure and optical payload unit (OPUk) frame structure. OTUk adopts a fixed-length frame structure, which does not change with the client signal rate, nor does it change with the OTU1, OTU2, OTU3, and OTU4 levels. When the client signal rate is high, the frame period is relatively shortened and the frame frequency is increased, but the data signal carried by each frame does not increase. To carry a frame of 10Gbit / sSDH signal, about 11 OTU2 optical channel frames are required, and to carry a frame of 2.5Gbit / sSDH signal, about 3 OTU1 optical channel frames are required.

[0056] refer to Figure 2 , which is a structural schematic diagram of an optical transmission unit frame provided according to an exemplary embodiment of the present disclosure.

[0057] Among them, the OTUk frame structure is a 4-row 4080-column structure, which mainly consists of three parts: OTUk overhead, OTUk payload, and OTUk forward error correction. In the figure, columns 1 to 14 in row 1 are OTUk overhead, columns 1 to 14 in rows 2 to 4 are ODUk overhead, columns 15 to 3824 in rows 1 to 4 are OTUk payload, and columns 3825 to 4080 in rows 1 to 4 are OTUk forward error correction codes.

[0058] After introducing the basic principles of the present disclosure, various non-limiting embodiments of the present disclosure are described in detail below.

[0059] refer to Figure 3 , which is a flow chart of a method for aggregating traffic provided by an exemplary embodiment of the present disclosure.

[0060] The traffic aggregation method includes the following steps:

[0061] Step S310: Obtain historical service traffic data on the IP side, and input the historical service traffic data into a pre-built service traffic prediction model to obtain the service traffic prediction result of the IP side in the target time period output by the service traffic prediction model.

[0062] The service traffic prediction model is used to learn the characteristics of historical service traffic on the IP side in the IP+optical network scenario, so as to predict the service traffic on the IP side in the target time period, and further use it to guide the IP traffic encapsulation and transmission on the OTN side.

[0063] In some exemplary embodiments, the traffic flow prediction result includes:

[0064] Service source-destination pairs, service categories (including data services and protection services), and service traffic bandwidth requirements (including the granularity, quantity, and arrival time domain of services).

[0065] In some exemplary embodiments, the traffic flow prediction model includes:

[0066] Gated recurrent neural network model.

[0067] refer to Figure 4 , which is a structural diagram of a gated recurrent neural network model provided according to an exemplary embodiment of the present disclosure.

[0068] Input of the GRU: input x at time t t and the hidden layer state h at time t-1 t-1 , this hidden layer state contains the relevant information of the previous node.

[0069] Output: The output y of the hidden node at time t t and the hidden state h passed to the next node t .

[0070] refer to Figure 5 , which is another structural schematic diagram of a gated recurrent neural network model provided according to an exemplary embodiment of the present disclosure.

[0071] Among them, the state h transmitted from the previous t-1 and the input x of the current node t To obtain the two gating states.

[0072] Reset Gate t =σ(x t W xr +h t-1 W hr +b r )

[0073] Update gate z t =σ(x t W xz +h t-1 W hz +b z )

[0074] σ is the sigmoid function, which can transform the data into a value in the range of 0-1, thus acting as a gating signal.

[0075] Candidate hidden layer states

[0076] Among them, h t-1 Contains information from the past, t is the reset gate and ⊙ is the element-wise multiplication.

[0077] By r t and ht-1 Multiply elements by element to determine whether past information is helpful in predicting the future:

[0078] When t When it approaches zero, the model will discard the past hidden information and only keep the current input information. t When it approaches 1, it is considered that past information is useful and is added to the current information.

[0079] Here It mainly contains the x input at time t t The information and the hidden layer state h at time t-1 are targeted t-1 reservation.

[0080] Final hidden state

[0081] Among them, h t-1 Contains information from the past. is the candidate hidden state, z t For updating the door.

[0082] The operation of this step is to forget the h passed down t-1 , and add some dimensional information of the current node input.

[0083] z t The range is 0 to 1. The closer the gating signal is to 1, the more past data is "remembered"; and the closer it is to 0, the more past data is "forgotten" (controlling the combination of past hidden states and current candidate information).

[0084] (1-z t )⊙h t-1 Indicates the selective "forgetting" of the originally hidden state. Indicates the selective "memory" of the candidate hidden states of the current node (and also forgets some unimportant things).

[0085] When z t When it tends to 1, it means that long-term dependence has always existed;

[0086] When z t When it approaches 0, it means that some unimportant information in the hidden information is forgotten.

[0087] The reset gate determines how new input information is combined with previous memory, and the update gate defines how much of the previous memory is preserved to the current time step.

[0088] When training traditional recurrent neural networks (RNNs), if the time interval is large, the gradient vanishing problem will occur, that is, the perception of the nodes in the later time period to the nodes in the previous time period will decrease. To solve this problem, there are roughly two types of methods: one is to improve or replace the traditional training model with a new method; the other is to design more sophisticated recurrent units, such as LSTM and GRU. LSTM (Long Short Term Memory) is a special type of RNN network that can learn information with a long time interval, usually called long-term dependency information. The core of the design of LSTM to achieve this ability is to selectively remember long-term important dependency information by designing a special network structure, while discarding long-term useless information. LSTM has three gates to transmit information, including forget gate, input gate, and output gate. Based on the above LSTM design ideas, many deep learning recurrent neural network unit variants have emerged. For example, the gated recurrent unit (GRU) merges the forget gate and input gate into an update gate, adds a reset gate, and simplifies the output. GRU is simpler than the LSTM model and has lower computational overhead. Moreover, GRU has fewer parameters, a relatively simple structure, and is easier to train. It only requires less data to train a better model.

[0089] In some exemplary embodiments, a method for constructing and training a traffic flow prediction model includes:

[0090] Construct a sample set including several samples; wherein the samples include: sample data and label data; the sample data includes historical service flow data of the training IP side; the label data includes service flow prediction results of the training IP side in the target time period corresponding to the historical service flow data of the training IP side;

[0091] According to the sample set, the business traffic prediction model is constructed and trained through a gated recurrent neural network algorithm.

[0092] The length of the target time period can be adjusted according to the traffic conditions. For example, in a high-traffic scenario, a smaller time period can be used for traffic prediction, making the prediction more precise and the service arrangement more compact, thus alleviating the congestion of high-traffic services.

[0093] Step S320: pre-arrange the service flow prediction result to obtain a pre-arrangement result.

[0094] In some exemplary embodiments, the pre-arranging the service flow prediction result to obtain the pre-arrangement result includes:

[0095] With the goal of fully loading the granular pipe on the optical network side, the service flow prediction result is pre-arranged to obtain the pre-arranged result.

[0096] Among them, combined with the prediction results, the small-granularity service traffic on the IP side is orchestrated to occupy the GE granular pipeline resources on the OTN side as much as possible, and the orchestration results are recorded in the controller.

[0097] In some exemplary embodiments, while pre-arranging the service flow prediction result to obtain the pre-arrangement result, the method further includes:

[0098] The primary and backup services on the IP side are respectively carried on physically separated optical network links to form a service multi-path carrying model.

[0099] In the orchestration process, the primary and backup IP services are respectively carried on physically separated optical network links, forming a multi-path carrying model for a service, so that the expected service protection effect will not be invalidated due to damage to the optical fiber link.

[0100] refer to Figure 6 , which is a schematic diagram of pre-arrangement provided according to an exemplary embodiment of the present disclosure.

[0101] As an example, the prediction time segment of the training model is set to 4 times the unit time τ, that is, t p =4τ, the figure exemplifies the orchestration effect of the services in the two time segments t1 and t2. It can be seen that through the predictive orchestration of each time segment, a better convergence effect of different granular services is achieved, and one ODU unit is saved at time t1 and time t2 respectively, and each ODU unit is transmitted in a fully loaded form, achieving a higher pipeline resource utilization rate; at the same time, based on the prediction results, the pre-orchestration plans the protection path for the IP service at time t1, and carries the protection service on the physically isolated path 2, realizing an IP service protection strategy with high resource utilization. Based on the above analysis, it can also be inferred that the pre-orchestration effect will increase with the increase of t p By lengthening the network, higher resource utilization can be achieved. More business granules can be included in the unified scheduling in one orchestration, and the original business and protection business can be autonomously physically isolated and transmitted to achieve better network anti-destruction capability.

[0102] Step S330, pre-load the pre-scheduling result to obtain the pre-loaded network resource status result, and adjust the business traffic prediction model according to the pre-loaded network resource status result to obtain the adjusted business traffic prediction model, and correct the pre-scheduling result through the adjusted business traffic prediction model to obtain the scheduling result.

[0103] The orchestration result is preloaded, and the carried network resource status result is used as input, with a small learning rate (optional, less than 10-4 ) Fine-tune the relevant models. At the same time, the business data generated in real time during the orchestration process is also added to the fine-tuning process as training data. This process can be continuously cycled with each business orchestration, so that the model gradually converges to the best prediction effect.

[0104] Step S340: According to the scheduling result, in the target time period, the service traffic on the IP side is aggregated to the optical network side.

[0105] In some exemplary embodiments, the step of aggregating the service traffic on the IP side to the optical network side in the target time period according to the scheduling result includes:

[0106] Allocate services of different granularities to corresponding optical network layer link channels.

[0107] Based on the fine-tuning process in step S330, the result of S320 is further corrected and predicted, and the network main controller schedules the service according to the established plan and allocates it to the corresponding link channel to complete the service bearing.

[0108] In some exemplary embodiments, after aggregating the service traffic on the IP side to the optical network side in the target time period according to the scheduling result, the method further includes:

[0109] Update path resource information.

[0110] From the above, it can be seen that the traffic aggregation method provided by the embodiment of the present disclosure is based on the recurrent neural network training module of the GRU unit, and the GRU neural network model is trained through historical traffic data to realize the prediction of IP service traffic as a reference for pipeline resource scheduling. First, the small-grained IP service traffic situation in the next period is predicted and output to the controller. The whole network controller pre-arranges the optical pipeline resources according to the prediction results, and adjusts the ODUk pipeline resource allocation strategy on the OTN side and the service traffic transmission timing on the IP side respectively, so that the core aggregation node in the network can optimize the resource integration of service groups from different network access sides, different time slots, and different granularities, and finally achieve ODUk pipeline carrying without resource congestion or waste, while minimizing transmission overhead. In addition, with the help of transfer learning technology, the services and service scheduling results arriving at the new time are used as input to fine-tune the model of the deep learning process, and the accuracy of the training model for traffic prediction and the utilization rate of pre-arranged resources are continuously corrected.

[0111] The present disclosure aims at the problem of unreasonable utilization of OTN pipeline resources when IP services converge traffic to the optical transmission network in the IP+optical network scenario, and proposes an integrated traffic convergence solution based on machine learning in the IP+optical network scenario. Specifically, the method combines the traffic prediction function of the recurrent neural network with the output fine-tuning method of transfer learning, and can give the predicted traffic arrival situation of the next time segment at the IP and OTN convergence nodes according to the characteristics of various services. Based on this, the resource arrangement of multi-granular services can be performed on the OTN side pipeline, the bandwidth utilization rate of the OTN pipeline can be improved, and the waste of pipeline resources and traffic congestion can be greatly reduced. At the same time, the protection service route can be dynamically planned according to the network topology, and the protection service can be effectively carried on the physically isolated path, which generally improves the intelligent level of service transmission of the IP+optical network.

[0112] It should be noted that the method of the embodiment of the present disclosure can be performed by a single device, such as a computer or a server. The method of the present embodiment can also be applied in a distributed scenario and completed by multiple devices cooperating with each other. In the case of such a distributed scenario, one of the multiple devices can only perform one or more steps in the method of the embodiment of the present disclosure, and the multiple devices will interact with each other to complete the described method.

[0113] It should be noted that the above describes some embodiments of the present disclosure. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the above embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0114] Based on the same inventive concept, corresponding to any of the above-mentioned embodiment methods, the present disclosure also provides a traffic convergence device.

[0115] refer to Figure 7 , the flow convergence device comprises:

[0116] The business traffic prediction module 710 is configured to obtain historical business traffic data on the IP side, and input the historical business traffic data into a pre-built business traffic prediction model to obtain the business traffic prediction result of the IP side in the target time period output by the business traffic prediction model.

[0117] In some exemplary embodiments, the traffic flow prediction model includes:

[0118] Gated Recurrent Neural Network Model.

[0119] In some exemplary embodiments, the traffic aggregation device is further configured to:

[0120] Construct a sample set including several samples; wherein the samples include: sample data and label data; the sample data includes historical service flow data of the training IP side; the label data includes service flow prediction results of the training IP side in the target time period corresponding to the historical service flow data of the training IP side;

[0121] According to the sample set, the business traffic prediction model is constructed and trained through a gated recurrent neural network algorithm.

[0122] The pre-arrangement module 720 is configured to pre-arrange the service flow prediction result to obtain a pre-arrangement result.

[0123] In some exemplary embodiments, the pre-arrangement module 720 is specifically configured to:

[0124] With the goal of fully loading the granular pipe on the optical network side, the service flow prediction result is pre-arranged to obtain the pre-arranged result.

[0125] In some exemplary embodiments, while pre-arranging the service flow prediction result to obtain the pre-arrangement result, the pre-arrangement module 720 is further configured to:

[0126] The primary and backup services on the IP side are respectively carried on physically separated optical network links to form a service multi-path carrying model.

[0127] The orchestration result correction module 730 is configured to pre-load the pre-orchestration result to obtain a pre-loaded network resource status result, and adjust the service traffic prediction model according to the pre-loaded network resource status result to obtain an adjusted service traffic prediction model, and correct the pre-orchestration result through the adjusted service traffic prediction model to obtain an orchestration result.

[0128] The service traffic scheduling module 740 is configured to converge the service traffic on the IP side to the optical network side in the target time period according to the scheduling result.

[0129] In some exemplary embodiments, the service traffic scheduling module 740 is specifically configured to:

[0130] Allocate services of different granularities to corresponding optical network layer link channels.

[0131] In some exemplary embodiments, after aggregating the service traffic on the IP side to the optical network side in the target time period according to the scheduling result, the traffic aggregation device is further configured to:

[0132] Update path resource information.

[0133] For the convenience of description, the above device is described by dividing it into various modules according to its functions. Of course, when implementing the present disclosure, the functions of each module can be implemented in the same or multiple software and / or hardware.

[0134] The device of the above embodiment is used to implement the corresponding traffic aggregation method in any of the above embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0135] Based on the same inventive concept, corresponding to any of the above-mentioned embodiments and methods, the present disclosure also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the traffic aggregation method described in any of the above embodiments is implemented.

[0136] Figure 8 A more specific schematic diagram of the hardware structure of an electronic device provided in this embodiment is shown, and the device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040 are connected to each other through the bus 1050 in the device.

[0137] The processor 1010 can be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0138] The memory 1020 may be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 may store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program codes are stored in the memory 1020 and are called and executed by the processor 1010.

[0139] The input / output interface 1030 is used to connect the input / output module to realize information input and output. The input / output module can be configured in the device as a component (not shown in the figure), or it can be externally connected to the device to provide corresponding functions. The input device may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device may include a display, a speaker, a vibrator, an indicator light, etc.

[0140] The communication interface 1040 is used to connect a communication module (not shown) to realize communication interaction between the device and other devices. The communication module can realize communication through a wired mode (such as USB, network cable, etc.) or a wireless mode (such as mobile network, WIFI, Bluetooth, etc.).

[0141] The bus 1050 includes a path that transmits information between the various components of the device (eg, the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040).

[0142] It should be noted that, although the above device only shows the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040 and the bus 1050, in the specific implementation process, the device may also include other components necessary for normal operation. In addition, it can be understood by those skilled in the art that the above device may also only include the components necessary for implementing the embodiments of the present specification, and does not necessarily include all the components shown in the figure.

[0143] The electronic device of the above embodiment is used to implement the corresponding traffic aggregation method in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be repeated here.

[0144] Based on the same inventive concept, corresponding to any of the above-mentioned embodiment methods, the present disclosure also provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the traffic aggregation method described in any of the above embodiments.

[0145] The above-mentioned non-transitory computer-readable storage medium can be any available medium or data storage device that can be accessed by a computer, including but not limited to magnetic storage (such as floppy disks, hard disks, magnetic tapes, magneto-optical disks (MO), etc.), optical storage (such as CD, DVD, BD, HVD, etc.), and semiconductor storage (such as ROM, EPROM, EEPROM, non-volatile memory (NAND FLASH), solid-state drive (SSD)), etc.

[0146] The computer instructions stored in the storage medium of the above embodiment are used to enable the computer to execute the traffic aggregation method described in any embodiment in the above exemplary method part, and have the beneficial effects of the corresponding method embodiment, which will not be repeated here.

[0147] Those skilled in the art will appreciate that the embodiments of the present disclosure may be implemented as a system, method, or computer program product. Therefore, the present disclosure may be specifically implemented in the following forms, namely: complete hardware, complete software (including firmware, resident software, microcode, etc.), or a combination of hardware and software, generally referred to herein as a "circuit," "module," or "system." In addition, in some embodiments, the present disclosure may also be implemented in the form of a computer program product in one or more computer-readable media, which contains computer-readable program code.

[0148] Any combination of one or more computer-readable media may be used. A computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive examples) of computer-readable storage media may include, for example: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device or device.

[0149] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0150] The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0151] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0152] It should be understood that each box in the flowchart and / or block diagram and the combination of boxes in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer or other programmable data processing device to produce a machine, and these computer program instructions are executed by a computer or other programmable data processing device to produce a device that implements the functions / operations specified in the boxes in the flowchart and / or block diagram.

[0153] These computer program instructions may also be stored in a computer-readable medium that enables a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable medium produce a product that includes an instruction device that implements the functions / operations specified in the blocks in the flowchart and / or block diagram.

[0154] Computer program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device so that a series of operational steps are performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby enabling the instructions executed on the computer or other programmable device to provide a process for implementing the functions / operations specified in the blocks in the flowchart and / or block diagram.

[0155] In addition, although the operations of the disclosed method are described in a particular order in the accompanying drawings, this does not require or imply that the operations must be performed in this particular order, or that all the operations shown must be performed to achieve the desired results. On the contrary, the steps depicted in the flow chart can be performed in a different order. Additionally or alternatively, some steps can be omitted, multiple steps can be combined into one step, and / or one step can be decomposed into multiple steps.

[0156] The use of the verbs "comprise", "include" and their conjugations mentioned in the application documents does not exclude the presence of elements or steps other than those recorded in the application documents. The article "a" or "an" before an element does not exclude the presence of a plurality of such elements.

[0157] Although the spirit and principle of the present disclosure have been described with reference to several specific embodiments, it should be understood that the present disclosure is not limited to the disclosed specific embodiments, and the division of various aspects does not mean that the features in these aspects cannot be combined to benefit, and such division is only for the convenience of expression. The present disclosure is intended to cover various modifications and equivalent arrangements included in the spirit and scope of the attached claims. The scope of the attached claims conforms to the broadest interpretation, thereby including all such modifications and equivalent structures and functions.

Claims

1. A method for converging traffic, characterized in that: include: Acquire historical service traffic data on the IP side, and input the historical service traffic data into a pre-built service traffic prediction model to obtain a service traffic prediction result of the IP side in a target time period output by the service traffic prediction model; Combined with the service traffic prediction result, the small-granular service traffic on the IP side is pre-arranged to occupy the bandwidth granular pipeline resources on the optical network side as much as possible, and the pre-arrangement result is recorded in the controller. In the pre-arrangement process, the primary and backup services on the IP side are respectively carried on physically separated optical network links to form a service multi-path bearing model; Pre-loading the pre-scheduling result to obtain a pre-loaded network resource status result, and adjusting the service flow prediction model according to the pre-loaded network resource status result to obtain an adjusted service flow prediction model, and correcting the pre-scheduling result by using the adjusted service flow prediction model to obtain a scheduling result; According to the scheduling result, in the target time period, the service traffic on the IP side is aggregated to the optical network side.

2. The method according to claim 1, characterized in that The business flow prediction model includes: Gated Recurrent Neural Network Model.

3. The method according to claim 1, characterized in that The method further comprises: Construct a sample set including several samples; wherein the samples include: sample data and label data; the sample data includes historical service flow data of the training IP side; the label data includes service flow prediction results of the training IP side in the target time period corresponding to the historical service flow data of the training IP side; According to the sample set, the business traffic prediction model is constructed and trained through a gated recurrent neural network algorithm.

4. The method according to claim 1, characterized in that The step of aggregating the service traffic on the IP side to the optical network side in the target time period according to the scheduling result includes: Allocate services of different granularities to corresponding optical network layer link channels.

5. The method according to claim 1, characterized in that After aggregating the service traffic on the IP side to the optical network side in the target time period according to the scheduling result, the method further includes: Update path resource information.

6. A flow convergence device, characterized in that: include: A service flow prediction module is configured to obtain historical service flow data of the IP side, input the historical service flow data into a pre-built service flow prediction model, and obtain a service flow prediction result of the IP side in a target time period output by the service flow prediction model; The pre-arrangement module is configured to pre-arrange the small-granularity service traffic on the IP side in combination with the service traffic prediction result, so as to make it occupy the bandwidth granular pipeline resources on the optical network side as much as possible, and record the pre-arrangement result in the controller. In the pre-arrangement process, the primary and backup services on the IP side are respectively carried on physically separated optical network links to form a service multi-path bearing model; an orchestration result correction module, configured to preload the pre-orchestrated result to obtain a pre-loaded network resource status result, and adjust the service flow prediction model according to the pre-loaded network resource status result to obtain an adjusted service flow prediction model, and correct the pre-orchestrated result by using the adjusted service flow prediction model to obtain an orchestration result; The service traffic scheduling module is configured to converge the service traffic on the IP side to the optical network side in the target time period according to the scheduling result.

7. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method according to any one of claims 1 to 5 is implemented.

8. A non-transitory computer-readable storage medium, characterized in that: The non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to cause a computer to execute the method according to any one of claims 1 to 5.

Images