CN113872703A - Method and system for predicting multi-network metadata in quantum communication network - Google Patents

Method and system for predicting multi-network metadata in quantum communication network Download PDF

Info

Publication number
CN113872703A
CN113872703A CN202111089520.0A CN202111089520A CN113872703A CN 113872703 A CN113872703 A CN 113872703A CN 202111089520 A CN202111089520 A CN 202111089520A CN 113872703 A CN113872703 A CN 113872703A
Authority
CN
China
Prior art keywords
network
model
lstm
dbn
quantum communication
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.)
Granted
Application number
CN202111089520.0A
Other languages
Chinese (zh)
Other versions
CN113872703B (en
Inventor
陈柱
夏晨臣
王平
陈德华
王旭东
张玲君
翟学锋
张远旸
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.)
Cas Quantum Network Co ltd
Original Assignee
Cas Quantum Network Co 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 Cas Quantum Network Co ltd filed Critical Cas Quantum Network Co ltd
Priority to CN202111089520.0A priority Critical patent/CN113872703B/en
Publication of CN113872703A publication Critical patent/CN113872703A/en
Application granted granted Critical
Publication of CN113872703B publication Critical patent/CN113872703B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B10/00—Transmission systems employing electromagnetic waves other than radio-waves, e.g. infrared, visible or ultraviolet light, or employing corpuscular radiation, e.g. quantum communication
    • H04B10/70—Photonic quantum communication
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H04L41/147—Network analysis or design for predicting network behaviour

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Molecular Biology (AREA)
  • Artificial Intelligence (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Electromagnetism (AREA)
  • Optics & Photonics (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The embodiment of the invention provides a method and a system for predicting multi-network metadata in a quantum communication network, a method for performing reliability evaluation prediction on the prediction method, and a method for evaluating the reliability evaluation prediction method. The method for predicting the multi-network metadata in the quantum communication network comprises the following steps: acquiring multi-network metadata in a quantum communication network; performing data preprocessing on the multi-network metadata; and predicting the preprocessed multi-network metadata by using the LSTM-DBN network model to obtain a predicted value of the multi-network element data at the next moment, wherein the LSTM-DBN network model comprises an LSTM model and a DBN model, and the predicted value of the multi-network element data at the next moment comprises the network state of quantum communication at the next moment and the multi-network metadata at the next moment. By the method, the characteristics of the LSTM model and the DBN model can be fully utilized, and the prediction accuracy is improved while the reliability of the network model is guaranteed.

Description

Method and system for predicting multi-network metadata in quantum communication network
Technical Field
The invention relates to the technical field of quantum communication networks, in particular to a method and a system for predicting multi-network metadata in a quantum communication network.
Background
A quantum communication network is a network employing a quantum communication system. As shown in fig. 1, the quantum communication network is composed of a plurality of subsystems, such as a quantum network subsystem, a transmission network subsystem, a classical network subsystem, a machine room environment subsystem, a secure network subsystem, and the like. Each subsystem also comprises a plurality of separated nodes, for example, the quantum network subsystem comprises an OKD network element, an OKM network element, an S600 network element, etc.; the transmission network subsystem includes transmission equipment, optical fiber lines, and the like. The quantum information is stored in the detection data (i.e. network element data) of these nodes.
Because network element data in the quantum communication network has time dependency and high-dimensional complexity, and the data characteristics of different network elements are different, the traditional machine learning method cannot be well applied to the quantum communication network. For example, artificial neural networks ANN, SVM, PCA and the like are relatively lack of data feature extraction capability; although the RNN can theoretically represent the dependency relationship between long-time step states, the memory capacity of the RNN in practical application is limited due to the existence of gradient explosion or gradient dispersion, and the dependency relationship of the RNN only can be learned to a short-time step; HMMs are limited to discrete hidden states and, like RNNs, have long-term dependency problems.
Therefore, a method and a system for predicting multi-network metadata in a quantum communication network with more stability and reliability are needed.
Disclosure of Invention
Therefore, an object of the embodiments of the present invention is to overcome the above-mentioned drawbacks of the related art, and provide a method and a system for predicting multi-network metadata in a quantum communication network, which can improve the prediction accuracy on the premise of ensuring reliability.
The above purpose is realized by the following technical scheme:
according to a first aspect of the embodiments of the present invention, there is provided a method for predicting multi-network metadata in a quantum communication network, including: acquiring multi-network metadata in a quantum communication network; performing data preprocessing on the multi-network metadata; and predicting the preprocessed multi-network metadata by using an LSTM-DBN network model to obtain a predicted value of the multi-network element data at the next moment, wherein the LSTM-DBN network model comprises an LSTM model and a DBN model, and the predicted value of the multi-network element data at the next moment comprises the network state of the quantum communication at the next moment and the multi-network metadata at the next moment.
Optionally, the performing data preprocessing on the multi-network element data includes: and carrying out normalization processing and consistency processing on the multi-network element data.
Optionally, the LSTM-DBN network model is obtained by training in a manner of obtaining a training set of multi-network metadata of a quantum communication network; and respectively training an LSTM model and a DBN model by utilizing the multi-network metadata training set to obtain the well-trained LSTM-DBN network model.
Optionally, the predicting the preprocessed multi-network element data by using the LSTM-DBN network model to obtain a predicted value of the multi-network element data at the next time includes: inputting the preprocessed multi-network metadata into the well-trained LSTM model to obtain network element data of the next moment predicted by the LSTM model; and integrating the network element data of the next moment predicted by the LSTM model, inputting the trained DBN model, and obtaining the network state of the quantum communication network of the next moment and the multi-network metadata of the next moment.
Optionally, the method further includes: carrying out reliability evaluation prediction on the LSTM-DBN network model; and optimizing the LSTM-DBN network model based on the reliability assessment prediction value.
In another aspect of the present invention, a method for performing reliability assessment and prediction on the LSTM-DBN network model is provided, which includes: obtaining a predicted value of multi-network metadata in the quantum communication network at the next moment based on the LSTM-DBN network model; obtaining a prediction curve of the LSTM-DBN network model based on the predicted value of the multi-network metadata; and performing reliability evaluation prediction on the LSTM-DBN network model based on the cycle number of the prediction curve reaching a preset standard threshold for the first time to obtain a reliability evaluation prediction value.
In a third aspect of the present invention, there is provided a method for evaluating the reliability evaluation prediction method, including: evaluating the reliability assessment prediction method using a scoring function that is:
Figure BDA0003266891660000031
wherein h isi＝RAPi′-RAPi，RAPi' evaluation of prediction value, RAP, for reliability of label of network element data of ith subsystem in quantum communication networkiEvaluating a true value for the reliability of a label of network element data of an ith subsystem in the quantum communication network, wherein N is the number of subsystems in the quantum communication network, and the coefficient b is less than a; score is the Score function value.
Optionally, the method further includes: evaluating the reliability evaluation prediction value using a root mean square error, the root mean square error being:
Figure BDA0003266891660000032
wherein h isi＝RAPi′-RAPi，RAPi' evaluation of prediction value, RAP, for reliability of label of network element data of ith subsystem in quantum communication networkiAnd evaluating a true value for the reliability of the label of the network element data of the ith subsystem in the quantum communication network, wherein N is the number of subsystems in the quantum communication network, and RMSE is the root mean square error.
According to a fourth aspect of embodiments of the present invention, there is provided a computer-readable storage medium having stored thereon a computer program which, when executed, implements any of the methods described above.
According to a fifth aspect of embodiments of the present invention, there is provided an electronic device comprising a processor and a memory, the memory having stored therein a computer program that, when executed by the processor, implements any of the methods described above.
The technical scheme of the embodiment of the invention can have the following beneficial effects:
on one hand, aiming at the characteristics of time dependence, high-dimensional and complex data and the like of multi-network metadata in the quantum network, the network element data at the next moment and the network state of the quantum communication network are accurately predicted through an LSTM-DBN network model by utilizing the advantages of the LSTM network and the DBN model in the aspects of sequence learning and complex data characteristic extraction respectively; on the other hand, the stability and the reliability of the LSTM-DBN network model are ensured through reliability evaluation and prediction, so that the accuracy of model prediction is improved on the premise of ensuring the reliability.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention, as claimed.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description, serve to explain the principles of the invention. It is obvious that the drawings in the following description are only some embodiments of the invention, and that for a person skilled in the art, other drawings can be derived from them without inventive effort. In the drawings:
fig. 1 shows a schematic diagram of a system architecture in a quantum communication network according to an embodiment of the invention;
FIG. 2 illustrates a flow diagram of a method of predicting multi-network metadata in a quantum communication network, according to one embodiment of the invention;
FIG. 3 illustrates a flow diagram of a method for reliability assessment prediction of an LSTM-DBN network model according to one embodiment of the invention;
fig. 4 shows a block diagram of a system for predicting multi-network metadata in a quantum communication network, according to one embodiment of the invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail by embodiments with reference to the accompanying drawings. It is to be understood that the embodiments described are only a few embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Furthermore, the described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to provide a thorough understanding of embodiments of the invention. One skilled in the relevant art will recognize, however, that the invention may be practiced without one or more of the specific details, or with other methods, components, devices, steps, and so forth. In other instances, well-known methods, devices, implementations or operations have not been shown or described in detail to avoid obscuring aspects of the invention.
The block diagrams shown in the figures are functional entities only and do not necessarily correspond to physically separate entities. I.e. these functional entities may be implemented in the form of software, or in one or more hardware modules or integrated circuits, or in different networks and/or processor means and/or microcontroller means.
The flow charts shown in the drawings are merely illustrative and do not necessarily include all of the contents and operations/steps, nor do they necessarily have to be performed in the order described. For example, some operations/steps may be decomposed, and some operations/steps may be combined or partially combined, so that the actual execution sequence may be changed according to the actual situation.
Long short-term memory (LSTM) is a special Recurrent Neural Network (RNN). In addition to including a standard loop layer, LSTM networks also introduce a threshold mechanism, i.e., a "memory" control gate, to control the rate of accumulation of information. The LSTM network standard module is divided into two parts, long-term state (ct) and short-term state (ht), with 3 control gates added along the state path: the input gate, the forgetting gate and the output gate are used for adjusting information. The LSTM network controls the transmission state by gating the state, remembering that it requires long time to remember, and forgetting unimportant information. Therefore, compared with the common RNN network, the LSTM network has better performance in longer sequences, and can avoid the problems of gradient extinction and gradient explosion in the training process of long sequences.
Deep Belief Network (DBN) is a semi-supervised learning method based on a Restricted Boltzmann Machine (RBM). The DBN network overall structure is a deep feedforward neural network and comprises an input layer, a plurality of RBM layers and an output layer, and model design is divided into two stages of pre-training and reverse fine tuning. The DBN network is a probabilistic generative model that builds a joint distribution between observed data and labels, as opposed to the neural network of the traditional discriminant model. By training the weights among its neurons, the entire neural network can be made to generate training data with maximum probability. Thus, a DBN network can establish a good mapping from raw data to low-dimensional abstract features through a deep network.
Based on the above research, an embodiment of the present invention provides a method for predicting multi-network metadata in a quantum communication network, where the method uses an LSTM-DBN network model to predict preprocessed multi-network metadata to obtain a predicted value of multi-network-element data at a next time, where the LSTM-DBN network model includes an LSTM model and a DBN model, and the predicted value of the multi-network-element data at the next time includes a network state of quantum communication at the next time and the multi-network metadata at the next time.
Fig. 2 shows a flow diagram of a method of predicting multi-network metadata in a quantum communication network, according to an embodiment of the invention. As shown in fig. 2, the method comprises the steps of:
s210, acquiring multi-network metadata in the quantum communication network.
As described above, a quantum communication network is composed of numerous subsystems such as a quantum network subsystem, a transmission network subsystem, a classical network subsystem, a room environment subsystem, and a secure network subsystem. Each subsystem also comprises a plurality of separated nodes, for example, the quantum network subsystem comprises an OKD network element, an OKM network element, an S600 network element, etc.; the transmission network subsystem includes transmission equipment, optical fiber lines, and the like. And acquiring network element data of one or more subsystems in the quantum communication network to predict and obtain the labels of the subsystems.
The method and system for predicting network element data in the present invention will be described below by taking OKD network element, OKM network element and S600 network element data in a quantum network subsystem as example network element data, but it should be understood by those skilled in the art that the method and system provided in the present invention are also applicable to network element data of other subsystems in a quantum communication network.
S220, data preprocessing is carried out on the multi-network metadata.
Because the original network element data directly acquired in the quantum communication network may have the problems of repetition, incompleteness, noise, consistency and the like, the acquired network element data can be subjected to data preprocessing operations such as duplication removal, completion, denoising, alignment and the like.
In one embodiment, since the monitoring data of each subsystem of the quantum communication network represents different characteristics, namely, dimensions and magnitude are different, and the monitoring data is divided into two types, namely, a network type index and a service type index, in order to eliminate the influence of data non-specification on the prediction effect, improve the prediction precision and the interpretability and monotonicity of a label, and perform normalization and uniformization processing on the original network element data.
The purpose of the data normalization process is to confine the original network element data to the range of [0, 1 ]. In one embodiment, the normalization process may be performed by equation (1):
Figure BDA0003266891660000071
wherein x isi，j(t) is the original monitoring data of the jth network element in the ith subsystem in the quantum communication network at the time t, min (x)i，j) And max (x)i，j) Respectively representing the minimum value and the maximum value, x, of the jth network element in the ith subsystem at all the time pointsi，jAnd (t) normalizing the original monitoring data of the jth network element in the ith subsystem at the time t.
In one embodiment, the normalized network element data may be subjected to a unification process by formula (2):
Figure BDA0003266891660000081
wherein x isi，j(t) is the normalized value of the original monitoring data of the jth network element in the ith subsystem at the time t,
Figure BDA0003266891660000082
is to xi，j(t) the value after the matching process.
In an embodiment, the normalized and unified OKD network element data, OKM network element data, and S600 network element data may be continuously intercepted within the same time period by using a fixed time length as a unit, so as to obtain a plurality of sets of multi-network metadata with time continuity, which are used for model prediction.
And S230, inputting the preprocessed multi-network metadata into the trained LSTM model to obtain the network element data at the next moment.
The LSTM-DBN network model includes an LSTM model and a DBN model. In one embodiment, a training set of the network model may be first constructed, the two models may be trained separately using the training set, and the output error may be minimized by adjusting the hyper-parameters to obtain the trained LSTM model and DBN model.
In LSTM model training, the data in the training set are sets of tagged net element data with temporal continuity. In one embodiment, a group of continuous-time network element data sequences can be intercepted according to a certain period length, and then the network element data sequences are sequentially slid backwards for one period until the end, so that the continuous-time network element data sequences of different periods related to the same network element are obtained. And training an LSTM model by taking the network element data sequence in a certain period as input and the network element data sequence in the next period as output, and finally adjusting model parameters by using a BPTT algorithm to obtain the trained LSTM model. In the LSTM model, the network element data in the training set are respectively input into the input layer unit of the LSTM model, and are processed and then introduced into the full connection layer, so that the time domain dimension of the full connection layer is reduced to be consistent with the data label in the training set.
In DBN model training, data labels in a training set may be set by fusing multi-mesh metadata. Compared with single network element data, the multi-network element data fusion can more comprehensively represent the running state of the quantum communication network, and meanwhile, the model parameters can be adjusted through reverse fine tuning in the process of training the DBN model, so that the accuracy of the model is improved. Thus, in one embodiment, the data labels in the DBN model training set can be set by equation (3):
Figure BDA0003266891660000091
wherein the content of the first and second substances,
Figure BDA0003266891660000092
is the value omega of the original monitoring data of the jth network element in the ith subsystem in the quantum communication network after being preprocessed at the time tjThe fusion coefficient of the jth network element data represents the proportion of the network element data in the data label, and M is the number of subsystems.
In one embodiment, to prevent overfitting, improve the generalization performance of the model, regularization is introduced in the two-stage training process, such as Dropout regularization or L2/L1 regularization.
And predicting different network element data in the preprocessed quantum communication network by using the trained LSTM model to obtain the network element data at the next moment.
And S240, integrating the network element data at the next moment, inputting the network element data into the trained DBN model, and obtaining the network state of the quantum communication network at the next moment and the multi-network element data at the next moment.
And integrating and inputting the prediction result of the LSTM model into the trained DBN model for feature extraction and prediction, so as to obtain the state of the quantum communication network and network element data in the next effective time period.
The embodiment aims at the characteristics of time dependency, high-dimensional and complex data and the like of multi-network metadata in the quantum network, and by utilizing the advantages of the LSTM network and the DBN model in the aspects of sequence learning and complex data characteristic extraction, network element data at the next moment and the network state of the quantum communication network can be accurately predicted through the LSTM-DBN network model.
In order to ensure the stability and Reliability of the LSTM-DBN network model, in an embodiment of the present invention, a Reliability Assessment Prediction (RAP) is also performed on the LSTM-DBN network model, and the LSTM-DBN network model is optimized based on the RAP value.
FIG. 3 shows a flowchart of a method for reliability assessment prediction of an LSTM-DBN network model according to an embodiment of the invention, the method comprising the steps of:
s310, obtaining a predicted value of multi-network metadata in the quantum communication network at the next moment based on the LSTM-DBN network model.
This step can be realized through the above steps S210-S240, which are not described herein again.
S320, obtaining a prediction curve of the LSTM-DBN network model based on the prediction value of the multi-network metadata.
And connecting the predicted values obtained by the LSTM-DBN network model prediction to obtain a prediction curve.
S330, based on the cycle number of the prediction curve reaching the preset standard threshold for the first time, reliability evaluation prediction is carried out on the LSTM-DBN network model, and an RAP value is obtained.
In the above embodiment, by performing reliability evaluation prediction on the LSTM-DBN network model, the model may be optimized according to the prediction result, so as to further improve the reliability and accuracy of the model. In one embodiment, during optimization, the predicted network element data at the next moment and the network state of the quantum communication network are combined, and corresponding optical cable optimization, node optimization and business level optimization are selected to improve the overall reliability of the network.
In order to objectively and effectively evaluate the performance of the reliability evaluation prediction method and ensure the generalization capability of the reliability evaluation prediction method, the invention also provides a method for evaluating the reliability evaluation prediction method.
In one embodiment, the reliability assessment prediction method described above may be evaluated using a scoring function. The scoring function is calculated as:
Figure BDA0003266891660000101
wherein h isi＝RAPi′-RAPi，RAPi' evaluation of prediction value, RAP, for reliability of label of network element data of ith subsystem in quantum communication networkiEvaluating a true value for the reliability of a label of network element data of an ith subsystem in the quantum communication network, wherein N is the number of subsystems in the quantum communication network, and the coefficient b is less than a; score is a Score function value, and the smaller the value, the better the RAP prediction effect.
Since the cost of catastrophic results caused by untimely operation and maintenance is far more than the cost of wasting resources due to excessive operation and maintenance, the scoring function makes different penalties for model prediction RAP errors, namely, for the case of overestimating RAP (namely, h)i≧ 0), which is more penalized than underestimating RAP (i.e., h)i< 0) to correspond to quantaReliability requirements of the communication network.
However, the evaluation of the overall prediction performance of the model is sometimes affected by the occurrence of an abnormal value (e.g., too large or too small) depending on the scoring function alone, and for this reason, in another embodiment, the RAP prediction effect may be evaluated by using Root Mean Square Error (RMSE) to avoid the phenomenon of artificially lowering the scoring function value. The root mean square error can be expressed as:
Figure BDA0003266891660000111
wherein h isi＝RAPi′-RAPi，RAPi' evaluation of prediction value, RAP, for reliability of label of network element data of ith subsystem in quantum communication networkiAnd (3) evaluating a true value for the reliability of the label of the network element data of the ith subsystem in the quantum communication network, wherein N is the number of subsystems in the quantum communication network, RMSE is root mean square error, and the smaller the numerical value is, the better the RAP prediction effect is.
Fig. 4 shows a block diagram of a system for predicting multi-network metadata in a quantum communication network, according to one embodiment of the invention. As shown in FIG. 4, the system 400 includes a receiving module 410, a preprocessing module 420, a prediction module 430, a training module 440, and an evaluation module 450. Although the block diagrams depict components in a functionally separate manner, such depiction is for illustrative purposes only. The components shown in the figures may be arbitrarily combined or separated into separate software, firmware, and/or hardware components. Moreover, regardless of how such components are combined or divided, they may execute on the same computing device or multiple computing devices, which may be connected by one or more networks.
The receiving module 410 is used for acquiring multi-network metadata in the quantum communication network. The preprocessing module 420 is used for performing data preprocessing on the multi-network metadata. The prediction module 430 is configured to predict the preprocessed multi-network metadata by using an LSTM-DBN network model, and obtain a predicted value of the multi-network-element data at the next time, where the LSTM-DBN network model includes an LSTM model and a DBN model, and the predicted value of the multi-network-element data at the next time includes a network state of the quantum communication at the next time and the multi-network-element data at the next time. The training module 440 is configured to train the LSTM model and the DBN model respectively using the multi-network metadata training set to obtain a trained LSTM-DBN network model. The evaluation module 450 is used for performing reliability evaluation prediction on the LSTM-DBN network model.
In another embodiment of the present invention, a computer-readable storage medium is further provided, on which a computer program or executable instructions are stored, and when the computer program or the executable instructions are executed, the technical solution as described in the foregoing embodiments is implemented, and the implementation principle thereof is similar, and is not described herein again. In embodiments of the present invention, the computer readable storage medium may be any tangible medium that can store data and that can be read by a computing device. Examples of computer readable storage media include hard disk drives, Network Attached Storage (NAS), read-only memory, random-access memory, CD-ROMs, CD-R, CD-RWs, magnetic tapes, and other optical or non-optical data storage devices. The computer readable storage medium may also include computer readable media distributed over a network coupled computer system so that computer programs or instructions may be stored and executed in a distributed fashion.
In another embodiment of the invention, the invention may be implemented in the form of an electronic device. The electronic device comprises a processor and a memory in which a computer program is stored which, when being executed by the processor, can be used for carrying out the method of the invention.
Reference in the specification to "various embodiments," "some embodiments," "one embodiment," or "an embodiment," etc., means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment. Thus, appearances of the phrases "in various embodiments," "in some embodiments," "in one embodiment," or "in an embodiment," or the like, in various places throughout this specification are not necessarily referring to the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. Thus, a particular feature, structure, or characteristic illustrated or described in connection with one embodiment may be combined, in whole or in part, with a feature, structure, or characteristic of one or more other embodiments without limitation, as long as the combination is not logical or operational.
The terms "comprises," "comprising," and "having," and similar referents in this specification, are intended to cover non-exclusive inclusions, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not limited to only those steps or elements but may alternatively include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus. The word "a" or "an" does not exclude a plurality. Additionally, the various elements of the drawings of the present application are merely schematic illustrations and are not drawn to scale.
Although the present invention has been described by the above embodiments, the present invention is not limited to the embodiments described herein, and various changes and modifications may be made without departing from the scope of the present invention.

Claims (11)

1. A method of predicting multi-network metadata in a quantum communication network, comprising:
acquiring multi-network metadata in a quantum communication network;
performing data preprocessing on the multi-network metadata; and
and predicting the preprocessed multi-network metadata by using an LSTM-DBN network model to obtain a predicted value of the multi-network element data at the next moment, wherein the LSTM-DBN network model comprises an LSTM model and a DBN model, and the predicted value of the multi-network element data at the next moment comprises the network state of the quantum communication at the next moment and the multi-network metadata at the next moment.
2. The method of claim 1, wherein the data preprocessing the multi-network element data comprises: and carrying out normalization processing and consistency processing on the multi-network element data.
3. The method of claim 1, wherein the LSTM-DBN network model is obtained by training in the following manner:
acquiring a training set of multi-network metadata of a quantum communication network; and
and respectively training an LSTM model and a DBN model by utilizing the multi-network metadata training set to obtain the well-trained LSTM-DBN network model.
4. The method of claim 1, wherein the predicting the preprocessed multi-network element data by using the LSTM-DBN network model to obtain a predicted value of the multi-network element data at the next time comprises:
inputting the preprocessed multi-network metadata into the well-trained LSTM model to obtain network element data of the next moment predicted by the LSTM model; and
and integrating the network element data of the next moment predicted by the LSTM model, and inputting the trained DBN model to obtain the network state of the quantum communication network of the next moment and the multi-network element data of the next moment.
5. The method of claim 1, further comprising,
carrying out reliability evaluation prediction on the LSTM-DBN network model; and
optimizing the LSTM-DBN network model based on the reliability assessment prediction value.
6. A method of making reliability assessment predictions for the LSTM-DBN network model of any of claims 1-5, comprising:
obtaining a predicted value of multi-network metadata in the quantum communication network at the next moment based on the LSTM-DBN network model;
obtaining a prediction curve of the LSTM-DBN network model based on the predicted value of the multi-network metadata; and
and performing reliability evaluation prediction on the LSTM-DBN network model based on the cycle number of the prediction curve reaching a preset standard threshold for the first time to obtain a reliability evaluation prediction value.
7. A method of evaluating the reliability assessment prediction method of claim 6, comprising: evaluating the reliability assessment prediction method using a scoring function that is:
Figure FDA0003266891650000021
wherein h isi＝RAPi′-RAPi，RAPi' evaluation of prediction value, RAP, for reliability of label of network element data of ith subsystem in quantum communication networkiEvaluating a true value for the reliability of a label of network element data of an ith subsystem in the quantum communication network, wherein N is the number of subsystems in the quantum communication network, and the coefficient b is less than a; score is the Score function value.
8. The method of claim 7, further comprising: evaluating the reliability evaluation prediction value using a root mean square error, the root mean square error being:
Figure FDA0003266891650000031
wherein h isi＝RAPi′-RAPi，RAPi' evaluation of prediction value, RAP, for reliability of label of network element data of ith subsystem in quantum communication networkiAnd evaluating a true value for the reliability of the label of the network element data of the ith subsystem in the quantum communication network, wherein N is the number of subsystems in the quantum communication network, and RMSE is the root mean square error.
9. A system for predicting multi-network metadata in a quantum communication network, comprising:
the receiving module is used for acquiring multi-network metadata in the quantum communication network;
the preprocessing module is used for preprocessing the data of the multi-network metadata;
the prediction module is used for predicting the preprocessed multi-network metadata by using an LSTM-DBN network model to obtain a predicted value of the multi-network element data at the next moment, wherein the LSTM-DBN network model comprises an LSTM model and a DBN model, and the predicted value of the multi-network element data at the next moment comprises the network state of the quantum communication at the next moment and the multi-network metadata at the next moment;
the training module is used for respectively training an LSTM model and a DBN model by utilizing a multi-network metadata training set so as to obtain the well-trained LSTM-DBN network model; and
and the evaluation module is used for carrying out reliability evaluation prediction on the LSTM-DBN network model.
10. A storage medium in which a computer program is stored which, when being executed by a processor, is operative to carry out the method of any one of claims 1-8.
11. An electronic device comprising a processor and a memory, the memory having stored therein a computer program which, when executed by the processor, is operable to carry out the method of any one of claims 1-8.
CN202111089520.0A 2021-09-16 2021-09-16 Method and system for predicting multi-network metadata in quantum communication network Active CN113872703B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111089520.0A CN113872703B (en) 2021-09-16 2021-09-16 Method and system for predicting multi-network metadata in quantum communication network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111089520.0A CN113872703B (en) 2021-09-16 2021-09-16 Method and system for predicting multi-network metadata in quantum communication network

Publications (2)

Publication Number Publication Date
CN113872703A true CN113872703A (en) 2021-12-31
CN113872703B CN113872703B (en) 2022-09-06

Family

ID=78996331

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111089520.0A Active CN113872703B (en) 2021-09-16 2021-09-16 Method and system for predicting multi-network metadata in quantum communication network

Country Status (1)

Country Link
CN (1) CN113872703B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116388864A (en) * 2023-05-31 2023-07-04 中诚华隆计算机技术有限公司 Quantum network device performance prediction method and device, electronic device and storage medium

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190354850A1 (en) * 2018-05-17 2019-11-21 International Business Machines Corporation Identifying transfer models for machine learning tasks
CN111260249A (en) * 2020-02-13 2020-06-09 武汉大学 Electric power communication service reliability assessment and prediction method and device based on LSTM and random forest mixed model
CN111934866A (en) * 2020-08-14 2020-11-13 国科量子通信网络有限公司 Multi-layer path automatic reduction method and system of quantum communication network
CN111934865A (en) * 2020-08-14 2020-11-13 国科量子通信网络有限公司 Method for evaluating operation index of quantum communication network based on entropy method

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190354850A1 (en) * 2018-05-17 2019-11-21 International Business Machines Corporation Identifying transfer models for machine learning tasks
CN111260249A (en) * 2020-02-13 2020-06-09 武汉大学 Electric power communication service reliability assessment and prediction method and device based on LSTM and random forest mixed model
CN111934866A (en) * 2020-08-14 2020-11-13 国科量子通信网络有限公司 Multi-layer path automatic reduction method and system of quantum communication network
CN111934865A (en) * 2020-08-14 2020-11-13 国科量子通信网络有限公司 Method for evaluating operation index of quantum communication network based on entropy method

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
LU LI，HE YIGANG: "Wind Turbine Planetary Gearbox Condition Monitoring Method Based on Wireless Sensor and Deep Learning Approach", 《IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT(VOLUME:70)》 *
王琴，陈以鹏: "机器学习在量子保密通信中的应用与研究", 《南京邮电大学学报(自然科学版)》 *
秦超等: "深度卷积记忆网络时空数据模型", 《自动化学报》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116388864A (en) * 2023-05-31 2023-07-04 中诚华隆计算机技术有限公司 Quantum network device performance prediction method and device, electronic device and storage medium
CN116388864B (en) * 2023-05-31 2023-08-11 中诚华隆计算机技术有限公司 Quantum network device performance prediction method and device, electronic device and storage medium

Also Published As

Publication number Publication date
CN113872703B (en) 2022-09-06

Similar Documents

Publication Publication Date Title
CN109804362B (en) Determining primary key-foreign key relationships by machine learning
Yoon et al. Semi-supervised learning with deep generative models for asset failure prediction
TW202139045A (en) Privacy protection-based target service model determination
US11586811B2 (en) Multi-layer graph-based categorization
CN110659742B (en) Method and device for acquiring sequence representation vector of user behavior sequence
CN111126576B (en) Deep learning training method
KR102444457B1 (en) Method for dialogue summarization with word graphs
CN113392209A (en) Text clustering method based on artificial intelligence, related equipment and storage medium
US20240028898A1 (en) Interpreting convolutional sequence model by learning local and resolution-controllable prototypes
CN113872703B (en) Method and system for predicting multi-network metadata in quantum communication network
CN115659244A (en) Fault prediction method, device and storage medium
Zhang et al. Feature relevance term variation for multi-label feature selection
Madhiarasan et al. Analysis of artificial neural network performance based on influencing factors for temperature forecasting applications
WO2021012263A1 (en) Systems and methods for end-to-end deep reinforcement learning based coreference resolution
CN111461862B (en) Method and device for determining target characteristics for service data
US20210174191A1 (en) Automated fine-tuning of a pre-trained neural network for transfer learning
Alam Recurrent neural networks in electricity load forecasting
Cebeci A recurrent neural network model for weather forecasting
CN116432810A (en) Traffic flow prediction model determination method, device, apparatus and readable storage medium
Watts et al. Local score dependent model explanation for time dependent covariates
US20220180180A1 (en) Data-driven neural network model compression
CN112463964A (en) Text classification and model training method, device, equipment and storage medium
US20240134937A1 (en) Method, electronic device, and computer program product for detecting model performance
US20220318627A1 (en) Time series retrieval with code updates
CN114898339B (en) Training method, device, equipment and storage medium of driving behavior prediction model

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