CN106789214B - A kind of network situation awareness method and device based on just remaining double string algorithms - Google Patents
A kind of network situation awareness method and device based on just remaining double string algorithms Download PDFInfo
- Publication number
- CN106789214B CN106789214B CN201611139491.3A CN201611139491A CN106789214B CN 106789214 B CN106789214 B CN 106789214B CN 201611139491 A CN201611139491 A CN 201611139491A CN 106789214 B CN106789214 B CN 106789214B
- Authority
- CN
- China
- Prior art keywords
- network
- flow
- time series
- neural network
- elman neural
- 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.)
- Expired - Fee Related
Links
Classifications
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/50—Monitoring users, programs or devices to maintain the integrity of platforms, e.g. of processors, firmware or operating systems
- G06F21/55—Detecting local intrusion or implementing counter-measures
- G06F21/552—Detecting local intrusion or implementing counter-measures involving long-term monitoring or reporting
-
- 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
-
- 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
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- 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
- H04L41/142—Network analysis or design using statistical or mathematical 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
- H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L63/00—Network architectures or network communication protocols for network security
- H04L63/20—Network architectures or network communication protocols for network security for managing network security; network security policies in general
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Computer Security & Cryptography (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Biomedical Technology (AREA)
- Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Computer Hardware Design (AREA)
- Algebra (AREA)
- Probability & Statistics with Applications (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of network situation awareness method and devices based on just remaining double string algorithms, this method comprises: using the threshold value and weight of just remaining double string algorithm adjustment Elman neural networks;Elman neural network is optimized, the time series of time series, the time series of network flow longitude and latitude and the network flow flow direction of network flow sample is obtained;Cluster normalized is carried out to the time series of the time series of network flow sample, the time series of network flow longitude and latitude and network flow flow direction, obtains the attack trend and situation flow direction of network flow.This method realizes the flowing situation for obtaining network flow.
Description
Technical field
The present invention relates to network technique fields, more particularly to a kind of network situation awareness side based on just remaining double string algorithms
Method and device.
Background technique
Currently, network situation awareness (Cyberspace Situation Awareness, CSA), i.e. network safety situation
Perceive (Network Security Situation Awareness, NSSA), this concept by Time Bass in 1999 for the first time
It proposes.Hereafter Stephen G.Batsen and Kokar M.M also proposed correlation model, purport on the basis of Time Bass
Apply mature situational awareness techniques and theory in network information security technology, to enhance the peace of computer network
Quan Xing.
The research direction of network situation awareness concentrates on the foundation of Situation Awareness model, the method for Situation Assessment and frame
Discussion and the raising of Tendency Prediction precision etc..In terms of network security situation awareness model, Jia Yan etc. is to collected network
Security postures Data Centralized Processing simultaneously carries out association analysis, and the prediction of situation is carried out by the index system built, is proposed
Model is suitable for larger network.It opens pellet etc. and proposes the NSSA model based on Autonomic computing, the model is first with certainly
It restrains feedback characteristics and extracts the factor for influencing network security, the foundation of Situation Evaluation Model then uses analytic hierarchy process (AHP) (AHP), passes through
Improved neural network method is used upper in terms of the prediction of situation.Liu Nian etc. is on the Research foundation of forefathers and is given birth to
The inspiration of object Immune System is proposed based on immune Situation Awareness method, and grey Markov model is applied to
Tendency Prediction, the model thus established are the preferable solutions of network security Initiative Defense.It opens brave building etc. and uses multisensor
Secure data in network is detected, thus obtains assets, threat data and the vulnerability data of network and to this progress
Game theory analysis proposes the NSSA method based on Markov betting model.The NSSA model of the propositions such as Liu Xiaowu will be from difference
The various information that sensor collection arrives carries out fusion treatment, and the optimal weights of D-S, the method are found with particle swarm optimization algorithm
Cyberthreat behavior can be identified well, and can accurately reflect the extent of the destruction of threat behavior.In networks security situation assessment side
Face, the Zhang Lijuan etc. of National Digital Switching System Engineering and Technology Research Center analyze the factor for influencing network safety situation, and
The Situation Evaluation Model of Fuzzy Level Analytic Approach is proposed based on this.Meng Jin etc. proposes the safe state of the hierarchical network with time parameter
Gesture assessment models are to merge the collected letter of multiple sensors on the basis of innovating to time-varying D-S evidence theory method
Breath is got.The stratification Situation Evaluation Model of the propositions such as Li Ling's Juan mainly uses gray relative analysis method to pay close attention to a certain specific time
Attack, and be thus associated in the network environment currently studied, realize the Situation Assessment of network security.Mostly
Several networks security situation assessments can all ignore the data distribution feature for influencing network security, and Li Fangwei etc. is then by improving tradition
Fuzzy Hierarchy Method, the networks security situation assessment model based on the method for proposition makes up this defect, and avoids carrying out
The problems such as excessively relying on expert when network security data pre-processes and causing pre-processed results excessively subjective.In network safety situation
Prediction aspect, selected prediction technique need to have very high sensibility to the historical evaluation data of target network.A Yuma man of virtue and ability
Deng two characteristics not only sensitive to historical data by Elman neural network but also with dynamic memory function, by the method
Method as Tendency Prediction is applied on the network safety system of its exploitation, and demonstrates the effective of the method with concrete instance
Property.In practical applications, often due to lacking sample data and making network safety situation prediction result accuracy decline, Xiang Xixi
The new Tendency Prediction method of equal propositions has well solved this using any sample is not needed come the Kalman Algorithm of training parameter
One problem.Chen Lei etc. analyzes the deficiency of traditional grey forecasting model, with adaptive grey parameter and waits dimension grey filling side
Method proposes the prediction that improved adaptive gray model carries out network safety situation.
But in the prior art, most of predictive situation cognitive methods only give the predictive situation value of whole network, pass
The Situation Awareness method of system only predicts some value, does not predict the longitude and latitude of flow and flow direction, can not provide whole
The flowing situation of a network network flow in a short period of time, i.e. network flow will wherefrom flow to where, allowing currently makes
User can not dispose in advance carries out Internet resources deployment preparation related to defending against network attacks.
Summary of the invention
The object of the present invention is to provide a kind of network situation awareness method and devices based on just remaining double string algorithms, to realize
Obtain the flowing situation of network flow.
In order to solve the above technical problems, the present invention provides a kind of network situation awareness method based on just remaining double string algorithms,
This method comprises:
Using the threshold value and weight of just remaining double string algorithm adjustment Elman neural networks;
Elman neural network is optimized, obtain the time series of network flow sample, network flow longitude and latitude when
Between sequence and network flow flow direction time series;
To the time of the time series of network flow sample, the time series of network flow longitude and latitude and network flow flow direction
Sequence carries out cluster normalized, obtains the attack trend and situation flow direction of network flow.
Preferably, the Elman neural network is the BP neural network with feedback.
Preferably, the threshold value and weight using just remaining double string algorithm adjustment Elman neural networks, comprising:
Training sample set is obtained, training sample set is trained using Elman neural network;
During being trained to training sample set, instruction is provided to just remaining double string algorithms by Elman neural network
Experienced mean square error MSE obtains the weight W and threshold value B of just remaining double string algorithm feedbacks, weight W and threshold value B is input to Elman
Neural network.
Preferably, the cluster normalized is the cluster normalized of Situation Awareness model.
The present invention also provides a kind of network situation awareness device based on just remaining double string algorithms, which includes:
Module is adjusted, for the threshold value and weight using just remaining double string algorithm adjustment Elman neural networks;
Optimization module obtains time series, the network of network flow sample for optimizing to Elman neural network
The time series of time series and the network flow flow direction of flow longitude and latitude;
Cluster module, time series and network for time series, network flow longitude and latitude to network flow sample
The time series of flux and flow direction carries out cluster normalized, obtains the attack trend and situation flow direction of network flow.
Preferably, the Elman neural network is the BP neural network with feedback.
Preferably, the adjustment module includes:
Training unit is trained training sample set using Elman neural network for obtaining training sample set;
Acquiring unit, for during being trained to training sample set, by Elman neural network to positive Yu Shuan
String algorithm provides the mean square error MSE of training, the weight W and threshold value B of just remaining double string algorithm feedbacks is obtained, by weight W and threshold value B
It is input to Elman neural network.
Preferably, the cluster normalized is the cluster normalized of Situation Awareness model.
A kind of network situation awareness method and device based on just remaining double string algorithms provided by the present invention, using positive Yu Shuan
The threshold value and weight of string algorithm adjustment Elman neural network;Elman neural network is optimized, network flow sample is obtained
Time series, the time series of network flow longitude and latitude and network flow flow direction time series;To network flow sample
The time series of time series, the time series of network flow longitude and latitude and network flow flow direction carries out cluster normalized,
Obtain the attack trend and situation flow direction of network flow.As it can be seen that adaptively adjusting Elman nerve by using just remaining double string algorithms
Network threshold and weight, and then optimization neural network model predicts network flow sample size in the short time, affiliated longitude and latitude
With the time sequential value of flow direction, then each attribute forecast value of acquisition is passed through at the cluster normalization of Situation Awareness model
Reason obtains attack trend and situation flow direction in the network short time, is conducive to the anti-of Internet resources deployment management and network attack
It is imperial.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
The embodiment of invention for those of ordinary skill in the art without creative efforts, can also basis
The attached drawing of offer obtains other attached drawings.
Fig. 1 is a kind of flow chart of network situation awareness method based on just remaining double string algorithms provided by the present invention;
Fig. 2 is positive remaining double string Optimum search figures;
Fig. 3 is SCA-Elman training Optimized model schematic diagram;
Fig. 4 is SCA-Elman prediction model schematic diagram;
Fig. 5 is Situation Awareness System model schematic;
Fig. 6 is a kind of structural representation of network situation awareness device based on just remaining double string algorithms provided by the present invention
Figure.
Specific embodiment
Core of the invention is to provide a kind of network situation awareness method and device based on just remaining double string algorithms, to realize
Obtain the flowing situation of network flow.
In order to enable those skilled in the art to better understand the solution of the present invention, below in conjunction in the embodiment of the present invention
Attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is only
A part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art
Every other embodiment obtained without making creative work, shall fall within the protection scope of the present invention.
Referring to FIG. 1, Fig. 1 is a kind of network situation awareness method based on just remaining double string algorithms provided by the present invention
Flow chart, this method comprises:
S11: using the threshold value and weight of just remaining double string algorithm adjustment Elman neural networks;
S12: optimizing Elman neural network, obtains time series, the network flow longitude and latitude of network flow sample
Time series and network flow flow direction time series;
S13: to the time series of network flow sample, the time series of network flow longitude and latitude and network flow flow direction
Time series carries out cluster normalized, obtains the attack trend and situation flow direction of network flow.
As it can be seen that this method adaptively adjusts Elman neural network threshold value and weight by using just remaining double string algorithms, in turn
Optimization neural network model predicts the time sequential value of network flow sample size in the short time, affiliated longitude and latitude and flow direction,
Then cluster normalized that each attribute forecast value of acquisition is passed through to Situation Awareness model, obtained in the network short time
Attack tends to and situation flow direction, is conducive to the defence of Internet resources deployment management and network attack.
Based on the above method, wherein just remaining double string algorithms (Sine Cosine Algorithm, SCA) are thought according to the Confucian school
Want a kind of novel swarm intelligence algorithm proposed.
For just remaining double string algorithms, specifically, the particle in population makees logarithmic spiral fortune by just remaining bispin both direction
Dynamic capture optimal solution, wherein the movement of cosine Optimum search is with optimum individual XbestAs the positioning coordinate of screw motion, quickening
Convergence rate, and sinusoidal Optimum search movement is with random individual XrandFor the positioning coordinate of screw motion, avoid individual to part
Very small region is drawn close, and the ability that algorithm jumps out local optimum is improved.And introduce chaos operator in the algorithm to control population at individual
Moving region, the calculating of just remaining double strings is respectively as shown in formula (1) and (2).
Formula (1) and formula (2) are the formula of sine spiral movement and cosine screw motion Optimum search respectively, wherein parameter r1
Effect be to control sinusoidal global search and regional scope that cosine is locally developed, calculate as shown in formula (3).Parameter r2It is base
In the Optimizing operator of cube chaotic maps, calculate as shown in formula (4).Parameter r3The random number on section [0,2], control with
Machine individual XrandWith optimum individual XbestInfluence apart from current individual X how far.
In formula (3), a is constant, and t is current iteration number, and T is total the number of iterations, r1With the increase of the number of iterations
It is adaptive to reduce, the optimizing regional scope of just remaining double strings is reduced, algorithm is converged on finally in the same optimal solution, ensure that calculation
The convergence of method.
In formula (4), by the randomness and ergodic of chaotic maps itself, it is adaptively adjusted the variation of population at individual
Degree enhances the ability that just remaining double strings jump out locally optimal solution in optimizing.
Random population individual X is used in the sinusoidal chaos screw motion of formula (1)randIt is right around sine for navigation coordinate
The track movement of number helical, global search find optimal solution, maintain the diversity of population, avoid population at individual concentrations, fall into
Enter locally optimal solution.Pass through the cosine chaos screw motion of formula (2), simultaneously with optimum individual XbestFor navigation coordinate, accelerate to seek
The speed of excellent positioning.
Fig. 2 illustrates particle individual respectively by just remaining double string mechanism middle progress within the scope of the different zones in the same space
Global search and local optimal searching.Sinusoidal chaos global search reduces the optimizing blind spot that cosine chaos is locally developed, and avoids potential
The case where optimal solution is lost.And locally exploitation makes up that sinusoidal chaos global search convergence rate is slow to be lacked for cosine chaos
It falls into, improves the efficiency of algorithm.And by introducing greedy mechanism, what more sinusoidal chaos predation and the predation of cosine chaos generated
Solution, preferentially retains.The double strings of just remaining chaos intersect optimizing, complement each other, promote individual information fast propagation in population, make to plant
Group's individual finally converges in the same optimal solution, on the one hand prevents algorithm precocious, improves solving precision, on the other hand accelerates to receive
Speed is held back, solution efficiency is improved.
Based on the above method, further, the Elman neural network is the BP neural network with feedback.
Further, the process of step S11 specifically includes:
S1: training sample set is obtained, training sample set is trained using Elman neural network;
S2: it during being trained to training sample set, is provided by Elman neural network to just remaining double string algorithms
Trained mean square error MSE obtains the weight W and threshold value B of just remaining double string algorithm feedbacks, weight W and threshold value B is input to
Elman neural network.
Detailed, this method is using just remaining double string algorithm (SCA) optimization Elman neural network threshold values and weight.
Elman neural network is a kind of BP network of band feedback, before having local memory unit, LOCAL FEEDBACK connection
To neural network and multilayered structure similar with Multilayer Feedforward Neural Networks.Compared with traditional BP neural network, Elman nerve net
Network has stronger dynamic behaviour and computing capability, the prediction model suitable for settling time sequence.However Elman and BP mind
Equally use momentum gradient descent method adjustment weight and threshold value through network, be easily trapped into local optimum, and when influence factor and
When learning sample increases, the calculation amount and weight number of neural network will be sharply increased, and cause convergence rate slow.Therefore exist herein
SCA algorithm is added in the training of Elman and carrys out optimization neural network, by SCA algorithm, makes the threshold value and power of Elman neural network
The adaptive adjustment of value, improves precision of prediction and speed.
Fig. 3 illustrates the training Optimized model of SCA-Elman, if neural network output layer vector is n dimension, hidden layer and holds
Connect layer vector be h dimension and output layer vector be m tie up.And IWhn、OWmhAnd CWhhInput layer respectively to hidden layer,
Hidden layer arrives the connection weight of hidden layer, and b to output layer and acceptingHh、bCh、bMmRespectively hidden layer, undertaking layer and output layer
Corresponding threshold values.Elman neural network starts to train after obtaining training sample set, provides training to SCA algorithm during its
Weight W from mean square error MSE, SCA algorithm to the corresponding training layer of Elman neural network feedback and threshold values B.Then Elman mind
The adaptive adjusting training of value fed back through network according to SCA algorithm, shown in evaluation fitness function such as formula (5), yk(w) andIt is the desired output and real output value of Elman neural network respectively, m is output layer training dimension.
The specific training process of CCGWO algorithm optimization Elman neural network is as follows:
Step 1 establishes Elman neural network, and basic parameter is arranged;
Step 2, initialization population, individual UVR exposure are as follows:
II=1 ..., N=[IW11...IWhnCW11...CWhhOW11...OWmhbH1...bHhbC1...bChbM1...bMm];
Step 3, training Elman neural network, the fitness value of population at individual is calculated according to training result using formula (5);
Step 4 carries out just remaining double string optimizing search exploitations according to formula (1) and formula (2), decodes to population at individual, transverse and longitudinal is double
To filial generation vie each other and preferentially leave, and adaptively adjust relevant weight W and threshold values B, feed back to Elman neural network;
Step 5 judges whether termination condition meets.If the number of iterations of population is greater than the maximum value or Elman of setting
When the value of fitness function fobj is less than 0.01, training terminates, and otherwise jumps to step 3 and continues to train optimization neural network.
Based on the above method, further, the cluster normalized is at the cluster normalization of Situation Awareness model
Reason.
This method predicts the time sequential value of network uninterrupted, affiliated longitude and latitude and flow direction in the short time, predicts
Journey is as indicated at 4.Fig. 4 is SCA-ELMAN prediction model schematic diagram.
Wherein, each attribute forecast value for obtaining flow passes through the cluster normalized of Situation Awareness model, network
Shown in the calculating such as formula (6) of flow cluster.
In formula (6), each composite attribute of network flow that prediction is obtained, according to ruleLongitude and latitude is in radius model
The network flow in r is enclosed according to the flow direction classification superposition of network flow, obtains the stream in a certain range in the network short time
Amount flows to trend, grasps the network flow situation of the whole network, Situation Awareness System is as shown in Figure 5.Fig. 5 is Situation Awareness System mould
Type schematic diagram.
The present invention optimizes Elman neural network using SCA swarm intelligence algorithm, using SCA-Elman model prediction network flow
Size, the time series of longitude and latitude and flow direction are measured, each attribute forecast value cluster normalized was obtained in the network short time
Flux and flow direction trend in a certain range grasps the network flow situation of the whole network.The present invention can provide whole network short
In the flowing situation of a certain range flow in period, i.e., flow will wherefrom flow where, allow currently used person can not
Internet resources deployment preparation related to defending against network attacks is carried out in deployment in advance.
Fig. 6 is a kind of structural representation of network situation awareness device based on just remaining double string algorithms provided by the present invention
Figure, the device include:
Module 101 is adjusted, for the threshold value and weight using just remaining double string algorithm adjustment Elman neural networks;
Optimization module 102 obtains time series, the net of network flow sample for optimizing to Elman neural network
The time series of time series and the network flow flow direction of network flow longitude and latitude;
Cluster module 103, time series and net for time series, network flow longitude and latitude to network flow sample
The time series of network flux and flow direction carries out cluster normalized, obtains the attack trend and situation flow direction of network flow.
As it can be seen that the device adaptively adjusts Elman neural network threshold value and weight by using just remaining double string algorithms, in turn
Optimization neural network model predicts the time sequential value of network flow sample size in the short time, affiliated longitude and latitude and flow direction,
Then cluster normalized that each attribute forecast value of acquisition is passed through to Situation Awareness model, obtained in the network short time
Attack tends to and situation flow direction, is conducive to the defence of Internet resources deployment management and network attack.
Based on above-mentioned apparatus, specifically, Elman neural network is the BP neural network with feedback.
Further, adjustment module includes:
Training unit is trained training sample set using Elman neural network for obtaining training sample set;
Acquiring unit, for during being trained to training sample set, by Elman neural network to positive Yu Shuan
String algorithm provides the mean square error MSE of training, the weight W and threshold value B of just remaining double string algorithm feedbacks is obtained, by weight W and threshold value B
It is input to Elman neural network.
Further, cluster normalized is the cluster normalized of Situation Awareness model.
Above to a kind of network situation awareness method and device progress based on just remaining double string algorithms provided by the present invention
It is discussed in detail.Used herein a specific example illustrates the principle and implementation of the invention, above embodiments
Explanation be merely used to help understand method and its core concept of the invention.It should be pointed out that for the common of the art
, without departing from the principle of the present invention, can be with several improvements and modifications are made to the present invention for technical staff, these
Improvement and modification are also fallen within the protection scope of the claims of the present invention.
Claims (8)
1. a kind of network situation awareness method based on just remaining double string algorithms characterized by comprising
Using the threshold value and weight of just remaining double string algorithm adjustment Elman neural networks;
Elman neural network is optimized, the time series and net of network flow sample size, network flow longitude and latitude are obtained
The time series of network flux and flow direction;
The time series of time series and network flow flow direction to network flow sample size, network flow longitude and latitude is gathered
Class normalized obtains the attack trend and situation flow direction of network flow.
2. the method as described in claim 1, which is characterized in that the Elman neural network is the BP neural network with feedback.
3. the method as described in claim 1, which is characterized in that described to adjust Elman neural network using just remaining double string algorithms
Threshold value and weight, comprising:
Training sample set is obtained, training sample set is trained using Elman neural network;
During being trained to training sample set, training is provided to just remaining double string algorithms by Elman neural network
Mean square error MSE obtains the weight W and threshold value B of just remaining double string algorithm feedbacks, weight W and threshold value B is input to Elman nerve
Network.
4. the method as described in any one of claims 1 to 3, which is characterized in that the cluster normalized is situation
The cluster normalized of sensor model.
5. a kind of network situation awareness device based on just remaining double string algorithms characterized by comprising
Module is adjusted, for the threshold value and weight using just remaining double string algorithm adjustment Elman neural networks;
Optimization module obtains network flow sample size, network flow longitude and latitude for optimizing to Elman neural network
Time series and network flow flow direction time series;
Cluster module, for network flow sample size, network flow longitude and latitude time series and network flow flow direction
Time series carries out cluster normalized, obtains the attack trend and situation flow direction of network flow.
6. device as claimed in claim 5, which is characterized in that the Elman neural network is the BP neural network with feedback.
7. device as claimed in claim 5, which is characterized in that the adjustment module includes:
Training unit is trained training sample set using Elman neural network for obtaining training sample set;
Acquiring unit, for being calculated by Elman neural network to just remaining double strings during being trained to training sample set
Method provides the mean square error MSE of training, obtains the weight W and threshold value B of just remaining double string algorithm feedbacks, weight W and threshold value B is inputted
To Elman neural network.
8. the device as described in any one of claim 5 to 7, which is characterized in that the cluster normalized is situation
The cluster normalized of sensor model.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611139491.3A CN106789214B (en) | 2016-12-12 | 2016-12-12 | A kind of network situation awareness method and device based on just remaining double string algorithms |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611139491.3A CN106789214B (en) | 2016-12-12 | 2016-12-12 | A kind of network situation awareness method and device based on just remaining double string algorithms |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106789214A CN106789214A (en) | 2017-05-31 |
CN106789214B true CN106789214B (en) | 2019-10-11 |
Family
ID=58875543
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201611139491.3A Expired - Fee Related CN106789214B (en) | 2016-12-12 | 2016-12-12 | A kind of network situation awareness method and device based on just remaining double string algorithms |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106789214B (en) |
Families Citing this family (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108205698A (en) * | 2017-06-15 | 2018-06-26 | 广东工业大学 | A kind of cloud resource load predicting method based on the double string whale optimization algorithms of just remaining chaos |
CN107729992B (en) * | 2017-10-27 | 2020-12-29 | 深圳市未来媒体技术研究院 | Deep learning method based on back propagation |
CN108830246B (en) * | 2018-06-25 | 2022-02-15 | 中南大学 | Multi-dimensional motion feature visual extraction method for pedestrians in traffic environment |
CN109040027B (en) * | 2018-07-12 | 2020-08-18 | 北京理工大学 | Active prediction method of network vulnerability node based on gray model |
CN109547431A (en) * | 2018-11-19 | 2019-03-29 | 国网河南省电力公司信息通信公司 | A kind of network security situation evaluating method based on CS and improved BP |
CN109507989A (en) * | 2018-12-24 | 2019-03-22 | 上海应用技术大学 | A kind of method of unmanned plane sensor fault diagnosis |
CN112085196A (en) * | 2020-09-10 | 2020-12-15 | 南京工业大学 | Ammonia injection adjusting method for SCR denitration system based on SCA algorithm optimization BP neural network |
CN112330637A (en) * | 2020-11-09 | 2021-02-05 | 哈尔滨理工大学 | Pulse coupling neural network skeletal muscle image processing method based on pixel values |
CN114978708A (en) * | 2022-05-25 | 2022-08-30 | 上海磐御网络科技有限公司 | Honeypot data-based graph neural network attack intention prediction method |
CN115695046B (en) * | 2022-12-28 | 2023-03-31 | 广东工业大学 | Network intrusion detection method based on reinforcement ensemble learning |
CN115981384B (en) * | 2023-02-07 | 2023-09-22 | 淮阴工学院 | Intelligent biomass ORC evaporation pressure control equipment |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103716180A (en) * | 2013-12-04 | 2014-04-09 | 国网上海市电力公司 | Network flow actual forecasting-based network abnormality pre-warning method |
CN103840988A (en) * | 2014-03-17 | 2014-06-04 | 湖州师范学院 | Network traffic measurement method based on RBF neural network |
CN105049279A (en) * | 2015-06-19 | 2015-11-11 | 国家电网公司 | Communication trend flexibility configuration method and system |
CN106130986A (en) * | 2016-06-30 | 2016-11-16 | 湘潭大学 | A kind of wind energy turbine set active safety defence method based on automated decision-making |
-
2016
- 2016-12-12 CN CN201611139491.3A patent/CN106789214B/en not_active Expired - Fee Related
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103716180A (en) * | 2013-12-04 | 2014-04-09 | 国网上海市电力公司 | Network flow actual forecasting-based network abnormality pre-warning method |
CN103840988A (en) * | 2014-03-17 | 2014-06-04 | 湖州师范学院 | Network traffic measurement method based on RBF neural network |
CN105049279A (en) * | 2015-06-19 | 2015-11-11 | 国家电网公司 | Communication trend flexibility configuration method and system |
CN106130986A (en) * | 2016-06-30 | 2016-11-16 | 湘潭大学 | A kind of wind energy turbine set active safety defence method based on automated decision-making |
Non-Patent Citations (2)
Title |
---|
PSO-BP网络在网络流量模拟在线预测中的应用研究;赵艳平,等;《兰州交通大学学报》;20121231;第31卷(第6期);第63~66页 * |
基于BP神经网络技术的网络流量预测模型;杜涛;《网络安全技术与应用》;20160731;第36卷(第7期);第55、57页 * |
Also Published As
Publication number | Publication date |
---|---|
CN106789214A (en) | 2017-05-31 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106789214B (en) | A kind of network situation awareness method and device based on just remaining double string algorithms | |
CN103164742B (en) | A kind of server performance Forecasting Methodology based on particle group optimizing neural network | |
CN106453293B (en) | A kind of network security situation prediction method based on improved BPNN | |
CN104899431B (en) | Based on ant colony and swarm of particles into LSSVM fluctuating wind speed Forecasting Methodologies | |
Asadnia et al. | Improved particle swarm optimization–based artificial neural network for rainfall-runoff modeling | |
Yang et al. | Pedestrian evacuation under guides in a multiple-exit room via the fuzzy logic method | |
CN106502799A (en) | A kind of host load prediction method based on long memory network in short-term | |
CN110008339A (en) | A kind of profound memory network model and its classification method for target emotional semantic classification | |
CN107045785B (en) | A method of the short-term traffic flow forecast based on grey ELM neural network | |
Xing et al. | Monthly mean streamflow prediction based on bat algorithm-support vector machine | |
Zhong et al. | Density-based evolutionary framework for crowd model calibration | |
CN106357456A (en) | Prediction method of network traffic and device thereof | |
CN106096115B (en) | A kind of crowd evacuation emulation method based on self-organizing sand ionization formula | |
Vieira et al. | Bio‐inspired algorithms for the optimization of offshore oil production systems | |
Tang et al. | Machine learning-based fast seismic risk assessment of building structures | |
Zhang et al. | Spark cloud-based parallel computing for traffic network flow predictive control using non-analytical predictive model | |
Jiang | BP neural network algorithm optimized by genetic algorithm and its simulation | |
Sun | Sports performance prediction based on chaos theory and machine learning | |
Duan et al. | Crowd evacuation under real data: a crowd congestion control method based on sensors and knowledge graph | |
Li et al. | Optimization of artificial CNN based on swarm intelligence algorithm | |
Lin et al. | PSO-BPNN-based prediction of network security situation | |
CN106339357A (en) | Optimizing parameterized network information security detection system and method | |
Cao et al. | Fuzzy Social Force Model for Pedestrian Evacuation under View‐Limited Condition | |
Balaji et al. | Deep Learning Based Energy Consumption Prediction on Internet of Things Environment. | |
Liao et al. | Simulation Research on Safe Flow Rate of Bidirectional Crowds Using Bayesian‐Nash Equilibrium |
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 | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20191011 Termination date: 20201212 |
|
CF01 | Termination of patent right due to non-payment of annual fee |