CN107241358A - A kind of smart home intrusion detection method based on deep learning - Google Patents
A kind of smart home intrusion detection method based on deep learning Download PDFInfo
- Publication number
- CN107241358A CN107241358A CN201710651758.5A CN201710651758A CN107241358A CN 107241358 A CN107241358 A CN 107241358A CN 201710651758 A CN201710651758 A CN 201710651758A CN 107241358 A CN107241358 A CN 107241358A
- Authority
- CN
- China
- Prior art keywords
- data
- neural network
- network
- deep learning
- smart home
- 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
Links
Classifications
-
- 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
- H04L63/14—Network architectures or network communication protocols for network security for detecting or protecting against malicious traffic
- H04L63/1408—Network architectures or network communication protocols for network security for detecting or protecting against malicious traffic by monitoring network traffic
- H04L63/1416—Event detection, e.g. attack signature detection
-
- 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
Abstract
The invention discloses a kind of smart home intrusion detection method based on deep learning, it is related to on-line system and is related to a kind of method that fuzzy neural network and deep learning be combined to judge that network whether there is intrusion behavior.This method organically combines deep learning and fuzzy neural network, and solution existing smart home Intrusion Detection Technique is difficult to handle the problem of a large amount of high dimensional datas, rate of false alarm are high, rate of failing to report is high, verification and measurement ratio is low.The present invention determines the operational factor of on-line system using off-line system, on-line system carries out real-time intrusion detection, compared with prior art, this is an actively monitoring model for being directed to intelligent home network attack, with higher verification and measurement ratio, relatively low rate of failing to report and rate of false alarm, and it is real-time the features such as.
Description
Technical field
The present invention relates to smart home security fields, more particularly to a kind of multilayer neural network invasion based on deep learning
Behavioral value method.
Background technology
With the fast development of technology of Internet of things, the Internet of product such as smart home is gradually gained popularity, however, intelligence at present
The security protection ability of energy equipment is universal more weak, and upgrade maintenance mechanism is unsound, smart machine security configuration is unreasonable etc.
Problem causes smart machine to there is more potential safety hazard.As country proposes and implements " internet+" action plan in recent years,
" plan of made in China 2025 ", smart city construction etc., substantial amounts of smart machine is continued to bring out, but corresponding safety precautions
Perfect not enough, smart home is moving towards increasing family, smart home system as a kind of emerging Internet of Things application
System includes a variety of smart machines such as camera, router, gateway, and these equipment have right and detour, refuse service, information and let out
The information security leaks such as dew, attacker relatively easily can be launched a offensive using these leaks to intelligent home network, so that
The privacy leakage of user, intelligent home network is caused normally to use, so that the problems such as other properties or personal safety.
Existing intelligent domestic system is perfect not enough in terms of security, and most of intelligent domestic system is using fire prevention
The technologies such as wall, certification or encryption improve its security, and these technologies belong to Passive Defence, for some specific attacks
Effect is preferable, it is impossible to is actively discovered attack and takes disposal or precautionary measures in time.
The content of the invention
In view of this, it is an object of the invention to provide a kind of rate of failing to report is low, verification and measurement ratio is high, detection rates are fast towards intelligence
The intrusion detection method of household.
The purpose of the present invention is achieved through the following technical solutions, a kind of smart home invasion based on deep learning
Detection method, specifically includes following steps：
S1 is initialized, and one content of generation is empty off-line system database, and the training that database includes tape label is tested
Data, data screening link parameter, three subdata bases of multitiered network parameter based on deep learning；
S2 is encoded the data on flows with label collected, normalized formation data to be tested, will be treated
In the training test data subdata base for detecting data Cun Chudao tape labels；
S3 is classified the data in the training test data subdata base of tape label according to the label of every data, shape
Into normal behaviour sample data set and intrusion behavior sample data set；Asked using K-means algorithms in two class sample data sets
Center value, two class sample datas of analysis concentrate the distance at each sample distance sample center and set a judgment threshold so that tool
There is the sample data set of certain category feature in such threshold range, center of a sample and threshold value are saved in data screening link ginseng
In number subdata base；Using in the training test data subdata base of tape label data training multilayer neural network weights and
Bias, the neural network parameter trained is saved in the multilayer neural network parameter subdata base based on deep learning,
Training link is completed, step S4 progress on-line systems is jumped to and monitors in real time；
S4, the data on flows without label collected encoded and normalized, form a testing data, meter
Testing data is calculated to the distance at two class sample data set centers in step S3, institute is right if the distance is less than the sample data set
The threshold value answered, then belong to the class behavior, otherwise jumps to step S5；
S5, by step S4 can not certain type of data to be tested be input in multilayer neural network and be identified, pass through
The output valve of multilayer neural network judges whether potential safety hazard, if there is potential safety hazard then drives smart home to alarm
Module is alarmed.
Further, in step s3, the training data of tape label in off-line system database is determined using K-means algorithms
The center of a sample of two class behaviors in storehouse, and calculate off-line system database midpoint to the Euclidean distance of center of a sample, to away from
From the distance threshold that data screening link is determined using Pauta criterion.
Further, it is described by step S4 can not certain type of data to be tested be input in multilayer neural network and known
Data Dimensionality Reduction Bao Kuo not be carried out using depth confidence network and fuzzy neural network is identified.
Further, described multilayer neural network includes depth confidence network and fuzzy neural network, depth confidence network
Output as fuzzy neural network input, wherein deep neural network by it is multiple limitation Boltzmann machines constitute.
Further, in step s 2, the data in the training test data subdata base using tape label train multilayer
In the weights and bias of neutral net, multilayer neural network is trained including the training to depth confidence network and obscured
The training of neutral net.
Further, unsupervised training from bottom to top is included to the training of depth confidence network and top-down has supervision
Small parameter perturbations；Training to fuzzy neural network uses gradient descent method.
Further, by the detection of reconstructed error, multilayer neural network that Boltzmann machine is limited in depth confidence network
Rate and detection time etc. build the depth that an assessment models determine multilayer neural network, that is, determine in depth confidence network
Limit the number of Boltzmann machine.
Further, when reconstructed error is more than 0.1, network depth increase by 1 is increasing by one in depth confidence network
Limit Boltzmann machine.
Further, if the reconstructed error of depth confidence network is less than 0.1, by the verification and measurement ratio for assessing multilayer neural network
And detection time, and the computing capability of combination intelligent domestic system server chooses the suitable number for limiting Boltzmann machine,
Determine the network depth of IDS Framework.
By adopting the above-described technical solution, the present invention has the advantage that：
The present invention is combined using on-line detecting system and off-line system, overcomes traditional detection method speed slowly and exists
The problem of larger hysteresis quality；Compared with traditional intrusion detection method, introduce deep learning and fuzzy neural network is combined
Multilayer neural network, is capable of detecting when some unknown intrusion behaviors, rejects some wrong reports brought due to human operational error
Behavior, i.e., have the advantage that rate of false alarm is low, verification and measurement ratio is high using such scheme.
Brief description of the drawings
In order that the object, technical solutions and advantages of the present invention are clearer, below in conjunction with accompanying drawing the present invention is made into
The detailed description of one step, wherein：
Fig. 1 is the flow chart of intrusion detection of the present invention；
Fig. 2 is intelligent domestic system structure chart of the present invention；
Fig. 3 is multilayer neural network training method schematic diagram of the present invention；
Fig. 4 is limitation Boltzmann machine structure chart of the present invention；
Fig. 5 is structure chart of the invention.
Embodiment
Below with reference to accompanying drawing, the preferred embodiments of the present invention are described in detail；It should be appreciated that preferred embodiment
Only for the explanation present invention, the protection domain being not intended to be limiting of the invention.
Overhaul flow chart shown in reference picture 1, smart home intrusion detection method, comprises the following steps：
101st, initialize, one content of generation is empty off-line system database, database include tape label training and
Test data, data screening link parameter, three subdata bases of multitiered network parameter based on deep learning；
102nd, intelligent domestic system is made up of sensor node, routing node, server, client etc., smart home
Composition is as shown in Fig. 2 using flow packet capturing software grabs smart home home server gateway with label data on flows, will adopt
The data collected are encoded, normalized formation data to be tested, are added to band mark in the off-line system database in 101
In the training test database of label, step 103 is jumped to.
103rd, the data with label are classified according to the label of every data in the off-line system database in 101,
Form normal behaviour sample data set and intrusion behavior sample data set.Two class sample data sets are asked using K-means algorithms
Central value and distance one judgment threshold of setting for analyzing each sample distance sample center so that the sample with certain category feature
Center of a sample and threshold value are saved in data screening link parameter subdata base by data set in such threshold range；Adopt
The weights and bias of multilayer neural network are trained with the data in the training test data subdata base of tape label, will be trained
Neural network parameter be saved in the multilayer neural network parameter subdata base based on deep learning, complete training link, jump
Step 104 progress on-line system is gone to monitor in real time.
104th, captured using flow packet capturing software grabs smart home server gateway without label data on flows, to grabbing
The data taken are encoded and normalized, form a testing data, are completed data screening link, that is, are calculated testing data
The distance at the two class sample data set centers in 103, is the class behavior if the threshold value that the distance is less than a certain class behavior,
Otherwise step 105 is jumped to.
105th, by step 104 can not certain type of data to be tested be input in multilayer neural network and be identified, its
Include carrying out Data Dimensionality Reduction using depth confidence network and fuzzy neural network is identified.If multilayer neural network output exists
[0,1.2], it is secure data to illustrate the data；If multitiered network be output as (1.2,2.5], illustrate the data exist safety it is hidden
Suffer from, alarm module can be driven to alarm, if output data then illustrates that network can not be known to data not in [0,2.5] is interval
Not, store data into off-line system, member to be managed checks and judges that the data whether there is potential safety hazard.
In step 103 in data screening link,
Parameters of the A needed for off-line system determines data screening link.In off-line system in the tranining database of tape label
The determination of the center of a sample of two class behaviors is to use K-means algorithms, and the algorithm is a kind of clustering algorithm, to there is the data of label
Cluster centre is asked for, and calculates offline rule base midpoint to the Euclidean distance of center of a sample
Its, wherein (X1..., Xk) it is center of a sample, (x1..., xk) it is testing data, adjusting the distance, it is accurate to be also referred to as 3 σ using Pauta criterion
Then it is determined the distance threshold of data screening link.
Data screening links of the B in on-line detecting system, is calculated in data to be tested and off-line system in two class samples
The distance of the heart, chooses less distance and enters with this apart from the classification where corresponding center of a sample in the threshold value that off-line system is set
Row compares, and then belongs to the category if less than the threshold value, the category is then not belonging to if greater than the threshold value, then carries out subsequently many
Layer neutral net detection.
Multilayer neural network algorithm in step 105, including：
A to being trained in off-line system to multilayer neural network parameters, its training method as shown in figure 3, its
Include the training to depth confidence network and the training of fuzzy neural network.The training of depth confidence network is included from lower
On unsupervised training and the top-down small parameter perturbations for having a supervision, the training to fuzzy neural network declined using gradient
Method.
B wherein depth confidence networks are made up of multiple limitation Boltzmann machines, and its model is an energy based on probability
Model, its principle such as Fig. 4 approaches the probability for obtaining hidden layer V by Gibbs sampling algorithms, and likelihood function logarithm is to parameter derivation
Method shows layer bias a and b to ask for the weights W and hidden layer of depth confidence network.It regard the output of depth confidence network as mould
The input of paste neutral net is classified to behavior, the numeral between output one [0,2.5].Sentenced according to the numeral of output
The classification broken belonging to it.
Reconstructed error, the verification and measurement ratio of multilayer neural network and inspection that C passes through limitation Boltzmann machine in depth confidence network
Survey time etc. builds the depth that an assessment models determine multilayer neural network, that is, determines to limit glass in depth confidence network
The number of the graceful machine of Wurz.When reconstructed error is more than 0.1, network depth increase by 1 is increasing by one in depth confidence network
Limit Boltzmann machine, if depth confidence network reconfiguration error be less than 0.1, by assess multilayer neural network verification and measurement ratio and
Detection time, and combine that the factors such as the computing capability of intelligent domestic system server choose suitable limitation Boltzmann machine
Number, determines the network depth of IDS Framework.
The present invention is applied to the intrusion detection for intelligent home network, uses intrusion detection side disclosed in this invention
Method, due to deep learning and fuzzy neural network being organically combined, can reach that rate of false alarm is low, false drop rate is low, verification and measurement ratio
High effect, also has preferable detectability for unknown intrusion behavior, and with preferable adaptive ability.
In traditional method, rate of false alarm is generally higher, and method of the invention can make rate of false alarm be reduced to less than 5%,
Simultaneously in terms of verification and measurement ratio, more than 95% can be reached.At the same time, for unknown new intrusion behavior, its verification and measurement ratio exists
More than 60%.
The preferred embodiments of the present invention are the foregoing is only, are not intended to limit the invention, it is clear that those skilled in the art
Member can carry out various changes and modification to the present invention without departing from the spirit and scope of the present invention.So, if the present invention
These modifications and variations belong within the scope of the claims in the present invention and its equivalent technologies, then the present invention is also intended to include these
Including change and modification.
Claims (9)
1. a kind of smart home intrusion detection method based on deep learning, it is characterised in that：Specifically include following steps：
S1 is initialized, and one content of generation is empty off-line system database, database include tape label training test data,
Data screening link parameter, three subdata bases of multitiered network parameter based on deep learning；
S2 is encoded the data on flows with label collected, normalized formation data to be tested, will be to be detected
In the training test data subdata base of data Cun Chudao tape labels；
S3 is classified the data in the training test data subdata base of tape label according to the label of every data, is formed just
Normal behavior sample data set and intrusion behavior sample data set；The central value of two class sample data sets is sought using K-means algorithms,
Analyze two class sample datas to concentrate the distance at each sample distance sample center and set a judgment threshold so that with certain class
Center of a sample and threshold value are saved in data screening link parameter subnumber by the sample data set of feature in such threshold range
According in storehouse；Weights and biasing using the data training multilayer neural network in the training test data subdata base of tape label
Value, the neural network parameter trained is saved in the multilayer neural network parameter subdata base based on deep learning, is completed
Link is trained, step S4 progress on-line systems is jumped to and monitors in real time；
S4, the data on flows without label collected encoded and normalized, form a testing data, calculating is treated
Data are surveyed to the distance at two class sample data set centers in step S3, if the distance is less than corresponding to the sample data set
Threshold value, then belong to the class behavior, otherwise jumps to step S5；
S5, by step S4 can not certain type of data to be tested be input in multilayer neural network and be identified, pass through multilayer
The output valve of neutral net judges whether potential safety hazard, then drives smart home alarm module if there is potential safety hazard
Alarm.
2. a kind of smart home intrusion detection method based on deep learning according to claim 1, it is characterised in that：
In step S3, the sample of two class behaviors in the tranining database of tape label in off-line system database is determined using K-means algorithms
This center, and off-line system database midpoint is calculated to the Euclidean distance of center of a sample, adjust the distance and use Pauta criterion
It is determined the distance threshold of data screening link.
3. a kind of smart home intrusion detection method based on deep learning according to claim 1, it is characterised in that：Institute
State by step S4 can not certain type of data to be tested be input in multilayer neural network and be identified including being put using depth
Communication network carries out Data Dimensionality Reduction and fuzzy neural network is identified.
4. a kind of smart home intrusion detection method based on deep learning according to claim 1, it is characterised in that：Institute
The multilayer neural network stated includes depth confidence network and fuzzy neural network, and the output of depth confidence network is used as fuzzy neural
The input of network, wherein deep neural network are made up of multiple limitation Boltzmann machines.
5. a kind of smart home intrusion detection method based on deep learning according to claim 1, it is characterised in that：
In step S2, weights of the data training multilayer neural network in the training test data subdata base using tape label and inclined
Put in value, multilayer neural network is trained including the training and the training of fuzzy neural network to depth confidence network.
6. a kind of smart home intrusion detection method based on deep learning according to claim 5, it is characterised in that：It is right
The training of depth confidence network includes unsupervised training from bottom to top and the top-down small parameter perturbations for having a supervision；To fuzzy
The training of neutral net uses gradient descent method.
7. a kind of smart home intrusion detection method based on deep learning according to claim 3, it is characterised in that：It is logical
Cross in depth confidence network and limit the structure such as reconstructed error, the verification and measurement ratio of multilayer neural network and detection time of Boltzmann machine
One assessment models determines the depth of multilayer neural network, that is, determines of limitation Boltzmann machine in depth confidence network
Number.
8. a kind of smart home intrusion detection method based on deep learning according to claim 7, it is characterised in that：
When reconstructed error is more than 0.1, network depth increase by 1 is increasing a limitation Boltzmann machine in depth confidence network.
9. a kind of smart home intrusion detection method based on deep learning according to claim 7, it is characterised in that：If
When the reconstructed error of depth confidence network is less than 0.1, by assessing the verification and measurement ratio and detection time of multilayer neural network, and combine
The computing capability of intelligent domestic system server chooses the number of suitable limitation Boltzmann machine, determines IDS Framework
Network depth.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710651758.5A CN107241358B (en) | 2017-08-02 | 2017-08-02 | Smart home intrusion detection method based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710651758.5A CN107241358B (en) | 2017-08-02 | 2017-08-02 | Smart home intrusion detection method based on deep learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107241358A true CN107241358A (en) | 2017-10-10 |
| CN107241358B CN107241358B (en) | 2020-04-07 |
Family
ID=59989480
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710651758.5A Active CN107241358B (en) | 2017-08-02 | 2017-08-02 | Smart home intrusion detection method based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107241358B (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107895171A (en) * | 2017-10-31 | 2018-04-10 | 天津大学 | A kind of intrusion detection method based on K averages Yu depth confidence network |
| CN108234500A (en) * | 2018-01-08 | 2018-06-29 | 重庆邮电大学 | A kind of wireless sense network intrusion detection method based on deep learning |
| CN108683658A (en) * | 2018-05-11 | 2018-10-19 | 上海交通大学 | Industry control network Traffic Anomaly recognition methods based on more RBM network structions benchmark models |
| CN108984904A (en) * | 2018-07-17 | 2018-12-11 | 北京理工大学 | A kind of Home Fashion & Design Shanghai method based on deep neural network |
| CN110581802A (en) * | 2019-08-27 | 2019-12-17 | 北京邮电大学 | fully-autonomous intelligent routing method and device based on deep belief network |
| CN110661781A (en) * | 2019-08-22 | 2020-01-07 | 中科创达软件股份有限公司 | DDoS attack detection method, device, electronic equipment and storage medium |
| CN111131069A (en) * | 2019-11-25 | 2020-05-08 | 北京理工大学 | Abnormal encryption flow detection and classification method based on deep learning strategy |
| CN112689281A (en) * | 2020-12-21 | 2021-04-20 | 重庆邮电大学 | Sensor network malicious node judgment method based on two-type fuzzy system |
| CN112769750A (en) * | 2020-12-11 | 2021-05-07 | 广东电力通信科技有限公司 | Protocol stack sending method suitable for intelligent gateway data management |
| CN113392403A (en) * | 2021-06-11 | 2021-09-14 | 连云港微部落网络技术有限公司 | Website security defense system and method with active defense function |
| CN113645231A (en) * | 2021-08-10 | 2021-11-12 | 北京易通信联科技有限公司 | Intrusion detection method, memory and processor of industrial control system |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6114976A (en) * | 1999-02-05 | 2000-09-05 | The Boeing Company | Vehicle emergency warning and control system |
| US20130116920A1 (en) * | 2011-11-07 | 2013-05-09 | International Business Machines Corporation | System, method and program product for flood aware travel routing |
| CN103336992A (en) * | 2013-06-27 | 2013-10-02 | 电子科技大学 | FNN learning algorithm |
| CN104777418A (en) * | 2015-05-11 | 2015-07-15 | 重庆大学 | Analog circuit fault diagnosis method based on depth Boltzman machine |
| CN104935600A (en) * | 2015-06-19 | 2015-09-23 | 中国电子科技集团公司第五十四研究所 | Mobile ad hoc network intrusion detection method and device based on deep learning |
| CN106911669A (en) * | 2017-01-10 | 2017-06-30 | 浙江工商大学 | A kind of DDOS detection methods based on deep learning |
-
2017
- 2017-08-02 CN CN201710651758.5A patent/CN107241358B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6114976A (en) * | 1999-02-05 | 2000-09-05 | The Boeing Company | Vehicle emergency warning and control system |
| US20130116920A1 (en) * | 2011-11-07 | 2013-05-09 | International Business Machines Corporation | System, method and program product for flood aware travel routing |
| CN103336992A (en) * | 2013-06-27 | 2013-10-02 | 电子科技大学 | FNN learning algorithm |
| CN104777418A (en) * | 2015-05-11 | 2015-07-15 | 重庆大学 | Analog circuit fault diagnosis method based on depth Boltzman machine |
| CN104935600A (en) * | 2015-06-19 | 2015-09-23 | 中国电子科技集团公司第五十四研究所 | Mobile ad hoc network intrusion detection method and device based on deep learning |
| CN106911669A (en) * | 2017-01-10 | 2017-06-30 | 浙江工商大学 | A kind of DDOS detection methods based on deep learning |
Non-Patent Citations (3)
| Title |
|---|
| GUANGZHEN ZHAO: "《Intrusion Detection using Deep Belief Network and Probabilistic Neural Network》", 《2017 IEEE INTERNATIONAL CONFERENCE ON COMPUTATIONAL SCIENCE AND ENGINEERING (CSE) AND IEEE INTERNATIONAL CONFERENCE ON EMBEDDED AND UBIQUITOUS COMPUTING (EUC)》 * |
| MD. ZAHANGIR ALOM: "Intrusion Detection using Deep Belief Networks", 《2015 NATIONAL AEROSPACE AND ELECTRONICS CONFERENCE》 * |
| 安琪: "基于深度置信网络的入侵检测研究", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107895171A (en) * | 2017-10-31 | 2018-04-10 | 天津大学 | A kind of intrusion detection method based on K averages Yu depth confidence network |
| CN108234500A (en) * | 2018-01-08 | 2018-06-29 | 重庆邮电大学 | A kind of wireless sense network intrusion detection method based on deep learning |
| CN108683658A (en) * | 2018-05-11 | 2018-10-19 | 上海交通大学 | Industry control network Traffic Anomaly recognition methods based on more RBM network structions benchmark models |
| CN108683658B (en) * | 2018-05-11 | 2020-11-03 | 上海交通大学 | Industrial control network flow abnormity identification method based on multi-RBM network construction reference model |
| CN108984904A (en) * | 2018-07-17 | 2018-12-11 | 北京理工大学 | A kind of Home Fashion & Design Shanghai method based on deep neural network |
| CN108984904B (en) * | 2018-07-17 | 2022-09-20 | 北京理工大学 | Home design method based on deep neural network |
| CN110661781B (en) * | 2019-08-22 | 2022-05-17 | 中科创达软件股份有限公司 | DDoS attack detection method, device, electronic equipment and storage medium |
| CN110661781A (en) * | 2019-08-22 | 2020-01-07 | 中科创达软件股份有限公司 | DDoS attack detection method, device, electronic equipment and storage medium |
| CN110581802A (en) * | 2019-08-27 | 2019-12-17 | 北京邮电大学 | fully-autonomous intelligent routing method and device based on deep belief network |
| CN111131069A (en) * | 2019-11-25 | 2020-05-08 | 北京理工大学 | Abnormal encryption flow detection and classification method based on deep learning strategy |
| CN111131069B (en) * | 2019-11-25 | 2021-06-08 | 北京理工大学 | Abnormal encryption flow detection and classification method based on deep learning strategy |
| CN112769750A (en) * | 2020-12-11 | 2021-05-07 | 广东电力通信科技有限公司 | Protocol stack sending method suitable for intelligent gateway data management |
| CN112689281A (en) * | 2020-12-21 | 2021-04-20 | 重庆邮电大学 | Sensor network malicious node judgment method based on two-type fuzzy system |
| CN112689281B (en) * | 2020-12-21 | 2022-08-05 | 重庆邮电大学 | Sensor network malicious node judgment method based on two-type fuzzy system |
| CN113392403A (en) * | 2021-06-11 | 2021-09-14 | 连云港微部落网络技术有限公司 | Website security defense system and method with active defense function |
| CN113645231A (en) * | 2021-08-10 | 2021-11-12 | 北京易通信联科技有限公司 | Intrusion detection method, memory and processor of industrial control system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107241358B (en) | 2020-04-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107241358A (en) | A kind of smart home intrusion detection method based on deep learning | |
| CN103581186B (en) | A kind of network security situational awareness method and system | |
| CN104486141B (en) | A kind of network security situation prediction method that wrong report is adaptive | |
| CN108566364B (en) | Intrusion detection method based on neural network | |
| CN104935600A (en) | Mobile ad hoc network intrusion detection method and device based on deep learning | |
| Aziz et al. | Genetic algorithm with different feature selection techniques for anomaly detectors generation | |
| CN107241226A (en) | Fuzz testing method based on industry control proprietary protocol | |
| CN109818798A (en) | A kind of wireless sensor network intruding detection system and method merging KPCA and ELM | |
| TW200849917A (en) | Detecting method of network invasion | |
| CN106789904A (en) | Internet of things intrusion detection method and device | |
| CN108322445A (en) | A kind of network inbreak detection method based on transfer learning and integrated study | |
| CN108632269A (en) | Detecting method of distributed denial of service attacking based on C4.5 decision Tree algorithms | |
| CN112565270A (en) | HTTP session abnormity detection method and detection system | |
| CN110445766A (en) | Ddos attack method for situation assessment and device | |
| CN110493142A (en) | Mobile applications Activity recognition method based on spectral clustering and random forests algorithm | |
| Chen et al. | An effective metaheuristic algorithm for intrusion detection system | |
| CN102195975A (en) | Intelligent NIPS (Network Intrusion Prevention System) framework for quantifying neural network based on mobile agent (MA) and learning vector | |
| CN112685272B (en) | Interpretable user behavior abnormity detection method | |
| Tran | Network anomaly detection | |
| Petrovic et al. | Labelling clusters in an intrusion detection system using a combination of clustering evaluation techniques | |
| Devaraju et al. | Performance comparison of intrusion detection system using various techniques–A review | |
| Salazar et al. | Monitoring approaches for security and safety analysis: application to a load position system | |
| Bahrepour et al. | Fire data analysis and feature reduction using computational intelligence methods | |
| Palomo et al. | An intrusion detection system based on hierarchical self-organization | |
| Manandhar | A practical approach to anomaly-based intrusion detection system by outlier mining in network traffic |
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 |