CN108769993A - Based on the communication network abnormal user detection method for generating confrontation network - Google Patents
Based on the communication network abnormal user detection method for generating confrontation network Download PDFInfo
- Publication number
- CN108769993A CN108769993A CN201810458873.5A CN201810458873A CN108769993A CN 108769993 A CN108769993 A CN 108769993A CN 201810458873 A CN201810458873 A CN 201810458873A CN 108769993 A CN108769993 A CN 108769993A
- Authority
- CN
- China
- Prior art keywords
- abnormal user
- generator
- data
- user data
- discriminator
- 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.)
- Pending
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W12/00—Security arrangements; Authentication; Protecting privacy or anonymity
- H04W12/12—Detection or prevention of fraud
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Security & Cryptography (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of based on the communication network abnormal user detection method for generating confrontation network, carries out regularization to the data of abnormal user first, obtains the dimension data consistent with magnitude；It is trained to generating confrontation network, i.e., generator and discriminator is trained, realizes the over-sampling to abnormal user；The data that generator is generated form training dataset with normal users, are classified to training dataset using the full Connection Neural Network of depth, judge user type；The present invention fights the mutual game formula training method in network between neural network by generation, and realization approaches abnormal user data distribution, realizes abnormal user detection, solves the technical issues of training data concentrates ratio unbalanced influence disaggregated model training effect.
Description
Technical field
The invention belongs to communication network abnormal user detection fields, and in particular to a kind of based on the communication for generating confrontation network
Network Abnormal user's detection method.
Background technology
Due to the opening of wireless channel, with the development of wireless communication technique, present in safety problem increasingly
It is more.If lacking effective countermeasure, it would be possible to cause immeasurable loss to cordless communication network and validated user.
Non-orthogonal multiple access (NOMA) technology has become one of the key technology of next generation mobile communication system (5G).With NOMA
Development, the safety problem in NOMA also begins to attract attention and study.Power domain NOMA is the channel status letter based on user
Cease (Channel State Information, CSI) and export corresponding power allocation scheme, on same frequency spectrum to user into
Row overlapped information transmits.Similar to there are frequency spectrum perception data to forge (SSDF) attack in cognitive radio frequency spectrum collaborative sensing,
Illegally occupy frequency spectrum resource.In NOMA, there is also malicious users to the CSI of base station feedback mistake, makes base station misjudgment, from
And gain larger distribution power by cheating to reach corresponding purpose, such as：Illegal act operation is carried out under high quality communication state.This
Kind fraud attack wastes the energy resource of communication significantly, and affects the communication quality of other users.Due to traditional solution
Scheme based on certain hypothesis, mathematical model is simple, the information dimension that uses is few, can only be individually to different types of attack mould
The defects of type is handled, this patent propose it is based on deep learning, with extensive and high latitude data-handling capacity
Solution.
In recent years, it is attacked for the fraud in cordless communication network, domestic and foreign scholars have carried out extensive research, and propose
Various solutions.But traditional solution mathematical model is simple, and data dimension is low, and data calculation amount is small, cannot be efficient
The different challenge models of joint confrontation malicious user, greatly reduce the safety of cordless communication network.Currently, having big data
Processing capacity, efficient feature extractability, the depth learning technology of module fusion faculty are expected to solve the defect of conventional method.
It is most important for the performance for promoting network security protection system to study abnormal user challenge model.In practical communication
In system, abnormal user only accounts for few part of total number of users, the ratio extreme of abnormal user and normal users under normal conditions
It is unbalanced, it is difficult to meet requirement of the deep learning model to training dataset, for using accuracy rate and recall ratio as performance indicator
Disaggregated model training in, affect the training effect of disaggregated model.
The apllied abnormal user challenge model structure that confrontation network is generated based on depth of the present invention, examines abnormal user
Survey problem is attributed to the pattern recognition classifier problem based on deep learning, and network is fought to a small amount of abnormal user number using generating
According to over-sampling is carried out, the equalization of data set is realized.
Invention content
It is an object of the invention to realize the efficient joint of the different challenge models of confrontation malicious user, propose that one kind is based on
The communication network abnormal user detection method of confrontation network is generated, realizes abnormal user detection, training data is solved and concentrates ratio
The technical issues of unbalanced influence disaggregated model training effect.
Requirement in view of deep learning model to data volume, the present invention use the thought of minority class sample over-sampling, research
Based on the abnormal user detection method for generating confrontation network, generates a large amount of abnormal user using this method and attacks data so that
The data volume of abnormal user and normal users is reached an agreement, and finally utilizes the full Connection Neural Network of depth to abnormal user and normal
User classifies.
It detects and applies specific to abnormal user, network (Generative Adversarial are fought using generating
Networks, GAN) distribution of abnormal user sample data is approached, main thought is to utilize the mutual of deep neural network
Game generates abnormal user data.
The present invention adopts the following technical scheme that, a kind of based on the communication network abnormal user detection side for generating confrontation network
Method is as follows：
1) regularization is carried out to the data of abnormal user, obtains the dimension data consistent with magnitude；
2) it is trained to generating confrontation network, i.e., to generator (Generator, G) and discriminator
(Discriminator, D) is trained, and realizes the over-sampling to abnormal user；
3) data for generating generator form training dataset with normal users, utilize the full Connection Neural Network pair of depth
Training dataset is classified, and judges user type.
Preferably, it is specially to be returned using minimax to different magnitude of data to carry out regularization to the data of abnormal user
One change mode carries out regularization, uses one-hot coding methods to be labeled with implementation rule label data.
Preferably, step 2) the specific steps are：
21) the full Connection Neural Network of two multilayers is used to constitute generator and discriminator；
22) it is trained for generator, generator simulates the distribution of true abnormal user data, generates the abnormal use of simulation
User data；Training goal is so that discriminator cannot be distinguished from out the simulation abnormal user number of true abnormal user data and generation
According to, and then reach the target of simulation abnormal user attack data；
23) the simulation abnormal user data that true abnormal user data and generator generate are inputted into discriminator respectively, it is right
Discriminator is trained；Trained purpose is that discriminator is enable to distinguish true abnormal user data and generator generation
Simulation abnormal user data；
24) training to generating confrontation network is reached to generator and discriminator progress alternating iteration.
Preferably, generator is G (z), and wherein z is a random noise, and generator G converts random noise z, mould
Intend the distribution of true abnormal user data, generates simulation abnormal user data.
Preferably, discriminator is D (x), is inputted as true abnormal user data XaThe simulation generated with generator is used extremely
User data Xg, a real number of discriminator D output 0-1 ranges, the simulation abnormal user data sample generated for judging generator G
This XgWith true abnormal user data sample XaProbability, Pr and Pg respectively represent true abnormal user data distribution and simulate different
Common user data distribution, the object function of discriminator are：
Wherein,Indicate that true abnormal user data are judged as the expectation of true abnormal user data,
Indicate that the simulation abnormal user data that generator generates are judged as simulating the expectation of abnormal user data.
Preferably, it is specially maximum-minimize object function to carry out alternating iteration to generator and discriminator, respectively to life
The G and discriminator D that grows up to be a useful person interacts iteration, optimization discriminator D when fixing generator G；Optimize generator G when fixed discriminator D,
Until process restrains；
The target of generator G is to make discriminator D that authentic specimen cannot be distinguished and generate sample, and global optimization function is as follows：
Wherein,Indicate that true abnormal user data are judged as the expectation of true abnormal user data,
Indicate that the simulation abnormal user data that generator generates are judged as simulating the expectation of abnormal user data；D (x) indicates to differentiate
Device.
The obtained deep neural network of training can export correspondence for the random noise that arbitrarily inputs, the deep neural network
Data, for simulating abnormal user data, which is known as generator.It is generated in large quantities by generator G
The over-sampling to abnormal user may be implemented in analogue data so that the balanced target that training data reaches.
The reached advantageous effect of invention：The present invention is a kind of based on the communication network abnormal user inspection for generating confrontation network
Survey method realizes abnormal user detection, solves training data and concentrates the unbalanced technology for influencing disaggregated model training effect of ratio
Problem；The distribution that training data can be approached well realizes the equal of lack of balance data set by the over-sampling to a few sample
Weighing apparatusization.The main innovation point of the present invention is to propose solves asking for abnormal user data balancing based on generation confrontation network
Topic, compared to conventional method, can preferably solve the problems, such as the lack of balance of the training dataset in disaggregated model training.
Description of the drawings
Fig. 1 is based on the abnormal user detection method flow chart for generating confrontation network；
Fig. 2 is the present invention based on the abnormal user challenge model schematic diagram for generating confrontation network.
Specific implementation mode
Below according to attached drawing and technical scheme of the present invention is further elaborated in conjunction with the embodiments.
The present invention adopts the following technical scheme that, a kind of based on the communication network abnormal user detection side for generating confrontation network
Method, Fig. 1 are the abnormal user detection method flow charts based on generation confrontation network, and GAN1, GAN2 and GAN3 in figure make a living
At confrontation network, it is as follows：
1) regularization is carried out to the data of abnormal user, obtains the dimension data consistent with magnitude；
2) it is trained to generating confrontation network, i.e., to generator (Generator, G) and discriminator
(Discriminator, D) is trained, and realizes the over-sampling to abnormal user；
3) data for generating generator form training dataset with normal users, utilize the full Connection Neural Network pair of depth
Training dataset is classified, and judges user type.
Preferably, it is specially to be returned using minimax to different magnitude of data to carry out regularization to the data of abnormal user
One change mode carries out regularization, uses one-hot coding methods to be labeled with implementation rule label data.
Preferably, step 2) the specific steps are：
21) the full Connection Neural Network of two multilayers is used to constitute generator and discriminator；
22) it is trained for generator, generator simulates the distribution of true abnormal user data, generates the abnormal use of simulation
User data；Training goal is so that discriminator cannot be distinguished from out the simulation abnormal user number of true abnormal user data and generation
According to, and then reach the target of simulation abnormal user attack data；
23) the simulation abnormal user data that true abnormal user data and generator generate are inputted into discriminator respectively, it is right
Discriminator is trained；Trained purpose is that discriminator is enable to distinguish true abnormal user data and generator generation
Simulation abnormal user data；
24) training to generating confrontation network is reached to generator and discriminator progress alternating iteration.
Preferably, generator is G (z), and wherein z is a random noise, and generator G converts random noise z, mould
Intend the distribution of true abnormal user data, generates simulation abnormal user data.
Preferably, discriminator is D (x), is inputted as true abnormal user data XaThe simulation generated with generator is used extremely
User data Xg, a real number of discriminator D output 0-1 ranges, the simulation abnormal user data sample generated for judging generator G
This XgWith true abnormal user data sample XaProbability, Pr and Pg respectively represent true abnormal user data distribution and simulate different
Common user data distribution, the object function of discriminator are：
Wherein,Indicate that true abnormal user data are judged as the expectation of true abnormal user data,
Indicate that the simulation abnormal user data that generator generates are judged as simulating the expectation of abnormal user data.
Preferably, it is specially maximum-minimize object function to carry out alternating iteration to generator and discriminator, respectively to life
The G and discriminator D that grows up to be a useful person interacts iteration, optimization discriminator D when fixing generator G；Optimize generator G when fixed discriminator D,
Until process restrains；
The target of generator G is to make discriminator D that authentic specimen cannot be distinguished and generate sample, and global optimization function is as follows：
Wherein,Indicate that true abnormal user data are judged as the expectation of true abnormal user data,
Indicate that the simulation abnormal user data that generator generates are judged as simulating the expectation of abnormal user data；D (x) indicates to differentiate
Device.
Claims (6)
1. based on the communication network abnormal user detection method for generating confrontation network, which is characterized in that include the following steps：
1) regularization is carried out to the data of abnormal user, obtains the dimension data consistent with magnitude；
2) it is trained, i.e., generator and discriminator is trained to generating confrontation network, the mistake of abnormal user is adopted in realization
Sample；
3) data for generating generator form training dataset with normal users, using the full Connection Neural Network of depth to training
Data set is classified, and judges user type.
2. according to claim 1 based on the communication network abnormal user detection method for generating confrontation network, feature exists
In it is specially to be normalized using minimax to different magnitude of data to carry out regularization to the data of abnormal user in step 1)
Mode carries out regularization, uses one-hot coding methods to be labeled with implementation rule label data.
3. according to claim 1 based on the communication network abnormal user detection method for generating confrontation network, feature exists
In, step 2) the specific steps are：
21) the full Connection Neural Network of two multilayers is used to constitute generator and discriminator；
22) it is trained for generator, generator simulates the distribution of true abnormal user data, generates simulation abnormal user number
According to；
23) the simulation abnormal user data that true abnormal user data and generator generate are inputted into discriminator respectively, to differentiating
Device is trained；
24) training to generating confrontation network is reached to generator and discriminator progress alternating iteration.
4. according to claim 3 based on the communication network abnormal user detection method for generating confrontation network, feature exists
In, in the step 22), generator is G (z), and wherein z is a random noise, and generator G converts random noise z,
The distribution of true abnormal user data is simulated, simulation abnormal user data are generated.
5. according to claim 3 based on the communication network abnormal user detection method for generating confrontation network, feature exists
In in the step 23), discriminator is D (x), is inputted as true abnormal user data XaThe simulation generated with generator is abnormal
User data Xg, a real number of discriminator D output 0-1 ranges, the simulation abnormal user data generated for judging generator G
Sample XgWith true abnormal user data sample XaProbability, Pr and Pg respectively represent true abnormal user data distribution and simulation
The object function of abnormal user data distribution, discriminator is：
Wherein,Indicate that true abnormal user data are judged as the expectation of true abnormal user data,It indicates
The simulation abnormal user data that generator generates are judged as the expectation of simulation abnormal user data.
6. according to claim 3 based on the communication network abnormal user detection method for generating confrontation network, feature exists
In, method that alternating iteration uses is carried out for maximum-minimize object function to generator and discriminator in step 24), it is right respectively
Generator G and discriminator D interacts iteration, optimization discriminator D when fixing generator G；Optimize generator when fixed discriminator D
G, until process restrains；
The target of generator G is to make discriminator D that authentic specimen cannot be distinguished and generate sample, and global optimization function is as follows：
Wherein,Indicate that true abnormal user data are judged as the expectation of true abnormal user data,It indicates
The simulation abnormal user data that generator generates are judged as the expectation of simulation abnormal user data；D (x) indicates discriminator.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810458873.5A CN108769993A (en) | 2018-05-15 | 2018-05-15 | Based on the communication network abnormal user detection method for generating confrontation network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810458873.5A CN108769993A (en) | 2018-05-15 | 2018-05-15 | Based on the communication network abnormal user detection method for generating confrontation network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108769993A true CN108769993A (en) | 2018-11-06 |
Family
ID=64006764
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810458873.5A Pending CN108769993A (en) | 2018-05-15 | 2018-05-15 | Based on the communication network abnormal user detection method for generating confrontation network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108769993A (en) |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109492193A (en) * | 2018-12-28 | 2019-03-19 | 同济大学 | Abnormal network data based on depth machine learning model generate and prediction technique |
| CN109584221A (en) * | 2018-11-16 | 2019-04-05 | 聚时科技(上海)有限公司 | A kind of abnormal image detection method generating confrontation network based on supervised |
| CN109685200A (en) * | 2018-11-19 | 2019-04-26 | 华东师范大学 | Industrial protocol construction method and building system are calculated based on the mist for generating confrontation network |
| CN109978550A (en) * | 2019-03-12 | 2019-07-05 | 同济大学 | A kind of credible electronic transaction clearance mechanism based on generation confrontation network |
| CN110009171A (en) * | 2018-11-27 | 2019-07-12 | 阿里巴巴集团控股有限公司 | Customer behavior modeling method, apparatus, equipment and computer readable storage medium |
| CN110505241A (en) * | 2019-09-17 | 2019-11-26 | 武汉思普崚技术有限公司 | A kind of network attack face detection method and system |
| CN110855654A (en) * | 2019-11-06 | 2020-02-28 | 中国移动通信集团广东有限公司 | Vulnerability risk quantitative management method and system based on flow mutual access relation |
| CN111090685A (en) * | 2019-12-19 | 2020-05-01 | 第四范式（北京）技术有限公司 | Method and device for detecting data abnormal characteristics |
| CN111507934A (en) * | 2019-01-30 | 2020-08-07 | 富士通株式会社 | Training apparatus, training method, and computer-readable recording medium |
| CN112036955A (en) * | 2020-09-07 | 2020-12-04 | 贝壳技术有限公司 | User identification method and device, computer readable storage medium and electronic equipment |
| CN112446002A (en) * | 2020-11-13 | 2021-03-05 | 天津大学 | Abnormity detection method for time sequence KPI data |
| CN112738092A (en) * | 2020-12-29 | 2021-04-30 | 北京天融信网络安全技术有限公司 | Log data enhancement method, classification detection method and system |
| CN112818407A (en) * | 2021-04-16 | 2021-05-18 | 中国工程物理研究院计算机应用研究所 | Video privacy protection method based on generation countermeasure network |
| CN113033656A (en) * | 2021-03-24 | 2021-06-25 | 厦门航空有限公司 | Interactive hole exploration data expansion method based on generation countermeasure network |
| WO2021130392A1 (en) | 2019-12-26 | 2021-07-01 | Telefónica, S.A. | Computer-implemented method for accelerating convergence in the training of generative adversarial networks (gan) to generate synthetic network traffic, and computer programs of same |
| CN113067653A (en) * | 2021-03-17 | 2021-07-02 | 北京邮电大学 | Spectrum sensing method and device, electronic equipment and medium |
| CN113378718A (en) * | 2021-06-10 | 2021-09-10 | 中国石油大学(华东) | Action identification method based on generation of countermeasure network in WiFi environment |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107368752A (en) * | 2017-07-25 | 2017-11-21 | 北京工商大学 | A kind of depth difference method for secret protection based on production confrontation network |
| CN107563355A (en) * | 2017-09-28 | 2018-01-09 | 哈尔滨工程大学 | Hyperspectral abnormity detection method based on generation confrontation network |
| CN108009628A (en) * | 2017-10-30 | 2018-05-08 | 杭州电子科技大学 | A kind of method for detecting abnormality based on generation confrontation network |
-
2018
- 2018-05-15 CN CN201810458873.5A patent/CN108769993A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107368752A (en) * | 2017-07-25 | 2017-11-21 | 北京工商大学 | A kind of depth difference method for secret protection based on production confrontation network |
| CN107563355A (en) * | 2017-09-28 | 2018-01-09 | 哈尔滨工程大学 | Hyperspectral abnormity detection method based on generation confrontation network |
| CN108009628A (en) * | 2017-10-30 | 2018-05-08 | 杭州电子科技大学 | A kind of method for detecting abnormality based on generation confrontation network |
Non-Patent Citations (3)
| Title |
|---|
| SHOUJIN WANG ECT.: "Training Deep Neural Networks on Imbalanced Data Sets", 《IEEE》 * |
| 王坤峰等: "生成式对抗网络GAN的研究进展与展望", 《自动化学报》 * |
| 袁辰、钱丽萍、张慧、张婷: "基于生成对抗网络的恶意域名训练数据生成", 《计算机应用研究》 * |
Cited By (27)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109584221A (en) * | 2018-11-16 | 2019-04-05 | 聚时科技(上海)有限公司 | A kind of abnormal image detection method generating confrontation network based on supervised |
| CN109584221B (en) * | 2018-11-16 | 2020-07-28 | 聚时科技(上海)有限公司 | Abnormal image detection method based on supervised generation countermeasure network |
| CN109685200A (en) * | 2018-11-19 | 2019-04-26 | 华东师范大学 | Industrial protocol construction method and building system are calculated based on the mist for generating confrontation network |
| CN109685200B (en) * | 2018-11-19 | 2023-04-25 | 华东师范大学 | Mist computing industrial protocol construction method and system based on generation countermeasure network |
| CN110009171A (en) * | 2018-11-27 | 2019-07-12 | 阿里巴巴集团控股有限公司 | Customer behavior modeling method, apparatus, equipment and computer readable storage medium |
| CN109492193A (en) * | 2018-12-28 | 2019-03-19 | 同济大学 | Abnormal network data based on depth machine learning model generate and prediction technique |
| CN111507934B (en) * | 2019-01-30 | 2024-04-26 | 富士通株式会社 | Training device, training method, and computer-readable recording medium |
| CN111507934A (en) * | 2019-01-30 | 2020-08-07 | 富士通株式会社 | Training apparatus, training method, and computer-readable recording medium |
| CN109978550A (en) * | 2019-03-12 | 2019-07-05 | 同济大学 | A kind of credible electronic transaction clearance mechanism based on generation confrontation network |
| CN110505241B (en) * | 2019-09-17 | 2021-07-23 | 武汉思普崚技术有限公司 | Network attack plane detection method and system |
| CN110505241A (en) * | 2019-09-17 | 2019-11-26 | 武汉思普崚技术有限公司 | A kind of network attack face detection method and system |
| CN110855654A (en) * | 2019-11-06 | 2020-02-28 | 中国移动通信集团广东有限公司 | Vulnerability risk quantitative management method and system based on flow mutual access relation |
| CN110855654B (en) * | 2019-11-06 | 2021-10-08 | 中国移动通信集团广东有限公司 | Vulnerability risk quantitative management method and system based on flow mutual access relation |
| CN111090685A (en) * | 2019-12-19 | 2020-05-01 | 第四范式（北京）技术有限公司 | Method and device for detecting data abnormal characteristics |
| CN111090685B (en) * | 2019-12-19 | 2023-08-22 | 第四范式（北京）技术有限公司 | Method and device for detecting abnormal characteristics of data |
| WO2021130392A1 (en) | 2019-12-26 | 2021-07-01 | Telefónica, S.A. | Computer-implemented method for accelerating convergence in the training of generative adversarial networks (gan) to generate synthetic network traffic, and computer programs of same |
| CN112036955B (en) * | 2020-09-07 | 2021-09-24 | 贝壳找房(北京)科技有限公司 | User identification method and device, computer readable storage medium and electronic equipment |
| CN112036955A (en) * | 2020-09-07 | 2020-12-04 | 贝壳技术有限公司 | User identification method and device, computer readable storage medium and electronic equipment |
| CN112446002A (en) * | 2020-11-13 | 2021-03-05 | 天津大学 | Abnormity detection method for time sequence KPI data |
| CN112446002B (en) * | 2020-11-13 | 2022-11-15 | 天津大学 | Abnormity detection method for time sequence KPI data |
| CN112738092A (en) * | 2020-12-29 | 2021-04-30 | 北京天融信网络安全技术有限公司 | Log data enhancement method, classification detection method and system |
| CN113067653A (en) * | 2021-03-17 | 2021-07-02 | 北京邮电大学 | Spectrum sensing method and device, electronic equipment and medium |
| CN113033656A (en) * | 2021-03-24 | 2021-06-25 | 厦门航空有限公司 | Interactive hole exploration data expansion method based on generation countermeasure network |
| CN113033656B (en) * | 2021-03-24 | 2023-12-26 | 厦门航空有限公司 | Interactive hole detection data expansion method based on generation countermeasure network |
| CN112818407B (en) * | 2021-04-16 | 2021-06-22 | 中国工程物理研究院计算机应用研究所 | Video privacy protection method based on generation countermeasure network |
| CN112818407A (en) * | 2021-04-16 | 2021-05-18 | 中国工程物理研究院计算机应用研究所 | Video privacy protection method based on generation countermeasure network |
| CN113378718A (en) * | 2021-06-10 | 2021-09-10 | 中国石油大学(华东) | Action identification method based on generation of countermeasure network in WiFi environment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108769993A (en) | Based on the communication network abnormal user detection method for generating confrontation network | |
| Bao et al. | Threat of adversarial attacks on DL-based IoT device identification | |
| CN110288979A (en) | A kind of audio recognition method and device | |
| CN110413707A (en) | The excavation of clique's relationship is cheated in internet and checks method and its system | |
| CN107547460A (en) | Radio communication Modulation Signals Recognition method based on deep learning | |
| CN107846392A (en) | A kind of intrusion detection algorithm based on improvement coorinated training ADBN | |
| CN112333194B (en) | GRU-CNN-based comprehensive energy network security attack detection method | |
| CN113922985B (en) | Network intrusion detection method and system based on ensemble learning | |
| CN110012019A (en) | A kind of network inbreak detection method and device based on confrontation model | |
| CN104052612B (en) | A kind of Fault Identification of telecommunication service and the method and system of positioning | |
| CN107835496A (en) | A kind of recognition methods of refuse messages, device and server | |
| CN107808358A (en) | Image watermark automatic testing method | |
| CN108684043B (en) | Abnormal user detection method of deep neural network based on minimum risk | |
| CN108076060A (en) | Neutral net Tendency Prediction method based on dynamic k-means clusters | |
| CN110267272A (en) | A kind of fraud text message recognition methods and identifying system | |
| CN106650828A (en) | Support vector machine-based intelligent terminal security level classification method | |
| CN108846405A (en) | Uneven medical insurance data classification method based on SSGAN | |
| CN114529819A (en) | Household garbage image recognition method based on knowledge distillation learning | |
| CN110020868A (en) | Anti- fraud module Decision fusion method based on online trading feature | |
| CN115422995A (en) | Intrusion detection method for improving social network and neural network | |
| CN103401626B (en) | Based on the collaborative spectrum sensing optimization method of genetic algorithm | |
| CN111917765A (en) | Network attack flow generation system based on generation type countermeasure network | |
| CN108566642A (en) | A kind of two-dimentional union feature authentication method based on machine learning | |
| CN108846476A (en) | A kind of intelligent terminal security level classification method based on convolutional neural networks | |
| CN107966678A (en) | Localization method, electronic device and storage medium based on signal data screening |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181106 |