CN111625785B - Time sequence data watermark comparison method based on data characteristic weight analysis - Google Patents
Time sequence data watermark comparison method based on data characteristic weight analysis Download PDFInfo
- Publication number
- CN111625785B CN111625785B CN202010375937.2A CN202010375937A CN111625785B CN 111625785 B CN111625785 B CN 111625785B CN 202010375937 A CN202010375937 A CN 202010375937A CN 111625785 B CN111625785 B CN 111625785B
- Authority
- CN
- China
- Prior art keywords
- data
- watermark
- sequence
- point
- data point
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 25
- 238000004364 calculation method Methods 0.000 claims description 10
- 238000000605 extraction Methods 0.000 claims description 5
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 238000011425 standardization method Methods 0.000 description 1
Images
Classifications
-
- 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
- G06F21/10—Protecting distributed programs or content, e.g. vending or licensing of copyrighted material ; Digital rights management [DRM]
- G06F21/16—Program or content traceability, e.g. by watermarking
-
- 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
- G06F21/10—Protecting distributed programs or content, e.g. vending or licensing of copyrighted material ; Digital rights management [DRM]
- G06F21/106—Enforcing content protection by specific content processing
- G06F21/1063—Personalisation
Abstract
The invention relates to the field of industrial Internet of things data security, and discloses a time sequence data watermark comparison method based on data characteristic weight analysis. And finally, judging whether the watermark sequence to be detected meets the judgment condition or not according to the same number of bit positions of the original watermark sequence and the watermark sequence to be detected. The method adopts data characteristic analysis, gives higher digital watermark weight to the high-value density data points, and ensures the data security of the high-value density area through weighted voting analysis.
Description
Technical Field
The invention relates to the technical field of industrial Internet of things data security, in particular to a time sequence data watermark comparison method based on data characteristic weight analysis.
Background
Industrial internet of things data is a major source of rapid expansion of industrial big data scale. Various internet of things sensors collect working state Data of devices where the sensors are located at an extremely high frequency, and the working state Data is usually a series of tuple sequences containing Data generation time stamps (timestamps) and collected Data (Data) in the form of (timestamps, Data), and the tuple sequences are called time sequences. The industrial time series data has the characteristics of wide application field, large data scale and high economic value, and contains huge commercial value, so that the safety of the industrial time series data is threatened by technical means such as hacking and the like adopted by lawless persons and non-technical means such as employment of commercial spying and the like.
Data owners usually adopt a prior method to protect data in a database, but the methods can only effectively prevent external personnel from carrying out illegal theft, and ways such as internal personnel theft cannot effectively inhibit. The digital watermarking is a main branch for solving the security problem of data in the transmission process, the general digital watermarking adopts a grouped majority voting method to improve the robustness of the algorithm, but the time sequence data generally has more noise and high-value data points are relatively centralized, so that a non-weighted voting algorithm can cause the failure of watermark judgment due to the interference of large-range noise.
The mainstream digital watermarking algorithm adopts packet embedding, and extracts and judges by a majority voting method, so that an attacker can not detect the watermark only by destroying more than 50% of watermark data. Time series data from industrial application scenarios are typically machine sensor data, with a large proportion of relatively stable values and a small number of outliers/special condition values that change at high speed, and with persistent environmental noise interference. Because the high-value data points are relatively concentrated, and the stable-value watermark is relatively easy to remove, the problem that the watermark in the high-value data points cannot be detected after the low-value data points are replaced by the low-value data points can be caused by adopting the common majority voting method.
Disclosure of Invention
In order to solve the problems, the invention provides a time sequence data watermark comparison method based on data characteristic weight analysis, which gives different weights to the grouped watermarks by analyzing and comparing the relative value density of time sequence data, ensures that the superposition of the watermarks in a high-value interval is easier to monitor, and greatly improves the detectability and the safety of the digital watermark method.
The invention discloses a time sequence data watermark comparison method based on data characteristic weight analysis, which comprises the following steps:
analyzing the data characteristic weight: analyzing the characteristics of the time series data, calculating coefficients of contribution degrees of different parts of data to important characteristics, obtaining relative weights of all data points on the time series, extracting digital watermarks on the same watermark bit positions on the time series data through weighted majority voting according to the relative weights, and forming a watermark sequence to be detected after characteristic weighting;
watermark comparison step: and judging whether the watermark sequence to be detected meets the judgment condition or not according to the same number of bits of the original watermark sequence and the watermark sequence to be detected.
Further, the data feature weight analyzing step comprises the following sub-steps:
s11, for a given data point in the time sequence data, selecting data points in a preset range before and after the given data point to perform change parameter calculation, and if the selected data point is just a local change point in the preset range, considering that the data point has change information of the time sequence data in a time dimension and has relatively high data value density;
s12, calculating the variation parameters of the whole time sequence data, obtaining the data value density coefficient of each data point through standardization, extracting the digital watermarks on the same watermark bit position on the time sequence data, performing weighted majority voting judgment according to the data value density coefficient of each data point, and selecting the digital watermark with the most weighted votes on the same watermark bit position as a digital watermark extraction value; and after arranging the extracted values of the digital watermarks, obtaining a watermark sequence to be detected after characteristic weighting.
Further, in step S11, the setting range is determined based on the frequency of the core feature of the time-series data.
Further, in step S11, the variation parameter calculation formula is:
wherein, d is the proximity range of given data point and takes the value of positive odd numberiFor values of data points within a proximity range, equation (1) refers to calculating the proximity of a given data pointThe mean of the data in the near range is divided by the absolute value of the difference from 1 of the mean of the other data not containing the data point.
Further, when the given data point is a change point, or a continuous interval with the length of β is a change interval with the given data point as a center, the deviation between the change parameter calculated by the formula (1) and 1 is large, otherwise, the deviation is close to 1, so as to extract the data characteristic parameter.
Further, in step S12, a data value density coefficient of the data point is obtained using a linear normalization calculation formula:
where δ(s) is a variation parameter of the entire time-series data, δ(s) [ [ δ [ ]1,δ2,…,δs]。
Further, in step S12, the formula for calculating the weighted majority vote determination is:
wherein θ(s) is a data value density coefficient of each data point, and θ(s) ([ θ ]1,θ2,…,θi,…θs]W(s) the extracted watermark bit value for each data point.
Further, in step S12, the digital watermark includes 0 and 1, and the data value density coefficient θ of the data pointiHas a value interval of [0,1 ]](ii) a When the weighted majority vote is judged, if the digital watermark on the same watermark bit is greater than 0.5, the digital watermark on the weighted majority on the watermark bit is 1, otherwise, the digital watermark is 0.
Further, the watermark comparing step comprises the sub-steps of:
s21, setting a confidence factor representing a threshold limit of probability of extracting an expected result due to random hit in time sequence data without the embedded watermark; the smaller the confidence factor is, the higher the credibility is, otherwise, the lower the credibility is;
s22, calculating the number of the same bits of the original watermark sequence and the watermark sequence to be detected, calculating that the two require at least the same number of bits as a bit threshold under a confidence factor, and if the number of the same bits of the original watermark sequence and the watermark sequence to be detected is greater than or equal to the bit threshold, determining that the watermark sequence to be detected meets a judgment condition.
Further, in step S21, a confidence factor is set to less than or equal to 1%.
The invention has the beneficial effects that:
the invention adopts data characteristic analysis, gives higher digital watermark weight to the high-value density data points, and ensures that the watermark loss of the noise section does not have overlarge influence on the judgment result during the watermark judgment through weighted voting analysis, thereby better protecting the data security of the high-value density area. In addition, the invention adjusts the confidence factor according to the data characteristics, and can reduce the false alarm rate of the watermarking method.
Drawings
FIG. 1 is a flow chart of a time series data watermark comparison method of the present invention;
reference numerals: 1-time sequence data to be verified, 2-data points and adjacent data points thereof, 3-carrying out watermark extraction on the data points according to an original watermark extraction method, 4-calculating a variation parameter according to the characteristics of the adjacent data points, 5-calculating a value density function of the whole time sequence, 6-carrying out weighted voting on each watermark position, 7-voting results are each watermark position results of the watermark sequence to be verified, 8-the watermark sequence to be verified, 9-the original watermark sequence, and 10-giving watermark matching judgment according to a set confidence factor.
Detailed Description
In order to more clearly understand the technical features, objects, and effects of the present invention, specific embodiments of the present invention will now be described. It should be understood that the detailed description and specific examples, while indicating the preferred embodiment of the invention, are intended for purposes of illustration only and are not intended to limit the scope of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present invention without making any creative effort, shall fall within the protection scope of the present invention.
As shown in fig. 1, the present embodiment provides a time series data watermark comparison method based on data feature weight analysis, which includes a data feature weight analysis step and a watermark comparison step, where:
the data feature weight analyzing step of the present embodiment includes the following sub-steps:
1. for a given data point, a time series of data points in the vicinity of β (the value of β can be selected according to the characteristics of the time series data) is selected for calculation of the variation parameter δ. The selection of the proximity range β may take into account the frequency of the data core characteristics.
2. The default calculation method of the variation parameter delta is as follows:
where β is the proximity of a given data point and takes the value of positive odd, diFor the values of data points in the vicinity, equation (1) refers to calculating the absolute value of the mean of the data in the vicinity of a given data point divided by the difference between the mean of the other data not containing the data point and 1. When the given data point is a change point or a continuous interval with the given data point as a center and the length of beta is a change interval, the deviation between the change parameter calculated by the formula (1) and 1 is large, otherwise, the deviation is close to 1, and therefore the data characteristic parameter is extracted.
3. Calculating a variation parameter delta(s) [ delta ] of the whole time sequence data1,δ2,…,δs]Then, a data value density coefficient theta(s) of each data point is obtained through a standardization method, and the data value density coefficient of each data point is obtained by default through a linear standardization calculation formula:
wherein the data value density coefficient theta of each data point(s)＝[θ1,θ2,…,θi,…θs]。
4. The extraction result of the digital watermark on the same watermark bit in the time sequence is judged by weighted majority voting, and the calculation mode is as follows:
where w(s) is the extracted watermark bit value for each data point. The data value density coefficient theta of the data point is that the digital watermark is 0 and 1iHas a value interval of [0,1 ]]Therefore, when the weighted majority vote is determined, if the digital watermark on the same watermark bit is greater than 0.5, it means that the digital watermark of the weighted majority on the watermark bit is 1, otherwise it is 0.
5. After arranging the digital watermarks voted by the weighted majority according to the bit sequence of the digital watermarks, the complete watermark sequence W (l) to be detected after the feature weighted majority is voted can be obtained.
The watermark comparison step of this embodiment comprises the following sub-steps:
1. the confidence factor alpha is chosen as desired. α is a parameter value between (0,1) and represents a threshold limit of the probability (p3) that the expected result is extracted due to random hits in the time-series data in which the watermark is not embedded. The smaller the alpha is, the higher the reliability is; the larger α, the lower the reliability. In the watermarking method, α may be set by a user, and a value of α ═ 0.01 or less is generally selected as the confidence factor.
2. Calculating the same bit number num of an original watermark sequence W (l) and a watermark sequence W (l) to be detected, then calculating the bit number b which needs to be at least the same of W (l) and W (l) under a confidence factor alpha, and returning a judgment result according to the size relationship between num and b.
3. If num is greater than or equal to b, if the same number of bits is num, the probability of random hit is less than α, and at this time, it can be considered that w (l) and w (l)' have a probability of 1- α, and return to True. Otherwise, it is considered that W (l) and W (l)' are not similar, and False is returned.
In summary, the present invention analyzes the data characteristics of the time sequence, calculates the coefficient of the contribution degree of different parts of data to the important characteristics, obtains the relative weight of each point on the time sequence, and changes the general simple majority voting into the weighted majority voting in the voting process, thereby ensuring that the watermark missing of the noise section does not affect the judgment result too much when the watermark is judged, and better protecting the data security of the high-value density area. In addition, the confidence factor is adjusted according to the data characteristics, so that the false alarm rate of the watermarking method can be reduced.
The foregoing is illustrative of the preferred embodiments of this invention, and it is to be understood that the invention is not limited to the precise form disclosed herein and that various other combinations, modifications, and environments may be resorted to, falling within the scope of the concept as disclosed herein, either as described above or as apparent to those skilled in the relevant art. And that modifications and variations may be effected by those skilled in the art without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (6)
1. A time series data watermark comparison method based on data characteristic weight analysis is characterized by comprising the following steps:
analyzing the data characteristic weight: analyzing the characteristics of the time series data, calculating coefficients of contribution degrees of different parts of data to important characteristics, obtaining relative weights of all data points on the time series, extracting digital watermarks on the same watermark bit positions on the time series data through weighted majority voting according to the relative weights, and forming a watermark sequence to be detected after characteristic weighting;
watermark comparison step: judging whether the watermark sequence to be detected meets the judgment condition or not according to the same number of bits of the original watermark sequence and the watermark sequence to be detected;
the data feature weight analyzing step includes the substeps of:
s11, for a given data point in the time sequence data, selecting data points in a preset range before and after the given data point to perform change parameter calculation, and if the selected data point is just a local change point in the preset range, considering that the data point has change information of the time sequence data in a time dimension and has relatively high data value density;
s12, calculating the variation parameters of the whole time sequence data, obtaining the data value density coefficient of each data point through standardization, extracting the digital watermarks on the same watermark bit position on the time sequence data, performing weighted majority voting judgment according to the data value density coefficient of each data point, and selecting the digital watermark with the most weighted votes on the same watermark bit position as a digital watermark extraction value; arranging the extracted values of the digital watermarks to obtain a watermark sequence to be detected after characteristic weighting;
in step S11, a setting range is determined based on the frequency of the core feature of the time-series data;
in step S11, the variation parameter calculation formula is:
where β is the proximity of a given data point and takes the value of positive odd, diFor the values of data points in the vicinity, equation (1) refers to calculating the absolute value of the mean of the data in the vicinity of a given data point divided by the difference between the mean of the other data not containing the data point and 1;
the watermark comparison step comprises the sub-steps of:
s21, setting a confidence factor representing a threshold limit of probability of extracting an expected result due to random hit in time sequence data without the embedded watermark; the smaller the confidence factor is, the higher the credibility is, otherwise, the lower the credibility is;
s22, calculating the number of the same bits of the original watermark sequence and the watermark sequence to be detected, calculating that the two require at least the same number of bits as a bit threshold under a confidence factor, and if the number of the same bits of the original watermark sequence and the watermark sequence to be detected is greater than or equal to the bit threshold, determining that the watermark sequence to be detected meets a judgment condition.
2. The method according to claim 1, wherein when the given data point is a change point or a continuous interval with a length β around the given data point is a change interval, the shift between the change parameter calculated by the formula (1) and 1 is larger, otherwise the shift is closer to 1, so as to extract the data feature parameter.
3. The method for comparing time series data watermarks based on data feature weight analysis according to claim 1, wherein in step S12, the data cost density coefficient of the data point is obtained by using a linear normalized calculation formula:
where δ(s) is a variation parameter of the entire time-series data, δ(s) [ [ δ [ ]1,δ2,…,δs]。
4. The method for comparing watermarks of time series data based on data characteristic weight analysis according to claim 3, wherein in step S12, the formula for calculating the weighted majority vote judgment is:
wherein θ(s) is a data value density coefficient of each data point, and θ(s) ([ θ ]1,θ2,…,θi,…θs]W(s) the extracted watermark bit value for each data point.
5. The method for comparing time series data watermarks based on data characteristic weight analysis as claimed in claim 4, wherein in step S12, the digital watermark comprises 0 and 1, and the data value density coefficient θ of the data point isiHas a value interval of [0,1 ]](ii) a When the weighted majority vote is judged, if the digital watermark on the same watermark bit is greater than 0.5, the digital watermark on the weighted majority on the watermark bit is 1, otherwise, the digital watermark is 0.
6. The method for comparing time-series data watermarks according to claim 1, wherein in step S21, the confidence factor is less than or equal to 1%.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010375937.2A CN111625785B (en) | 2020-05-07 | 2020-05-07 | Time sequence data watermark comparison method based on data characteristic weight analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010375937.2A CN111625785B (en) | 2020-05-07 | 2020-05-07 | Time sequence data watermark comparison method based on data characteristic weight analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111625785A CN111625785A (en) | 2020-09-04 |
| CN111625785B true CN111625785B (en) | 2022-03-01 |
Family
ID=72259009
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010375937.2A Active CN111625785B (en) | 2020-05-07 | 2020-05-07 | Time sequence data watermark comparison method based on data characteristic weight analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111625785B (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1870132A (en) * | 2006-06-26 | 2006-11-29 | 南京邮电大学 | Didigal audio waterprint method based on low density parity check code |
| CN101169779A (en) * | 2007-11-30 | 2008-04-30 | 清华大学 | Method for embedding and extracting frequency domain water mark in English text |
| CN101493928A (en) * | 2009-02-10 | 2009-07-29 | 国网信息通信有限公司 | Digital watermarking embedding, extracting and quantizing step size coordinating factor optimizing method and device |
| CN103198448A (en) * | 2013-04-11 | 2013-07-10 | 江苏大学 | Three-dimensional model digital watermarking embedding method and blind detection method based on vertex curvature |
| CN106780286A (en) * | 2017-01-23 | 2017-05-31 | 天津大学 | A kind of particle group optimizing water mark method extracted based on blind watermatking |
| CN107911698A (en) * | 2017-11-03 | 2018-04-13 | 清华大学深圳研究生院 | The embedding grammar and detection method of video digital watermark information |
| CN108614961A (en) * | 2017-12-21 | 2018-10-02 | 神龙汽车有限公司 | A kind of Database Digital Watermark is embedded and guard method |
| CN109872267A (en) * | 2019-02-19 | 2019-06-11 | 哈尔滨工业大学（深圳） | A kind of packet-based digital watermark method of robustness |
| CN110418153A (en) * | 2019-07-30 | 2019-11-05 | 北京达佳互联信息技术有限公司 | Watermark adding method, device, equipment and storage medium |
| CN110598851A (en) * | 2019-08-29 | 2019-12-20 | 北京航空航天大学合肥创新研究院 | Time series data abnormity detection method fusing LSTM and GAN |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8156414B2 (en) * | 2007-11-30 | 2012-04-10 | Seiko Epson Corporation | String reconstruction using multiple strings |
| CN101482965B (en) * | 2009-01-05 | 2011-01-05 | 深圳大学 | Digital watermark imbedding and detecting method and device |
| CN108229113B (en) * | 2017-12-21 | 2020-09-04 | 神龙汽车有限公司 | Database digital watermark extraction method |
-
2020
- 2020-05-07 CN CN202010375937.2A patent/CN111625785B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1870132A (en) * | 2006-06-26 | 2006-11-29 | 南京邮电大学 | Didigal audio waterprint method based on low density parity check code |
| CN101169779A (en) * | 2007-11-30 | 2008-04-30 | 清华大学 | Method for embedding and extracting frequency domain water mark in English text |
| CN101493928A (en) * | 2009-02-10 | 2009-07-29 | 国网信息通信有限公司 | Digital watermarking embedding, extracting and quantizing step size coordinating factor optimizing method and device |
| CN103198448A (en) * | 2013-04-11 | 2013-07-10 | 江苏大学 | Three-dimensional model digital watermarking embedding method and blind detection method based on vertex curvature |
| CN106780286A (en) * | 2017-01-23 | 2017-05-31 | 天津大学 | A kind of particle group optimizing water mark method extracted based on blind watermatking |
| CN107911698A (en) * | 2017-11-03 | 2018-04-13 | 清华大学深圳研究生院 | The embedding grammar and detection method of video digital watermark information |
| CN108614961A (en) * | 2017-12-21 | 2018-10-02 | 神龙汽车有限公司 | A kind of Database Digital Watermark is embedded and guard method |
| CN109872267A (en) * | 2019-02-19 | 2019-06-11 | 哈尔滨工业大学（深圳） | A kind of packet-based digital watermark method of robustness |
| CN110418153A (en) * | 2019-07-30 | 2019-11-05 | 北京达佳互联信息技术有限公司 | Watermark adding method, device, equipment and storage medium |
| CN110598851A (en) * | 2019-08-29 | 2019-12-20 | 北京航空航天大学合肥创新研究院 | Time series data abnormity detection method fusing LSTM and GAN |
Non-Patent Citations (2)
| Title |
|---|
| An Information-Preserving Watermarking Scheme for Right Protection of EMR Systems;M. Kamran等;《IEEE Transactions on Knowledge and Data Engineering》;20111018;第24卷(第11期);第1950-1962页 * |
| 基于变换域的多媒体数字水印关键技术研究;莫佳;《中国优秀博硕士学位论文全文数据库(博士)信息科技辑》;20160315(第3期);第I138-5页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111625785A (en) | 2020-09-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107493277B (en) | Large data platform online anomaly detection method based on maximum information coefficient | |
| CN104199410B (en) | A kind of bridge structural health monitoring versatility acquisition control system | |
| CN101458810B (en) | Vector map watermark method based on object property characteristic | |
| CN105208040A (en) | Network attack detection method and device | |
| CN111092891A (en) | Method, system and related device for detecting abnormal point in network | |
| CN103607391B (en) | SQL injection attack detection method based on K-means | |
| CN112329713A (en) | Network flow abnormity online detection method, system, computer equipment and storage medium | |
| CN108801437B (en) | Distributed optical fiber vibration sensing positioning method and device based on disturbance signal feature extraction | |
| CN102664013A (en) | Audio digital watermark method of discrete cosine transform domain based on energy selection | |
| CN110493181A (en) | User behavior detection method, device, computer equipment and storage medium | |
| CN103973697A (en) | Intrusion detecting method of internet-of-things sensing layer | |
| CN116627707A (en) | Detection method and system for abnormal operation behavior of user | |
| CN111625785B (en) | Time sequence data watermark comparison method based on data characteristic weight analysis | |
| CN103377457A (en) | Vector geographic data accurate authentication fragile watermarking method | |
| CN113743486B (en) | Method for predicting heading coal and gas outburst risk by using post-blasting gas concentration | |
| CN103886864B (en) | A kind of secret information detection method is embedded in by DWT domains distortion-compensated Quantisation | |
| CN102024244A (en) | Method and device for embedding and detecting watermarks based on image characteristic region | |
| CN108881277A (en) | The method, device and equipment of monitoring wireless sensor network node invasion | |
| CN111651735B (en) | Time sequence data watermarking method with timestamp alignment function | |
| Dang et al. | Trend-adaptive multi-scale PCA for data fault detection in IoT networks | |
| CN115150206B (en) | Intrusion detection safety early warning system and method for information safety | |
| CN115052304A (en) | GCN-LSTM-based industrial sensor network abnormal data detection method | |
| CN117240602B (en) | Identity authentication platform safety protection method | |
| Rong et al. | Dam safety monitoring data anomaly recognition using multiple-point model with local outlier factor | |
| CN100416595C (en) | Blind detecting system and method for digital watermarking flooding |
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 | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20221115 Address after: Area A, Tianfu New Economic Industrial Park, No. 366, North Section of Hupan Road, Tianfu New District, Chengdu, Sichuan 610000 Patentee after: SICHUAN ENERGY INTERNET RESEARCH INSTITUTE, TSINGHUA University Patentee after: TSINGHUA University Address before: Tianfu maker block, no.2039, south section of Tianfu Avenue, Tianfu New District, Chengdu, Sichuan 610000 Patentee before: SICHUAN ENERGY INTERNET RESEARCH INSTITUTE, TSINGHUA University |
|
| TR01 | Transfer of patent right | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20200904 Assignee: BEIJING TONGTECH Co.,Ltd. Assignor: TSINGHUA University Contract record no.: X2023990000347 Denomination of invention: A Time Series Data Watermarking Comparison Method Based on Data Feature Weight Analysis Granted publication date: 20220301 License type: Common License Record date: 20230327 |
|
| EE01 | Entry into force of recordation of patent licensing contract |