CN107124329A - Outlier detection method and system based on low water level sliding time window - Google Patents
Outlier detection method and system based on low water level sliding time window Download PDFInfo
- Publication number
- CN107124329A CN107124329A CN201710284487.4A CN201710284487A CN107124329A CN 107124329 A CN107124329 A CN 107124329A CN 201710284487 A CN201710284487 A CN 201710284487A CN 107124329 A CN107124329 A CN 107124329A
- Authority
- CN
- China
- Prior art keywords
- water level
- low water
- data
- data processing
- time window
- 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
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
- H04L43/0852—Delays
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/06—Management of faults, events, alarms or notifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/16—Threshold monitoring
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Environmental & Geological Engineering (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention discloses the outlier detection method and system based on low water level sliding time window;Including:Data distribution:External data flow is received, external data flow is then distributed to each data processing node;Data processing:Data processing node is handled the external data flow received;Define low water level sliding time window, using timestamp as horizontal axis, elapse over time, low water level sliding time window is constantly mobile from left to right in timestamp horizontal axis, point at any time, it is for reduced data above low water level sliding time window horizontal axis below untreatment data, horizontal axis;Then the position in the range of low water level sliding time window is stabbed according to current data processing time to find whether current data processing is Outlier Data;Data aggregate:The result of data processing collect to be exported.Discardable data, Outlier Data and normal pending data are distinguished, data processing reliability is improved, acceleration disturbance is recovered.
Description
Technical field
The present invention relates to a kind of outlier detection method, more particularly to the number that peels off based on low water level sliding time window
It is found that method and system.
Background technology
Stream process is that the data flow being continually changing is calculated in real time.In order to tackle instant place of the user to mass data
The challenge brought is managed, solution tradition MapReduce is the batch processing mode of representative in the bottleneck problem handled in real time, emerging stream
Processing method, important application value is respectively provided with terms of risk management, marketing management, advertisement putting, socialization are recommended.
The data source of stream process due in network delay, system in the reason, homogeneous data such as concurrently it cannot be guaranteed that strict
Data processing node is reached according to timestamps ordering, data occurs and produce inconsistent to peel off with reaching data processing node priority
Data.A large amount of Outlier Datas, its processing speed is slow, and data handling failure is judged to produce interference, the failure erroneous judgement of increase stream process
Probability.
Prior art is mainly replicated by daily record, heat, upstream is backed up etc., and method is realized fault-tolerant, and Outlier Data is not discussed.
Daily record and heat replicate fault-tolerance approach and use synchronous protocol incremental replication, therefore a large amount of Outlier Datas can seriously wear reproduction process down;
Upstream backup fault-tolerance approach can be by Outlier Data as troubleshooting, it will start the fault recovery of mistake.
Prior art D-Stream has found Outlier Data using parallel recovery method, and it is extensive to perform progress failure using supposition
Multiple, it analyzes storehouse dependent on batching data.Prior art gives a kind of out of order arrival processing method, passes through punctuation mark
With the explicit method such as heartbeat mechanism by out of order data ordering.Prior art MillWheel systems are proposed on this basis
The lowest limit of low water level representation of concept pending data, can quilt when the data that timestamp is less than low water level reach data processing
Directly abandon.The decision method for losing data is this method give, but does not provide the decision method of Outlier Data, the time is only used
Point represents that low water level can not strictly distinguish Outlier Data.Prior art Trident is kept away by the strict demand in order of pending data
Exempt to produce Outlier Data, this method depends on transaction framework, produce a large amount of extra expenses.
The content of the invention
The purpose of the present invention is exactly that there is provided the Outlier Data based on low water level sliding time window in order to solve the above problems
It was found that method and system, effectively distinguish discardable data, Outlier Data and normal pending data, data processing is improved reliable
Property, acceleration disturbance is recovered.
To achieve these goals, the present invention is adopted the following technical scheme that:
Outlier detection method based on low water level sliding time window, including:
Step (1):Data distribution:External data flow is received, external data flow is then distributed to each data processing section
Point;
Step (2):Data processing:Data processing node is handled the external data flow received;
Low water level sliding time window is defined, the timestamp of low water level sliding time window originates in low water level initial value, low
The width of water level sliding time window is w;The timestamp scope of low water level sliding time window is [low water level initial value, low water level
The width w of initial value+low water level sliding time window];
Using timestamp as horizontal axis, elapse over time, low water level sliding time window is in timestamp horizontal coordinate
Constantly moved from left to right on axle, at any time point, be untreated number above low water level sliding time window horizontal axis
According to horizontal axis lower section is reduced data;Then according to current data processing time stamp in low water level sliding time window
In the range of position come find current data processing whether be Outlier Data;
Step (3):Data aggregate:The result of data processing collect to be exported.
The data flow from different keywords can be concurrently on different data processing nodes at progress in step (2)
Reason.
The obtaining step of the low water level initial value is:Identify not processed packet earliest in current data processing
Timestamp is the low water level that current data is handled;Identify earliest not processed in the upstream data processing of current data processing
The timestamp of packet is the low water level that upstream data is handled;Then compare at the low water level and current data of current data processing
The size of both low water levels of the upstream data processing of reason, the low water level that small low water level is handled as current data, then,
Upstream data processing is reviewed along stream processing network topology, eventually through recurrence, the earliest untreated of whole stream processing network is found
Data, and low water level initial value is used as using the earliest untreatment data timestamp of whole stream processing network.
Stab the position in the range of low water level sliding time window to find current data according to current data processing time
Whether processing be the step of be Outlier Data:
If current data processing time stamp is not [low water level, the scope of low water level+sliding time window width/2), then locate
Reason packet is Outlier Data;
If current data processing time is stabbed in [low water level+sliding time window width/2, low water level+time slip-window
Mouth width] scope, then untreatment data bag is normal pending data;
If current data processing time stamp is less than low water level, untreatment data is discardable data.
Low water level sliding time window size w is to reach the time to set according to the patient maximum data delay of data processing
Fixed.
Outlier detection system based on low water level sliding time window, including:
Data distribution module:External data flow is received, external data flow is then distributed to each data processing node;
Data processing module:Data processing node is handled the external data flow received;
Low water level sliding time window is defined, the timestamp of low water level sliding time window originates in low water level initial value, low
The width of water level sliding time window is w;The timestamp scope of low water level sliding time window is [low water level initial value, low water level
The width w of initial value+low water level sliding time window];
Using timestamp as horizontal axis, elapse over time, low water level sliding time window is in timestamp horizontal coordinate
Constantly moved from left to right on axle, at any time point, be untreated number above low water level sliding time window horizontal axis
According to horizontal axis lower section is reduced data;Then according to current data processing time stamp in low water level sliding time window
In the range of position come find current data processing whether be Outlier Data;
Data aggregate module:The result of data processing collect to be exported.
Data flow from different keywords in data processing module can be concurrently enterprising in different data processing nodes
Row processing.
The obtaining step of the low water level initial value is:Identify not processed packet earliest in current data processing
Timestamp is the low water level that current data is handled;Identify earliest not processed in the upstream data processing of current data processing
The timestamp of packet is the low water level that upstream data is handled;Then compare at the low water level and current data of current data processing
The size of both low water levels of the upstream data processing of reason, the low water level that small low water level is handled as current data, then,
Upstream data processing is reviewed along stream processing network topology, eventually through recurrence, the earliest untreated of whole stream processing network is found
Data, and low water level initial value is used as using the earliest untreatment data timestamp of whole stream processing network.
Stab the position in the range of low water level sliding time window to find current data according to current data processing time
Whether processing be the step of be Outlier Data:
If current data processing time stamp is not [low water level, the scope of low water level+sliding time window width/2), then locate
Reason packet is Outlier Data;
If current data processing time is stabbed in [low water level+sliding time window width/2, low water level+time slip-window
Mouth width] scope, then untreatment data bag is normal pending data;
If current data processing time stamp is less than low water level, untreatment data is discardable data.
Low water level sliding time window size w is to reach the time to set according to the patient maximum data delay of data processing
Fixed.
Due to the control of low water level sliding time window, data are abandoned few.
Explanation on technical term:
Stream process, is that the data flow being continually changing is calculated in real time.
The packet handled during data flow, stream process, packet is by keyword, key value, timestamp triple group
Into.Stream calculation can be distributed on multiple stream process nodes according to the scope of the keyword of packet and run.
Beneficial effects of the present invention:
1st, delayed to reach and data processing failure, reduction stream process in the intraoral effective district divided data of low water level time slip-window
During data processing fault recovery erroneous judgement number of times;
2nd, discardable data, Outlier Data and the normal pending data in the processing of effective district divided data;
3rd, the Outlier Data of Data processing is found, data processing reliability can be improved, acceleration disturbance is recovered.
4th, step (2) low water level is defined based on the data flow between data processing, it is ensured that current data processing is not
Generation time stabs packet earlier again.Data processing processing time stabs the packet bigger than low water level, and timestamp is than low water
The small packet in position is directly abandoned.
Brief description of the drawings
Fig. 1 stream processing networks topology;
Fig. 2 low water level sliding time windows;
Fig. 3 low water level sliding time window embodiments;
Low water level sliding time window snapshot of Fig. 4 data processings three time points.
Embodiment
The invention will be further described with embodiment below in conjunction with the accompanying drawings.
As shown in figure 1, stream processing network is topological, including data distribution 101, various data processings 102,103,104,105
With data aggregate 106.The data distribution 101 is forwarded to follow-up various data processings for receiving external data flow.The number
It is the computing unit of stream process according to processing 102,103,104,105.The data aggregate 106, the result of data processing is carried out
Collect and exported.Data flow from different keywords can be performed concurrently on different data processing nodes.
The low water level is defined based on the data flow between data processing, identify current data processing in earliest not
The timestamp of processed packet, it is ensured that the current data processing no longer packet of generation time stamp earlier.Data processing is only
Processing time stabs the packet bigger than low water level, and the timestamp packet smaller than low water level is directly abandoned.
It is A upstream data processing that data-oriented, which handles A and B, B, and data processing A low water level recursive definition is
Lowwatermark (A)=min (oldestwork (A), lowwatermark (B)), wherein oldestwork function tables registration
According to the timestamp of the minimum untreatment data of timestamp in processing A.Upstream data processing is reviewed along stream processing network topology, can be with
The earliest untreatment data of whole stream process is found out, and low water level initial value is used as using its timestamp.It can be tolerated according to data processing
Maximum data delay reach time setting low water level sliding time window size w.
As shown in Fig. 2 the low water level sliding time window 200, timestamp originates in low water level 201, width for w when
Between window.The timestamp scope of low water level sliding time window is [low water level, low water level+sliding time window width].
Outlier detection method based on low water level sliding time window:
Current data processing time stamp is in [low water level, the untreated number of the scope of low water level+sliding time window width/2)
It is Outlier Data according to bag;
Current data processing time, stamp was [low water level+sliding time window width/2, low water level+sliding time window is wide
Degree] scope untreatment data bag be normal pending data;
Current data processing time stamp is discardable data less than the untreatment data of low water level, when being slided due to low water level
Between window control, abandon data it is few.
As shown in figure 3, low water level sliding time window specific embodiment, A is discardable data, B, D, F are processed number
According to C is Outlier Data, and E, G are normal pending data.
As shown in figure 4, low water level sliding time window snapshot of the data processing three time points, rectangle represents low water level
Sliding time window, the left vertical line of rectangle is low water level, is elapsed with system time, low water level sliding time window is constantly moved to right.Often
Data distribution in individual data processing snapshot is in horizontal time axis.Point, low water level sliding time window trunnion axis at any time
It is reduced data below untreatment data, trunnion axis that top, which is,.Low water level sliding time window timestamp scope is
[lowwatermark,lowwatermark+w].Intraoral in low water level time slip-window, timestamp is in [lowwatermark+w/
2, lowwatermark+w] untreatment data is non-Outlier Data, and timestamp is in [lowwatermark, lowwatermark+
W/2 untreatment data) is Outlier Data (such as Fig. 4 data I).It is not straight in the intraoral untreatment data of low water level time slip-window
Discarding (such as Fig. 4 data H) is connect, due to the control of low water level sliding time window, data are abandoned few.
Although above-mentioned the embodiment of the present invention is described with reference to accompanying drawing, not to present invention protection model
The limitation enclosed, one of ordinary skill in the art should be understood that on the basis of technical scheme those skilled in the art are not
Need to pay various modifications or deform still within protection scope of the present invention that creative work can make.
Claims (10)
1. the outlier detection method based on low water level sliding time window, it is characterized in that, including:
Step (1):Data distribution:External data flow is received, external data flow is then distributed to each data processing node;
Step (2):Data processing:Data processing node is handled the external data flow received;
Low water level sliding time window is defined, the timestamp of low water level sliding time window originates in low water level initial value, low water level
The width of sliding time window is w;The timestamp scope of low water level sliding time window for [low water level initial value, low water level initial value+
The width w of low water level sliding time window];
Using timestamp as horizontal axis, elapse over time, low water level sliding time window is in timestamp horizontal axis
Constantly move from left to right, at any time point, be untreatment data, water above low water level sliding time window horizontal axis
It is reduced data below flat reference axis;Then according to current data processing time stamp in the range of low water level sliding time window
Position come find current data processing whether be Outlier Data;
Step (3):Data aggregate:The result of data processing collect to be exported.
2. the outlier detection method as claimed in claim 1 based on low water level sliding time window, it is characterized in that,
The data flow from different keywords can be handled concurrently on different data processing nodes in step (2).
3. the outlier detection method as claimed in claim 1 based on low water level sliding time window, it is characterized in that,
The obtaining step of the low water level initial value is:Identify the time of not processed packet earliest in current data processing
Stab the low water level for current data processing;Identify not processed data earliest in the upstream data processing of current data processing
The timestamp of bag is the low water level that upstream data is handled;What the low water level and current data for then comparing current data processing were handled
The size of both low water levels of upstream data processing, the low water level that small low water level is handled as current data, then, along stream
Processing network topology reviews upstream data processing, eventually through recurrence, finds the earliest untreatment data of whole stream processing network,
And low water level initial value is used as using the earliest untreatment data timestamp of whole stream processing network.
4. the outlier detection method as claimed in claim 1 based on low water level sliding time window, it is characterized in that,
The position in the range of low water level sliding time window is stabbed according to current data processing time to find that current data is handled
The step of whether being Outlier Data is:
If current data processing time stamp is in [low water level, the scope of low water level+sliding time window width/2), then untreated number
It is Outlier Data according to bag;
If current data processing time, stamp was [low water level+sliding time window width/2, low water level+sliding time window is wide
Degree] scope, then untreatment data bag is normal pending data;
If current data processing time stamp is less than low water level, untreatment data is discardable data.
5. the outlier detection method as claimed in claim 1 based on low water level sliding time window, it is characterized in that,
Low water level sliding time window size w is to reach the time to set according to the patient maximum data delay of data processing
's.
6. the outlier detection system based on low water level sliding time window, it is characterized in that, including:
Data distribution module:External data flow is received, external data flow is then distributed to each data processing node;
Data processing module:Data processing node is handled the external data flow received;
Low water level sliding time window is defined, the timestamp of low water level sliding time window originates in low water level initial value, low water level
The width of sliding time window is w;The timestamp scope of low water level sliding time window for [low water level initial value, low water level initial value+
The width w of low water level sliding time window];
Using timestamp as horizontal axis, elapse over time, low water level sliding time window is in timestamp horizontal axis
Constantly move from left to right, at any time point, be untreatment data, water above low water level sliding time window horizontal axis
It is reduced data below flat reference axis;Then according to current data processing time stamp in the range of low water level sliding time window
Position come find current data processing whether be Outlier Data;
Data aggregate module:The result of data processing collect to be exported.
7. the outlier detection system as claimed in claim 6 based on low water level sliding time window, it is characterized in that,
Data flow from different keywords in data processing module can be concurrently on different data processing nodes at progress
Reason.
8. the outlier detection system as claimed in claim 6 based on low water level sliding time window, it is characterized in that,
The obtaining step of the low water level initial value is:Identify the time of not processed packet earliest in current data processing
Stab the low water level for current data processing;Identify not processed data earliest in the upstream data processing of current data processing
The timestamp of bag is the low water level that upstream data is handled;What the low water level and current data for then comparing current data processing were handled
The size of both low water levels of upstream data processing, the low water level that small low water level is handled as current data, then, along stream
Processing network topology reviews upstream data processing, eventually through recurrence, finds the earliest untreatment data of whole stream processing network,
And low water level initial value is used as using the earliest untreatment data timestamp of whole stream processing network.
9. the outlier detection system as claimed in claim 6 based on low water level sliding time window, it is characterized in that,
The position in the range of low water level sliding time window is stabbed according to current data processing time to find that current data is handled
The step of whether being Outlier Data is:
If current data processing time stamp is in [low water level, the scope of low water level+sliding time window width/2), then untreated number
It is Outlier Data according to bag;
If current data processing time, stamp was [low water level+sliding time window width/2, low water level+sliding time window is wide
Degree] scope, then untreatment data bag is normal pending data;
If current data processing time stamp is less than low water level, untreatment data is discardable data.
10. the outlier detection system as claimed in claim 6 based on low water level sliding time window, it is characterized in that,
Low water level sliding time window size w is to reach the time to set according to the patient maximum data delay of data processing
's.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710284487.4A CN107124329B (en) | 2017-04-25 | 2017-04-25 | Outlier data discovery method and system based on low-water-level sliding time window |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710284487.4A CN107124329B (en) | 2017-04-25 | 2017-04-25 | Outlier data discovery method and system based on low-water-level sliding time window |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107124329A true CN107124329A (en) | 2017-09-01 |
CN107124329B CN107124329B (en) | 2020-05-05 |
Family
ID=59724825
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710284487.4A Active CN107124329B (en) | 2017-04-25 | 2017-04-25 | Outlier data discovery method and system based on low-water-level sliding time window |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107124329B (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110069331A (en) * | 2019-04-24 | 2019-07-30 | 北京百度网讯科技有限公司 | A kind of data processing method, device and electronic equipment |
CN110460495A (en) * | 2019-08-01 | 2019-11-15 | 北京百度网讯科技有限公司 | A kind of water level propulsion method, device, calculate node and storage medium |
CN111478949A (en) * | 2020-03-25 | 2020-07-31 | 中国建设银行股份有限公司 | Data processing method and device |
CN113807632A (en) * | 2021-01-21 | 2021-12-17 | 北京沃东天骏信息技术有限公司 | Wind control data processing method and device |
US11809430B2 (en) | 2019-10-23 | 2023-11-07 | Nec Corporation | Efficient stream processing with data aggregations in a sliding window over out-of-order data streams |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101908065A (en) * | 2010-07-27 | 2010-12-08 | 浙江大学 | On-line attribute abnormal point detecting method for supporting dynamic update |
CN103400152A (en) * | 2013-08-20 | 2013-11-20 | 哈尔滨工业大学 | High sliding window data stream anomaly detection method based on layered clustering |
CN105764162A (en) * | 2016-05-10 | 2016-07-13 | 江苏大学 | Wireless sensor network abnormal event detecting method based on multi-attribute correlation |
-
2017
- 2017-04-25 CN CN201710284487.4A patent/CN107124329B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101908065A (en) * | 2010-07-27 | 2010-12-08 | 浙江大学 | On-line attribute abnormal point detecting method for supporting dynamic update |
CN103400152A (en) * | 2013-08-20 | 2013-11-20 | 哈尔滨工业大学 | High sliding window data stream anomaly detection method based on layered clustering |
CN105764162A (en) * | 2016-05-10 | 2016-07-13 | 江苏大学 | Wireless sensor network abnormal event detecting method based on multi-attribute correlation |
Non-Patent Citations (3)
Title |
---|
SEYED HESAMODIN KARIMIAN等: ""I-IncLOF: Improved Incremental Local Outlier Detection for Data Streams"", 《THE 16TH CSI INTERNATIONAL SYMPOSIUM ON ARTIFICIAL INTELLIGENCE AND SIGNAL PROCESSING (AISP 2012)》 * |
TYLER AKIDAU等: ""MillWheel: Fault-Tolerant Stream Processing at Internet Scale"", 《VLDB ENDOWMENT》 * |
赵学良: ""基于滑动窗口模型的数据流离群点检测研究"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110069331A (en) * | 2019-04-24 | 2019-07-30 | 北京百度网讯科技有限公司 | A kind of data processing method, device and electronic equipment |
CN110460495A (en) * | 2019-08-01 | 2019-11-15 | 北京百度网讯科技有限公司 | A kind of water level propulsion method, device, calculate node and storage medium |
CN110460495B (en) * | 2019-08-01 | 2024-02-23 | 北京百度网讯科技有限公司 | Water level propelling method and device, computing node and storage medium |
US11809430B2 (en) | 2019-10-23 | 2023-11-07 | Nec Corporation | Efficient stream processing with data aggregations in a sliding window over out-of-order data streams |
CN111478949A (en) * | 2020-03-25 | 2020-07-31 | 中国建设银行股份有限公司 | Data processing method and device |
CN111478949B (en) * | 2020-03-25 | 2022-05-24 | 中国建设银行股份有限公司 | Data processing method and device |
CN113807632A (en) * | 2021-01-21 | 2021-12-17 | 北京沃东天骏信息技术有限公司 | Wind control data processing method and device |
Also Published As
Publication number | Publication date |
---|---|
CN107124329B (en) | 2020-05-05 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107124329A (en) | Outlier detection method and system based on low water level sliding time window | |
US9838215B2 (en) | Ethernet ring protection node with node-level redundancy | |
WO2019101241A3 (en) | System and method for ending view change protocol | |
US9239864B2 (en) | Distributing and processing streams over one or more networks | |
Wan et al. | Distributed node‐to‐node consensus of multi‐agent systems with stochastic sampling | |
CN103843286A (en) | Triggering a redundant router master/backup status change based on switch connectivity | |
CN103763155A (en) | Multi-service heartbeat monitoring method for distributed type cloud storage system | |
CN103986129B (en) | Bus bar protecting method based on voltage lock phase networking sampling | |
CN106873945A (en) | Data processing architecture and data processing method based on batch processing and Stream Processing | |
CN106656589A (en) | Server dual hot backup system | |
Wang et al. | Meteor shower: A reliable stream processing system for commodity data centers | |
CN107276916B (en) | Switch flow table management method based on protocol non-perception forwarding technology | |
CN106844083A (en) | A kind of fault-tolerance approach and system perceived towards stream calculation system exception | |
Shang | Consensus recovery from intentional attacks in directed nonlinear multi-agent systems | |
US8948002B2 (en) | Method and device for preventing repeated refreshing of address table of Ethernet ring network node | |
CN102761448A (en) | Cluster monitoring and early warning method | |
Wang et al. | Resilient consensus for multi-agent systems under adversarial spreading processes | |
US10735247B2 (en) | Spanning tree protocol traffic handling system | |
CN106656584B (en) | Distributed system invalid node judgment method | |
CN107491359A (en) | A kind of distributed magnanimity real-time stream disaster recovery system and method | |
US20210021665A1 (en) | Apparatus, system, and method for flow-level switchover of video streams | |
CN103607288B (en) | The linear lossless network guard method of data network | |
Altaha et al. | Improvement of high-availability seamless redundancy (HSR) unicast traffic performance using enhanced port locking (EPL) approach | |
CN108055321B (en) | High-reliability cluster construction method based on localization platform | |
CN102186204A (en) | Heterogeneous wireless sensor network and special self-recovery method thereof |
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 |