CN115526365A - Index optimization method, server and computer-readable storage medium - Google Patents
Index optimization method, server and computer-readable storage medium Download PDFInfo
- Publication number
- CN115526365A CN115526365A CN202110707644.4A CN202110707644A CN115526365A CN 115526365 A CN115526365 A CN 115526365A CN 202110707644 A CN202110707644 A CN 202110707644A CN 115526365 A CN115526365 A CN 115526365A
- Authority
- CN
- China
- Prior art keywords
- optimization
- server
- index
- performance
- model
- 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
- 238000005457 optimization Methods 0.000 title claims abstract description 245
- 238000000034 method Methods 0.000 title claims abstract description 87
- 230000002159 abnormal effect Effects 0.000 claims abstract description 24
- 230000008569 process Effects 0.000 claims abstract description 23
- 230000015654 memory Effects 0.000 claims description 30
- 238000004590 computer program Methods 0.000 claims description 11
- 230000000694 effects Effects 0.000 description 10
- 238000007726 management method Methods 0.000 description 10
- 238000004891 communication Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 3
- 230000007246 mechanism Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000012546 transfer Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 230000004931 aggregating effect Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 238000013145 classification model Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 230000002427 irreversible effect Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000007723 transport mechanism Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/20—Ensemble learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/067—Enterprise or organisation modelling
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Theoretical Computer Science (AREA)
- Entrepreneurship & Innovation (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Game Theory and Decision Science (AREA)
- Educational Administration (AREA)
- Software Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides an index optimization method, a server and a computer-readable storage medium, wherein the method applied to a first server comprises the steps of obtaining a first optimization result and service characteristics which are respectively sent by encryption by each second server, wherein the first optimization result is determined by first log information, the first log information carries operation characteristics of a first performance index in the process of recovering from an abnormal state to a normal state, the service characteristics are used for representing environment information for executing the operation characteristics, and the first optimization result comprises a first optimization model aiming at the first performance index; constructing a federal learning model according to all the first optimization results and all the service characteristics, and determining a second optimization model according to the federal learning model; and respectively encrypting and sending the second optimization model to each second server. According to the embodiment of the invention, the success rate of optimizing the performance index of the second server can be improved, the performance index can be optimized under the condition of not leaking or transferring data, and the safety is high.
Description
Technical Field
The embodiment of the invention relates to the technical field of artificial intelligence, in particular to an index optimization method, a server and a computer-readable storage medium.
Background
With the continuous development of society, a series of data protection laws and regulations are continuously provided in various countries at present, aiming at protecting the data privacy of citizens and enterprises, and it is seen that the protection of user data and the requirement of data security will become the irreversible trend in the future. The establishment of these laws and regulations to varying degrees presents new challenges to the traditional data processing model of artificial intelligence, since data in different regions cannot be transferred at will. At present, operation and maintenance personnel of a telecommunication operator need to continuously observe key performance indexes for a network management system and optimize related performance indexes, generally speaking, operations influencing the performance indexes are many, such as avoiding internal or external interference, modifying adjacent cell configuration, modifying engineering parameter configuration and the like, the operations are recorded in log information, performance indexes in related technologies are optimized, namely a set of algorithm models are constructed according to the operations to optimize the indexes, but the actual optimization effect is very common, and the success rate of optimizing the performance indexes is relatively low.
Disclosure of Invention
The following is a summary of the subject matter described in detail herein. This summary is not intended to limit the scope of the claims.
The embodiment of the invention provides an index optimization method, a server and a computer readable storage medium, which can improve the success rate of performance index optimization.
In a first aspect, an embodiment of the present invention provides an index optimization method, which is applied to a first server, and the method includes:
acquiring first optimization results and service characteristics which are sent by each second server in an encrypted manner and correspond to first performance indexes, wherein the first optimization results are determined by first log information of the second servers, the first log information carries operation characteristics of the first performance indexes in a process of recovering from an abnormal state to a normal state, the service characteristics are used for representing environment information for executing the operation characteristics, and the first optimization results comprise first optimization models aiming at the first performance indexes;
establishing a federal learning model by taking all the first optimization results and all the business characteristics as input data, and determining a second optimization model according to the federal learning model;
and respectively encrypting and sending the second optimization model to each second server, so that each second server updates the first optimization model according to the second optimization model.
In a second aspect, an embodiment of the present invention further provides an index optimization method, which is applied to a second server, where a plurality of second servers are provided, and the method includes:
acquiring first log information and service characteristics, wherein the first log information carries operation characteristics of a first performance index in the process of recovering from an abnormal state to a normal state, and the service characteristics are used for representing environment information for executing the operation characteristics;
determining a first optimization result corresponding to the first performance index according to the operation characteristic, wherein the first optimization result comprises a first optimization model aiming at the first performance index;
encrypting and transmitting the first optimization result and the business features to a first server so that the first server determines a second optimization model according to a federal learning model, wherein the federal learning model is constructed by using the first optimization result and the business features respectively sent by a plurality of second servers as input data by the first server;
and receiving the second optimization model sent by the first server in an encryption mode, and updating the first optimization model according to the second optimization model.
In a third aspect, an embodiment of the present invention further provides a server, including: a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the indicator optimization method of the first aspect as described above when executing the computer program.
In a fourth aspect, an embodiment of the present invention further provides a server, including: a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the index optimization method of the second aspect as described above when executing the computer program.
In a fifth aspect, an embodiment of the present invention further provides a computer-readable storage medium, which stores computer-executable instructions for performing the index optimization method of the first aspect described above, or performing the index optimization method of the second aspect described above.
The embodiment of the invention comprises the following steps: the index optimization method applied to the first server comprises the steps of obtaining a first optimization result and service characteristics, corresponding to a first performance index, sent by each second server in an encryption mode, wherein the first optimization result is determined by first log information of the second server, the first log information carries operation characteristics of the first performance index in the process of recovering from an abnormal state to a normal state, the service characteristics are used for representing environment information for executing the operation characteristics, and the first optimization result comprises a first optimization model aiming at the first performance index; all the first optimization results and all the service characteristics are used as input data to construct a federal learning model, and a second optimization model is determined according to the federal learning model; and respectively encrypting and sending the second optimization model to each second server so that each second server updates the first optimization model according to the second optimization model. According to the scheme provided by the embodiment of the invention, the first optimization results and the service characteristics sent by the second servers are obtained, so that the federal learning model is constructed by combining the service characteristics on the basis of the determined first optimization results, and the second optimization model with higher accuracy than the first optimization model can be obtained on the basis of the federal learning model.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
Drawings
The accompanying drawings are included to provide a further understanding of the present invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the example serve to explain the principles of the invention and do not constitute a limitation thereof.
FIG. 1 is a schematic diagram of a network topology for performing a metric optimization method according to an embodiment of the present invention;
FIG. 2 is a flow chart of an index optimization method provided by an embodiment of the invention;
FIG. 3 is a schematic illustration of operational features provided by one embodiment of the present invention;
FIG. 4 is a schematic diagram of a service feature provided by one embodiment of the present invention;
FIG. 5 is a flow chart of a metric optimization method provided by another embodiment of the present invention;
FIG. 6 is a flowchart of determining a first optimization result in an index optimization method according to an embodiment of the present invention;
fig. 7 is a flowchart of transmitting a first optimization result and service characteristics in an index optimization method according to an embodiment of the present invention;
FIG. 8 is a schematic diagram of a server provided by one embodiment of the present invention;
fig. 9 is a schematic diagram of a server provided by another embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
It is noted that while functional block divisions are provided in device diagrams and logical sequences are shown in flowcharts, in some cases, steps shown or described may be performed in sequences other than block divisions within devices or flowcharts. The terms first, second and the like in the description and in the claims, and the drawings described above, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order.
The invention provides an index optimization method, a server and a computer readable storage medium, wherein a federal learning model is constructed by combining service features on the basis of a determined first optimization result by obtaining the first optimization result and the service features sent by each second server, so that a second optimization model with higher accuracy than the first optimization model can be obtained on the basis of the federal learning model.
The embodiments of the present invention will be further explained with reference to the drawings.
As shown in fig. 1, fig. 1 is a schematic diagram of a network topology for performing an index optimization method according to an embodiment of the present invention.
In the example of fig. 1, the network topology includes, but is not limited to: the present invention relates to a wireless communication system, and more particularly, to a wireless communication system including a first server 100 and a second server 200, wherein the first server 100 may be provided as one, the second server 200 may be provided as a plurality of servers, each second server 200 may be provided in different countries, regions, areas, etc., and each second server 200 may be connected to the first server 100 through an intranet switch 300, which may be a wireless network, for example, a wireless communication network under the condition of a base station, or a wired network, etc., without limitation.
In an embodiment, the second servers 200 may be, but not limited to, used for monitoring user data and performing performance optimization on related indicators in the user data, where a source of the user data is not limited, and theoretically, the second servers 200 only need to be able to obtain the user data, for example, referring to fig. 1, each second server 200 is connected to the local network management system 400 through a corresponding switch, that is, each second server 200 can obtain the first log information from the corresponding local network management system 400, so as to obtain the user data of the corresponding local network management system 400 according to the first log information, and thus, the network management performance indicator optimization can be achieved through the second server 200, or, without being limited to the local network management system 400, if there is a security, privacy, or anomaly problem in the user data of the area where the second server 200 is located, the second server 200 may also perform performance optimization on the user data, in other words, a source of the user data obtained by the second server 200 may not be limited.
In an embodiment, the first server 100, as a core of federal Learning, may establish a virtual common model based on data information obtained from each second server 200 through a parameter exchange manner under an encryption mechanism, that is, without violating a data privacy regulation, and accordingly, a federal learned confidence model is deployed on each second server 200 as a working node, and each second server 200 performs performance index optimization according to the federal learned confidence model, in this process, no private data is leaked, and the security is high, where federal Learning (federal Learning) is a machine Learning framework, which can effectively help a plurality of organizations perform data usage and machine Learning modeling under the condition of meeting requirements of user privacy protection, data security and government regulations, and the functional principle of the framework is briefly described by a basic example:
assuming that there are two different and unrelated enterprises 1 and 2 that have different data, for example, enterprise 1 is configured with user characteristic data and enterprise 2 is configured with product characteristic data and annotation data, the respective data of the two enterprises cannot be simply and directly merged according to the relevant data regulation. In view of solving the above problems, one approach is to create a task model for each of the two parties, each task model being either a classification model or a prediction model, which may be approved by users of the respective enterprises when obtaining data, but presents a new problem of how to create high quality models on the respective sides of enterprise 1 and enterprise 2. The problem can be solved by adopting a federal learning system, under the condition that the own data of each enterprise can not be locally generated, the federal learning system can establish a virtual common model in a parameter exchange mode under an encryption mechanism, namely under the condition that data privacy regulations are not violated, the virtual common model is equivalent to aggregating the respective data of the enterprise 1 and the enterprise 2 together, but when the virtual model is established, the respective data of the enterprise 1 and the enterprise 2 does not move, so that the privacy and the data compliance cannot be affected, and under the federal learning mechanism, the established model only serves local targets in the respective areas of the enterprise 1 and the enterprise 2, so that the data of each enterprise can not be affected when in use.
In one embodiment, the User data is terminal data, where a terminal may be configured in multiple numbers, and each terminal may be referred to as an access terminal, a User Equipment (UE), a subscriber unit, a subscriber station, a mobile station, a remote terminal, a mobile device, a User terminal, a wireless communication device, a User agent, or a User Equipment. For example, each terminal may be a cellular phone, a cordless phone, a Session Initiation Protocol (SIP) phone, a Wireless Local Loop (WLL) station, a Personal Digital Assistant (PDA), a handheld device with a Wireless communication function, a computing device or other processing device connected to a Wireless modem, a vehicle-mounted device, a wearable device, a terminal device in a 5G network or a future 5G or higher network, and the like, and this embodiment is not particularly limited thereto.
The first server 100 and the second server 200 may each include a memory and a processor, respectively, wherein the memory and the processor may be connected by a bus or other means.
The memory, as a non-transitory computer-readable storage medium, may be used to store non-transitory software programs as well as non-transitory computer-executable programs. Further, the memory may include high speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, the memory optionally includes memory located remotely from the processor, and these remote memories may be connected to the processor through a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The network topology and the application scenario described in the embodiment of the present invention are for more clearly illustrating the technical solution of the embodiment of the present invention, and do not form a limitation on the technical solution provided in the embodiment of the present invention, and it is known to a person skilled in the art that the technical solution provided in the embodiment of the present invention is also applicable to similar technical problems with the evolution of the network topology and the occurrence of a new application scenario.
Those skilled in the art will appreciate that the network topology shown in fig. 1 is not meant to limit embodiments of the present invention, and may include more or fewer components than those shown, or some components may be combined, or a different arrangement of components.
In the network topology shown in fig. 1, the first server 100 and the second server 200 may respectively call the performance index optimization programs stored therein to execute the index optimization method.
Based on the structure of the network topology, various embodiments of the index optimization method of the present invention are proposed.
As shown in fig. 2, fig. 2 is a flowchart of an index optimization method according to an embodiment of the present invention, which may be applied to, but is not limited to, a first server in a network topology as shown in the embodiment of fig. 1, and the index optimization method includes, but is not limited to, steps S100 to S300.
And S100, acquiring a first optimization result and a service characteristic which are respectively sent by each second server in an encryption manner and correspond to the first performance index, wherein the first optimization result is determined by first log information of the second server, the first log information carries the operation characteristic of the first performance index in the process of recovering from the abnormal state to the normal state, the service characteristic is used for representing environment information for executing the operation characteristic, and the first optimization result comprises a first optimization model aiming at the first performance index.
In an embodiment, the first optimization results and the business characteristics sent by the second servers are obtained and used as basic data for constructing the federal learning model, so that the federal learning model can be constructed in subsequent steps.
It should be noted that the first performance index is not uniquely determined, and may be selected according to an actual scenario, for example, in a network parameter configuration scenario, a network configuration situation is desired to be known, so that the network neighboring situation is relatively concerned, and in this scenario, the network neighboring index is determined to be the first performance index, and further, the network neighboring index is further analyzed and processed; the first log information may be obtained by the second server in real time, or may be obtained by the second server in advance and stored, and as can be seen from the foregoing embodiments regarding the network topology, the carrier of the first log information may be various, and may be, for example, a network management system, a service system, and the like, which is not limited.
In an embodiment, the first log information carries an operation characteristic of the first performance indicator in the process of returning from the abnormal state to the normal state, and since the operation characteristic is accompanied with the state change process of the first performance indicator, the operation characteristic can be considered to have an influence on the optimization of the first performance indicator, and therefore, the first optimization result corresponding to the first performance indicator can be further confirmed based on the operation characteristic.
It is to be understood that the operation characteristic may correspond to all of the operation characteristic of the first performance indicator during the process of recovering from the abnormal state to the normal state, or may correspond to a part of the operation characteristic of the first performance indicator during the process of recovering from the abnormal state to the normal state, which is not limited.
In one embodiment, as shown in fig. 3, the first log information holds a record of operational characteristics in the form of a stored list, the operational characteristics including at least one of the following types:
marking the network element corresponding to the first performance index as Neid;
the Object corresponding to the first performance index is recorded as Object;
the Name of the Operation executed aiming at the first performance index is marked as Operation Name;
the numerical Value of the first performance index Before the first performance index is recovered from the abnormal state to the normal state is marked as Beforee Value;
the Value of the first performance index after the first performance index is recovered from the abnormal state to the normal state is marked as a Modify Value;
the Result of the operation performed on the first performance indicator is denoted as Result.
It should be noted that, in the above various types of operation characteristics, the network element and the object represent the target of the first performance index, the name and the result of the operation performed on the first performance index represent the specific meaning of the operation characteristic, and the numerical values before and after the first performance index is restored from the abnormal state to the normal state may represent the degree of the operation characteristic, so that the operation influence on the first performance index may be represented by the operation characteristic, and therefore, the first optimization result may be further determined based on the operation characteristic.
In an embodiment, the first optimization result comprises a first optimization model for the first performance indicator, and further comprises at least one of the following types:
a success rate of prediction optimization for the first performance index;
actual optimization success rate for the first performance index;
a confidence level corresponding to a success rate of predictive optimization.
The prediction optimization success rate is calculated after relevant characteristic parameters are input into the first optimization model, and is different from the actual optimization success rate, and the accuracy degree of the first optimization model can be determined by comparing the difference between the two success rates; the confidence corresponding to the success rate of the predictive optimization can represent the possibility of achieving the success rate of the predictive optimization, that is, the success rate of the predictive optimization is subjected to interval estimation, so that the stability of the result of the predictive optimization can be reflected, and therefore, the confidence is considered as one content of the first optimization result.
It is understood that the first optimization result may also include, but is not limited to, other types, such as the related feature parameters mentioned as the sample, other parameters involved in the optimization process, and the like, which is not limited.
And S200, constructing a federal learning model by taking all the first optimization results and all the service characteristics as input data, and determining a second optimization model according to the federal learning model.
In an embodiment, since all the first optimization results and all the service features are transmitted to the first server in an encrypted manner, the data security of all the first optimization results and all the service features is ensured, so that the first server obtains a second optimization model which is secure and stable according to the encrypted data.
In an embodiment, the business features are used for representing environment information for executing the operation features, and reflecting the influence of the environment information on the operation features, and the operation features can influence a first performance index in a process of recovering from an abnormal state to a normal state, so the environment information can also indirectly influence the optimization of the first performance index, and further the first optimization result and the business features are used as input data to construct a federal learning model on the basis of the first optimization result, so that the influence of the business features on the optimization of the first performance index can be further reflected, that is, a second optimization model with higher accuracy can be determined on the basis of the federal learning model.
It should be noted that the service features corresponding to different second servers do not affect each other, that is, the service features corresponding to different second servers may be the same or different, which is determined according to the local situation of the second servers.
In one embodiment, the environmental information may be diversified, i.e., all other factors affecting the operational characteristics may be considered as environmental information.
Example 1
Referring to fig. 4, different types of recorded parameters of service features are shown in a list form, and it is understood that, as a base station of a network node, since a communication network provided by the base station may affect the performance of an operation feature, relevant parameters of the base station may be used as the service features to represent corresponding environment information, for example, the service features may be, but are not limited to, one or more of longitude of the base station, latitude of the base station, antenna angle of the base station, altitude of the base station, and length of normal operation time of the base station; or, considering the influence of the scale capacity of the network management system itself, when the scale of the network management system is small, the execution speed and the fluency of the operation characteristics may be influenced, that is, the service characteristics may also be but not limited to one or more of the scale of the network management system or the region, the population density of the region, the ratio of the cell to the neighboring cell, and the like.
And S300, encrypting and sending the second optimization model to each second server respectively so that each second server updates the first optimization model according to the second optimization model.
In an embodiment, the second optimization models are respectively sent to the second servers, so that the second servers update the first optimization models according to the second optimization models, performance index optimization can be performed under the condition of no data leakage or data transfer, the security is higher, and the success rate of performance index optimization of the second servers can be improved.
In an embodiment, since the second optimization model is obtained through federal learning, the second optimization model is obtained by integrating different data based on different second servers, and can meet performance index optimization conditions of the different second servers.
Example two
The index optimization method provided by the embodiment of the invention can be implemented by the following processes:
assuming that two second servers (denoted as a and B) exist, when a and B are determined, a and B extract operation features from corresponding first log information respectively and confirm to obtain a first optimization result, encrypt and send the respective operation features and the first optimization result to corresponding intermediate nodes (the first intermediate node corresponds to a, and the second intermediate node corresponds to B), such as switches, transfer stations, and the like (may be intranet switches as shown in the embodiment in fig. 1), encrypt and send the respective obtained corresponding service features to the corresponding intermediate nodes, the first server obtains the operation features, the first optimization result and the service features from the first intermediate node and the second intermediate node respectively, constructs a federal learning model by using the obtained first optimization result and the service features as input data, determines a second optimization model according to the federal learning model, and finally returns the second optimization model to a and B through the first intermediate node and the second intermediate node respectively, so that a and B update the first optimization model according to the second optimization index, thereby increasing success rate of a and B.
It is understood that steps S100 to S300 are dynamically and cyclically executable, that is, each second server continuously updates the first optimization model according to the second optimization model, and re-determines the optimization results and the business features based on the updated optimization model, so that the first server continuously constructs the federal learning model with all the first optimization results and all the business features as input data, and determines the continuously optimized second optimization model according to the federal learning model.
In addition, in the above process, according to the difference of the optimization results determined by the second servers under different conditions, the influence degree of the service features may be further determined correspondingly, so as to determine whether to perform encryption transmission on the service features of the previous time, for example, the prediction optimization success rate in the first optimization result is 79%, after the first optimization model is updated by the second optimization model, a service feature is newly added on the basis of the original service feature, in this case, the prediction optimization success rate of the updated first optimization result obtained for the second time is 85%, it may be inferred that the added new service feature has a significant influence on improving the prediction optimization success rate, so that the newly added service feature may be retained in the subsequent performance index optimization process, and similarly, if the prediction optimization success rate of the first optimization result is relatively lowered after the service feature is adjusted, the corresponding service feature may be selected to be removed.
As shown in fig. 5, fig. 5 is a flowchart of an index optimization method according to another embodiment of the present invention, where the index optimization method may be applied to a second server in the network topology in the embodiment shown in fig. 1, where the second server is provided in multiple numbers, and the method includes, but is not limited to, steps S400 to S700.
Step S400, acquiring first log information and service characteristics, wherein the first log information carries operation characteristics of a first performance index in the process of recovering from an abnormal state to a normal state, and the service characteristics are used for representing environment information for executing the operation characteristics;
in an embodiment, by acquiring the first log information and the service characteristic, an operation characteristic in a process of recovering the first performance index from an abnormal state to a normal state and environment information of the operation characteristic under an execution condition can be determined, so that the federal learning model can be constructed in a subsequent step based on the operation characteristic and the environment information.
In an embodiment, the operational characteristics include at least one of the following types:
a network element corresponding to the first performance index;
an object corresponding to the first performance metric;
a name of an operation performed for the first performance indicator;
the first performance index is a value before the first performance index is recovered from the abnormal state to the normal state;
a value of the first performance indicator after the first performance indicator is restored from the abnormal state to the normal state;
a result of performing the operation on the first performance indicator.
It should be noted that the operational features in this embodiment have the same technical principles and the same technical effects as those of the embodiment shown in fig. 3, and for the technical principles and the technical effects of this embodiment, reference may be made to the description related to the embodiment shown in fig. 3, and redundant description is not repeated here to avoid redundancy.
And S500, determining a first optimization result corresponding to the first performance index according to the operation characteristics, wherein the first optimization result comprises a first optimization model aiming at the first performance index.
In an embodiment, the first log information carries an operation characteristic in a process of returning the first performance indicator from the abnormal state to the normal state, and since the operation characteristic is accompanied by a state change process of the first performance indicator, the operation characteristic may be considered to have an influence on optimization of the first performance indicator, and therefore, a first optimization result corresponding to the first performance indicator may be further confirmed based on the operation characteristic.
In an embodiment, the first optimization result comprises a first optimization model for the first performance indicator, and further comprises at least one of the following types:
a success rate of prediction optimization for the first performance index;
actual optimization success rate for the first performance index;
a confidence level corresponding to a success rate of predictive optimization.
It is understood that the first optimization result may also include, but is not limited to, other types, such as related characteristic parameters as a sample, other parameters involved in the optimization process, and the like, which is not limited.
In the example of fig. 6, in the case where the operation feature is plural, and the first log information further includes an index-affecting feature, the index-affecting feature being used to characterize a correlation between the plural operation features, step S500 includes, but is not limited to, steps S510 to S520.
Step S510, determining a plurality of operation characteristics with correlation from the operation characteristics according to the index influence characteristics;
step S520, determining a first optimization result corresponding to the first performance index according to the plurality of operation characteristics with the correlation.
In an embodiment, when the operation characteristics are multiple, it means that different operation characteristics may be executed simultaneously, in this case, if the cooperation effect between different operation characteristics is better, it is beneficial to optimize the first performance index more accurately, therefore, a concept of an index influence characteristic is introduced, the correlation between multiple operation characteristics may be determined by the index influence characteristic, so as to extract multiple operation characteristics with correlation to execute cooperatively, and it is possible to avoid the influence of irrelevant operation characteristics, so that the execution effect of the operation characteristics will be better, wherein the index influence characteristic may be determined by sampling statistics, and may be represented in the form of a correlation coefficient or a level, and a threshold may be set during the determination to select relevant operation characteristics, for example, for the operation characteristics provided in the embodiment shown in fig. 3, if the index influence characteristic represents a correlation coefficient between the number of occurrences of the network element and the operation object and the operation result, a correlation coefficient threshold is preset, and when the correlation coefficient is not less than the correlation coefficient threshold, it may be determined that the correlation between the network element, the operation object and the operation result are a group of the network element, and the operation result is determined as a group of the first performance index, and the first performance index.
In an embodiment, when index optimization prediction is performed on a plurality of selected operation characteristics, by providing the network element, the operation object, and the operation result as input data to the first optimization model, the first optimization model records the content related to the input information, and outputs a success rate of prediction optimization for the first performance index, and based on the above information, the entire content of the first optimization result can be determined.
Step S600, a first optimization result and service characteristics are transmitted to a first server in an encrypted mode, so that the first server determines a second optimization model according to a federal learning model, and the federal learning model is constructed by the first server by taking the first optimization result and the service characteristics which are respectively sent by a plurality of second servers as input data;
in an embodiment, the first optimization results and the business features are transmitted to the first server in an encryption mode, data security of all the first optimization results and all the business features is guaranteed, the second optimization model determined by the first server has data security, and meanwhile due to the fact that the business features and the first optimization results are transmitted in a matched mode, namely the first optimization results and the business features are used as input data to construct the federal learning model, the influence of the business features on optimization of the first performance index can be further embodied, namely the second optimization model with higher accuracy can be determined based on the federal learning model.
In the example of fig. 7, step S600 includes, but is not limited to, step S610 to step S620.
And S610, determining a first gradient corresponding to the first optimization result and the service characteristic according to the first optimization result and the service characteristic.
Step S620, the first gradient is transmitted to the first server in an encrypted manner.
In an embodiment, since the gradient is used as a vector parameter, which represents both the magnitude and the direction, the gradient has reliable stability as a data parameter, and therefore, the first gradient is used to carry the first optimization result and the service feature, which can simplify the transmission difficulty.
The specific means for encryption in the above embodiments is not limited, and may be set by itself according to the actual application scenario, for example, differential privacy, secret sharing, and the like.
Step S700: and receiving the second optimization model sent by the first server in an encryption mode, and updating the first optimization model according to the second optimization model.
In an embodiment, by receiving the second optimization model, the first optimization model can be updated according to the second optimization model, so that performance index optimization can be performed without data leakage or data transfer, the security is higher, and the success rate of performance index optimization of the second server can be improved.
In addition, the index optimization method may be applied to the second server and the first server in the network topology in the embodiment shown in fig. 1, and the method includes, but is not limited to, steps S800 to S1200.
Step S800, the second server acquires first log information and service characteristics, wherein the first log information carries operation characteristics of the first performance index in the process of recovering from an abnormal state to a normal state, and the service characteristics are used for representing environment information for executing the operation characteristics;
step S900, the second server determines a first optimization result corresponding to the first performance index according to the operation characteristics, and transmits the first optimization result and the service characteristics to the first server in an encrypted manner, wherein the first optimization result comprises a first optimization model aiming at the first performance index;
step S1000, the first server takes the first optimization results and the service characteristics of all the second servers as input data to construct a federal learning model, and determines a second optimization model according to the federal learning model;
and S1100, encrypting and sending the second optimization model to a second server.
And step S1200, the second server updates the first optimization model according to the second optimization model.
It should be noted that steps S800 to S1200 in this embodiment have the same technical principles and the same technical effects as steps S100 to S300 in the embodiment shown in fig. 2 and steps S400 to S700 in the embodiment shown in fig. 5, and the difference between the different embodiments is that the execution main body is different, where the execution main body in the embodiment shown in fig. 2 is the first server, the execution main body in the embodiment shown in fig. 5 is the second server, and the execution main bodies in this embodiment are the first server and the second server. With regard to the technical principle and the technical effect of the present embodiment, reference may be made to the description related to the embodiments shown in fig. 2 and fig. 5, and details are not repeated here to avoid redundant contents.
In addition, referring to fig. 8, an embodiment of the present invention also provides a server, including: a first memory, a first processor, and a computer program stored on the first memory and executable on the first processor.
The first processor and the first memory may be connected by a first bus or otherwise.
It should be noted that the server in this embodiment may be applied to, for example, the first server in the embodiment shown in fig. 1, the server in this embodiment can form a part of, for example, a network topology in the embodiment shown in fig. 1, and all of these embodiments belong to the same inventive concept, so that these embodiments have the same implementation principle and technical effect, and detailed description is omitted here.
The non-transitory software programs and instructions required to implement the index optimization method of the above embodiments are stored in the first memory and, when executed by the first processor, perform the index optimization method of the above embodiments, e.g., perform the above-described method steps S100 to S300 in fig. 2.
In addition, referring to fig. 9, an embodiment of the present invention further provides a server, including: a second memory, a second processor, and a computer program stored on the second memory and executable on the second processor.
The second processor and the second memory may be connected by a second bus or otherwise.
It should be noted that the server in this embodiment may be applied to, for example, the second server in the embodiment shown in fig. 1, the server in this embodiment can form a part of, for example, a network topology in the embodiment shown in fig. 1, and these embodiments all belong to the same inventive concept, so that these embodiments have the same implementation principle and technical effect, and are not described in detail herein.
The non-transitory software programs and instructions required to implement the index optimization method of the above embodiments are stored in the second memory, and when executed by the second processor, perform the index optimization method of the above embodiments, for example, performing the above-described method steps S400 to S700 in fig. 5, the method steps S510 to S520 in fig. 6, or the method steps S610 to S620 in fig. 7.
In addition, an embodiment of the present invention also provides a network node, including: a third memory, a third processor, and a computer program stored on the third memory and executable on the third processor.
The third processor and the third memory may be connected by a third bus or otherwise.
It should be noted that the network node in this embodiment may be applied to, for example, the first server and the second server in the embodiment shown in fig. 1, the network node in this embodiment can form a part of, for example, a network topology in the embodiment shown in fig. 1, and all of these embodiments belong to the same inventive concept, so that these embodiments have the same implementation principle and technical effect, and detailed description is omitted here.
The non-transitory software programs and instructions required to implement the index optimization method of the above embodiments are stored in the third memory, and when executed by the third processor, perform the index optimization method of the above embodiments, e.g., perform the above-described method steps S800 to S1200.
The above-described embodiments of the apparatus are merely illustrative, wherein the units illustrated as separate components may or may not be physically separate, i.e. may be located in one place, or may also be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment.
Furthermore, an embodiment of the present invention also provides a computer-readable storage medium storing computer-executable instructions, which are executed by a first processor, a second processor, a third processor or a controller, for example, by a first processor, a second processor or a third processor in the above-mentioned apparatus embodiment, and can cause the first processor, the second processor or the third processor to execute the index optimization method in the above-mentioned embodiment, for example, execute the above-mentioned method steps S100 to S300 in fig. 2, or execute the above-mentioned method steps S400 to S700 in fig. 5, the method steps S510 to S520 in fig. 6, or the method steps S610 to S620 in fig. 7, or execute the above-mentioned method steps S800 to S1200.
One of ordinary skill in the art will appreciate that all or some of the steps, systems, and methods disclosed above may be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). The term computer storage media includes volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data, as is well known to those of ordinary skill in the art. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital Versatile Disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can accessed by a computer. In addition, communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media as known to those skilled in the art.
While the preferred embodiments of the present invention have been described in detail, it will be understood by those skilled in the art that the foregoing and various other changes, omissions and deviations in the form and detail thereof may be made without departing from the scope of this invention.
Claims (11)
1. An index optimization method is applied to a first server, and comprises the following steps:
acquiring a first optimization result and service characteristics which are respectively sent by each second server in an encryption manner and correspond to a first performance index, wherein the first optimization result is determined by first log information of the second server, the first log information carries operation characteristics of the first performance index in a process of recovering from an abnormal state to a normal state, the service characteristics are used for representing environment information for executing the operation characteristics, and the first optimization result comprises a first optimization model aiming at the first performance index;
establishing a federal learning model by taking all the first optimization results and all the business characteristics as input data, and determining a second optimization model according to the federal learning model;
and respectively encrypting and sending the second optimization model to each second server, so that each second server updates the first optimization model according to the second optimization model.
2. An index optimization method according to claim 1, said first optimization result further comprising at least one of the following types:
a predictive optimization success rate for the first performance indicator;
actual optimization success rate for the first performance index;
a confidence corresponding to the prediction optimization success rate.
3. An index optimization method according to claim 1 or 2, said operational characteristics comprising at least one of the following types:
a network element corresponding to the first performance index;
an object corresponding to the first performance indicator;
a name of an operation performed on the first performance metric;
the first performance indicator is a value before the first performance indicator is restored from the abnormal state to the normal state;
a value of the first performance indicator after the first performance indicator is restored from an abnormal state to a normal state;
a result of performing an operation on the first performance metric.
4. An index optimization method is applied to a second server, wherein a plurality of second servers are arranged, and the method comprises the following steps:
acquiring first log information and service characteristics, wherein the first log information carries operation characteristics of a first performance index in the process of recovering from an abnormal state to a normal state, and the service characteristics are used for representing environment information for executing the operation characteristics;
determining a first optimization result corresponding to the first performance index according to the operation characteristic, wherein the first optimization result comprises a first optimization model aiming at the first performance index;
encrypting and transmitting the first optimization result and the business characteristics to a first server so that the first server determines a second optimization model according to a federated learning model, wherein the federated learning model is constructed by the first server by using the first optimization result and the business characteristics respectively sent by a plurality of second servers as input data;
and receiving the second optimization model sent by the first server in an encryption mode, and updating the first optimization model according to the second optimization model.
5. The index optimization method according to claim 4, wherein the operation features are plural, the first log information further includes index influence features for characterizing correlation between the plural operation features;
the determining a first optimization result corresponding to the first performance index according to the operation characteristic includes:
determining a plurality of the operation characteristics having correlation from the respective operation characteristics according to the index influence characteristics;
and determining a first optimization result corresponding to the first performance index according to a plurality of operation characteristics with correlation.
6. The method of claim 4, wherein the step of transmitting the first optimization result and the traffic characteristic to the first server in an encrypted manner comprises:
determining a first gradient corresponding to the first optimization result and the service characteristic according to the first optimization result and the service characteristic;
the first gradient is transmitted to a first server in an encrypted manner.
7. The index optimization method of claim 4, wherein the first optimization result further comprises at least one of the following types:
a predictive optimization success rate for the first performance indicator;
actual optimization success rate for the first performance index;
a confidence corresponding to the predictive optimization success rate.
8. An index optimization method according to claim 4 or 7, said operational characteristics comprising at least one of the following types:
a network element corresponding to the first performance index;
an object corresponding to the first performance metric;
a name of an operation performed on the first performance metric;
the first performance index is a value before the first performance index is restored from the abnormal state to the normal state;
a value of the first performance indicator after the first performance indicator is restored from an abnormal state to a normal state;
a result of performing an operation on the first performance indicator.
9. A server, comprising: memory, processor and computer program stored on the memory and executable on the processor, characterized in that the processor implements the index optimization method according to any one of claims 1 to 3 when executing the computer program.
10. A server, comprising: memory, processor and computer program stored on the memory and executable on the processor, characterized in that the processor implements the index optimization method according to any one of claims 4 to 8 when executing the computer program.
11. A computer-readable storage medium storing computer-executable instructions for performing the index optimization method of any one of claims 1 to 3, or performing the index optimization method of any one of claims 4 to 8.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110707644.4A CN115526365A (en) | 2021-06-24 | 2021-06-24 | Index optimization method, server and computer-readable storage medium |
| PCT/CN2022/097172 WO2022267870A1 (en) | 2021-06-24 | 2022-06-06 | Index optimization method, server, and computer-readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110707644.4A CN115526365A (en) | 2021-06-24 | 2021-06-24 | Index optimization method, server and computer-readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115526365A true CN115526365A (en) | 2022-12-27 |
Family
ID=84544103
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110707644.4A Pending CN115526365A (en) | 2021-06-24 | 2021-06-24 | Index optimization method, server and computer-readable storage medium |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN115526365A (en) |
| WO (1) | WO2022267870A1 (en) |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180089587A1 (en) * | 2016-09-26 | 2018-03-29 | Google Inc. | Systems and Methods for Communication Efficient Distributed Mean Estimation |
| CN109905287B (en) * | 2018-05-21 | 2021-02-12 | 华为技术有限公司 | Performance index calibration method and device |
| US20220052925A1 (en) * | 2018-12-07 | 2022-02-17 | Telefonaktiebolaget Lm Ericsson (Publ) | Predicting Network Communication Performance using Federated Learning |
| WO2021073726A1 (en) * | 2019-10-15 | 2021-04-22 | Telefonaktiebolaget Lm Ericsson (Publ) | Method for dynamic leader selection for distributed machine learning |
| CN111709534A (en) * | 2020-06-19 | 2020-09-25 | 深圳前海微众银行股份有限公司 | Federal learning method, device, equipment and medium based on evolution calculation |
| CN112926747B (en) * | 2021-03-25 | 2022-05-17 | 支付宝(杭州)信息技术有限公司 | Method and device for optimizing business model |
-
2021
- 2021-06-24 CN CN202110707644.4A patent/CN115526365A/en active Pending
-
2022
- 2022-06-06 WO PCT/CN2022/097172 patent/WO2022267870A1/en active Application Filing
Also Published As
| Publication number | Publication date |
|---|---|
| WO2022267870A1 (en) | 2022-12-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Gao et al. | Location privacy in database-driven cognitive radio networks: Attacks and countermeasures | |
| US20200244691A1 (en) | Risk-informed autonomous adaptive cyber controllers | |
| US20220043920A1 (en) | Blockchain-based secure federated learning | |
| CN113591119B (en) | Cross-domain identification analysis node data privacy protection and safety sharing method and system | |
| CN111901309B (en) | Data security sharing method, system and device | |
| Aysal et al. | Sensor data cryptography in wireless sensor networks | |
| CN112800472B (en) | Industrial internet identification data protection system based on micro-service architecture | |
| CN113515760B (en) | Horizontal federal learning method, apparatus, computer device, and storage medium | |
| US11228423B2 (en) | Method and device for security assessment of encryption models | |
| CN105340240A (en) | Methods and systems for shared file storage | |
| An et al. | Enhancement of opacity for distributed state estimation in cyber–physical systems | |
| KR20110083937A (en) | Method and apparatus for securely communicating between mobile devices | |
| CN115549888A (en) | Block chain and homomorphic encryption-based federated learning privacy protection method | |
| CN111669795B (en) | Ad hoc network mobile access switching method based on block chain security attribute | |
| CN110991905A (en) | Risk model training method and device | |
| CN113011632B (en) | Enterprise risk assessment method, device, equipment and computer readable storage medium | |
| CN110933040B (en) | Block chain based data uplink method, device, equipment and medium | |
| Letafati et al. | On learning-assisted content-based secure image transmission for delay-aware systems with randomly-distributed eavesdroppers | |
| CN116112175A (en) | Service processing method, device and medium of digital twin network based on block chain | |
| CN115526365A (en) | Index optimization method, server and computer-readable storage medium | |
| Seo et al. | Heimdallr: Fingerprinting SD-WAN Control-Plane Architecture via Encrypted Control Traffic | |
| EP3866527A1 (en) | Paging method and apparatus, equipment, and storage medium | |
| CN115859371A (en) | Privacy calculation method based on block chain, electronic device and storage medium | |
| CN105183740A (en) | Apparatus And Method For Data Taint Tracking | |
| CN115051835A (en) | Method, electronic device, storage medium and system for processing data |
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 |