CN111654855B - Trust updating method in underwater wireless sensor network based on AHP - Google Patents
Trust updating method in underwater wireless sensor network based on AHP Download PDFInfo
- Publication number
- CN111654855B CN111654855B CN202010498521.XA CN202010498521A CN111654855B CN 111654855 B CN111654855 B CN 111654855B CN 202010498521 A CN202010498521 A CN 202010498521A CN 111654855 B CN111654855 B CN 111654855B
- Authority
- CN
- China
- Prior art keywords
- trust
- node
- matrix
- evidence
- moment
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W12/00—Security arrangements; Authentication; Protecting privacy or anonymity
- H04W12/009—Security arrangements; Authentication; Protecting privacy or anonymity specially adapted for networks, e.g. wireless sensor networks, ad-hoc networks, RFID networks or cloud networks
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W12/00—Security arrangements; Authentication; Protecting privacy or anonymity
- H04W12/08—Access security
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W84/00—Network topologies
- H04W84/18—Self-organising networks, e.g. ad-hoc networks or sensor networks
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Engineering & Computer Science (AREA)
- Computer Security & Cryptography (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Mobile Radio Communication Systems (AREA)
- Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)
Abstract
The invention discloses an AHP-based underwater wireless sensor network trust updating method, which comprises the following four steps: 1. selecting trust attributes and initializing node trust levels; 2. selecting trust evidence according to trust attributes, and establishing an underwater wireless sensor network trust mechanism hierarchical structure model based on an AHP algorithm; 3. constructing a judgment matrix of trust attributes and trust evidences, and carrying out consistency test, if the judgment matrix passes the consistency test, constructing a trust vector and a multidimensional trust matrix; 4. constructing a time weight matrix, obtaining a multidimensional weight matrix, calculating a comprehensive trust level by combining the multidimensional trust level matrix, and judging and updating the node trust level according to the comprehensive trust level. The invention can effectively calculate and update the comprehensive trust level of the node, thereby ensuring the safety of the underwater wireless sensor network.
Description
Technical Field
The invention belongs to the field of Internet of things safety, and particularly relates to a trust updating method in an underwater wireless sensor network based on AHP.
Background
With the rapid development and wide application of computer network technology and communication technology (broadband technology, wireless technology, etc.), network security problems are also increasingly prominent. In recent years, some students at home and abroad begin to introduce trust means aiming at the network security problem, and some preliminary achievements are obtained, so that trust provides a new thought for solving the access control problem in an open dynamic network. The trust model plays an important role in resisting internal attack, identifying malicious nodes, selfish nodes, low-competitiveness nodes and the like, and can be applied to researches on security routing, security positioning, data fusion and the like so as to improve the security, reliability and fairness of the system.
Different from trust calculation in a land wireless sensor network, the calculation of the comprehensive trust of the nodes in the underwater wireless sensor network is quite complex, and the trust evidence to be considered is not only the capacity of the nodes, but also the volatility, stability and the like of the underwater environment. There is a degree of difficulty in dealing with sophisticated and diverse trust evidence under water.
The analytic hierarchy process, AHP for short, refers to a decision making process of decomposing elements related to decision making into target, rule, scheme and other layers and performing qualitative and quantitative analysis based on the target, rule, scheme and other layers. The method is a hierarchical weight decision analysis method provided by a method for teaching the Satty application network system theory and the multi-objective comprehensive evaluation method by the university of Pittsburgh of an operation student in the United states. The trust mechanism of the underwater wireless sensor network is evaluated by referring to the thought of the analytic hierarchy process, and a more reliable scheme is provided for constructing a trust model of the trusted network.
At present, domestic and foreign scholars have studied trust calculation by utilizing AHP, and related documents are as follows:
1.Jie MA,Yongsheng ZHANG et al in publication Research on Trusted Evaluation Method of User Behavior Based on AH P Algorithm of 2015, propose a user behavior evaluation method based on analytic hierarchy process, firstly, combine user behavior activities and reward factors, secondly, add an improved indirect credibility calculation method, and finally evaluate the user behavior.
2.Jingpei Wang,Jie Liu,Zi Xing et al in the article Credibility Evaluation of Trust Models based on Fuzzy Quantization and AHP in Ad hoc Scene published in 2016, propose an evaluation method for trust management models for Ad hoc networks. Firstly, some attributes are extracted from the trusted requests of specific scenes in the ad hoc network, and qualitative analysis is carried out on some traditional ad hoc trust models. And then, quantitatively calculating the overall evaluation value of the candidate trust model hierarchy attribute by using a fuzzy theory and a hierarchical analysis method, and selecting an optimal scheme according to the ordered evaluation result. The method can effectively evaluate the trust model and select the best trust model for a decision maker in a specific scene.
3.Eghbal Ghazizadeh,Brian Cusack et al, in 2017 publication Based on Tree Structure and AHP Algorithm Procotol in Wireless Sensor Networks, propose a theoretical solution for trust gap between cloud identity provider and cloud identity client, in which multi-standard decision (MCDM) was introduced to prioritize attributes of the cloud identity trust framework, dividing the overall trust assessment into two parts: trust analysis and quantification of trust by federated identity management systems, inputs are provided to a hierarchical analysis process (AHP) using MCDM methods to complete trust evaluation.
Disclosure of Invention
Aiming at the problems, the invention provides a trust updating method in an underwater wireless sensor network based on AHP, which solves the problem that the weight value is difficult to be accurately valued in the traditional method, effectively calculates and updates the node trust in real time, and comprises the following steps:
a trust updating method in an underwater wireless sensor network based on AHP comprises the following four main steps:
(1) Initializing node trust level and dividing node trust attribute;
(2) Selecting trust evidence according to trust attributes, and constructing an underwater wireless sensor network trust level model based on AHP;
(3) Constructing a judgment matrix of trust attributes and trust evidences, carrying out consistency test, and if the consistency test is passed, constructing a trust vector and a multidimensional trust matrix;
(4) Constructing a time weight matrix, obtaining a multi-dimensional weight matrix, calculating a comprehensive trust level by combining the multi-dimensional trust matrix, and judging and updating the node trust level according to the comprehensive trust level.
The step (1) specifically comprises the following steps:
(1.1) node trust level initialization
Node trust is initialized to five levels, full trust [0.8,1], high trust [0.6,0.8), ordinary trust [0.4,0.6), low trust [0.2, 0.4), no trust [0,0.2);
(1.2) node trust attribute partitioning
Node trust attributes are divided into three classes: availability, reliability, security.
Selecting trust evidences corresponding to the trust attributes of the nodes in the step (2): the availability evidence is selected as node available memory, and node residual energy is obtained; the reliability evidence is selected as the successful interaction times of the nodes, and the transmission delay of the nodes; the security evidence is selected as the consistency of the node recommending capability and the data packet.
The step (3) specifically includes:
(3.1) judgment matrix construction of trust attributes
Firstly, calculating the importance degree of each trust attribute of the node j at the m moment according to the trust evidence of the node j at the m moment;
secondly, the importance degrees of the three trust attributes of the node j at the m moment are respectively compared to obtain a trust attribute judgment matrix of the node j at the m moment;
calculating an attribute weight vector according to the attribute judgment matrix and carrying out consistency check on the attribute weight vector;
(3.2) judgment matrix construction of trust evidence
After the judgment matrix of the trust attribute passes the verification, the importance degree of the trust evidence corresponding to each trust attribute of the node j at the m moment is respectively compared to obtain a trust evidence judgment matrix of the node j at the m moment;
finally, calculating weight vectors of the trust evidences according to the judgment matrix of the trust evidences and performing consistency verification;
(3.3) construction of a confidence vector matrix
Firstly, combining trust evidence weight vectors passing through consistency test at the mth moment to form a trust evidence weight matrix;
secondly, calculating a trust degree vector of the node j at m moment by using a trust evidence weight matrix of the node j at m moment;
(3.4) construction of a Multi-dimensional Trust matrix
And forming a multidimensional trust matrix of the node j according to the trust vector of the node j at the m moment and the trust vectors of the node j at N moments before the m moment.
The step (4) specifically includes:
(4.1) construction of a time weight matrix
According to the principle that the trust weight at the latest moment is the largest, the weight occupied by trust evidences at the m-th moment and N moments before the m-th moment of the node j is set, and a time weight matrix is constructed;
multiplying the trust evidence weight vector by a time weight matrix to obtain a multidimensional weight matrix of the node j;
(4.2) calculation of comprehensive Trust
Calculating the comprehensive trust degree of the node j by using the multi-dimensional trust degree matrix and the multi-dimensional weight matrix of the node j;
(4.3) updating node trust levels
The trust factor is set as a supporting factor and two rejection factors; and when the trust level of the node is increased, calculating by adopting a supporting factor, and when the trust level is reduced, selecting a rejection factor according to the malicious behavior interaction proportion.
The invention has the beneficial effects that:
(1) In the invention, the trust attribute and the trust evidence are compared by utilizing the thought of AHP, and the weight matrix is obtained by combining the importance of the trust evidence, so that the weight among the trust evidence can be conveniently measured, and the comprehensive trust degree is more accurate.
(2) The method for updating the trust by utilizing the AHP in the invention combines time weight, thereby being convenient for processing the dynamic update of the trust.
(3) The node trust attribute is divided differently from the prior art, and trust evidences are further divided from three layers of usability, reliability and safety, so that the decision of the AHP method can be better realized, and the final trust decision result is facilitated.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a hierarchical structure of a trust model of an underwater wireless sensor network.
Detailed Description
The present invention will be described in further detail with reference to the following examples in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
The principle of application of the invention is described in detail below with reference to the accompanying drawings.
As shown in fig. 1, a trust update method in an underwater wireless sensor network based on AHP includes the following steps: (1) And initializing a node trust level and dividing node trust attributes.
(2) And selecting trust evidence according to the trust attribute, and constructing an underwater wireless sensor network trust hierarchy model based on the AHP.
(3) Constructing a judgment matrix of trust attributes and trust evidences, carrying out consistency test, and if the consistency test is passed, constructing a trust vector and a multidimensional trust matrix.
(4) Constructing a time weight matrix, obtaining a multidimensional weight matrix, calculating a comprehensive trust level by combining the multidimensional trust level matrix, and judging and updating the node trust level according to the comprehensive trust level.
The step (1) specifically comprises the following steps:
(1.1) initialization of node trust levels: node trust is initialized to five levels, full trust 0.8,1, high trust 0.6,0.8, normal trust 0.4,0.6, low trust 0.2, 0.4), no trust 0,0.2, respectively.
(1.2) partitioning of node trust attributes: node trust attributes are divided into three classes: availability, reliability, security.
The step (2) specifically comprises:
(2.1) selecting trust evidence corresponding to the trust attribute of the node: the availability evidence is selected as node available memory (represented by FM) and node residual energy (represented by RE); the reliability evidence is selected as the successful interaction times (indicated by NS) of the node, and the transmission delay (indicated by TD) of the node; the security evidence is chosen as node recommendation capability (denoted by RA) and consistency of the data packets (denoted by PC).
And (2.2) constructing a trust hierarchy structure of the underwater wireless sensor network, dividing the trust hierarchy structure into three layers according to different trust behavior characteristics contained in trust evidences, and establishing a trust hierarchy model according to an AHP (advanced high performance) idea according to the usability, the reliability and the safety as shown in fig. 2, wherein trust attributes and the trust evidences are in one-to-one correspondence.
The step (3) specifically comprises:
(3.1) judgment matrix construction of trust attributes
Firstly, calculating the importance degree of each trust attribute of the node j at the m moment according to the trust evidence of the node j at the m moment;
secondly, the importance degrees of the three trust attributes of the node j at the m moment are respectively compared to obtain a trust attribute judgment matrix of the node j at the m moment;
calculating an attribute weight vector according to the attribute judgment matrix and carrying out consistency check on the attribute weight vector;
(3.1.1) trust attribute C for each level k (k=1, 2,., m) comparing peer trust attribute pairs to form a decision matrix C as:
wherein c ij The comparison result of each factor i with respect to the factor j is shown. Wherein nine-level classification is used for comparison of importance between attributes, table 1 shows a general numerical setting standard group.
TABLE 1 numerical setting criteria group (3.1.2)
The consistency in consistency test refers to the logic consistency of judging logic thinking, and the maximum characteristic root of the matrix is as follows:
wherein lambda is max Is the largest characteristic root of matrix A, w i For each vector element in A, corresponding to lambda in the judgment matrix max The feature vector of (a) is normalized and then is marked as W (the sum of elements in the vector is 1).
The consistency index is denoted CI and is calculated as follows:
(3.2) judgment matrix construction of trust evidence
Wherein n is the number of vectors in A.
After the judgment matrix of the trust attribute passes the verification, the importance degree of the trust evidence corresponding to each trust attribute of the node j at the m moment is respectively compared to obtain a trust evidence judgment matrix E of the node j at the m moment;
wherein e ij Representing all levels of trust evidence corresponding to the trust attribute at the moment m.
Finally, calculating weight vectors of the trust evidences according to the judgment matrix of the trust evidences and performing consistency verification; (3.2.1) performing geometric averaging on the weight calculation of the trust evidence according to each row vector of the matrix of the trust evidence, and then normalizing to obtain a feature vector W and a weight matrix W of the trust evidence, wherein the feature vector W and the weight matrix W are expressed as follows:
(3.3) construction of a confidence vector matrix
Firstly, combining trust evidence weight vectors passing through consistency test at the mth moment to form a trust evidence weight matrix;
secondly, calculating a trust degree vector of the node j at m moment by using a trust evidence weight matrix of the j at m moment;
(3.3.1) calculating the trust evidence matrix C and the weight matrix W to obtain a trust degree vector of the node trust evidence at the ith moment, and setting the trust degree vector as t i I.e.
Finally, forming a multidimensional trust matrix of the node j according to the trust vector of the node j at N times before the mth time;
(3.3.2) the trust vector of each trust evidence of the node at the i-th moment and the trust vector at the i-1-th moment and the i-2-th moment, and the trust vector at the i-3-th moment form a multidimensional trust matrix tm.
Wherein t is FMi For the trust level of the memory available to the node at the moment i, t REi The remaining energy trust level of the node at the moment i, t NSi Trust level, t for successful interaction times of node at moment i TDi To transmit the time delay trust level at the node at the moment i, t RAi Recommending capability trust level for node at moment i, t PCi The consistency trust level of the data packet at the moment i is the consistency trust level of the node; i-1 is the previous instant of i, and so on.
The step (4) specifically comprises:
(4.1) time-based matrix construction: according to the principle that the trust weight at the latest moment is the largest, the weight occupied by the trust attribute at the ith moment and N moments before the ith moment of the node j is set, a time weight matrix is constructed, and the trust evidence weight vector w is multiplied by the time weight matrix to obtain a multidimensional weight matrix of the node j;
(4.1.1) setting the time weight to be
tw m =[tw i-3 ,tw i-2 ,tw i-1 ,tw i ] T (8)
Multiplying the time weight by the belief evidence weight vector to arrive at a usable time-dependent weight matrix tw m 。
(4.2) calculation of comprehensive trust: and calculating the comprehensive trust degree of the node j by using the multi-dimensional trust degree matrix and the multi-dimensional weight matrix of the node j.
The comprehensive trust degree is obtained by multiplying a multi-dimensional trust degree matrix and a multi-dimensional weight matrix
ET=rw m *tm (9)。
(4.3) updating of trust level: the trust factor is set to one support factor, two rejection factors. Node trust is generally changed in two ways, the level is increased or decreased, when the trust level is increased, a supporting factor is adopted for calculation, and when the trust level is decreased, a rejection factor is selected according to the interaction proportion of malicious behaviors. a is a support factor defined as:
wherein λ is the regulator, λ ε [0,1].
For the change of the unreliable evidence, two rejection factors are set, the update of the trusted evidence is related to the interaction proportion of malicious behaviors, when the proportion of the malicious behaviors is smaller than r, the punishment factor is smaller, and when the proportion of the malicious behaviors is larger than r, the punishment is doubled, so that the trusted evidence is updated as follows:
wherein the time slot is kth, h is the next time slot, T i l Other trust levels within the time slot.
b is a rejection factor defined as:
Claims (4)
1. the trust updating method in the underwater wireless sensor network based on the AHP is characterized by comprising the following four main steps:
(1) Initializing node trust level and dividing node trust attribute;
(2) Selecting trust evidence according to trust attributes, and constructing an underwater wireless sensor network trust level model based on AHP;
(3) Constructing a judgment matrix of trust attributes and trust evidences, carrying out consistency check, and if the consistency check is passed, constructing a trust vector and a multidimensional trust matrix;
the step (3) specifically comprises:
(3.1) judgment matrix construction of trust attributes
First according to the nodejIn the first placemTime trust evidence, computing nodejAt the position ofmImportance degree of each trust attribute at moment;
second to atmTime nodejThe importance degree of the three trust attributes of the node is respectively compared to obtain the nodejFirst, themA trust attribute judgment matrix of time;
then calculating weight vectors of all the attributes according to the attribute judgment matrix and carrying out consistency check on the weight vectors;
(3.2) judgment matrix construction of trust evidence
After the judgment matrix of the trust attribute passes the verification, the nodes are respectively comparedjAt the position ofmObtaining nodes according to importance degrees of trust evidences corresponding to trust attributes at momentjAt the position ofmA trust evidence judgment matrix at the moment;
finally, calculating weight vectors of the trust evidences according to the judgment matrix of the trust evidences and performing consistency verification;
(3.3) construction of a confidence vector matrix
First pass the consistency checkmCombining the trust evidence weight vectors at the moment to form a trust evidence weight matrix;
second, use the nodejAt the position ofmTime trust evidence weight matrix and computing nodejAt the position ofmA confidence vector for the moment;
(3.4) construction of a Multi-dimensional Trust matrix
According to the nodejAt the position ofmTime trust vector and nodejIn the first placemBefore the momentNTrust vector forming node for each momentjIs a multi-dimensional confidence matrix;
(4) Constructing a time weight matrix, obtaining a multidimensional weight matrix, calculating a comprehensive trust level by combining the multidimensional trust level matrix, and judging and updating the node trust level according to the comprehensive trust level.
2. The trust update method in an AHP-based underwater wireless sensor network of claim 1, wherein: the step (1) specifically comprises the following steps:
(1.1) node trust level initialization
Node trust is initialized to five levels, full trust [0.8,1], high trust [0.6,0.8), ordinary trust [0.4,0.6), low trust [0.2, 0.4), no trust [0,0.2);
(1.2) node trust attribute partitioning
Node trust attributes are divided into three classes: availability, reliability, security.
3. The trust updating method in an underwater wireless sensor network based on AHP according to claim 1, wherein the selection of trust evidence corresponding to the trust attribute of the node in the step (2) is: the availability evidence is selected as node available memory, and node residual energy is obtained; the reliability evidence is selected as the successful interaction times of the nodes, and the transmission delay of the nodes; the security evidence is selected as the consistency of the node recommending capability and the data packet.
4. The trust update method in an AHP-based underwater wireless sensor network according to claim 1, wherein the step (4) specifically comprises:
(4.1) construction of a time weight matrix
According to the principle of maximum trust weight at the latest moment, setting nodesjIs the first of (2)mTime of day and the firstmThe trust attributes of N times before the time occupy weights, and a time weight matrix is constructed;
multiplying the trust evidence weight vector by the time weight matrix to obtain a nodejIs a multi-dimensional weight matrix of (a);
(4.2) calculation of comprehensive Trust
Calculating the comprehensive trust degree of the node j by using the multi-dimensional trust degree matrix and the multi-dimensional weight matrix of the node j;
(4.3) updating node trust levels
The trust factor is set as a supporting factor and two rejection factors; and when the trust level of the node is increased, calculating by adopting a supporting factor, and when the trust level is reduced, selecting a rejection factor according to the malicious behavior interaction proportion.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010498521.XA CN111654855B (en) | 2020-06-04 | 2020-06-04 | Trust updating method in underwater wireless sensor network based on AHP |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010498521.XA CN111654855B (en) | 2020-06-04 | 2020-06-04 | Trust updating method in underwater wireless sensor network based on AHP |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111654855A CN111654855A (en) | 2020-09-11 |
CN111654855B true CN111654855B (en) | 2023-05-16 |
Family
ID=72348284
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010498521.XA Active CN111654855B (en) | 2020-06-04 | 2020-06-04 | Trust updating method in underwater wireless sensor network based on AHP |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111654855B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115409318B (en) * | 2022-07-22 | 2024-03-19 | 南方海洋科学与工程广东省实验室(广州) | Natural-based water purification scheme optimization method integrating fuzzy AHP and MDS |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104009993A (en) * | 2014-05-29 | 2014-08-27 | 安徽师范大学 | Trust evaluation method based on fuzzy filtration |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106850613B (en) * | 2017-01-24 | 2019-10-25 | 中国科学院信息工程研究所 | A kind of user behavior method for evaluating trust and system based on advanced AHP |
CN107832621B (en) * | 2017-11-16 | 2021-01-05 | 成都艾尔普科技有限责任公司 | AHP-based weight calculation method for behavior trust evidence |
CN110572822B (en) * | 2019-08-29 | 2022-09-02 | 河海大学常州校区 | Trust updating method in underwater wireless sensor network based on decision tree |
-
2020
- 2020-06-04 CN CN202010498521.XA patent/CN111654855B/en active Active
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104009993A (en) * | 2014-05-29 | 2014-08-27 | 安徽师范大学 | Trust evaluation method based on fuzzy filtration |
Also Published As
Publication number | Publication date |
---|---|
CN111654855A (en) | 2020-09-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Sun et al. | Dynamic digital twin and federated learning with incentives for air-ground networks | |
CN108009710A (en) | Node test importance appraisal procedure based on similarity and TrustRank algorithms | |
Ding et al. | A GAN-based intrusion detection model for 5G enabled future metaverse | |
CN104539601B (en) | Dynamic network attack process analysis method for reliability and system | |
Li et al. | A location and reputation aware matrix factorization approach for personalized quality of service prediction | |
CN108270723A (en) | A kind of acquisition methods in electric power networks Forecast attack path | |
Wang et al. | Anomaly-aware network traffic estimation via outlier-robust tensor completion | |
CN110543978A (en) | Traffic flow data prediction method and device based on wavelet neural network | |
CN113988441A (en) | Power wireless network link quality prediction and model training method and device | |
CN111654855B (en) | Trust updating method in underwater wireless sensor network based on AHP | |
CN110807508B (en) | Bus peak load prediction method considering complex weather influence | |
CN105933138B (en) | Space-time dimension combined cloud service credibility situation assessment and prediction method | |
Xu et al. | Cooperative swarm learning for distributed cyclic edge intelligent computing | |
Luo et al. | A Study on Many‐Objective Optimization Using the Kriging‐Surrogate‐Based Evolutionary Algorithm Maximizing Expected Hypervolume Improvement | |
Yang et al. | Energy scheduling for DoS attack over multi-hop networks: Deep reinforcement learning approach | |
Li et al. | Cyber performance situation awareness on fuzzy correlation analysis | |
Yi et al. | FedPE: Adaptive Model Pruning-Expanding for Federated Learning on Mobile Devices | |
CN113177675A (en) | Air conditioner cold load prediction method based on optimization neural network of longicorn group algorithm | |
Kong et al. | Taprank: A time-aware author ranking method in heterogeneous networks | |
Li et al. | DTCMF: Dynamic trust-based context-aware matrix factorization for collaborative filtering | |
CN114065220A (en) | Dual-level analysis situation assessment method based on distributed system | |
CN112926198A (en) | Credibility evaluation method and system of MSaaS simulation architecture | |
Campos et al. | Federated Transfer Learning for Energy Efficiency in Smart Buildings | |
Mao et al. | QoS trust rate prediction for Web services using PSO-based neural network | |
Liu et al. | Recommendation algorithm fusing implicit similarity of users and trust |
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 |