CN102306336A - Service selecting frame based on cooperative filtration and QoS (Quality of Service) perception - Google Patents
Service selecting frame based on cooperative filtration and QoS (Quality of Service) perception Download PDFInfo
- Publication number
- CN102306336A CN102306336A CN 201110155599 CN201110155599A CN102306336A CN 102306336 A CN102306336 A CN 102306336A CN 201110155599 CN201110155599 CN 201110155599 CN 201110155599 A CN201110155599 A CN 201110155599A CN 102306336 A CN102306336 A CN 102306336A
- Authority
- CN
- China
- Prior art keywords
- service
- qos
- user
- services
- similarity
- 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
Images
Abstract
The invention discloses a service selecting frame based on cooperative filtration and QoS (Quality of Service) perception. Services are combined by users through a service design module; the users select one service or the combination of multiple services for achieving an equivalent function; each service can be achieved by one or more concrete services of the equivalent function; and candidate services are selected from each group of {b}concrete services of the equivalent function selected based on QoS service so as to satisfy an end-to-end function requirement and optimize the combination of QoS. In a QoS predicting method adopted in the invention, two predicting methods based on the users and based on the services are mixed; a more accurate QoS predicting effect can be provided; and moreover, an optimal mixing proportion can be regulated according to data types of different characteristics by regulating system parameters. Similarity is computed by a provided cosine modifying algorithm; and the execution efficiency and a computing effect of the similarity are enhanced compared with a traditional Persson related coefficient.
Description
Technical field
The invention belongs to the web service field, mainly realize a kind of services selection framework based on collaborative filtering and QoS perception.
Background technology
Service-oriented calculating (SOC) provides a kind of method that the simple function application seamless is polymerized to the coarsegrain value-added service for us.Nowadays, Services Combination has received the concern of industrial community, and has been applied to every field, such as Workflow Management, finance, ecommerce or the like.On the internet; A large amount of optional services is distributing; They all provide identical functions, but exist difference qualitatively in nonfunctional characteristics and service, select best candidate service anabolic process so the problem of Services Combination changes into according to quality of services for users (QoS) gradually.
In recent years, the researchs about services selection are arranged a lot, but these researchs based on a common prerequisite, that is exactly the user for the QoS of all candidate service all is known.But this is also unrealistic in real environment for use.Such as, because geographic position and network environment is different, different user possibly have a long way to go for the QoS of same service.Moreover a user is difficult to guarantee to use all candidate service, that is to say that for some service, the user can't provide QoS information.
Summary of the invention
The present invention is based on 2 problems in the background technology, provide a kind of QoS information to predict and, carry out the services selection framework of perception of the services selection of QoS perception based on complete QoS information to the unknown.
The technical scheme that the present invention solves existing issue is: a kind of services selection framework based on collaborative filtering and QoS perception,
1) user is through the composite services voluntarily of Service Design module; The user selects the combination of one or more services to realize equivalent functions; Every kind of service maybe be by the concrete service implementing of one or more equivalent functions; Based on the QoS services selection from the concrete service of every group of equivalent functions; Select candidate service to satisfy end-to-end functional requirement, optimal combination QoS;
2) targeted customer calls candidate service for the first time; The QoS selective system is that targeted customer and destination service are extracted proper vector; Through the similarity computing function; Obtain the similarity of training data in targeted customer and destination service and the database; Through the screening function; Obtain similar users and similar service, utilize their historical QoS to characterize, the target of prediction user is to the QoS of candidate service;
3) the target of prediction user selects the optimal service combined recommendation to give the user to the QoS of candidate service, carries out the screening to the concrete service of candidate by the beta pruning module.
As further improvement of the present invention, step 2) proper vector extracted of targeted customer and destination service calculates similarity, predicts respectively with service two aspects from the user, make up according to system's weight then.
The present invention is a further improvement, the system weights the system parameters
confidence and prediction of two parts.
As further improvement of the present invention; Beta pruning module in the step 3) utilizes the skyline algorithm to select the domination service according to multidimensional QoS for the candidate service data set; To arrange import of services services selection module, the beta pruning module is calculated through utility function and is selected optimal service.
As further improvement of the present invention, the utilization of described similarity computing function be to revise cosine-algorithm (Adjusted Cosine).Described screening function uses and mixes Top-K (Hybrid Top-K) algorithm, and described algorithm is:
Input:u: targeted customer; T (u): other users' set;
P: similarity threshold; K: the maximum amount of data of output candidate user
Output:S (u): the set of similar users
1): the every other user set of traversal except that the targeted customer joins all similarities in the targeted customer in the preliminary election formation greater than the user who both establishes threshold value P;
2): number of users
in the record preliminary election formation; And, the preliminary election formation is sorted according to similarity order from high to low;
In such cases, illustrate that similarity is abundant greater than the number of users of both establishing threshold value, be no less than K, so the similarity in the preliminary election formation is sorted from high to low, a preceding K user inserts the similar users set, joins among the S (u);
In this case, both instructions set the similarity threshold value is greater than the number of users does not reach the set of K, in order to avoid low similarity user information on the final predicted interference, only the pre-selection queue
user as output, was added to S (u) in;
5):?if?
=0
In such cases, all similar users similarity of targeted customer is all low excessively, and these data are excessive for the interference that predicts the outcome, so the output empty set, promptly the targeted customer does not have similar users;
6):return?S(u)。
As further improvement of the present invention, described proper vector comprises that described user's QoS proper vector has been described the QoS eigenwert to the difference service based on user's QoS proper vector with based on the QoS proper vector of serving, and the user fixes; Described QoS proper vector based on service has been described the eigenwert of service to different user, and service is fixed.
The present invention compared with prior art, its beneficial effect is to recommend for the user provides more accurately more efficiently Services Combination, further:
(1) used in the present invention, QoS prediction method, using the user-based mixture of two simulations service, QoS can provide more accurate prediction results, and can adjust the system parameters
, for different data types to adjust the characteristics of the most excellent mixing ratio.
(2) the correction cosine-algorithm that proposes among the present invention (Adjusted Cosine) calculates similarity, carries out on efficient and the similarity calculating effect increasing than existing Pei Ersen related coefficient (Pearson Correlation Coefficient).
(3) on screening function link, traditional top-K algorithm is unfavorable for calculating to a nicety, and does not consider the quality of choosing the result, if the low excessively user data of similarity as the data source of forecasting process, can influence the accuracy of experimental result on the contrary.Can well address this problem and propose to mix Top-K (Hybrid Top-K) algorithm among the present invention.
Description of drawings
Fig. 1 selects partial schematic diagram based on the dynamic Service of QoS.
Fig. 2 is the architectural framework figure based on the services selection framework of collaborative filtering.
Fig. 3 is a QoS predicted entire process flow diagram.
Embodiment
Referring to Fig. 1-3, a kind of services selection framework based on collaborative filtering and QoS perception, this case study on implementation comprises the steps:
1) user is through the composite services voluntarily of Service Design module; The user selects the combination of one or more services to realize equivalent functions; Every kind of service maybe be by the concrete service implementing of one or more equivalent functions; Based on the QoS services selection from the concrete service of every group of equivalent functions; Select candidate service to satisfy end-to-end functional requirement, optimal combination QoS;
2) targeted customer calls candidate service for the first time; The QoS selective system is that targeted customer and destination service are extracted proper vector; Through the similarity computing function; Obtain the similarity of training data in targeted customer and destination service and the database; Through the screening function; Obtain similar users and similar service, utilize their historical QoS to characterize, the target of prediction user is to the QoS of candidate service;
3) the target of prediction user selects the optimal service combined recommendation to give the user to the QoS of candidate service, carries out the screening to the concrete service of candidate by the beta pruning module.
Wherein, step 2) proper vector extracted of targeted customer and destination service is calculated similarity, predicts respectively with service two aspects from the user, makes up according to system's weight then.System's weight is made up of systematic parameter
and forecast confidence two parts.And proper vector comprises based on user's QoS proper vector with based on the QoS proper vector of serving, and described user's QoS proper vector has been described the QoS eigenwert to the difference service, and the user fixes; Described QoS proper vector based on service has been described the eigenwert of service to different user, and service is fixed.
Beta pruning module in the step 3) utilizes the skyline algorithm to select the domination service according to multidimensional QoS for the candidate service data set, will arrange import of services services selection module, and the beta pruning module is calculated through utility function and selected optimal service.
The utilization of similarity computing function be to revise cosine-algorithm (Adjusted Cosine).The screening function uses and mixes Top-K (Hybrid Top-K) algorithm, and described algorithm is:
Input:u: targeted customer; T (u): other users' set;
P: similarity threshold; K: the maximum amount of data of output candidate user
Output:S (u): the set of similar users
1): the every other user set of traversal except that the targeted customer joins all similarities in the targeted customer in the preliminary election formation greater than the user who both establishes threshold value P;
2):? Recording preselect the number of users in the queue
, and follow the descending order of similarity, sorts the queue for pre-selection;
In such cases, illustrate that similarity is abundant greater than the number of users of both establishing threshold value, be no less than K, so the similarity in the preliminary election formation is sorted from high to low, a preceding K user inserts the similar users set, joins among the S (u);
4):?if?0?<
<?K
In such cases, illustrate that similarity does not reach K of setting greater than the number of users of both establishing threshold value.In order to avoid the low similarity user information on the final result of the interference prediction, just select the pre-queue
user as output, added to the S (u) in;
5):?if?
=0
In such cases, all similar users similarity of targeted customer is all low excessively, and these data are excessive for the interference that predicts the outcome, so the output empty set, promptly the targeted customer does not have similar users;
6):return?S(u)。
The advantage of this algorithm is to introduce threshold concept, can be accurately and choose candidate's similar users rapidly, and guarantee the quality of similar users, and then improved the application accuracy of prediction algorithm.
Below in conjunction with concrete application of the present invention, further produce the present invention.
Fig. 1 has showed the practical implementation architectural schemes based on the services selection framework of collaborative filtering and QoS perception.
(1) sets up the QoS data set of historical user to numerous candidate service; This data set can be used (U, S, QoS) vector representation; Representative " User IP, service-number, user QoS " respectively, User IP and service-number can uniquely be confirmed a user and a service.This data gathering system provides training data for the integrity service Selection Framework.Utilize reptile can obtain the wsdl document of available service, service ID is added the single concrete service list of services selection framework in service search engine (like seekda.com etc.).Supervision is obtained different user at the QoS that calls these different services, and " User IP, service-number, user QoS " imported database as a QoS message block, utilizes the data set of these data as whole services selection framework.
(2) user can realize Services Combination through the Service Design process.Fig. 2 has showed the conceptual framework based on the QoS services selection, and user's request possibly satisfied by existing single service, and at this moment, the user can select composite services to realize equivalent functions.Composite services can be made up of a plurality of abstract service, and each abstract service can realize through the identical concrete service of one or more functions.Based on the final goal of QoS services selection is from the concrete service of function such as every group, selects candidate service to satisfy end-to-end functional requirement, and the maximum QoS after the optimal combination.
(3) according to user and concrete service ID, use and revise cosine-algorithm (Adjusted Cosine), all users that computational data is concentrated and targeted customer's similarity, and all services and targeted customer's similarity.Utilize and mix Top-K algorithm selection optimal candidate similar users and similar service, these QoS message block are input to the QoS prediction module.
(4) in the QoS prediction module, utilize the QoS message block of input to calculate respectively, and generate the confidence level that each predicts the outcome based on user's predicted value with based on the predicted value of serving.If there is not similar users, then use predicted value conduct to predict the outcome based on service; If there is not similar service, then use predicted value conduct to predict the outcome based on the user; Otherwise, calculate based on the user with based on the predicted value weight separately of service according to reliability forecasting, and generate result of calculation that final QoS predicts the outcome thus as predicted value.After carrying out disappearance QoS forecasting process, use all update data set that predicts the outcome.
Fig. 3 has showed the idiographic flow of QoS prediction.
Main flow process and algorithm are following:
At first variable and the formula that needs are used carries out some definition.
Definition 1. : the user
QoS vector for set of service S.For example, S = {
,?
,?
},?
QoS for them are {(1,? 3),? (0,? 0),? (1? 4)}, then
.
Definition 6.S (u): all similar users of user u.
Definition 7.S (s): all similar services of service s.
Concrete steps:
A) be each user, according to his the QoS generation user characteristics vector for all services, if the QoS disappearance, then respective element is changed to 0 in the vector;
B) be each service, according to the QoS generation service characteristic vector of all users to it, if the QoS disappearance, then respective element is changed to 0 in the vector;
C) according to the user characteristics vector, utilize and revise cosine-algorithm (Adjusted Cosine), calculate the similarity between user to be predicted and other users.
Similarity between the user is calculated:
D) according to service characteristic vector, utilize and revise cosine-algorithm (Adjusted Cosine), calculate the similarity between service to be predicted and other services.
Similarity between the service is calculated:
E) use mixing Top-K algorithm, choose k similar users and similar service respectively.
F) according to the QoS of similar users and similar service calculate respectively based on the user predict the outcome (user-based prediction) and based on (service-based prediction) value that predicts the outcome of service.
Predict the outcome (user-based prediction) based on the user:
Predict the outcome (service-based prediction) based on service:
G) according to predict the outcome (the user-based prediction) based on the user, based on predict the outcome (the service-based prediction) of service, and the final QoS of the calculating of confidence level separately predicts the outcome.If there is no similar users, use the
as a predictor of outcome; if there is no similar service, use the
as a predictor of outcome; otherwise, the use of mixed results as the predicted value.
The confidence level that predicts the outcome based on the user
The confidence level that predicts the outcome based on service
Mix the QoS predicted value
(5) because data bulk is too huge, carry out screening operation to the concrete service of candidate by the beta pruning module.This module utilizes the skyline algorithm to select the domination service according to multidimensional QoS for the candidate service data set.To arrange import of services services selection module, this module is calculated through utility function and is selected optimal service.
In the composite services of user's design, utilizing said method is the concrete service set selection of the candidate optimal service of each abstract service, gives the user with combined recommendation of these concrete services.
Claims (7)
1. services selection framework based on collaborative filtering and QoS perception, it is characterized in that: step is following,
1) targeted customer is through the composite services voluntarily of Service Design module; Select the combination of one or more services to realize equivalent functions; Every kind of service maybe be by the concrete service implementing of one or more equivalent functions; Based on the QoS services selection from the concrete service of function such as every group; Select candidate service to satisfy end-to-end functional requirement, optimal combination QoS;
2) targeted customer calls candidate service for the first time; The QoS selective system is that targeted customer and destination service are extracted proper vector; Through the similarity computing function; Obtain the similarity of training data in targeted customer and destination service and the database; Through the screening function; Obtain similar users and similar service, utilize their historical QoS to characterize, the target of prediction user is to the QoS of candidate service;
3) the target of prediction user selects the optimal service combined recommendation to give the user to the QoS of candidate service, carries out the screening to the concrete service of candidate by the beta pruning module.
2. the services selection framework based on collaborative filtering and QoS perception as claimed in claim 1; It is characterized in that: step 2) proper vector extracted of targeted customer and destination service calculates similarity; Predict respectively with service two aspects from the user, make up according to system's weight then.
4. the services selection framework based on collaborative filtering and QoS perception as claimed in claim 1; It is characterized in that: the beta pruning module in the step 3) utilizes the skyline algorithm to select the domination service according to multidimensional QoS for the candidate service data set; To arrange import of services services selection module, the beta pruning module is calculated through utility function and is selected optimal service.
5. the services selection framework based on collaborative filtering and QoS perception as claimed in claim 1 is characterized in that: the utilization of described similarity computing function be to revise cosine-algorithm (Adjusted Cosine).
6. the services selection framework based on collaborative filtering and QoS perception as claimed in claim 1 is characterized in that: described screening function uses and mixes Top-K (Hybrid Top-K) algorithm, and described algorithm is:
Input:u: targeted customer; T (u): other users' set;
P: similarity threshold; K: the maximum amount of data of output candidate user
Output:S (u): the set of similar users
1): the every other user set of traversal except that the targeted customer joins all similarities in the targeted customer in the preliminary election formation greater than the user who both establishes threshold value P;
2):? Recording preselect the number of users in the queue
, and follow the descending order of similarity, sorts the queue for pre-selection;
3):?if?
In such cases, illustrate that similarity is abundant greater than the number of users of both establishing threshold value, be no less than K, so the similarity in the preliminary election formation is sorted from high to low, a preceding K user inserts the similar users set, joins among the S (u);
In this case, both instructions set the similarity threshold value is greater than the number of users does not reach the set of K, in order to avoid low similarity user information on the final predicted interference, only the pre-selection queue
user as output added to S (u) in;
5):?if?
=0
In such cases, all similar users similarity of targeted customer is all low excessively, and these data are excessive for the interference that predicts the outcome, so the output empty set, promptly the targeted customer does not have similar users;
6):return?S(u)。
7. the services selection framework based on collaborative filtering and QoS perception as claimed in claim 1 or 2; It is characterized in that: described proper vector; Comprise based on user's QoS proper vector with based on the QoS proper vector of serving; Described user's QoS proper vector has been described the QoS eigenwert to the difference service, and the user fixes; Described QoS proper vector based on service has been described the eigenwert of service to different user, and service is fixed.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 201110155599 CN102306336A (en) | 2011-06-10 | 2011-06-10 | Service selecting frame based on cooperative filtration and QoS (Quality of Service) perception |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 201110155599 CN102306336A (en) | 2011-06-10 | 2011-06-10 | Service selecting frame based on cooperative filtration and QoS (Quality of Service) perception |
Publications (1)
Publication Number | Publication Date |
---|---|
CN102306336A true CN102306336A (en) | 2012-01-04 |
Family
ID=45380196
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN 201110155599 Pending CN102306336A (en) | 2011-06-10 | 2011-06-10 | Service selecting frame based on cooperative filtration and QoS (Quality of Service) perception |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102306336A (en) |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103793505A (en) * | 2014-01-27 | 2014-05-14 | 西安理工大学 | Network service collaborative filtering method based on user-service characteristics |
CN104778636A (en) * | 2015-04-08 | 2015-07-15 | 中国人民解放军海军工程大学 | Ship equipment maintenance support information service recommendation method |
CN105357043A (en) * | 2015-11-09 | 2016-02-24 | 南京邮电大学 | SLA system framework for automatic feedback adjustment in cloud computing environment and adjustment method |
WO2017161646A1 (en) * | 2016-03-23 | 2017-09-28 | 南京华苏科技有限公司 | Method for dynamically selecting optimal model by three-layer association for large data volume prediction |
CN109002903A (en) * | 2018-06-12 | 2018-12-14 | 昆明理工大学 | A kind of Optimization Scheduling of printed circuit board surface mounting line |
CN110232151A (en) * | 2019-05-22 | 2019-09-13 | 温州大学 | A kind of construction method of the QoS prediction model of mixing probability distribution detection |
CN110266539A (en) * | 2019-06-24 | 2019-09-20 | 公安部第一研究所 | A kind of Internet of Things service QoS prediction technique based on collaborative filtering |
CN110489223A (en) * | 2019-08-26 | 2019-11-22 | 北京邮电大学 | Method for scheduling task, device and electronic equipment in a kind of isomeric group |
WO2020063963A1 (en) * | 2018-09-30 | 2020-04-02 | 华为技术有限公司 | Communication method and related device |
RU2787009C2 (en) * | 2018-09-30 | 2022-12-28 | Хуавэй Текнолоджиз Ко., Лтд. | Communication method and related device |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101741650A (en) * | 2010-01-04 | 2010-06-16 | 北京航空航天大学 | QoS prediction-based service combination method and QoS prediction-based service combination device |
-
2011
- 2011-06-10 CN CN 201110155599 patent/CN102306336A/en active Pending
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101741650A (en) * | 2010-01-04 | 2010-06-16 | 北京航空航天大学 | QoS prediction-based service combination method and QoS prediction-based service combination device |
Cited By (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103793505A (en) * | 2014-01-27 | 2014-05-14 | 西安理工大学 | Network service collaborative filtering method based on user-service characteristics |
CN103793505B (en) * | 2014-01-27 | 2017-02-15 | 西安理工大学 | Network service collaborative filtering method based on user-service characteristics |
CN104778636A (en) * | 2015-04-08 | 2015-07-15 | 中国人民解放军海军工程大学 | Ship equipment maintenance support information service recommendation method |
CN104778636B (en) * | 2015-04-08 | 2017-11-07 | 中国人民解放军海军工程大学 | Method is recommended in the information service of apparel maintenance support |
CN105357043A (en) * | 2015-11-09 | 2016-02-24 | 南京邮电大学 | SLA system framework for automatic feedback adjustment in cloud computing environment and adjustment method |
CN105357043B (en) * | 2015-11-09 | 2018-06-05 | 南京邮电大学 | The SLA system frameworks and adjusting method that automatic feedback is adjusted under cloud computing environment |
WO2017161646A1 (en) * | 2016-03-23 | 2017-09-28 | 南京华苏科技有限公司 | Method for dynamically selecting optimal model by three-layer association for large data volume prediction |
CN109002903A (en) * | 2018-06-12 | 2018-12-14 | 昆明理工大学 | A kind of Optimization Scheduling of printed circuit board surface mounting line |
CN109002903B (en) * | 2018-06-12 | 2022-02-08 | 昆明理工大学 | Optimized scheduling method for printed circuit board surface mounting production line |
CN110972200A (en) * | 2018-09-30 | 2020-04-07 | 华为技术有限公司 | Communication method and related equipment |
WO2020063963A1 (en) * | 2018-09-30 | 2020-04-02 | 华为技术有限公司 | Communication method and related device |
RU2787009C2 (en) * | 2018-09-30 | 2022-12-28 | Хуавэй Текнолоджиз Ко., Лтд. | Communication method and related device |
US11646968B2 (en) | 2018-09-30 | 2023-05-09 | Huawei Technologies Co., Ltd. | Communication method and related device |
CN110972200B (en) * | 2018-09-30 | 2023-09-26 | 华为技术有限公司 | Communication method and related equipment |
CN110232151A (en) * | 2019-05-22 | 2019-09-13 | 温州大学 | A kind of construction method of the QoS prediction model of mixing probability distribution detection |
CN110232151B (en) * | 2019-05-22 | 2022-07-15 | 温州大学 | Construction method of QoS (quality of service) prediction model for mixed probability distribution detection |
CN110266539A (en) * | 2019-06-24 | 2019-09-20 | 公安部第一研究所 | A kind of Internet of Things service QoS prediction technique based on collaborative filtering |
CN110266539B (en) * | 2019-06-24 | 2022-03-01 | 公安部第一研究所 | Internet of things service QoS prediction method based on collaborative filtering |
CN110489223A (en) * | 2019-08-26 | 2019-11-22 | 北京邮电大学 | Method for scheduling task, device and electronic equipment in a kind of isomeric group |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102306336A (en) | Service selecting frame based on cooperative filtration and QoS (Quality of Service) perception | |
Wang et al. | Minimizing the age-of-critical-information: an imitation learning-based scheduling approach under partial observations | |
CN110533183A (en) | The model partition and task laying method of heterogeneous network perception in a kind of assembly line distribution deep learning | |
CN104735166A (en) | Skyline service selection method based on MapReduce and multi-target simulated annealing | |
CN104219319B (en) | A kind of distributed network flow self-organizing dispatching method | |
Lin et al. | AI-driven collaborative resource allocation for task execution in 6G-enabled massive IoT | |
Li et al. | Adaptive service function chaining mappings in 5G using deep Q-learning | |
Li et al. | Adaptive resource allocation based on the billing granularity in edge-cloud architecture | |
Ding et al. | Multiobjective optimization model for service node selection based on a tradeoff between quality of service and resource consumption in mobile crowd sensing | |
CN104954278A (en) | Bee colony optimization based network traffic scheduling method under multiple QoS (quality of service) constraints | |
CN109919532A (en) | Logistics node determination method and device | |
Jia et al. | An IoT service aggregation method based on dynamic planning for QoE restraints | |
Chen et al. | A hybrid task scheduling scheme for heterogeneous vehicular edge systems | |
CN115481748A (en) | Federal learning freshness optimization method and system based on digital twin assistance | |
Bahadur et al. | A novel method for optimizing energy consumption in wireless sensor network using genetic algorithm | |
CN114301935A (en) | Reputation-based method for selecting edge cloud collaborative federated learning nodes of Internet of things | |
CN115907038A (en) | Multivariate control decision-making method based on federated split learning framework | |
Seo et al. | Resource-efficient federated learning with non-iid data: An auction theoretic approach | |
Dangi et al. | 5G network traffic control: a temporal analysis and forecasting of cumulative network activity using machine learning and deep learning technologies | |
CN112650579B (en) | Cloud resource distribution scheduling method based on gridding | |
CN113596868A (en) | 5G network slice resource management mechanism based on SDN and NFV | |
Shi et al. | CoLEAP: Cooperative learning-based edge scheme with caching and prefetching for DASH video delivery | |
CN109711913A (en) | A kind of electric car supply and demand difference prediction technique and charging pile method of allocation plan | |
CN101616179B (en) | Dynamic member service selection method based on descending factors in service combination | |
CN109889573A (en) | Based on the Replica placement method of NGSA multiple target in mixed cloud |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20120104 |