CN110266539A - A kind of Internet of Things service QoS prediction technique based on collaborative filtering - Google Patents
A kind of Internet of Things service QoS prediction technique based on collaborative filtering Download PDFInfo
- Publication number
- CN110266539A CN110266539A CN201910550459.1A CN201910550459A CN110266539A CN 110266539 A CN110266539 A CN 110266539A CN 201910550459 A CN201910550459 A CN 201910550459A CN 110266539 A CN110266539 A CN 110266539A
- Authority
- CN
- China
- Prior art keywords
- service
- user
- qos
- similarity
- time
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/147—Network analysis or design for predicting network behaviour
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/50—Network service management, e.g. ensuring proper service fulfilment according to agreements
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/12—Protocols specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of, and the Internet of Things based on collaborative filtering services QoS prediction technique, first against k group (T1, T2..., Tk) QoS data calculates separately the similarity of user and service, further by combining k group similarity to obtain final user and servicing similarity.Most like TopN user and service are chosen, the collaborative forecasting method based on user and based on service is blended, final prediction obtains missing QoS.The method of the present invention is based on CF algorithm, introduces Time Perception mechanism, to realize that QoS is predicted.
Description
Technical field
The present invention relates to internet of things field, and in particular to a kind of Internet of Things service QoS prediction based on collaborative filtering
Method.
Background technique
Internet of Things (Internet of Things, IoT) Yao Shixian large-scale development, network complexity necessarily sharply increase
It is long.Services Oriented Achitecture (Service-oriented Architecture, SOA) is the building of such complication system
And application integration provides ideal solution.SOA is that the business logic modules that will be applied turn to different functional units (title
For service), by defining good interface and contract connects between these services.The appearance of Service-Oriented Architecture Based changes
Cooperation mode between traditional Internet of Things application, so that between different applications money can be shared by service discovery and combination
Source.
There are multiple service providers externally to provide service under actual Internet of Things Service-Oriented Architecture Based environment.Therefore, right
In identical function, there may be multiple services realizes.This has just drawn services selection (i.e. how in one group of alternative services
Select optimal service) the problem of.By introducing non-functional description --- service quality (Quality of Service,
QoS), the identical service of one group of function can be distinguished, i.e. the identical one group of service of function has respectively different QoS attributes, such as
Service price, service response time, service reliability etc..Service requester not only has its Functional Requirement also when requesting service
There is its qos requirement, the purpose of services selection is exactly selection while the optimal clothes for meeting requestor's Functional Requirement and qos requirement
Business.
However, same user's different time calls same service QoS to be possible to different, and different users calls same clothes
The QoS that business obtains also almost always is different.Therefore, the QoS for individually calculating each service is required for any user.It is real
In the situation of border, user only called sub-fraction service, what the QoS of many services was missing from, needed to carry out missing QoS data
Prediction.Missing QoS prediction usually has three kinds of matrix decomposition, time series forecasting and collaborative filtering methods.Wherein, collaborative filtering
Since its good performance is widely used.Jiang et al. points out that changing little QoS attribute between different user is calculating use
Lesser weight should be endowed when the similarity of family.Liu et al. people has found similar users that traditional collaborative filtering obtains not phase
Seemingly, and only because they, which happen to be in certain service invocation procedures, obtains similar QoS.Therefore, they consider user and
The location information of service, improves collaborative filtering.Ma et al. thinks similar two users, and similarity is once servicing
Front and back is called to remain unchanged.To obtain missing QoS by solving quadratic equation.However, scenes of internet of things lower network dynamic
Larger, QoS information has stronger timeliness, and the prior art is simply averaged history qos value, does not consider QoS particulate
The time attribute of degree causes QoS prediction accuracy lower.
Bibliography
[1]Y.Hu,Q.Peng,X.Hu,and R.Yang,“Web service recommendation based on
time series forecasting and collaborative filtering,”in Proceedings of IEEE
International Conference on Web Services.IEEE,2015,pp.233–240.
[2]Y.Koren,“Factor in the neighbors:Scalable and accurate
collaborative filtering,”ACM Transactions on Knowledge Discovery from Data,
vol.4,no.1,pp.1–24,2010.
[3]G.Adomavicius and A.Tuzhilin,“Toward the next generation of
recommender systems:A survey of the state-of-the-art and possible
extensions,”IEEE Transactions on Knowledge and Data Engineering,vol.17,no.6,
pp.734–749,2005.
[4]J.S.Breese,D.Heckerman,and C.Kadie,“Empirical analysis of
predictive algorithms for collaborative filtering,”in Proceedings of
International Conference on Uncertainty in Artificial Intelligence.ACM,1998,
pp.43–52.
[5]W.Lo,J.Yin,Y.Li,and Z.Wu,“Efficient web service qos prediction
using local neighborhood matrix factorization,”Engineering Applications of
Artificial Intelligence,vol.38,pp.14–23,2015.
[6]G.Xue,C.Lin,Q.Yang,W.Xi,H.Zeng,Y.Yu,and Z.Chen,“Scalable
collaborative filtering using cluster-based smoothing,”in Proceedings of ACM
Sigir Conference on Information Retrieval.ACM,2005,pp.114–121.
[7]B.Sarwar,G.Karypis,J.Konstan,and J.Riedl,“Item-based collaborative
filtering recommendation algorithms,”in Proceedings of IEEE International
Conference on World Wide Web.IEEE,2001,pp.285–295.
[8]J.Wang,A.P.Vries,and M.J.Reinders,“Unifying user-based and item-
based collaborative filtering approaches by similarity fusion,”in Proceedings
of ACM Sigir Conference on Information Retrieval.ACM,2006,pp.501–508.
[9]“Pearson correlation coefficient,”https://en.wikipedia.org/wiki/
Pearson correlation coefficient,2018.
[10]Y.Jiang,J.Liu,M.Tang,and X.Liu,“An effective web service
recommendation method based on personalized collaborative filtering,”in
Proceedings of IEEE International Conference on Web Services,2011,pp.211–218.
[11]J.Wu,L.Chen,Y.Feng,Z.Zheng,M.C.Zhou,and Z.Wu,“Predicting quality
of service for selection by neighborhood-based collaborative filtering,”IEEE
Transactions on Systems,Man,and Cybernetics:Systems,vol.43,no.2,pp.428–439,
2013.
[12]J.Liu,M.Tang,Z.Zheng,X.F.Liu,and S.Lyu,“Locationaware and
personalized collaborative filtering for web service recommendation,”IEEE
Transactions on Services Computing,vol.9,no.5,pp.686–699,2016.
[13]Y.Ma,S.Wang,P.C.K.Hung,C.Hsu,Q.Sun,and F.Yang,“A highly accurate
prediction algorithm for unknown web service qos values,”IEEE Transactions on
Services Computing,vol.9,no.4,pp.511–523,2016.
[14]B.Cavallo,M.D.Penta,and G.Canfora,“An empirical comparison of
methods to support qos-aware service selection,”in Proceedings of
International Workshop on Principles of Engineering Service-Oriented Systems,
2010,pp.64–70.
[15]C.Wu,W.Qiu,X.Wang,Z.Zheng,and X.Yang,“Time-aware and sparsity-
tolerant qos prediction based on collaborative filtering,”in Proceedings of
IEEE International Conference on Web Services.IEEE,2016,pp.637–640.
Summary of the invention
In view of the deficiencies of the prior art, the present invention is intended to provide a kind of Internet of Things service QoS prediction based on collaborative filtering
Method may be implemented to improve QoS prediction accuracy.
To achieve the goals above, the present invention adopts the following technical scheme:
A kind of Internet of Things service QoS prediction technique based on collaborative filtering, includes the following steps:
S1, the QoS data that each user calls each service in nearest T time is obtained by filtration;
S2, time T is divided into k group T1, T2..., Tk, the 1st group of time T1It is closest with current time;Every group of time
Span is T/k, then each element is that some user calls a certain all QoS numbers serviced in every group of time in QoS matrix
According to average value;
S3, T is predicted based on the QoS matrix in nearest T time1The missing qos value of time:
S3.1, the QoS prediction based on user:
3.1.1 phase between target user and the user of other each users) is calculated separately based on the QoS data in every group of time
Like degree;
3.1.2 it) for similarity between target user and the user of other each users, is calculated as follows respectively:
Similarity distributes weight, and the power based on distribution between k user is calculated for the QoS data based on the k group time
Value calculates similarity between the synthetic user between target user and corresponding user;
3.1.3) based on similarity between the synthetic user between the target user being calculated and other each users, selection
With the maximum top n user of target user's similarity, the QoS predicted value based on user is further obtained;
S3.2, the QoS prediction based on service:
3.2.1 the QoS data) based on every group of time calculates separately similar between destination service and the service of other each services
Degree;
3.2.2 it) for similarity between destination service and the service of each service, is calculated as follows respectively:
Similarity distributes weight between k service being calculated for the QoS data based on the k group time, and based on distribution
Weight computing obtains similarity between the integrated service between destination service and corresponding with service;
3.2.3 the QoS prediction based on service) is calculated:
Based on similarity between the service being calculated between destination service and other each services, selection and destination service se
The maximum preceding M service of similarity, further obtains the QoS predicted value based on service;
What the QoS predicted value based on user and step S3.2 being calculated in S3.3, combining step S3.1 were calculated
QoS predicted value based on service calculates the QoS predicted value based on collaborative filtering.
Further, in step S1, filtering calculation formula is as follows:
r(ui,sj)={ r (ui,sj,t)|tcurrent-T≤t≤tcurrent, i=1,2 ..., m, j=1,2 ..., n;
Wherein, r (ui,sj, t) and indicate user uiSometime t selection service s in T timejQoS data, r (ui,
sj) indicate user uiThe selection service s in T timejAll QoS datas;tcurrentIndicate current time, m is number of users, n
For quantity of service.
Further, in step S2, the calculation formula of each element of QoS matrix is as follows:
Wherein, NθFor the number of QoS data in the θ group time, tbAnd tuAt the time of indicating two endpoints of θ group time
Value.
Further, step 3.1.1) the specific method is as follows:
The similarity between target user and other each users is carried out as the following formula:
V=1,2 ..., m and e ≠ v;
Wherein, ueFor target user, uvFor other users,It indicates based on the QoS number in the θ group time
According to the user u being calculatedeWith user uvSimilarity;Sθ(ue)、Sθ(uv) respectively indicate user ueAnd user uvIn θ group
The set of the service of interior calling;r'θ(ue,sj) indicate target user ueService s is called within the θ group timejAll QoS numbers
According to average value, r'θ(uv,sj) difference user uvService s is called within the θ group timejAll QoS datas average value;
Respectively indicate user uv、ueThe set S for the service called within the θ group timeθ(ue)∩Sθ
(uv) corresponding QoS data average value;
N(Sθ(ue)∩Sθ(uv)) it is Sθ(ue)∩Sθ(uv) in contained service quantity.
Further, step 3.1.2) in, the power of similarity between the user that the QoS data in the θ group time is calculated
Value calculating method is as follows:
δ is parameter;
Based on the weight being calculated, similarity between the synthetic user between target user and each user is obtained:
V=1,2 ..., m and e ≠ v.
Further, step 3.1.3) in, the QoS predicted value based on user is calculated as follows:
Wherein, seFor destination service;SU(ue) indicate and user ueThe maximum preceding N user's set of similarity.
Further, step 3.2.1) in, the QoS data based on every group of time calculates separately destination service and other are each
The specific calculating process of similarity is as follows between service between service:
F=1,2 ..., n and e ≠ f;
Wherein, seFor destination service, sfIt is serviced for other,It indicates based on QoS data in the θ group time
Obtained destination service seWith service sfSimilarity;r'θ(ui,se) indicate user uiService s is called within the θ group timee
All QoS datas average value, r'θ(ui,sf) indicate user uiService s is called within the θ group timefAll QoS datas
Average value;Uθ(se) and Uθ(sf) be illustrated respectively in the θ group time and call service seAnd sfUser set;
It is illustrated respectively in user's set U in the θ group timeθ(se)∩Uθ(sf) call service sf、se
When QoS data average value:
N(Uθ(se)∩Uθ(sf)) it is set Uθ(se)∩Uθ(sf) in number of users.
Further, step 3.2.2) in, similarity distributes different weights between k service being calculated as follows:
δ is parameter;
Weight based on distribution, similarity is as follows between obtaining the service between destination service and other each services:
F=1,2 ..., n and e ≠ f.
Further, step 3.2.3) in, the QoS prediction based on service calculates as follows:
Wherein, SS (se) indicate and service seThe set of the maximum preceding M service of similarity.
Further, the specific calculating of step S3.3 carries out as the following formula:
Wherein,Indicate the QoS predicted value based on collaborative filtering,Indicate that the QoS based on user is pre-
Measured value,For the QoS predicted value based on service, λ is parameter factors, 0≤λ≤1.
The beneficial effects of the present invention are: the present invention considers the dynamic feature of QoS, introduces Time Perception mechanism, is based on
Collaborative filtering obtains really similar user or service, to improve QoS prediction accuracy.
Detailed description of the invention
Fig. 1 is the method flow schematic diagram of the embodiment of the present invention 1;
Fig. 2 is the QoS matrix schematic diagram in the embodiment of the present invention 1;
Fig. 3 is the network topology schematic diagram of the embodiment of the present invention 2.
Specific embodiment
Below with reference to attached drawing, the invention will be further described, it should be noted that the present embodiment is with this technology side
Premised on case, the detailed implementation method and specific operation process are given, but protection scope of the present invention is not limited to this reality
Apply example.
Embodiment 1
The present embodiment provides a kind of, and the Internet of Things based on collaborative filtering services QoS prediction technique, as shown in Figure 1, including such as
Lower step:
S1, the timeliness to ensure QoS data used in QoS prediction, are obtained by filtration each user in nearest T first
Interior QoS data;It is as follows to filter calculation formula:
r(ui,sj)={ r (ui,sj,t)|tcurrent-T≤t≤tcurrent, i=1,2 ..., m, j=1,2 ..., n;
Wherein, r (ui,sj, t) and indicate user uiSometime t selection service s in T timejQoS data, r (ui,
sj) indicate user uiThe selection service s in T timejAll QoS datas;tcurrentIndicate current time, m is number of users, n
For quantity of service;
S2, time T is divided into k group T1, T2..., Tk, the 1st group of time T1It is closest with current time;Every group of time
Span is T/k, then in QoS matrix each element be the QoS data in every group of time average value;Calculation formula is as follows:
Wherein, NθFor the number of QoS data in the θ group time, tbAnd tuAt the time of indicating two endpoints of θ group time
Value.
S3, as shown in Fig. 2, the QoS matrix that is namely based in nearest T time of QoS prediction carrys out predicted time T1Missing QoS
Value.
The QoS prediction technique proposed in the present embodiment is first against k group period (T1, T2..., Tk) QoS data difference
The similarity of user and service is calculated, further by combining k group similarity to obtain final user and service similarity.
Selection and the most like top n user of target user and service, by the collaborative forecasting method phase based on user and based on service
Fusion, final prediction obtain missing QoS.CF (Collaborative Filtering, collaborative filtering) is using emerging in simple terms
The hobby of the group congenial, that possess common experience of interest speculates the interested information of user, due to the preferable speed of CF and stalwartness
Property, it is widely used.The present embodiment method is based on CF algorithm, introduces Time Perception mechanism, to realize that QoS is predicted.
S3.1, the QoS prediction based on user
3.1.1) the similarity calculation of each group
History QoS data is accordingly divided into k group, is calculated separately based on the QoS data in every group of time similar between user
Degree:
V=1,2 ..., m and e ≠ v
Wherein, ueFor target user,Indicate the use being calculated based on the QoS data in the θ group time
Family ueWith user uvSimilarity;Sθ(ue)、Sθ(uv) respectively indicate user ueAnd user uvThe service called within the θ group time
Set;
Respectively indicate user uv、ueThe set S for the service called within the θ group timeθ(ue)∩Sθ
(uv) corresponding QoS data average value;
N(Sθ(ue)∩Sθ(uv)) it is Sθ(ue)∩Sθ(uv) in contained service quantity.
3.1.2) similarity calculation between synthetic user
Similarity has different confidence levels between the user that QoS data based on different groups (periods) is calculated, and is
This, needs to distribute different weights for similarity between k user being calculated.The calculating of weight considers two factors: (1) away from
User's similarity that the group closer from current time is calculated has higher confidence level;(2) clothes called jointly between user
Being engaged in, number is more, and similarity has higher confidence level between the user being thus calculated.QoS data in the θ group time calculates
The weight calculation method of similarity is as follows between obtained user:
δ is parameter, and default value is 2.
Based on the weight being calculated, similarity between synthetic user is obtained:
V=1,2 ..., m and e ≠ v;
3.1.3) QoS prediction
Based on similarity between target user and the synthetic user of each user is calculated, select and target user's similarity
Maximum top n user further obtains the QoS predicted value based on user:
Wherein, seFor destination service;SU(ue) indicate and user ueThe maximum preceding N user's set of similarity;
S3.2, the QoS prediction based on service
3.2.1 similarity calculation between) servicing:
QoS data based on every group of time calculates separately similarity between service:
F=1,2 ..., n and e ≠ f;
Wherein,Indicate the destination service s being calculated based on QoS data in the θ group timeeWith service sf
Similarity;Uθ(se) indicate to call service s within the θ group timeeUser set, Uθ(sf) indicate to adjust within the θ group time
With service sfUser set;
It is illustrated respectively in user's set U in the θ group timeθ(se)∩Uθ(sf) call service sf、se
When QoS data average value:
N(Uθ(se)∩Uθ(sf)) it is set Uθ(se)∩Uθ(sf) in number of users;
3.2.2) comprehensive similarity calculating
Similarity has different confidence levels between the service that QoS data based on different groups (periods) is calculated, and is
This, needs the similarity between be calculated k service to distribute different weights:
δ is parameter, and default value is 2.
Based on the weight being calculated, similarity between integrated service is obtained:
F=1,2 ..., n and e ≠ f;
3.2.3) QoS is predicted:
Based on the integrated service similarity between the target user and each user being calculated, selection and destination service se
The maximum preceding M service of similarity, further obtains the QoS predicted value based on service:
Wherein, SS (se) indicate and service seThe set of the maximum preceding M service of similarity;
S3.3, the QoS prediction based on collaborative filtering is calculated:
Wherein, λ is parameter factors, and 0≤λ≤1, λ are obtained according to real data training fitting, usual number of users it is more and
When quantity of service is less, λ > 0.5;Number of users is less and when quantity of service is less, λ < 0.5.
Embodiment 2
The present embodiment is by comparing existing method and 1 the method for embodiment so that Internet of Things services as an example, this reality
The network topology that example uses is applied as shown in figure 3, QoS matrix is as shown in table 1.
Table 1
As shown in table 1, U is shared1, U2, U3Three users and S1, S2, S3Three service, and the data in table indicate corresponding and use
The QoS of corresponding with service is called at family ,/indicate corresponding user's corresponding service of never call within the corresponding period.T1When being nearest
Between group, T2It is time group farther out, T3It is farthest time group.
1, in the existing method based on CF, QoS prediction is as follows:
In T3Time, user U1, U2, U3QoS set are as follows:
R(U1)={ 100,80,30 }
R(U2)={ 100,78,32 }
R(U3)={ 70,50,60 }
User U1With user U2Similarity are as follows:
User U1With user U3Similarity are as follows:
In the T2 time, congestion occurs for router R1, causes to increase by 30, in the T3 time, road by the service response time of R1
Continued to keep congestion by device R1, causes to increase by 60 by the service response time of R1.
In T1Time, user U1, U2, U3QoS gather change are as follows:
R(U1)={ 160,80,30 }
R(U2)={ 100,138,32 }
R(U3)={ 130,80,60 }
User U1With user U2Similarity are as follows:
User U1With user U3Similarity are as follows:
As shown in figure 3, user U1, U2It exists together in a local area network, it should similarity with higher, but due to existing side
Method does not account for time attribute, the U being calculated1With U2Similarity is lower, and U1With U3Similarity be but significantly increased, with reality
Border is not inconsistent.
2,1 method of embodiment considers that time attribute, QoS prediction technique are as follows:
In T3Time, user U1With user U2In T3Similarity be 0.999, user U1With user U3In T3Similarity be
0.229。
In T2Time, user U1, U2, U3QoS set are as follows:
R(U1)={ 130, ×, 30 }
R(U2)={ ×, 108,32 }
R(U3)={ ×, 80,60 }
User U1With user U2In T2Similarity be 0.5, user U1With user U3In T2Similarity be 0.5.
In T1Time, user U1, U2, U3QoS set are as follows:
R(U1)={ 160, ×, 30 }
R(U2)={ ×, 138,32 }
R(U3)={ 130, ×, 60 }
User U1With user U2In T1Similarity be 0.5, user U1With user U3In T1Similarity be 0.5.
Finally, user U1With user U2Weighted Similarity are as follows:
Sim(U1,U2)=0.999*0.73+0.5*0.15+0.5*0.12=0.864
User U1With user U3Weighted Similarity are as follows:
Sim(U1,U3)=0.229*0.73+0.5*0.15+0.5*0.12=0.302
Final result is more conform with reality.
For those skilled in the art, it can be provided various corresponding according to above technical solution and design
Change and modification, and all these change and modification, should be construed as being included within the scope of protection of the claims of the present invention.
Claims (10)
1. a kind of Internet of Things based on collaborative filtering services QoS prediction technique, which comprises the steps of:
S1, the QoS data that each user calls each service in nearest T time is obtained by filtration;
S2, time T is divided into k group T1, T2..., Tk, the 1st group of time T1It is closest with current time;Every group of time span is
T/k, then each element is that some user calls the flat of a certain all QoS datas serviced in every group of time in QoS matrix
Mean value;
S3, T is predicted based on the QoS matrix in nearest T time1The missing qos value of time:
S3.1, the QoS prediction based on user:
3.1.1 it) is calculated separately based on the QoS data in every group of time similar between target user and the user of other each users
Degree;
3.1.2 it) for similarity between target user and the user of other each users, is calculated as follows respectively:
Similarity distributes weight between k user is calculated for the QoS data based on the k group time, and based on the weight of distribution
Similarity between synthetic user between calculation target user and corresponding user;
3.1.3) based on similarity between the synthetic user between the target user being calculated and other each users, selection and mesh
The maximum top n user of user's similarity is marked, the QoS predicted value based on user is further obtained;
S3.2, the QoS prediction based on service:
3.2.1 the QoS data) based on every group of time calculates separately similarity between destination service and the service of other each services;
3.2.2 it) for similarity between destination service and the service of each service, is calculated as follows respectively:
Similarity distributes weight, and the weight based on distribution between k service being calculated for the QoS data based on the k group time
Similarity between the integrated service between destination service and corresponding with service is calculated;
3.2.3 the QoS prediction based on service) is calculated:
Based on similarity between the service being calculated between destination service and other each services, selection and destination service seIt is similar
Maximum preceding M service is spent, the QoS predicted value based on service is further obtained;
The QoS predicted value based on user and step S3.2 being calculated in S3.3, combining step S3.1 be calculated based on
The QoS predicted value of service calculates the QoS predicted value based on collaborative filtering.
2. the method according to claim 1, wherein filtering calculation formula is as follows in step S1:
r(ui,sj)={ r (ui,sj,t)|tcurrent-T≤t≤tcurrent, i=1,2 ..., m, j=1,2 ..., n;
Wherein, r (ui,sj, t) and indicate user uiSometime t selection service s in T timejQoS data, r (ui,sj) table
Show user uiThe selection service s in T timejAll QoS datas;tcurrentIndicate current time, m is number of users, and n is clothes
Business quantity.
3. according to the method described in claim 2, it is characterized in that, in step S2, the calculation formula of each element of QoS matrix
It is as follows:
Wherein, NθFor the number of QoS data in the θ group time, tbAnd tuIndicate value at the time of two endpoints of θ group time.
4. the method according to claim 1, wherein step 3.1.1) the specific method is as follows:
The similarity between target user and other each users is carried out as the following formula:
V=1,2 ..., m and e ≠ v;
Wherein, ueFor target user, uvFor other users,It indicates to calculate based on the QoS data in the θ group time
Obtained user ueWith user uvSimilarity;Sθ(ue)、Sθ(uv) respectively indicate user ueAnd user uvIt is adjusted within the θ group time
The set of service;r'θ(ue,sj) indicate target user ueService s is called within the θ group timejAll QoS datas it is flat
Mean value, r'θ(uv,sj) difference user uvService s is called within the θ group timejAll QoS datas average value;
Respectively indicate user uv、ueThe set S for the service called within the θ group timeθ(ue)∩Sθ(uv)
The average value of corresponding QoS data;
N(Sθ(ue)∩Sθ(uv)) it is Sθ(ue)∩Sθ(uv) in contained service quantity.
5. according to the method described in claim 4, it is characterized in that, step 3.1.2) in, the QoS data meter in the θ group time
The weight calculation method of similarity is as follows between obtained user:
δ is parameter;
Based on the weight being calculated, similarity between the synthetic user between target user and each user is obtained:
V=1,2 ..., m and e ≠ v.
6. according to the method described in claim 5, it is characterized in that, step 3.1.3) in, the QoS predicted value based on user is pressed
Formula calculates:
Wherein, seFor destination service;SU(ue) indicate and user ueThe maximum preceding N user's set of similarity.
7. the method according to claim 1, wherein step 3.2.1) in, the QoS data based on every group of time point
Not Ji Suan between the service between destination service and other each services similarity specific calculating process it is as follows:
F=1,2 ..., n and e ≠ f;
Wherein, seFor destination service, sfIt is serviced for other,It indicates to calculate based on QoS data in the θ group time
The destination service s arrivedeWith service sfSimilarity;r'θ(ui,se) indicate user uiService s is called within the θ group timeeInstitute
There are the average value of QoS data, r'θ(ui,sf) indicate user uiService s is called within the θ group timefAll QoS datas it is flat
Mean value;Uθ(se) and Uθ(sf) be illustrated respectively in the θ group time and call service seAnd sfUser set;
It is illustrated respectively in user's set U in the θ group timeθ(se)∩Uθ(sf) call service sf、seWhen
The average value of QoS data:
N(Uθ(se)∩Uθ(sf)) it is set Uθ(se)∩Uθ(sf) in number of users.
8. the method according to the description of claim 7 is characterized in that step 3.2.2) in, k be calculated as follows services
Between similarity distribute different weights:
δ is parameter;
Weight based on distribution, similarity is as follows between obtaining the service between destination service and other each services:
F=1,2 ..., n and e ≠ f.
9. according to the method described in claim 8, it is characterized in that, step 3.2.3) in, the QoS prediction based on service calculates such as
Under:
Wherein, SS (se) indicate and service seThe set of the maximum preceding M service of similarity.
10. the method according to claim 1, wherein the specific calculating of step S3.3 carries out as the following formula:
Wherein,Indicate the QoS predicted value based on collaborative filtering,Indicate the QoS predicted value based on user,For the QoS predicted value based on service, λ is parameter factors, 0≤λ≤1.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910550459.1A CN110266539B (en) | 2019-06-24 | 2019-06-24 | Internet of things service QoS prediction method based on collaborative filtering |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910550459.1A CN110266539B (en) | 2019-06-24 | 2019-06-24 | Internet of things service QoS prediction method based on collaborative filtering |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110266539A true CN110266539A (en) | 2019-09-20 |
CN110266539B CN110266539B (en) | 2022-03-01 |
Family
ID=67921014
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910550459.1A Active CN110266539B (en) | 2019-06-24 | 2019-06-24 | Internet of things service QoS prediction method based on collaborative filtering |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110266539B (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102130955A (en) * | 2011-03-18 | 2011-07-20 | 东北大学 | System and method for generating alternative service set of composite service based on collaborative filtering |
CN102170449A (en) * | 2011-04-28 | 2011-08-31 | 浙江大学 | Web service QoS prediction method based on collaborative filtering |
CN102306336A (en) * | 2011-06-10 | 2012-01-04 | 浙江大学 | Service selecting frame based on cooperative filtration and QoS (Quality of Service) perception |
EP3349500A1 (en) * | 2017-07-26 | 2018-07-18 | smartTECH GmbH | Enhancement of a cellular mobile network |
-
2019
- 2019-06-24 CN CN201910550459.1A patent/CN110266539B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102130955A (en) * | 2011-03-18 | 2011-07-20 | 东北大学 | System and method for generating alternative service set of composite service based on collaborative filtering |
CN102170449A (en) * | 2011-04-28 | 2011-08-31 | 浙江大学 | Web service QoS prediction method based on collaborative filtering |
CN102306336A (en) * | 2011-06-10 | 2012-01-04 | 浙江大学 | Service selecting frame based on cooperative filtration and QoS (Quality of Service) perception |
EP3349500A1 (en) * | 2017-07-26 | 2018-07-18 | smartTECH GmbH | Enhancement of a cellular mobile network |
Non-Patent Citations (1)
Title |
---|
陈唯: "基于协同过滤的Web服务QoS预测", 《科技创新与应用》 * |
Also Published As
Publication number | Publication date |
---|---|
CN110266539B (en) | 2022-03-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Yu et al. | A Web service QoS prediction approach based on time-and location-aware collaborative filtering | |
Wang et al. | Service composition in cyber-physical-social systems | |
Gao et al. | Online quality-aware incentive mechanism for mobile crowd sensing with extra bonus | |
Tran et al. | How geo-distributed data centers do demand response: A game-theoretic approach | |
Wu et al. | Predicting quality of service for selection by neighborhood-based collaborative filtering | |
Costa-Montenegro et al. | Which App? A recommender system of applications in markets: Implementation of the service for monitoring users’ interaction | |
CN104813284B (en) | Generic resource provider for cloud service | |
CN103108031B (en) | Cloud-edge topologies | |
Shen et al. | Optimal nonlinear pricing for a monopolistic network service provider with complete and incomplete information | |
Zhu et al. | Towards pricing for sensor-cloud | |
Yu et al. | Time-aware collaborative filtering for QoS-based service recommendation | |
Tran et al. | Reward-to-reduce: An incentive mechanism for economic demand response of colocation datacenters | |
Min et al. | An approach to resource and QoS-aware services optimal composition in the big service and Internet of Things | |
Huang et al. | Toward efficient data trading in AI enabled reconfigurable wireless sensor network using contract and game theories | |
CN105243080B (en) | The adaptive adaption system of resource perception based on ESB frameworks | |
Shi et al. | QoS prediction for smart service management and recommendation based on the location of mobile users | |
CN117273242B (en) | Virtual power plant management system and method based on block chain | |
CN102880725B (en) | Recommending method based on demand-based service organization | |
CN109086976B (en) | Task allocation method for crowd sensing | |
Zong et al. | A broker-assisting trust and reputation system based on artificial neural network | |
Faruk et al. | A genetic PSO algorithm with QoS-aware cluster cloud service composition | |
CN110266539A (en) | A kind of Internet of Things service QoS prediction technique based on collaborative filtering | |
Milocco et al. | Evaluating the upper bound of energy cost saving by proactive data center management | |
Li et al. | Cultural distance-aware service recommendation approach in mobile edge computing | |
Shao et al. | A scheduling algorithm for applications in a cloud computing system with communication changes |
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 |