CN109743727A - Big data collection method is perceived based on the efficient secret protection that mist calculates - Google Patents
Big data collection method is perceived based on the efficient secret protection that mist calculates Download PDFInfo
- Publication number
- CN109743727A CN109743727A CN201811596901.6A CN201811596901A CN109743727A CN 109743727 A CN109743727 A CN 109743727A CN 201811596901 A CN201811596901 A CN 201811596901A CN 109743727 A CN109743727 A CN 109743727A
- Authority
- CN
- China
- Prior art keywords
- data
- mist
- matrix
- node
- encryption
- 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
Abstract
The invention discloses a kind of efficient secret protections calculated based on mist to perceive big data collection method; mist by devising layering calculates auxiliary data and collects framework; calculating task is handled on local device or network edge device; to avoid long haul communication with cloud center, correlation provides effective support when sky to explore perception data.Meanwhile encryption method is disturbed by sampling, protect data-privacy from the infringement of listener-in and active attack person, this encryption method will not destroy the correlation of data, and simplify the decryption and reconstruction operation to encryption sampled data.Meanwhile the model of the mist node data tupe of design and observing matrix optimization, redundant data transmissions amount is greatly reduced, spatial coherence is effectively explored, it is ensured that data can accurately be rebuild.
Description
Technical field
The invention belongs to cordless communication networks, wireless sensor network technology field, and in particular to one kind is calculated based on mist
Efficient secret protection perceive big data collection method.
Background technique
For wireless sensor network (Wireless sensor network, WSN), ad hoc and Internet of Things
For (Internetofthings, IoT) application, data collection is indispensable.In general, the component in these networks is (such as
Sensor etc.) it is limitation by resource, therefore how to overcome these restraining factors, develop energy-efficient data collection and structure
Green communications network is built as a huge challenge.The scheme for realizing efficient data collection currently having proposed can substantially divide
At following three types: based on mathematical model, based on compressed sensing (Compressive sensing, CS) and based on query driven
Approximate data collection scheme.However, these traditional data handling procedures or structure, with being unable to satisfy under big data scene
The demand of efficient data collection and processing.Large-scale data is usually along with data collection, processing and the great burden of storage.Such as
It is a huge challenge that, which is alleviated these pressure, improves service quality,.
It is the new mould by executing data processing in network edge (local) equipment that mist, which calculates (to the supplement of cloud computing),
Type.It can mitigate the data processing load at cloud center, reduce telecommunication expense, significantly improve user experience.What mist calculated
Advantage allows to be applied to every field, proposes following three kinds of representational schemes thus:
The service-oriented mist computing architecture of first kind conceptual design is analyzed for remote health mining data, and is assessed
Different data is excavated and influence of the compress technique to data collection.But this kind of scheme is not studied specific data and is dug
Pick and compression method, and do not account for the highly important Privacy Protection for health data.
A kind of method of data capture calculated based on mobile mist of the second class conceptual design, in conjunction with mist structure design, minimum
Cost dispatch algorithm and routing algorithm reduce propagation delay time to the maximum extent, reduce transmission energy consumption.But the program has ignored
Using perception data sky when correlation, by reduce sensor read transmission quantity, further reduce energy consumption.
Third class scheme considers the temporal correlation of perception data, by CS theory and low complex degree sensing frame, network
Coding, energy consumption model and Clustering combine, and significantly reduce the communication overhead of data acquisition.Second class scheme and third class
Scheme has very big performance advantage, but the secret protection to sensitive data is had ignored in the transmission and treatment process of data.
Summary of the invention
In view of the above technical problems, the present invention devises a kind of efficient secret protection perception big data receipts calculated based on mist
Set method can reduce the transmission quantity of redundant data, reduce network data transmission energy consumption and improve the mist meter of data reconstruction precision
Calculate the efficient secret protection big data cognitive method of auxiliary.
The present invention solves its technical problem and is achieved through the following technical solutions:
A kind of efficient secret protection perception big data collection method calculated based on mist, is included the following steps:
(1) the perception mist computing architecture of design layering, the perception mist computing architecture is divided into from bottom to top mutually to be interconnected
Logical sensing layer, mist layer and client layer；
Sensing layer is made of K cluster, and each cluster includes L sensing node, and the sampling period of sensing node is N；
Mist layer corresponds to each other the mist node i carried out data transmission with cluster by K and forms；
(2) sensing node obtains compressed data in data sampling process, and is sent to after encrypting to compressed data
Mist layer；
(3) the mist node i in mist layer executes space compression operation, and mist node to the encryption data received from sensing layer
Obtained space compression data are sent client layer by i；
(4) client layer is rebuild and is decrypted to the space compression data being collected into from mist layer.
Further, step (2) specifically:
The initial data of sensing node p ∈ { 1,2 ..., L } its time dimension assuming that cluster i ∈ in { 1,2 ... K } is
Xi,p=[xi,p1xi,p2...xi,pN], wherein xi,pq, q ∈ { 1,2 ..., N } indicates in cluster i sensing node p in time slot q
Data；According to compression sampling theory, sensing node p can obtain compressed data Y in data sampling processi,p=Φ Xi,p=
Φ·Ψ·Θi,p, wherein Φ ∈ Rn×NFor observing matrix, Ψ ∈ RN×NFor the sparse basis of corresponding Φ, Θi,p∈RNFor sparse system
Number vector, wherein R indicates that set of real numbers, n indicate the size of data after compression sampling on time dimension, is a natural number；
For compressed data Yi,pIt carries out sampling disturbance encryption and obtains encryption data Y 'i,p, the sampling disturbance encryption method
It is as follows:
Wherein Y 'i,p∈RnIndicate encryption data,Indicate that key E and observing matrix Φ does the knot of matrix multiplication
Fruit, i.e.,The element of key E meets Gaussian Profile N (0,1/n), finally, sensing node p is by encryption data Y 'i,pHair
Mist node i is sent to be further processed.
Further, the encryption data Y ' received mist node i in step (3)i,pIt is arranged as matrix Y 'i=[Y 'i,1
Y′i,2 ... Y′i,L], i.e., each cluster has L sensing node, and Y ' ∈ Rn×L, mist node i is to the matrix Y ' receivediIt executes
Space compression operation, as follows:
Wherein,Indicate Y 'iTransposition, Φ ' ∈ Rl×LRepresentation space observing matrix, l are indicated by pressing on Spatial Dimension
The size of L after contracting sampling, Ψ ' ∈ RL×LIndicate the sparse basis of corresponding Φ ', Θ 'i,q∈RL, q ' ∈ 1,2 ..., and n } indicate empty
Between dimension sparse coefficient vector,Indicate the space compression data on n time slot.
Further, based on spatial observation matrix Φ ' and the sparse basis Ψ ' of corresponding Φ ' in step (3), pass through minimum
Change the correlation of matrix Φ ' and Ψ ' to construct following optimization problem:
Wherein, H and HμIt is to meet the matrix of above-mentioned condition, and habIndicate the element of H a row b column；
When H is fixed, the solution of above-mentioned optimization problem can be converted following format, i.e., optimal observing matrix
Φ′*Solve problems:By gradient descent method can in the hope of with sparse basis Ψ '
The minimum optimal observing matrix Φ ' of correlation*, thus the space compression data Y updatedi″TIt is sent to client layer.
It further, can be in the hope of the optimal observing matrix minimum with sparse basis Ψ ' correlation by gradient descent method
Φ′*, specifically: the objective function optimized will be needed to be defined as following format firstIt is then right
Φ ' seeks local derviation, i.e.,Then the iteration of observing matrix Φ ' is updated is defined as:
- 4 β Φ ' (t ', t) Ψ ' (Ψ ' of Φ ' (t ', t+1)=Φ ' (t ', t)TΦ′T(t′,t)Φ(t′,t)Ψ-H)Ψ′T,
In, t and t ' are the number of inner iterative and outer iteration respectively, and symbol beta indicates iteration step length, in order to reduce between Φ ' and Ψ '
Correlation, the upper limit of the off diagonal element of G is set, i.e., | gab|≤μ, μ are given threshold value, and the update of H is depending on μ's
Value, and the renewal process is described as follows:
Wherein, gabIt is G=Ψ 'TΦ′TThe element of a row b column of Φ ' Ψ '；
This method obtains optimal observing matrix Φ '*, the space compression data of mist node i are updated to Yi″T=Φ '*·Yi′T
=Φ '*·Ψ′·Θ′i。
Further, step (4) specifically:
Firstly, to rebuild data Y 'iIt need to solve following optimization problem:
Wherein,Representation space compressed data Yi″TThe q ' column, time slot q ' ∈ { 1,2 ..., n }, due to vector
Θ′i,q′It is sparse, and space compression data Yi″TWith Φ '*Ψ ' is known, therefore can use CS algorithm for reconstructing
Reconstructed coefficients matrix Θ 'i；Above-mentioned optimization problem has been solved, Θ ' is obtainediApproximationWhen, Y ' is obtained by following equalitiesi
Approximation
Wherein Θ 'iFor sparse coefficient vector Θ 'i,q′By column composition matrix,Indicate that all encryptions are adopted in cluster i
The reconstruction data of sample data rebuild data in exportAfterwards, it is decrypted and rebuilds initial data Xi,p, following optimize need to be solved
Problem:
Wherein,Due to coefficient vector Θi,pIt is sparse, and key E, observing matrix
Φ, corresponding Φ sparse basis Ψ andBe it is known, can use compressed sensing reconstruction algorithm and solve approximation coefficient vector
The invention has the benefit that
(1) the hierarchical perception mist computing architecture that the present invention designs, avoids the length between sensing node and cloud server
Distance communication effectively reduces communication energy consumption.In addition, compressing original perception data respectively from time and Spatial Dimension, reduce
The transmission of redundant data, further reduced communication overhead；
(2) the sampling disturbance encryption method that the present invention designs, realizes secret protection, and will not destroy phase between data
Guan Xing；
(3) the mist node data tupe that the present invention designs, significantly reduces redundant data transmissions amount.The observation of building
Matrix optimizing model ensure that the successful reconstitution of initial data, and further improve the reconstruction precision of data.
Detailed description of the invention
Fig. 1 is that the individual-layer data for calculating auxiliary for the mist under big data sensing network that the present invention designs collects framework；
Fig. 2 is that the comparison between sampling disturbance encryption method institute sampled data and original sampling data is utilized in the present invention；
Fig. 3 is the comparison of data reconstruction result between user and attacker in the present invention；
Fig. 4 is design method of the present invention compared with other two schemes are on opposite reconstruction error, wherein Privacy
Preserving data collection indicates method proposed by the invention.
Specific embodiment
Below by specific embodiment, the invention will be further described, and it is not limit that following embodiment, which is descriptive,
Qualitatively, this does not limit the scope of protection of the present invention.
As shown in Figure 1, a kind of efficient secret protection calculated based on mist perceives big data collection method, including walk as follows
It is rapid:
(1) the perception mist computing architecture of design layering, perception mist computing architecture are divided into interconnected from bottom to top
Sensing layer, mist layer and client layer.
Sensing layer is made of various awareness apparatus, for monitoring specified region.According to the location information of awareness apparatus, by it
It is divided into multiple clusters.I.e. sensing layer is made of K cluster, and each cluster includes L sensing node, and the sampling period of sensing node is N.
Awareness apparatus carries out disturbance encryption sampling to data, the squeeze operation of initial data is realized, later, by the sampling data transmitting of encryption
It is sent to the corresponding node of mist layer, i.e. mist node and sensing node belongs to the same cluster, so mist layer is mutually right with cluster by K
The mist node i composition that should carry out data transmission.
Each cluster in mist layer can dispose mist node, and mist node serves as following two role.1) data processing: when mist node
When receiving data harvesting request from the user in local user or other regions, it can be acquired from local sensing node encrypted
Sampled data, and data sended over using the spatial observation matrix compression optimized from local sensing node.Finally, will
Encrypted compressed data is sent to user terminal or next mist node.2) data relay: when some mist node is from other mists
It, can be by the data forwarding received to next mist node, until data reach and request user when node receives data
Belong to the mist node of the same area.
Client layer is made of a large number of users terminal in network monitoring region.User terminal is connected with mist layer, has one
Fixed computing capability.If user wants to obtain perception data from self zone or near zone, request is committed to local
Mist node.Later, mist node will provide corresponding data service.Due to the data received from mist node be it is encrypted, because
(reconstruction) operation need to be decrypted in this user by the ciphertext that the key pair shared with awareness apparatus receives.
(2) assume the initial data of sensing node p ∈ { 1,2 ..., L } its time dimension in cluster i ∈ { 1,2 ... K }
For Xi,p=[xi,p1 xi,p2 ... xi,pN], wherein xi,pq, q ∈ { 1,2 ..., N } be expressed as in cluster i sensing node p when
Data when gap q.According to compression sampling theory, sensing node p can obtain compressed data Y in data sampling processi,p=
Φ·Xi,p=Φ Ψ Θi,p, wherein Φ ∈ Rn×NFor observing matrix, Ψ ∈ RN×NFor the sparse basis of corresponding Φ, Θi,p∈RN
For sparse coefficient vector, wherein R indicates that set of real numbers, n indicate the size of data after compression sampling on time dimension, is one
A natural number.For the privacy of the sensitive data of guarantee user's request, i.e. the confidentiality of data, the present invention devises a kind of sampling
Encryption method is disturbed, to prevent the acquisition of listener-in and active attack person to data.For compressed data Yi,pCarry out sampling disturbance
Encryption obtains encryption data Y 'i,p, it is as follows which disturbs encryption method:
Wherein Y 'i,p∈RnIt indicates encryption data (i.e. ciphertext),Indicate that key E and observing matrix Φ does Matrix Multiplication
Method as a result, i.e.The element of key E meets Gaussian Profile N (0,1/n), finally, sensing node p is by encryption data
Y′i,pMist node i is sent to be further processed.
(3) mist node i is introduced into network with reduction and cloud long haul communication energy consumption, and by mist node
The redundant data transmissions amount that data are handled to be further reduced between mist node and user.According to adding for sensing node p
Close sampled result, the encryption data Y ' that mist node i is receivedi,pIt is arranged as Y 'i=[Y 'i,1 Y′i,2 ... Y′i,L], i.e., often
A cluster i has L sensing node, and Y ' ∈ Rn×L.In order to be further reduced redundant data transmissions amount, mist node i is to receiving
Data Y 'iSpace compression operation is executed, as follows:
Wherein,Indicate Y 'iTransposition, Φ ' ∈ Rl×LRepresentation space observing matrix, l are indicated by pressing on Spatial Dimension
The size of L, Ψ ' ∈ R after contracting samplingL×LIndicate the sparse basis of corresponding Φ ', Θ 'i,q∈RL, q ' ∈ 1,2 ..., and n } representation space
The sparse coefficient vector of dimension,Indicate the space compression data on n time slot.
The height of data reconstruction precision is wanted since the spatial observation matrix Φ ' of above-mentioned setting is not able to satisfy partial task
It asks, the present invention constructs and solved an observing matrix optimization problem thus.Based on spatial observation matrix Φ ' with corresponding Φ's '
Sparse basis Ψ ' constructs following optimization problem by minimizing the correlation of matrix Φ ' and Ψ ':
Wherein, H and HμIt is to meet the matrix of above-mentioned condition, and habIndicate the element of H a row b column；
When H is fixed, the solution of above-mentioned optimization problem can be converted following format, i.e., optimal observing matrix
Φ′*Solve problems:By gradient descent method can in the hope of with sparse basis Ψ '
The minimum optimal observing matrix Φ ' of correlation*, thus the space compression data Y updatedi″TIt is sent to client layer.
The objective function optimized will be needed to be defined as following format firstThen to Φ '
Local derviation is sought, i.e.,Then the iteration of observing matrix Φ ' is updated is defined as: Φ '
- 4 β Φ ' (t ', t) Ψ ' (Ψ ' of (t ', t+1)=Φ ' (t ', t)TΦ′T(t′,t)Φ(t′,t)Ψ-H)Ψ′T, wherein t
It is the number of inner iterative and outer iteration respectively with t ', symbol beta indicates iteration step length.In order to reduce the phase between Φ ' and Ψ '
The upper limit of the off diagonal element of G is arranged in Guan Xing, i.e., | gab|≤μ, μ are given threshold value, and the update of H depends on the value of μ, and
The renewal process is described as follows:
Wherein, gabIt is G=Ψ 'TΦ′TThe element of a row b column of Φ ' Ψ '.
This method obtains optimal observing matrix Φ '*, the space compression data of mist node i are updated to Yi″T=Φ '*·Yi′T
=Φ '*·Ψ′·Θ′i, mist node i is by the space compression data Y of updatei″TIt is sent to client layer.
(4) for user convenience from the data of mist node i, the encryption data for needing to be collected into user carries out weight
It builds and decrypts.Firstly, to rebuild data Y 'iIt need to solve following optimization problem:
Wherein Θ 'iFor sparse coefficient vector Θ 'i,q′By column composition matrix,Representation space compressed data Yi″T
The q ' column, time slot q ' ∈ { 1,2 ..., n }, due to vector theta 'i,q′It is sparse, and space compression data Yi″TWith
Φ′*Ψ ' is known, therefore can use CS algorithm for reconstructing reconstructed coefficients matrix Θ 'i；Above-mentioned optimization problem has been solved, has been obtained
To Θ 'iApproximationWhen, Y ' is obtained by following equalitiesiApproximation
Wherein Θ 'iFor sparse coefficient vector Θ 'i,q′By column composition matrix,Indicate that all encryptions are adopted in cluster i
The reconstruction data of sample data rebuild data in exportAfterwards, it is decrypted and rebuilds initial data Xi,p, following optimize need to be solved
Problem:
Wherein,Due to coefficient vector Θi,pIt is sparse, and key E, observing matrix
Φ, sparse basis Ψ andBe it is known, can use suitable compressed sensing reconstruction algorithm and solve approximation coefficient vector
For current research scheme that there are energy consumptions is high, Privacy Safeguarding is low and temporal correlation there are the performance deficiencies such as error
Problem, we incorporate encryption sampling, mist calculates and CS is theoretical, and the invention discloses a kind of efficient privacy guarantors calculated based on mist
Shield perception big data collection method.Mist by devising layering calculates auxiliary data and collects framework, so that calculating task can be with
It is handled on local device or network edge device, so that the long haul communication with cloud center is avoided, to explore perception data
Sky when correlation provide effective support.Meanwhile encryption method is disturbed by sampling, protect data-privacy from listener-in
With the infringement of active attack person, this encryption method will not destroy the correlation of data, and simplify to encryption sampled data
Decryption and reconstruction operation.Meanwhile the model of the mist node data tupe of design and observing matrix optimization, it greatly reduces
Redundant data transmissions amount, effectively explores spatial coherence, it is ensured that data can accurately be rebuild.
The application of the efficient secret protection perception big data collection method that mist calculates in practice, as shown in Fig. 2, of the invention
Design method shows: existing between encryption data and original sampling data compared with large disturbances, it means that eavesdropping can effectively be defendd to attack
It hits；Encryption data and initial data keep like attribute, it means that it is in plain text or close that the data obtained, which cannot be distinguished, in listener-in
Text.Therefore, listener-in is obscured while this programme can provide the secret protection of data.In addition, as shown in figure 3, institute of the present invention
The reconstruction data and initial data of user terminal are several in the efficient secret protection perception big data collection method that the mist of design calculates
Consistent, i.e., reconstruction precision is higher；And the reconstructed results rapid fluctuation of attacker, and differ greatly with initial data.This meaning
Taste, the sampling disturbance encryption method of this programme design can also preferably protect data while will not destroying data dependence
Privacy.Moreover, as shown in figure 4, this programme (Privacy preserving data collection) and remaining two
Kind scheme (Clustered compressive data collection, Big spatio-temporal
Datacollection it) compares, there is lower Relative reconstruction error, while the application effect of data is influenced smaller.
The above is only a preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications are also answered
It is considered as protection scope of the present invention.
Claims (6)
1. a kind of efficient secret protection calculated based on mist perceives big data collection method, characterized by the following steps:
(1) the perception mist computing architecture of design layering, the perception mist computing architecture is divided into interconnected from bottom to top
Sensing layer, mist layer and client layer；
Sensing layer is made of K cluster, and each cluster includes L sensing node, and the sampling period of sensing node is N；
Mist layer corresponds to each other the mist node i carried out data transmission with cluster by K and forms；
(2) sensing node obtains compressed data in data sampling process, and mist layer is sent to after encrypting to compressed data；
(3) the mist node i in mist layer executes space compression operation to the encryption data received from sensing layer, and mist node i will
Obtained space compression data are sent to client layer；
(4) client layer is rebuild and is decrypted to the space compression data being collected into from mist layer.
2. a kind of efficient secret protection calculated based on mist as described in claim 1 perceives big data collection method, feature
It is: step (2) specifically:
The initial data of sensing node p ∈ { 1,2 ..., L } its time dimension assuming that cluster i ∈ in { 1,2 ... K } is Xi,p=
[xi,p1xi,p2…xi,pN], wherein xi,pq, q ∈ 1,2 ..., and N } indicate data of the sensing node p in time slot q in cluster i；Root
According to compression sampling theory, sensing node p can obtain compressed data Y in data sampling processi,p=Φ Xi,p=Φ
Ψ·Θi,p, wherein Φ ∈ Rn×NFor observing matrix, Ψ ∈ RN×NFor the sparse basis of corresponding Φ, Θi,p∈RNFor sparse coefficient to
Amount, wherein R indicates that set of real numbers, n indicate the size of data after compression sampling on time dimension, is a natural number；
For compressed data Yi,pIt carries out sampling disturbance encryption and obtains encryption data Y 'i,p, it is as follows that the sampling disturbs encryption method:
Wherein Y 'i,p∈RnIndicate encryption data,Indicate key E and observing matrix Φ do matrix multiplication as a result, i.e.The element of key E meets Gaussian Profile N (0,1/n), finally, sensing node p is by encryption data Y 'i,pIt is sent to mist
Node i is to be further processed.
3. a kind of efficient secret protection calculated based on mist as claimed in claim 2 perceives big data collection method, feature
It is: the encryption data Y ' for receiving mist node i in step (3)i,pIt is arranged as matrix Yi'=[Y 'i,1 Y′i,2 … Y
′i,L], i.e., each cluster has L sensing node, and Y ' ∈ Rn×L, mist node i is to the matrix Y receivedi' execute space compression behaviour
Make, as follows:
Yi″T=Φ ' Yi′T
=Φ ' [Y 'i,1 Y′i,2 … Y′i,L]T
=Φ ' Ψ ' Θ 'i
=Φ ' Ψ ' [Θ 'i,1 Θ′i,2 … Θ′i,n],
Wherein, Yi′TIndicate Yi' transposition, Φ ' ∈ Rl×LRepresentation space observing matrix, l indicate to adopt by compressing on Spatial Dimension
The size of L after sample, Ψ ' ∈ RL×LIndicate the sparse basis of corresponding Φ ', Θ 'i,q∈RL, q ' ∈ 1,2 ..., and n } representation space dimension
The sparse coefficient vector of degree, Yi″T∈Rl×nIndicate the space compression data on n time slot.
4. a kind of efficient secret protection calculated based on mist as claimed in claim 3 perceives big data collection method, feature
Be: based on spatial observation matrix Φ ' and the sparse basis Ψ ' of corresponding Φ ' in step (3), by minimum matrix Φ ' and
The correlation of Ψ ' constructs following optimization problem:
Wherein, H and HμIt is to meet the matrix of above-mentioned condition, and habIndicate the element of H a row b column；
When H is fixed, the solution of above-mentioned optimization problem can be converted following format, i.e., optimal observing matrix Φ '*'s
Solve problems:By gradient descent method can in the hope of with sparse basis Ψ ' correlation
Minimum optimal observing matrix Φ '*, thus the space compression data Y updatedi″TIt is sent to client layer.
5. a kind of efficient secret protection calculated based on mist as claimed in claim 4 perceives big data collection method, feature
It is: can be in the hope of the optimal observing matrix Φ ' minimum with sparse basis Ψ ' correlation by gradient descent method*, specifically: it is first
First the objective function optimized will be needed to be defined as following formatLocal derviation then is asked to Φ ', i.e.,Then the iteration of observing matrix Φ ' is updated is defined as: Φ ' (t ', t+1)
- 4 β Φ ' (t ', t) Ψ ' (Ψ ' of=Φ ' (t ', t)TΦ′T(t′,t)Φ(t′,t)Ψ-H)Ψ′T, wherein t and t ' is respectively
It is the number of inner iterative and outer iteration, symbol beta indicates iteration step length, in order to reduce the correlation between Φ ' and Ψ ', if
Set the upper limit of the off diagonal element of G, i.e., | gab|≤μ, μ are given threshold value, and the update of H depends on the value of μ, and this is updated
Journey is described as follows:
Wherein, gabIt is G=Ψ 'TΦ′TThe element of a row b column of Φ ' Ψ '；
This method obtains optimal observing matrix Φ '*, the space compression data of mist node i are updated to Yi″T=Φ '*·Yi′T=
Φ′*·Ψ′·Θ′i。
6. a kind of efficient secret protection calculated based on mist as claimed in claim 5 perceives big data collection method, feature
It is: step (4) specifically:
Firstly, to rebuild data Yi' following optimization problem need to be solved:
Wherein,Representation space compressed data Yi″TThe q ' column, time slot q ' ∈ { 1,2 ..., n }, due to vector
Θ′i,q′It is sparse, and space compression data Yi″TWith Φ '*Ψ ' is known, therefore can use CS algorithm for reconstructing
Reconstructed coefficients matrix Θ 'i；Above-mentioned optimization problem has been solved, Θ ' is obtainediApproximationWhen, Y is obtained by following equalitiesi′
Approximation
Wherein Θ 'iFor sparse coefficient vector Θ 'i,q′By column composition matrix,Indicate all encryption hits in cluster i
According to reconstruction data, export rebuild dataAfterwards, it is decrypted and rebuilds initial data Xi,p, following optimization problem need to be solved:
Wherein,Due to coefficient vector Θi,pIt is sparse, and key E, observing matrix Φ, right
Answer Φ sparse basis Ψ andBe it is known, can use compressed sensing reconstruction algorithm and solve approximation coefficient vector
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811596901.6A CN109743727B (en) | 2018-12-26 | 2018-12-26 | Efficient privacy protection perception big data collection method based on fog calculation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811596901.6A CN109743727B (en) | 2018-12-26 | 2018-12-26 | Efficient privacy protection perception big data collection method based on fog calculation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109743727A true CN109743727A (en) | 2019-05-10 |
| CN109743727B CN109743727B (en) | 2022-06-10 |
Family
ID=66361238
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811596901.6A Active CN109743727B (en) | 2018-12-26 | 2018-12-26 | Efficient privacy protection perception big data collection method based on fog calculation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109743727B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110300159A (en) * | 2019-06-10 | 2019-10-01 | 华侨大学 | A kind of sensing cloud data safety low cost storage method based on edge calculations |
| CN110519776A (en) * | 2019-08-07 | 2019-11-29 | 东南大学 | Balanced cluster and federated resource distribution method in a kind of mist computing system |
| CN110839028A (en) * | 2019-11-14 | 2020-02-25 | 南京邮电大学 | Privacy protection method for fog-assisted industrial Internet of things |
| CN111835830A (en) * | 2020-06-16 | 2020-10-27 | 北京邮电大学 | Data perception system, method and device |
| CN111954236A (en) * | 2020-07-27 | 2020-11-17 | 河海大学 | Hierarchical edge calculation unloading method based on priority |
| CN113033026A (en) * | 2021-04-26 | 2021-06-25 | 上海交通大学 | Industrial edge sensing method with observability guarantee |
| CN113742779A (en) * | 2021-09-18 | 2021-12-03 | 湖北工业大学 | Service customization system and method with privacy protection function |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130070624A1 (en) * | 2010-01-11 | 2013-03-21 | Research In Motion Limited | Sensor-based wireless communication systems using compressed sensing with sparse data |
| CN105025498A (en) * | 2015-06-08 | 2015-11-04 | 南京邮电大学 | A sensing network clustering type space time compression method based on network coding and compression sensing |
| CN107786959A (en) * | 2017-10-18 | 2018-03-09 | 北京京航计算通讯研究所 | The compressed data collection method based on adaptive measuring in wireless sensor network |
-
2018
- 2018-12-26 CN CN201811596901.6A patent/CN109743727B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130070624A1 (en) * | 2010-01-11 | 2013-03-21 | Research In Motion Limited | Sensor-based wireless communication systems using compressed sensing with sparse data |
| CN105025498A (en) * | 2015-06-08 | 2015-11-04 | 南京邮电大学 | A sensing network clustering type space time compression method based on network coding and compression sensing |
| CN107786959A (en) * | 2017-10-18 | 2018-03-09 | 北京京航计算通讯研究所 | The compressed data collection method based on adaptive measuring in wireless sensor network |
Non-Patent Citations (2)
| Title |
|---|
| SIGUANG CHEN ET AL.: "Fog Computing Assisted Efficient Privacy Preserving Data Collection for Big Sensory Data", 《2018 IEEE GLOBAL COMMUNICATIONS CONFERENCE (GLOBECOM)》 * |
| SIGUANG CHEN ET AL.: "Fog Computing based Optimized Compressive Data Collection for Big Sensory Data", 《2018 IEEE INTERNATIONAL CONFERENCE ON COMMUNICATIONS (ICC)》 * |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110300159B (en) * | 2019-06-10 | 2021-08-31 | 华侨大学 | Sensing cloud data safe low-cost storage method based on edge computing |
| CN110300159A (en) * | 2019-06-10 | 2019-10-01 | 华侨大学 | A kind of sensing cloud data safety low cost storage method based on edge calculations |
| CN110519776A (en) * | 2019-08-07 | 2019-11-29 | 东南大学 | Balanced cluster and federated resource distribution method in a kind of mist computing system |
| CN110519776B (en) * | 2019-08-07 | 2021-09-17 | 东南大学 | Balanced clustering and joint resource allocation method in fog computing system |
| CN110839028A (en) * | 2019-11-14 | 2020-02-25 | 南京邮电大学 | Privacy protection method for fog-assisted industrial Internet of things |
| CN110839028B (en) * | 2019-11-14 | 2022-04-05 | 南京邮电大学 | Privacy protection method for fog-assisted industrial Internet of things |
| CN111835830A (en) * | 2020-06-16 | 2020-10-27 | 北京邮电大学 | Data perception system, method and device |
| CN111835830B (en) * | 2020-06-16 | 2022-06-21 | 北京邮电大学 | Data perception system, method and device |
| CN111954236B (en) * | 2020-07-27 | 2021-11-09 | 河海大学 | Hierarchical edge calculation unloading method based on priority |
| CN111954236A (en) * | 2020-07-27 | 2020-11-17 | 河海大学 | Hierarchical edge calculation unloading method based on priority |
| CN113033026A (en) * | 2021-04-26 | 2021-06-25 | 上海交通大学 | Industrial edge sensing method with observability guarantee |
| CN113033026B (en) * | 2021-04-26 | 2022-03-15 | 上海交通大学 | Industrial edge sensing method with observability guarantee |
| CN113742779A (en) * | 2021-09-18 | 2021-12-03 | 湖北工业大学 | Service customization system and method with privacy protection function |
| CN113742779B (en) * | 2021-09-18 | 2024-03-22 | 湖北工业大学 | Service customization system and method with privacy protection function |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109743727B (en) | 2022-06-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109743727A (en) | Big data collection method is perceived based on the efficient secret protection that mist calculates | |
| Chen et al. | Efficient privacy preserving data collection and computation offloading for fog-assisted IoT | |
| CN105812126A (en) | Lightweight back-up and efficient restoration method of health block chain data encryption keys | |
| Wang et al. | Achieve privacy-preserving priority classification on patient health data in remote eHealthcare system | |
| Liu et al. | Mobile-edge computing framework with data compression for wireless network in energy internet | |
| CN103259787B (en) | Safety interactive method of cloud and terminal based on JSON protocol packet in medical field | |
| KR20150115778A (en) | Privacy against interference attack for large data | |
| CN112600697B (en) | QoS prediction method and system based on federal learning, client and server | |
| Song et al. | Autoregressive integrated moving average model–based secure data aggregation for wireless sensor networks | |
| CN115766159A (en) | Private data processing method and device and electronic equipment | |
| CN110300159A (en) | A kind of sensing cloud data safety low cost storage method based on edge calculations | |
| CN108566369A (en) | Data collecting system based on industrial big data and method | |
| Wei et al. | Image compression scheme based on PCA for wireless multimedia sensor networks | |
| Ifzarne et al. | Secure data collection for wireless sensor network | |
| Kuldeep et al. | Multi-class privacy-preserving cloud computing based on compressive sensing for IoT | |
| CN110855512A (en) | Ultra-large-scale DPI data processing system based on edge calculation | |
| Chen et al. | Fog computing based optimized compressive data collection for big sensory data | |
| CN106470212A (en) | A kind of method based on LZW compression algorithm, EIGRP protocol massages being compressed and encrypting | |
| Chen et al. | Fog computing assisted efficient privacy preserving data collection for big sensory data | |
| Sukanya et al. | Integration of wireless sensor networks and mobile cloud-A survey | |
| CN115134080B (en) | Data transmission method and device based on security encryption chip | |
| CN106792876A (en) | End to end network perception evaluating method and system | |
| Yu et al. | Verifiable, privacy-assured, and accurate signal collection for cloud-assisted wireless sensor networks | |
| Hung et al. | Privacy-preserving data collection and recovery of compressive sensing | |
| CN103327363A (en) | System and method for realizing control over video information encryption on basis of semantic granularity |
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 |