CN106341842A - Method and device for transmitting data of wireless sensor network - Google Patents
Method and device for transmitting data of wireless sensor network Download PDFInfo
- Publication number
- CN106341842A CN106341842A CN201610712279.5A CN201610712279A CN106341842A CN 106341842 A CN106341842 A CN 106341842A CN 201610712279 A CN201610712279 A CN 201610712279A CN 106341842 A CN106341842 A CN 106341842A
- Authority
- CN
- China
- Prior art keywords
- data
- cluster
- node
- prediction model
- matrix
- 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
- 238000000034 method Methods 0.000 title claims abstract description 79
- 230000005540 biological transmission Effects 0.000 claims abstract description 30
- 239000011159 matrix material Substances 0.000 claims description 66
- 238000004364 calculation method Methods 0.000 claims description 40
- 238000003860 storage Methods 0.000 claims description 38
- 230000008569 process Effects 0.000 claims description 36
- 241000854291 Dianthus carthusianorum Species 0.000 claims description 27
- 238000007906 compression Methods 0.000 claims description 21
- 230000006835 compression Effects 0.000 claims description 21
- 230000008447 perception Effects 0.000 claims description 19
- 230000002776 aggregation Effects 0.000 claims description 14
- 238000004220 aggregation Methods 0.000 claims description 14
- 238000005070 sampling Methods 0.000 claims description 7
- 238000013500 data storage Methods 0.000 claims description 6
- 238000000354 decomposition reaction Methods 0.000 claims description 6
- 238000006116 polymerization reaction Methods 0.000 claims description 5
- 230000009466 transformation Effects 0.000 claims description 5
- 235000013399 edible fruits Nutrition 0.000 claims 2
- 230000005094 fruit set Effects 0.000 claims 1
- 238000005265 energy consumption Methods 0.000 abstract description 7
- 230000002401 inhibitory effect Effects 0.000 abstract 1
- 230000002035 prolonged effect Effects 0.000 abstract 1
- 238000012546 transfer Methods 0.000 description 13
- 238000012545 processing Methods 0.000 description 7
- 238000004422 calculation algorithm Methods 0.000 description 6
- 230000007246 mechanism Effects 0.000 description 6
- 238000004891 communication Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 230000007613 environmental effect Effects 0.000 description 4
- 230000006872 improvement Effects 0.000 description 4
- 238000005259 measurement Methods 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 4
- 238000011084 recovery Methods 0.000 description 4
- 230000004913 activation Effects 0.000 description 3
- 230000003044 adaptive effect Effects 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000009826 distribution Methods 0.000 description 3
- 241001269238 Data Species 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- 238000012417 linear regression Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 108700020469 14-3-3 Proteins 0.000 description 1
- 102000004899 14-3-3 Proteins Human genes 0.000 description 1
- 241000282374 Puma concolor Species 0.000 description 1
- 241001515806 Stictis Species 0.000 description 1
- 239000000654 additive Substances 0.000 description 1
- 230000000996 additive effect Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000013144 data compression Methods 0.000 description 1
- 230000006837 decompression Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 239000003337 fertilizer Substances 0.000 description 1
- 238000003973 irrigation Methods 0.000 description 1
- 230000002262 irrigation Effects 0.000 description 1
- 230000002045 lasting effect Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 230000005055 memory storage Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000011664 signaling Effects 0.000 description 1
- 239000002689 soil Substances 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000000153 supplemental effect Effects 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 208000011580 syndromic disease Diseases 0.000 description 1
- 238000011426 transformation method Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W28/00—Network traffic management; Network resource management
- H04W28/02—Traffic management, e.g. flow control or congestion control
- H04W28/06—Optimizing the usage of the radio link, e.g. header compression, information sizing, discarding information
-
- 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
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/56—Provisioning of proxy services
- H04L67/568—Storing data temporarily at an intermediate stage, e.g. caching
-
- 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
- H04L41/142—Network analysis or design using statistical or mathematical methods
-
- 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
- H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention discloses a method and a device for transmitting data of a wireless sensor network, which are applied to a clustering wireless sensor network. A prediction model is used to perform prediction approximation on data collected by nodes, and the energy consumption of node equipment is reduced by inhibiting node data transmission. Through the scheme, under the premise of meeting application requirements, the amount of data transmitted inside a network can be reduced effectively, the energy consumption of nodes can be reduced, and the life cycle of a network can be prolonged.
Description
Technical field
The present invention relates to the data transmission method of technical field of data transmission, more particularly, to wireless sensor network and dress
Put.
Background technology
At present, wireless sensor network (wsn) is widely used to smart home, agricultural planting, environment monitoring and chemical industry
The fields such as production.
Wireless sensor network (wsn) is to pass through radio communication institute by the sensor node of a large amount of spatially Discrete Distribution
The self-organizing network of composition, as shown in figure 1, be traditional wireless sensor networks system structure diagram.Wherein, task management section
Point is connected with aggregation node by communication network, carries out wireless between the sensor node in aggregation node and Target monitoring area
Communication.
Each sensor node mutually work in coordination with to perceive, to gather, processing and transmission network institute overlay area in various mesh
Mark information, is upper layer application and user's decision-making provides network connection data to support service.
In actual applications, in order to ensure service quality, wsn application purpose region is entered usually through node dense deployment
Row monitoring.Therefore node institute gathered data is needed to be forwarded by some intermediate nodes and reaches destination node, and intermediate node will be at it
The most suitable next-hop (next hop) node is selected among neighbor node.
At present, in wsn widely used clustering route protocol realizing the transmission of data.Make biography according to sensor in routing table
Sensor
Clustering route protocol is data-centered to be divided into several clusters (cluster) by the sensor node in network.
Each cluster is made up of leader cluster node (cluster head) and some cluster interior nodes (cluster member), and leader cluster node should
In responsible cluster the collection of data be polymerized process, also between cluster to be undertaken data routing forwarding.Typical Clustering Routing has
leach(low energy adaptive clusterig hierarchy)、gaf(geographical adaptive
Fidelity), cougar agreement etc., as shown in Fig. 2 be clustering routing structural representation.
With the integrated of electronic equipment and miniaturization, using the battery power supply mode of finite energy more than sensor node,
It is thus desirable to data reasonable selection to reduce network energy consumption from the path of source sensor node to the aggregation node sending data.
Powered using battery in view of sensor node it is difficult to realize the lasting supply to node energy, and each biography
The storage capacity of sensor node and computing capability are also very limited, and many application scenarios need to meet by node dense distribution answers
Use demand.
Studied point out sensor node by 1 bit data transmit 100 meters needed for energy about can execute 3000 meters
Calculate instruction.Often there is dependency between each node institute gathered data of dense distribution, cause the redundancy of some related datas,
And same node there is also certain redundancy so that data transfer energy is high according to the data that time cycle property is gathered, cause
The waste of Internet resources.
Therefore, need a kind of technology badly, using clustering routing structure, can during data transfer, solve redundancy and
Data transfer energy consumes high problem.
Content of the invention
The purpose of the embodiment of the present invention is to provide a kind of data transmission method of wireless sensor network and device, permissible
Reducing data redundancy, reducing the energy that data transfer is consumed, thus extending Network morals.
For reaching above-mentioned purpose, the embodiment of the invention discloses a kind of transmission method of wireless sensor network data, should
For the cluster interior nodes in cluster structured, described cluster structured in also include leader cluster node, wherein, comprise the steps:
S101) prediction model parameterses are determined according to historical data, skew number of times k and offset threshold j are initialized, and
Transmit this prediction model parameters to leader cluster node, so that leader cluster node goes out prediction data according to prediction model parameterses and Time Calculation
And be stored in the data buffer storage corresponding to this node;Described forecast model is to carry out curve plan previously according to historical data
Close acquisition;
S102) gathered data, by the currently practical data collecting and the current predictive obtaining according to prediction model parameterses
Data is compared;
S103) if currently practical data is less than or equal to error threshold ε with the difference ε ' of prediction data, cache this actual number
According to offset threshold j adds 1, return to step s102);
S104) if currently practical data is more than error threshold ε with the difference ε ' of prediction data, cache this real data, partially
Move number of times k and add 1, offset threshold j halves, and transmits this real data to leader cluster node;And number of times k and offset threshold j will be offset
It is compared, if k >=j, return to step s101);If k < j, return to step s102).
Preferably, described method, wherein, forecast model is unitary linear prediction model: y '=α t+ β
Wherein, t represents sampling time point, and y ' is the predictive value in corresponding moment;
Cluster interior nodes are with same intervals t0Continuous n sampling is deposited in buffer queue, using the number in time serieses
According to calculating parameter (α, β).
The embodiment of the present invention also provides a kind of transmitting device of wireless sensor network data, be applied to cluster structured in
Cluster interior nodes, described cluster structured in also include leader cluster node, wherein, comprising:
Prediction model parameterses transport module, for determining prediction model parameterses according to historical data, to skew number of times k with partially
Move threshold value j to be initialized, and transmit this prediction model parameters to leader cluster node, so that leader cluster node is according to prediction model parameterses
Go out prediction data with Time Calculation and be stored in the data buffer storage corresponding to this node;Described forecast model is root in advance
Carry out curve fitting acquisition according to historical data;
Data acquisition module, for gathered data, the currently practical data collecting is obtained with according to prediction model parameterses
The current predictive data obtaining is compared;
First cache module, is less than or equal to the feelings of error threshold ε for the difference ε ' in currently practical data and prediction data
Under condition, cache this real data, offset threshold j adds 1, trigger described data acquisition module;
Second cache module, in the case of being more than error threshold ε for the difference ε ' in currently practical data and prediction data,
Cache this real data, skew number of times k adds 1, and offset threshold j halves, and transmits this real data to leader cluster node;And will offset
Number of times k and offset threshold j are compared, if k >=j, trigger described prediction model parameterses transport module;If k < j, triggering is described
Data acquisition module.
The embodiment of the present invention also provides a kind of method of reseptance of wireless sensor network data, be applied to cluster structured in
Leader cluster node, described cluster structured in also include cluster interior nodes, including following steps:
S201) prediction model parameterses that cluster interior nodes send are received;
S202) judge currently whether receive the real data of cluster interior nodes transmission;
S203) if it is, it stores in this data buffer storage corresponding to cluster interior nodes;
S204) if it is not, then going out prediction data and as data in cluster according to prediction model parameterses and Time Calculation
Store in this data buffer storage corresponding to cluster interior nodes.
Preferably, methods described, wherein, also includes polymeric compressing and processes, comprise the steps:
B1) receive the data that in cluster, each node uploads, and with independent spatial cache, data is carried out to each cluster interior nodes
Caching;
B2) each being uploaded the data storage that in cycle intra-cluster, each node uploads is data matrix x, wherein x=in cluster
[x1,x2,…,xn], n counts out for cluster internal segment, xi=[xi1,xi2,…,xin] for i-th cluster interior nodes according to certain time between
Data every acquisition order;
B3) joint sparse process is carried out to data matrix x in cluster, and by each cluster internal segment point data be decomposed into common ground and
Independent sector, obtains common ground z to after x decompositioncAnd its independent sector zn, decomposition result collection is combined into z=[zc,z1,z2,…,
zn];
B4) each component data in decomposition result set is compressed respectively with perception process, obtains final compression and gather
Close results set r=[rc,r1,r2,…,rn] and calculation matrix set φ=[{ φc1,φc2},{φ11,φ12},…,{φn1,
φn2}];
B5) final compression polymerization result set is sent to aggregation node with calculation matrix set along cluster-level routing or answers
Use terminal.
Preferably, methods described, wherein, independent spatial cache, is delayed with prediction model parameterses including data buffer storage space
Deposit space;
If the data buffer storage space being raw sensed data, directly storing that data into corresponding node receiving
In;If prediction model parameterses data, then update the prediction model parameterses caching of corresponding node, and according to the prediction after updating
Model parameter and Time Calculation go out in the data buffer storage space that prediction data stores corresponding node.
Preferably, methods described, wherein, joint sparse is processed, that is, to each cluster internal segment point data xiIt is respectively adopted sparse
Random matrix carries out sparse transformation.
Preferably, methods described, wherein, sparse random matrix adopts Bernoulli Jacob/Mach random matrix, and it is defined as follows:
In above formula, φijRepresent the element of optional position in Bernoulli Jacob/Mach random matrix, p is that the value of this element is general
Rate, i.e. the random value in { -1,0 ,+1 } of Probability p shown in element above formula in matrix, wherein parameter s control the sparse of matrix
Degree, s value is bigger, and in matrix, nonzero element is less, and sparse degree is higher.
Preferably, methods described, wherein, compressed sensing is processed, and comprises the steps:
A1) according to cluster internal segment count out n, perception compression after data length m and sparse control parameter s, calculating parameter a=
m/2s、Produce initially empty calculation matrix φ1、φ2;
A2 two random 1 × n dimension matrix t) are generated1=[t11,t12,…,t1n]、t2=[t21,t22,…,t2n], wherein
t1nWith t2nIt is the positive integer of random value in [1, b], by matrix t1、t2Store calculation matrix respectively as data line
φ1、φ2In.
A3) find out t1Intermediate value is the position of the element of i, and correspondence position element in primary signal is added;Find out t2Intermediate value is
The position of the element of i, correspondence position element in primary signal is added;Deduct second value of calculation, institute with first value of calculation
Obtain result and be deposited into corresponding compressed data riIn;
A4) repeat step a2) and step a3) a time;
A5) make b=m-a × b, repeat step a3), a4), obtain final compression polymerization result data riAnd calculation matrix
The embodiment of the present invention also provides a kind of reception device of wireless sensor network data, be applied to cluster structured in
Leader cluster node, described cluster structured in also include cluster interior nodes, this data sink includes:
Receiver module, with receiving the prediction model parameterses that cluster interior nodes send;
Judge module, for judging currently whether receive the real data of cluster interior nodes transmission;
First memory module, for when judged result is to be, by actual data storage to corresponding to this cluster interior nodes
In data buffer storage;
Second memory module, for judged result for no when, go out to predict number according to prediction model parameterses and Time Calculation
According to and as in the data buffer storage corresponding to this cluster interior nodes of data Cun Chudao in cluster.
Therefore, the present invention program is directed to the resource-constrained feelings of many application of higher wireless sensor network scene interior joint
Condition, based on existing common clustering routing structure it is proposed that a kind of data transmission mechanism of Energy Efficient.This mechanism is directed to node
The feature of energy constraint, in cluster, member node is predicted approaching to node data using forecast model, suppresses node data
Transmission is thus reduce energy expenditure.And, on leader cluster node, by improved compression sensing method, data in cluster is joined
Combined pressure contracting is processed, the volume of transmitted data of compression cluster head.This programme can effectively reduce in net under the premise of meeting application demand
Volume of transmitted data, thus reducing node energy consumption, extends network lifecycle.
Brief description
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
Have technology description in required use accompanying drawing be briefly described it should be apparent that, drawings in the following description be only this
Some embodiments of invention, for those of ordinary skill in the art, on the premise of not paying creative work, acceptable
Other accompanying drawings are obtained according to these accompanying drawings.
Fig. 1 is the structural representation of traditional wireless sensor networks application system;
Fig. 2 is the clustering routing structural representation of existing wireless sensor network;
The compressed sensing measurement process schematic diagram that Fig. 3 adopts for technical solution of the present invention;
The data acquisition transfer process figure of the cluster interior nodes that Fig. 4 provides for the embodiment of the present invention 1;
The data transfer of the cluster interior nodes that Fig. 5 provides for the embodiment of the present invention 1 adjusts schematic diagram with model;
Fig. 6 is a kind of structural representation of data transmission device provided in an embodiment of the present invention;
The data receiver flow chart of the leader cluster node that Fig. 7 provides for the embodiment of the present invention 2;
Fig. 8 is a kind of structural representation of data sink provided in an embodiment of the present invention.
Specific embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation describes it is clear that described embodiment is only a part of embodiment of the present invention, rather than whole embodiments.Based on this
Embodiment in invention, the every other reality that those of ordinary skill in the art are obtained under the premise of not making creative work
Apply example, broadly fall into the scope of protection of the invention.
The present invention is the wireless sensor network based on clustering routing structure, and data compression perception theory passes to data
The improvement that defeated process is carried out.
Clustering route protocol plane road by the basis of node division is member node in leader cluster node and cluster, and bright
The really respective division of labor.Cluster head is responsible for and member communication in cluster, collects information about firms data and is responsible for data forwarding;Become in cluster
Member gathers the target information in monitored area and is sent to cluster head.Typically carry out equilibrium energy using periodicity cluster head rotation mechanism to disappear
Consumption.This clustering algorithm splits the network into different logical subfield domains, has more preferable extensibility, can be to data in cluster
Carry out further dispatch deal.
Node energy mainly consumes on data transmit-receive, so the embodiment of the present invention is by being compressed locating to data in net
Manage and to realize energy efficient utilization.Compressive sensing theory is a kind of Signal Compression and high accuracy recovery technology, is initially mainly used in
Process the digital signal of discrete domain.Compressive sensing theory is pointed out: if signal x=[x1,x2,…,xn] a certain conversion base ψ=
[ψ1,ψ2,…,ψn] under can carry out k rarefaction representation, that is, meet element number in x=ψ s and vectorial s and be much smaller than x, if element in s
Number is that the energy of k then primary signal x all concentrates on this k nonzero element, and s is called the k rank rarefaction representation of signal x.Compression
Perception is as shown in Figure 3.When signal x meets above-mentioned condition, by signal vector y=[y1,y2,…,ym] can recover
X, wherein y=φ x=φ ψ s, the length of signal vector y are much smaller than original signal x, and that is, y is that x is compressed with the knot after processing
Really, and claim φ ∈ rm×nObserving matrix for x.
In extensive time multiplexed signal, can be by the convex optimization problem of solution:
Obtain optimal solution s, then utilize x=ψ s, realize the Exact recovery to primary signal x.Solving above formula needs to travel through
All nonzero value combinations in s, so above formula 0- norm can be converted into 1- norm approximate solution, obtain as drag:
Consider error and noise factor, also above formula can be changed further:
On stricti jurise, most of signal is not sparse, but signal can be made dilute under transform domain by suitable conversion base ψ
Relieving the exterior syndrome shows.Gaussian random matrix is frequently as calculation matrix;Common sparse alternative approach has Fourier transformation, wavelet transformation.To survey
Amount signaling protein14-3-3 algorithm has orthogonal matching pursuit (omp) algorithm, base to follow the trail of (bp) algorithm, Adaptive matching tracing algorithm etc.,
Respectively there is quality.
The present invention is based on clustering routing structured radio sensor network, for member node in leader cluster node and cluster not
With division characteristic it is proposed that a kind of data aggregate transmission of energy efficient is tactful.In this data transmission policies, member and cluster in cluster
Head node can carry out polymeric compressing process respectively according to application demand to transmission data, reduces the transmission of redundant data in network,
Save energy thus extending network normal working hourss.It is emphasized that the present invention program passes mainly for the data in cluster
Defeated processing procedure, is not related to cluster-level routing and sets up and data transfer.Particular content can be divided into cluster internal segment by the type according to node
Point treatment mechanism and leader cluster node treatment mechanism two parts.It is described in detail by way of example separately below.
Embodiment 1
The present embodiment provides a kind of cluster interior nodes data transmission method based on cluster structured wireless sensor network, such as
Shown in Fig. 4:
S101) prediction model parameterses are determined according to historical data, skew number of times k and offset threshold j are initialized, and
Transmit this prediction model parameters to leader cluster node, so that leader cluster node goes out prediction data according to prediction model parameterses and Time Calculation
And be stored in the data buffer storage corresponding to this node;Described forecast model is to carry out curve plan previously according to historical data
Close acquisition;
S102) gathered data, by the currently practical data collecting and the current predictive obtaining according to prediction model parameterses
Data is compared;
S103) if currently practical data is less than or equal to error threshold ε with the difference ε ' of prediction data, cache this actual number
According to offset threshold j adds 1, return to step s102);
S104) if currently practical data is more than error threshold ε with the difference ε ' of prediction data, cache this real data, partially
Move number of times k and add 1, offset threshold j halves, and transmits this real data to leader cluster node;And number of times k and offset threshold j will be offset
It is compared, if k >=j, return to step s101);If k < j, return to step s102).
Specifically, in common clustering routing structured radio sensor network, cluster interior nodes are led to as network tip
Often only it is responsible for the object information data in collection monitored area the node that continues thereto transmission, cluster interior nodes only use self-responsibility
The adjacent cluster interior nodes in the domain of same district are not needed to communicate.By being analyzed to environmental objects such as common temperature, humidity,
This kind of environmental data typically exhibits cyclically-varying feature, and presents linear variability law in certain period of time, that is, save
Put gathered environmental data and there is between the time linear relationship.Cluster interior nodes a period of time in gathered environmental data and when
Between linear correlation, then the time can be regarded as independent variable, corresponding data as dependent variable, using linear regression model (LRM) come structure
Make anticipation function to simulate the relation of both dynamic changes.
Define time serieses for cluster interior nodes sample in chronological order, have intervals time data to collection
Close, be designated as s={ (t1,y1),(t2,y2),…,(tn,yn)}.Wherein (tn,yn) represent in time tiMoment collection data be
yi, n is sequence length.Set up the piecewise linearity forecast model of dynamic adjustment using the calculating storage capacity of node come to adopting data
Number carries out approximation timates.
Initial phase includes the process of setting up of network cluster dividing routing infrastructure, and cluster interior nodes are in this stage and cluster head section
Point establishes a connection, and receives control information, including but not limited to error threshold ε, offset threshold j, buffer storage length, collection mesh
Mark, collection period, data upload cycle etc..
Subsequently cluster interior nodes enter into the data acquisition transmission stage, and this stage can be refined as building forecast model, data
Transmission and model set-up procedure.
Forecast model is set up
Consider that node calculates limited by storage capacity, set up unitary linear prediction model in cluster interior nodes:
Y '=α t+ β
Wherein, t represents sampling time point, and y ' is the predictive value in corresponding moment, and (α, β) is constant.
According to buffer storage length, node is with same intervals t0Continuous n sampling is deposited in caching sequence, using caching sequence
Data in row calculates parameter (α, β), and parameter is sent to leader cluster node by cluster interior nodes, and the data that it is subsequently gathered should edge
Time shafts are distributed near anticipation function.
Data transfer is adjusted with model
Referring to Fig. 5, in each data acquisition transmitting procedure, node calculates actual acquired data y and predicted estimate value y '
Between difference ε ', it is compared with default error threshold ε, when gathered data change slow when, i.e. ε ' < ε, cluster interior nodes
Do not transmit this secondary data, cluster head calculates the approximate data of this time point according to existing prediction model parameterses;When perception data becomes
When changing violent, i.e. ε ' > ε, then in real time this secondary data is sent to cluster head.
Node can monitor its data situation of change while data transfer, if the long-term deviation of forecast model is truly adopted
Sample, illustrates that parameter is inapplicable, need to be adjusted.Error ε ' is exceeded the number of times of threshold value ε and offset threshold is compared to
Determine parameter the need of adjustment.Offset threshold is not definite value, if predictive value meets requirement when being compared every time, will
Offset threshold adds 1, otherwise halves offset threshold.
In order to prevent the generation of " seemingly-dead node ", in time cycle t (t=t0× n) in, cluster interior nodes need to be forced to cluster head
Node sends a secondary data.In order that the new parameter calculating every time meets current state, need time serieses are carried out in real time
Update operation, so in node software design, maintaining a spatial cache storage time sequence.Entering with data acquisition
It is assumed that the model parameter of current time is (α, β), caching sequence is sc to row, and model offset number of times is k, and offset threshold is j, number
As follows with model method of adjustment according to transmitting:
Step 1:
Node gathered data yn, according to time point tnCalculate prediction data y with parameter (α, β)n' and forecast error ε '
if(ε′<ε)
By ynIt is added to queue sc tail of the queue, if sc length is full, remove queue heads;
J=j+1;
Go to step 1
if(ε′>ε)
By ynIt is added to queue sc tail of the queue, if sc length is full, remove queue heads;
K=k+1;
J=j/2;
Node sends perception data;
if(k≥j)
Go to step 2;
Step 2:
K=0;
J=1;(certainly, this offset threshold j can also be initialized as any be not 0 constant)
According to method of least square, calculate new parameter (α, β) using the data in sc, and send it to leader cluster node, turn
To step 1.
By the present invention program accompanying method, cluster interior nodes can be predicted to gathered data simulating, and need not send all
Gathered data, thus effectively reducing data is activation amount, reducing energy expenditure, extend node working life.
Below according to Fig. 7, a kind of data transmission device of wireless sensor network of the present invention is illustrated, this data
Transmitting device be applied to cluster structured in cluster interior nodes, described cluster structured in also include leader cluster node.This data transfer fills
Put including prediction model parameterses transport module 101, data acquisition module 102, the first cache module 103 and the second cache module
104.
Wherein, prediction model parameterses transport module 101, for determining prediction model parameterses according to historical data, to skew
Number of times k and offset threshold j are initialized, and transmit this prediction model parameters to leader cluster node, so that leader cluster node is according to pre-
Survey model parameter and Time Calculation goes out prediction data and is stored in the data buffer storage corresponding to this node;Described prediction mould
Type is to carry out curve fitting acquisition previously according to historical data;
Wherein, data acquisition module 102, for gathered data, by the currently practical data collecting and according to prediction mould
The current predictive data that shape parameter obtains is compared;
Wherein, the first cache module 103, is less than or equal to error door for the difference ε ' in currently practical data and prediction data
In the case of limit ε, cache this real data, offset threshold j adds 1, trigger described data acquisition module;
Wherein, the second cache module 104, is more than error threshold ε for the difference ε ' in currently practical data and prediction data
In the case of, cache this real data, skew number of times k adds 1, and offset threshold j halves, and transmits this actual number to leader cluster node
According to;And skew number of times k is compared with offset threshold j, if k >=j, trigger described prediction model parameterses transport module;If k <
J, then trigger described data acquisition module.
Embodiment 2
The present embodiment is directed to the data transmission method of the cluster interior nodes that embodiment 1 provides, and provides one kind based on cluster structured
The leader cluster node of wireless sensor network data receiver method, as shown in Figure 6:
S201) prediction model parameterses that cluster interior nodes send are received;
S202) judge currently whether receive the real data of cluster interior nodes transmission;
S203) if it is, it stores in this data buffer storage corresponding to cluster interior nodes;
S204) if it is not, then going out prediction data and as data in cluster according to prediction model parameterses and Time Calculation
Store in this data buffer storage corresponding to cluster interior nodes.
Specifically, based in cluster structured wireless sensor network, each cluster is equivalent to an independent network
Node region, as the manager in this region and coordinator, the data forwarding amount being undertaken is than member node in cluster for leader cluster node
More, energy consumption is bigger.
In order to reduce the load pressure of leader cluster node, reduce the energy expenditure that data transfer is brought, the technology of the present invention
Scheme carries out further polymeric compressing process in leader cluster node to data.As shown in fig. 6, the method also includes step s205)
Polymeric compressing process is carried out to the data of each the predetermined period memory storage in data cached.
In initial phase, leader cluster node needs the foundation completing cluster-level routing to be connected and distribute with the foundation of cluster interior nodes
Time slot, also will receive application or the control information of aggregation node simultaneously, including but not limited to upload cycle, purpose data, application
Tolerance thresholding etc..Work and concrete application scene close association is completed needed for this stage.After completing node initializing, cluster head
Node can start waiting for receiving data in cluster.
From embodiment 1, cluster interior nodes are predicted approaching to gathered data by forecast model, reduce data is activation
Amount.Cluster head node needs the model parameter being sent by each node offer spatial cache in this cluster in order to memory node respectively
Data and perception data, cluster head needs and cluster interior nodes keep clock synchronous simultaneously.
Leader cluster node receives each node data in cluster, if receive is raw sensed data, directly deposits this data
Store up in the data buffer storage of corresponding node;If prediction model parameterses data, leader cluster node updates this cluster interior nodes pair first
The parameter cache answered, subsequently goes out prediction data according to new parameter and Time Calculation and is stored to the data corresponding to this node
In caching.In each uplink time point, leader cluster node needs the data buffer storage of all nodes to be sent to aggregation node, clearly simultaneously
Empty caching.
In cluster, each node has certain association on geographical position, and its gathered data there is also certain dependency, this
Bright embodiment is directed to cluster head and sends the excessive problem of data volume it is proposed that new compression processing method carries out joint to data in cluster
Compression polymerization, thus reduce leader cluster node and sending data volume, saving energy expenditure.
Improved data compressing method
Compressed sensing may insure the reconstruction primary signal of high probability in the case of minimizing hits, can be used for wireless sensing
In device network, data is compressed.But because sensor node is resource-constrained, being compressed perception needs to process calculation matrix
With the multiplying of initial data, the storage of sensor node and computing capability are required higher.The present invention program is with regard to this problem
Propose and be suitably modified work.
At present, gaussian random calculation matrix is widely applied in compressive sensing theory, but it realizes process needs
Random number generator meets, to produce, the m × n random number requiring.Produce the mistake that calculation matrix and matrix are multiplied with initial data
Journey complexity is higher, is not suitable for resource-constrained sensor node.
This programme adopts sparse random matrix to obtain the main information of compressible data, to make up the deficiency of this respect, primary
As common sparse matrix, it is defined as follows Nu Li/Mach random matrix:
In above formula, φijRepresent the element of optional position in Bernoulli Jacob/Mach random matrix, p is that the value of this element is general
Rate, i.e. the random value in { -1,0 ,+1 } of Probability p shown in element above formula in matrix, wherein parameter s control the sparse of matrix
Degree, s value is bigger, and in matrix, nonzero element is less, and sparse degree is higher.If data length is respectively after initial data and compression
For n, m, then still need to produce and store the sparse matrix of m × n using said method.
Sparse random matrix is split by the present invention program according to element is positive and negative, then produced by way of piecemeal with
Machine number is representing the coordinate of nonzero element in calculation matrix, then is combined into improved sparse random measurement matrix by each coordinate block
φ1、φ2Two matrixes, φ1、φ2Do not store true calculation matrix element data, but store the position of its nonzero element.
Meanwhile, matrix multiplication is reduced to additive operation, completes the compressed sensing to primary signal while generating calculation matrix and survey
Amount process, reduces the complexity of data processing, improves the real-time performance of wireless sensor network system.
It is assumed that to length, data x for n is compressed the data length after perception is processed is m, then improved compressed sensing
Processing procedure is as follows:
(1) ignore normalization operator, the positive element of each column and negative element value ± 1, number is a, each column is divided into a group,
Meet a=m/2s, calculateInitial φ1、φ2It is empty matrix.
(2) two random 1 × n dimension matrix t are generated1=[t11,t12,…,t1n]、t2=[t21,t22,…,t2n], t1With
t2Represent element position in calculation matrix.Wherein t1nWith t2nIt is the positive integer of random value in [1, b], need to make t1nWith
t2nDo not take identical value it is ensured that the positive element of every string only one of which of each matrix in block form and negative element.Then, by matrix t1、t2
Store matrix φ respectively as data line1、φ2Among.
(3) subsequently carry out initial data x and be circulated perception process.In processing each time, find out t first1Intermediate value is i unit
The position of element, the value of relevant position element in primary signal x is added, subsequently equally finds out t2Intermediate value is the position of i element, will
In primary signal x, the value of relevant position element is added, and is multiplied by after finally subtracting each other above-mentioned two result againObtain result ri,
And by result riIt is stored in last compressed data r.I initial value is 1, and after each circulation, cumulative 1, i is equal to b is last locating
Reason.
(4) process 2 and process 3 are repeated a time;
(5) finally make b=m-a × b, carry out once-through operation according to process (3), process (4) and result is stored in riIn.
By said process, data r after initial data x is compressed and corresponding improvement are just obtained
Calculation matrix
From said process, node only needs uniformly random number generator and the memory space of 2 (a+1) × n, spatial complex
Degree reduces;Also by matrix multiplication, addition is reduced to the measurement process of primary signal, time complexity reduces;Product by calculation matrix
The raw process with data perception measurement combines, in hgher efficiency.
Cluster head handling process
Distributed compression perception theory is thought to multiple signals or data, their time domain can be utilized openness and spatial domain
Dependency tectonic syntaxis sparse model, is carried out combined coding and is recovered with combined decoding, so than independent, signal can be processed
There are more excellent energy consumption characteristics, also can preferably recover primary signal.
Because node each in cluster has seriality on geographical position, it can thus be assumed that the gathered target data of each node
There is certain dependency.Node data each in cluster is decomposed into the common ground that all nodes all comprise and each by the present invention program
From distinctive independent sector.In the cluster of supposition cluster head caching within each upload cycle, the data of i-th sensor node is
xi=[xi1,xi2,…,xin], xi=[xi1,xi2,…,xin] it is i-th cluster interior nodes according to intervals acquisition order
Data, then in this time period, all data of cluster head caching are x=[x1,x2,…,xn], n counts out for cluster internal segment.Due to
Not each cluster interior nodes can send perception data to cluster head in each sampling period, and this just correspondingly reduces communication in cluster and opens
Pin.
The data aggregate process of haptophore interior nodes, the data processing of leader cluster node, it is described in detail below:
(1) leader cluster node remains working condition and prepares to receive cluster internal segment point data, and is each cluster interior nodes
Distribute corresponding data buffer storage space.Cluster interior nodes send data and are divided into prediction model parameterses and true perception data two
Class.
(2) the prediction model parameterses caching of corresponding node, meter when cluster head receives data, if supplemental characteristic, are then updated
The prediction data calculating the corresponding moment is deposited into data buffer storage;If perception data, then it is directly stored in corresponding node
Data buffer storage.
(3) often through upload cycle t, it data cached is generated as signal vector x according to cluster interior nodes id by cluster headi=
[x1i,x2i,…,xni]tIf cluster internal segment is counted as n, the data matrix of cluster head storage is x=[x1,x2,…,xn].
(4) utilize distributed compression perception theory, by common sparse transformation method described previously, this cluster head is stored
Data matrix x carry out joint sparse process.The purpose entering to exercise joint sparse is so that the data after conversion is always sparse
Degree is minimum, so needing to meet optimal condition:
min(||zc||0+||z1||0+||z2||0+…+||zn||0) i=1,2 ... n
By above-mentioned condition to xiCarrying out the data after sparse transformation is ai=[a1i,a2i,…,ani]t, ai=zc+zn, wherein
zcFor the common ground of all data, znFor aiDistinctive independent sector.
(5) common ground of each signal and independent sector can be combined and be expressed as z=[zc,z1,z2,…,zn], each component root
Determine the data length after perception compression according to its degree of rarefication.Using previously described data compressing method, sparse data is entered
Obtaining last compression aggregated data after row compression is r=[rc,r1,r2,…,rn] and corresponding improvement calculation matrix
φ=[{ φc1,φc2},{φ11,φ12},…,{φn1,φn2}].
Leader cluster node uploads at each and will compress aggregated data in cycle t with calculation matrix data is activation to convergence section
Point, then data recovery is carried out by compressed sensing recovery algorithms by aggregation node.
Embodiment 3
The present embodiment is so that agriculture Internet of Things is applied as a example.In large-scale field application scene, need to be grasped the life of crops
Long ambient condition is realizing more scientific breed of crop.Typically temperature, humidity, illumination need to be collected by wireless sensor node
Etc. data for applying fertilizers scientifically, high-efficient irrigation provide data support, realize green agriculture.
Will not be too high to the required precision of the ambient parameter datas such as humiture in this kind of application scenarios, can there is certain mistake
Difference tolerance.Due to the sensor node that is positioned in farmland using battery or solar powered by the way of, and node cost is unsuitable
Too high, therefore its energy and calculating storage capacity are limited.Using the present invention program, can on the premise of meeting application demand,
Reduce the energy expenditure that data transfer is brought, extend the working time of each node in farmland, save production cost.
User determines the destination object needing monitoring by terminal, and obtains the target farmland being gathered by sensor node
Interior object data, these destination objects include but is not limited to temperature, humidity, intensity of illumination, something concentration etc. in soil.Tool
Body application process is as follows:
1. the sensor node in target area completes cluster structured construction process, and completes cluster-level routing foundation;Pass
Sensor node reports the positional information of oneself, id etc. to aggregation node
2. each leader cluster node is all joint structure spatial caches in this cluster.
3. user sends control information to aggregation node, including but not limited to collection target, acquisition time interval, upload week
Phase, error in data thresholding, leader cluster node rotational cycle, node time sequence length etc..
4. control information is sent to each sensor node by aggregation node, and to user terminal feedback node information.
5. sensor node starts data acquisition, and each node utilizes initial gathered data structure according to length of time series
Make prediction model parameterses, parameter is sent to leader cluster node.Subsequently according to cluster interior nodes treatment mechanism described previously, according to default
Error threshold examination data collection send work.
6. leader cluster node keeps receiving data in cluster, and when each uploads the moment in cycle and reaches, according to cluster head datatron
System carries out polymeric compressing process to data in cluster, and it is sent to aggregation node along cluster-level routing.
7. after aggregation node receives data, both can be immediately sent to user side, carry out reducing by user side extensive
Multiple;User side can also be sent it to again after aggregation node carries out decompression reduction.
8. often through a cluster head rotational cycle, the wireless sensor network of target area needs to carry out cluster head again to select
Lift so that energy uniformly consumes.
9. user can adjust the control parameters such as error threshold according to real needs.
Below according to Fig. 8, a kind of data transmission device of wireless sensor network of the present invention is illustrated, this data
Reception device be applied to cluster structured in leader cluster node, described cluster structured in also include cluster interior nodes.This data transfer fills
Put including receiving module 201, judge module 202, the first memory module 203 and the second memory module 204.
Wherein, receiver module 201, with receiving the prediction model parameterses that cluster interior nodes send;
Wherein, judge module 202, for judging currently whether receive the real data of cluster interior nodes transmission;
Wherein, the first memory module 203, for when judged result is to be, by actual data storage to this cluster interior nodes
In corresponding data buffer storage;
Wherein, the second memory module 204, for judged result for no when, according to prediction model parameterses and Time Calculation
Go out prediction data and as in the data buffer storage corresponding to this cluster interior nodes of data Cun Chudao in cluster.
As seen from the above-described embodiment, the present invention program is directed to the resource-constrained feelings of application of higher wireless sensor network interior joint
Condition, based on existing clustering routing structure it is proposed that a kind of data transmission method of Energy Efficient.For member node in cluster,
Using forecast model, node data is predicted approaching, suppression node data transmits thus reducing energy expenditure.For cluster head
Node, carries out joint compression by improved compression sensing method and processes, reduce the volume of transmitted data of cluster head to data in cluster.This
Scheme can effectively reduce volume of transmitted data in net under the premise of meeting application demand, thus reducing node energy consumption, prolongs
Long network lifecycle.
It should be noted that herein, such as first and second or the like relational terms are used merely to a reality
Body or operation are made a distinction with another entity or operation, and not necessarily require or imply these entities or deposit between operating
In any this actual relation or order.And, term " inclusion ", "comprising" or its any other variant are intended to
Comprising of nonexcludability, wants so that including a series of process of key elements, method, article or equipment and not only including those
Element, but also include other key elements being not expressly set out, or also include for this process, method, article or equipment
Intrinsic key element.In the absence of more restrictions, the key element that limited by sentence "including a ..." it is not excluded that
Also there is other identical element including in the process of described key element, method, article or equipment.
Each embodiment in this specification is all described by the way of related, identical similar portion between each embodiment
Divide mutually referring to what each embodiment stressed is the difference with other embodiment.Real especially for system
For applying example, because it is substantially similar to embodiment of the method, so description is fairly simple, referring to embodiment of the method in place of correlation
Part illustrate.
The foregoing is only presently preferred embodiments of the present invention, be not intended to limit protection scope of the present invention.All
Any modification, equivalent substitution and improvement made within the spirit and principles in the present invention etc., are all contained in protection scope of the present invention
Interior.
Claims (10)
1. a kind of transmission method of wireless sensor network data, be applied to cluster structured in cluster interior nodes, described sub-clustering knot
Also leader cluster node is included it is characterised in that comprising the steps: in structure
S101) prediction model parameterses are determined according to historical data, skew number of times k and offset threshold j are initialized, and to cluster
Head node transmits this prediction model parameters, so that leader cluster node goes out prediction data according to prediction model parameterses and Time Calculation and incites somebody to action
It stores in the data buffer storage corresponding to this node;Described forecast model is to carry out curve fitting previously according to historical data to obtain
?;
S102) gathered data, by the currently practical data collecting and the current predictive data obtaining according to prediction model parameterses
It is compared;
S103) if currently practical data is less than or equal to error threshold ε with the difference ε ' of prediction data, cache this real data, partially
Move threshold value j and add 1, return to step s102);
S104) if currently practical data is more than error threshold ε with the difference ε ' of prediction data, this real data, skew time are cached
Number k adds 1, and offset threshold j halves, and transmits this real data to leader cluster node;And skew number of times k is carried out with offset threshold j
Relatively, if k >=j, return to step s101);If k < j, return to step s102).
2. method according to claim 1 is it is characterised in that described forecast model is unitary linear prediction model: y '=α
t+β
Wherein, t represents sampling time point, and y ' is the predictive value in corresponding moment;
Cluster interior nodes are with same intervals t0Continuous n sampling is deposited in buffer queue, is calculated using the data in time serieses
Go out parameter (α, β).
3. a kind of transmitting device of wireless sensor network data, be applied to cluster structured in cluster interior nodes, described sub-clustering knot
Also leader cluster node is included it is characterised in that including in structure:
Prediction model parameterses transport module, for determining prediction model parameterses according to historical data, to skew number of times k and skew threshold
Value j is initialized, and transmits this prediction model parameters to leader cluster node so that leader cluster node according to prediction model parameterses and when
Between calculate prediction data and be stored in the data buffer storage corresponding to this node;Described forecast model is previously according to going through
History data carries out curve fitting acquisition;
Data acquisition module, for gathered data, by the currently practical data collecting and according to prediction model parameterses acquisition
Current predictive data is compared;
First cache module, in the case of being less than or equal to error threshold ε for the difference ε ' in currently practical data and prediction data,
Cache this real data, offset threshold j adds 1, trigger described data acquisition module;
Second cache module, in the case of being more than error threshold ε for the difference ε ' in currently practical data and prediction data, caching
This real data, skew number of times k adds 1, and offset threshold j halves, and transmits this real data to leader cluster node;And number of times will be offset
K is compared with offset threshold j, if k >=j, triggers described prediction model parameterses transport module;If k < j, trigger described data
Acquisition module.
4. a kind of method of reseptance of wireless sensor network data, be applied to cluster structured in leader cluster node, described sub-clustering knot
Cluster interior nodes are also included it is characterised in that comprising the steps: in structure
S201) prediction model parameterses that cluster interior nodes send are received;
S202) judge currently whether receive the real data of cluster interior nodes transmission;
S203) if it is, it stores in this data buffer storage corresponding to cluster interior nodes;
S204) if it is not, then going out prediction data and as data storage in cluster according to prediction model parameterses and Time Calculation
In data buffer storage corresponding to this cluster interior nodes.
5. method according to claim 4 is processed it is characterised in that also including polymeric compressing, comprises the steps:
B1) receive the data that in cluster, each node uploads, and with independent spatial cache, data buffer storage is carried out to each cluster interior nodes;
B2) each being uploaded the data storage that in cycle intra-cluster, each node uploads is data matrix x, wherein x=[x in cluster1,
x2,…,xn], n counts out for cluster internal segment, xi=[xi1,xi2..., xin] suitable according to intervals for i-th cluster interior nodes
The data of sequence collection;
B3) joint sparse process is carried out to data matrix x in cluster, and each cluster internal segment point data is decomposed into common ground and independence
Part, obtains common ground z to after x decompositioncAnd its independent sector zn, decomposition result collection is combined into z=[zc,z1,z2,…,zn];
B4) each component data in decomposition result set is compressed respectively with perception process, obtains final compression polymerization knot
Fruit set r=[rc,r1,r2,…,rn] and calculation matrix set φ=[{ φc1,φc2},{φ11,φ12},…,{φn1,
φn2}];
B5) final polymeric compressing results set is sent to aggregation node with calculation matrix set along cluster-level routing or application is whole
End.
6. method according to claim 5 is it is characterised in that described independent spatial cache, including data buffer storage space
With prediction model parameterses spatial cache;
If receive is raw sensed data, directly store that data in the data buffer storage space of corresponding node;As
Fruit is prediction model parameterses data, then update the prediction model parameterses caching of corresponding node, and according to the forecast model after updating
Parameter and Time Calculation go out in the data buffer storage space that prediction data stores corresponding node.
7. method according to claim 5, it is characterised in that described joint sparse is processed, is counted to each cluster internal segment
According to xiIt is respectively adopted sparse random matrix and carry out sparse transformation.
8. method according to claim 7 is it is characterised in that described sparse random matrix adopts Bernoulli Jacob/Mach random
Matrix, it is defined as follows:
In above formula, φijRepresent the element of optional position in Bernoulli Jacob/Mach random matrix, p is the probability of this element, that is,
Probability p shown in element above formula in matrix random value in { -1,0 ,+1 }, wherein parameter s control the sparse degree of matrix,
S value is bigger, and in matrix, nonzero element is less, and sparse degree is higher.
9. method according to claim 5, it is characterised in that described compressed sensing is processed, comprises the steps:
A1) according to cluster internal segment count out n, perception compression after data length m and sparse control parameter s, calculating parameter a=m/
2s、Produce initially empty calculation matrix φ1、φ2;
A2 two random 1 × n dimension matrix t) are generated1=[t11,t12,…,t1n]、t2=[t21,t22,…,t2n], wherein t1nWith
t2nIt is the positive integer of random value in [1, b], by matrix t1、t2Store calculation matrix φ respectively as data line1、
φ2In.
A3) find out t1Intermediate value is the position of the element of i, and correspondence position element in primary signal is added;Find out t2Intermediate value is i's
The position of element, correspondence position element in primary signal is added;Deduct second value of calculation with first value of calculation, gained is tied
Fruit is deposited into corresponding compressed data riIn;
A4) repeat step a2) and step a3) a time;
A5) make b=m-a × b, repeat step a3), a4), obtain final compression polymerization result data riAnd calculation matrix
10. a kind of reception device of wireless sensor network data, be applied to cluster structured in leader cluster node, described sub-clustering knot
Cluster interior nodes are also included it is characterised in that including in structure:
Receiver module, with receiving the prediction model parameterses that cluster interior nodes send;
Judge module, for judging currently whether receive the real data of cluster interior nodes transmission;
First memory module, for judged result be when, by actual data storage to this cluster interior nodes corresponding to data
In caching;
Second memory module, for judged result for no when, go out prediction data simultaneously according to prediction model parameterses and Time Calculation
As in the data buffer storage corresponding to this cluster interior nodes of data Cun Chudao in cluster.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610712279.5A CN106341842B (en) | 2016-08-23 | 2016-08-23 | Method and device for transmitting wireless sensor network data |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610712279.5A CN106341842B (en) | 2016-08-23 | 2016-08-23 | Method and device for transmitting wireless sensor network data |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106341842A true CN106341842A (en) | 2017-01-18 |
CN106341842B CN106341842B (en) | 2020-05-01 |
Family
ID=57824975
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610712279.5A Expired - Fee Related CN106341842B (en) | 2016-08-23 | 2016-08-23 | Method and device for transmitting wireless sensor network data |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106341842B (en) |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106961656A (en) * | 2017-02-23 | 2017-07-18 | 南京邮电大学 | A kind of wireless sensor network data Forecasting Methodology |
CN107071800A (en) * | 2017-03-01 | 2017-08-18 | 北京邮电大学 | A kind of cluster wireless sensor network method of data capture and device |
CN107105394A (en) * | 2017-05-11 | 2017-08-29 | 江苏食品药品职业技术学院 | Building safety monitoring system based on wireless sensor network |
CN108347709A (en) * | 2017-01-21 | 2018-07-31 | 中山大学 | A kind of system adaptive recognition method and system of super low-power consumption traceability system |
CN110505597A (en) * | 2019-07-31 | 2019-11-26 | 北京邮电大学 | A kind of data transmission method of wireless sensor network |
CN110996304A (en) * | 2019-12-04 | 2020-04-10 | 深圳信可通讯技术有限公司 | Method and system for inquiring and calculating edge data of low-power-consumption Internet of things |
CN111010704A (en) * | 2019-12-03 | 2020-04-14 | 沈阳化工大学 | Underwater wireless sensor network data prediction optimization method based on exponential smoothing |
CN111262948A (en) * | 2020-02-18 | 2020-06-09 | 广东大橘果业有限公司 | Planting intelligent management system based on internet |
CN111311427A (en) * | 2020-02-18 | 2020-06-19 | 广东大橘果业有限公司 | Planting intelligent management system based on internet |
CN111726768A (en) * | 2020-06-16 | 2020-09-29 | 天津理工大学 | Edge-oriented computation reliable data collection method based on compressed sensing |
CN112039934A (en) * | 2019-06-03 | 2020-12-04 | 大唐移动通信设备有限公司 | Information feedback method, feedback information processing method and device |
CN115695564A (en) * | 2022-12-30 | 2023-02-03 | 深圳市润信数据技术有限公司 | Efficient transmission method for data of Internet of things |
CN115835060A (en) * | 2023-02-17 | 2023-03-21 | 云南瀚哲科技有限公司 | Data acquisition method for industrial and agricultural Internet of things |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130097128A1 (en) * | 2010-04-26 | 2013-04-18 | Shoji Suzuki | Time-series data diagnosing/compressing method |
CN104469797A (en) * | 2014-11-28 | 2015-03-25 | 北京农业信息技术研究中心 | Method for generating sequence prediction on basis of farmland wireless network intra-cluster data sparsity |
-
2016
- 2016-08-23 CN CN201610712279.5A patent/CN106341842B/en not_active Expired - Fee Related
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130097128A1 (en) * | 2010-04-26 | 2013-04-18 | Shoji Suzuki | Time-series data diagnosing/compressing method |
CN104469797A (en) * | 2014-11-28 | 2015-03-25 | 北京农业信息技术研究中心 | Method for generating sequence prediction on basis of farmland wireless network intra-cluster data sparsity |
Non-Patent Citations (2)
Title |
---|
WEI CHEN AND IAN J.WASSELL: "energy efficient signal acquisition via compressive sensing in wireless sensor networks", 《INTERNATIONAL SYMPOSIUM ON WIRELESS AND PERVASIVE COMPUTING》 * |
刘少强等: "基于节点间输出比值稳定的WSN分布式数据压缩", 《PROCEEDINGS OF THE 32ND CHINESE CONTROL CONFERENCE》 * |
Cited By (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108347709B (en) * | 2017-01-21 | 2020-07-24 | 中山大学 | Self-adaptive data transmission method and system of ultra-low power consumption traceability system |
CN108347709A (en) * | 2017-01-21 | 2018-07-31 | 中山大学 | A kind of system adaptive recognition method and system of super low-power consumption traceability system |
CN106961656A (en) * | 2017-02-23 | 2017-07-18 | 南京邮电大学 | A kind of wireless sensor network data Forecasting Methodology |
CN106961656B (en) * | 2017-02-23 | 2020-04-07 | 南京邮电大学 | Wireless sensor network data prediction method |
CN107071800B (en) * | 2017-03-01 | 2019-10-18 | 北京邮电大学 | A kind of cluster wireless sensor network method of data capture and device |
CN107071800A (en) * | 2017-03-01 | 2017-08-18 | 北京邮电大学 | A kind of cluster wireless sensor network method of data capture and device |
CN107105394B (en) * | 2017-05-11 | 2020-06-30 | 江苏食品药品职业技术学院 | Building safety monitoring system based on wireless sensor network |
CN107105394A (en) * | 2017-05-11 | 2017-08-29 | 江苏食品药品职业技术学院 | Building safety monitoring system based on wireless sensor network |
CN112039934A (en) * | 2019-06-03 | 2020-12-04 | 大唐移动通信设备有限公司 | Information feedback method, feedback information processing method and device |
CN110505597A (en) * | 2019-07-31 | 2019-11-26 | 北京邮电大学 | A kind of data transmission method of wireless sensor network |
CN111010704A (en) * | 2019-12-03 | 2020-04-14 | 沈阳化工大学 | Underwater wireless sensor network data prediction optimization method based on exponential smoothing |
CN110996304A (en) * | 2019-12-04 | 2020-04-10 | 深圳信可通讯技术有限公司 | Method and system for inquiring and calculating edge data of low-power-consumption Internet of things |
CN111262948A (en) * | 2020-02-18 | 2020-06-09 | 广东大橘果业有限公司 | Planting intelligent management system based on internet |
CN111311427A (en) * | 2020-02-18 | 2020-06-19 | 广东大橘果业有限公司 | Planting intelligent management system based on internet |
CN111726768A (en) * | 2020-06-16 | 2020-09-29 | 天津理工大学 | Edge-oriented computation reliable data collection method based on compressed sensing |
CN115695564A (en) * | 2022-12-30 | 2023-02-03 | 深圳市润信数据技术有限公司 | Efficient transmission method for data of Internet of things |
CN115695564B (en) * | 2022-12-30 | 2023-03-10 | 深圳市润信数据技术有限公司 | Efficient transmission method of Internet of things data |
CN115835060A (en) * | 2023-02-17 | 2023-03-21 | 云南瀚哲科技有限公司 | Data acquisition method for industrial and agricultural Internet of things |
CN115835060B (en) * | 2023-02-17 | 2023-05-05 | 云南瀚哲科技有限公司 | Data acquisition method of industrial and agricultural Internet of things |
Also Published As
Publication number | Publication date |
---|---|
CN106341842B (en) | 2020-05-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106341842A (en) | Method and device for transmitting data of wireless sensor network | |
CN106304191B (en) | A kind of data receiver method and device based on cluster structured radio sensor network | |
CN107871164B (en) | Fog computing environment personalized deep learning method | |
Li et al. | Minimizing packet expiration loss with path planning in UAV-assisted data sensing | |
Li et al. | INCOME: Practical land monitoring in precision agriculture with sensor networks | |
CN105959987A (en) | Data fusion algorithm for improving energy utilization rate and service performance of wireless sensor network | |
Geng et al. | An Agricultural Monitoring System Based on Wireless Sensor and Depth Learning Algorithm. | |
CN102244882A (en) | Mobility-agent-based intelligent data acquisition method for wireless sensor network | |
CN102612080B (en) | Minimum-energy-consumption self-adaption convergent routing method on basis of second-generation wavelet zero tree encoding | |
CN108924897A (en) | A kind of mobile sink paths planning method based on deeply learning algorithm | |
CN115530054A (en) | Irrigation control method and device, electronic equipment and storage medium | |
CN103210817A (en) | Automatic farmland irrigation system | |
Shouyi et al. | Design of wireless multi-media sensor network for precision agriculture | |
CN103957582A (en) | Wireless sensor network self-adaptation compression method | |
CN107332631A (en) | A kind of method of use multi-attribute group decision making theoretical evaluation link-quality assessment models | |
CN104486807A (en) | Routing method of small-scale wireless sensor network | |
CN108289285A (en) | A kind of ocean wireless sensor network is lost data and is restored and reconstructing method | |
Sinha et al. | A multi-level strategy for energy efficient data aggregation in wireless sensor networks | |
Yao et al. | Research and development precision irrigation control system in agricultural | |
CN102065510A (en) | Wireless sensor network node adaptive power control system and method based on fuzzy logic | |
Palanisamy et al. | Urban wastewater treatment for high yielding in agriculture through smart irrigation system | |
CN108376099A (en) | A kind of mobile terminal computation migration method of optimization time delay and efficiency | |
CN117634738B (en) | Internet of things agricultural environment-friendly integrated platform based on artificial intelligence | |
CN101394418B (en) | General computation data processing control method based on mobile proxy | |
CN104469797A (en) | Method for generating sequence prediction on basis of farmland wireless network intra-cluster data sparsity |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200501 |