CN113660614B - Distributed compressed sensing method of wireless sensor network based on Internet of things - Google Patents
Distributed compressed sensing method of wireless sensor network based on Internet of things Download PDFInfo
- Publication number
- CN113660614B CN113660614B CN202110965561.5A CN202110965561A CN113660614B CN 113660614 B CN113660614 B CN 113660614B CN 202110965561 A CN202110965561 A CN 202110965561A CN 113660614 B CN113660614 B CN 113660614B
- Authority
- CN
- China
- Prior art keywords
- sensor network
- node
- wireless sensor
- cluster head
- cluster
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/02—Services making use of location information
- H04W4/029—Location-based management or tracking services
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W40/00—Communication routing or communication path finding
- H04W40/24—Connectivity information management, e.g. connectivity discovery or connectivity update
- H04W40/32—Connectivity information management, e.g. connectivity discovery or connectivity update for defining a routing cluster membership
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W52/00—Power management, e.g. TPC [Transmission Power Control], power saving or power classes
- H04W52/02—Power saving arrangements
- H04W52/0209—Power saving arrangements in terminal devices
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W84/00—Network topologies
- H04W84/18—Self-organising networks, e.g. ad-hoc networks or sensor networks
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Mobile Radio Communication Systems (AREA)
- Arrangements For Transmission Of Measured Signals (AREA)
Abstract
The application provides a distributed compressed sensing method of a wireless sensor network based on the Internet of things, which comprises the steps of dividing a plurality of sensor network nodes of the wireless sensor network into a plurality of clusters; aiming at each cluster head node, enabling the sensor network node in the cluster where the cluster head node is located to perform compressed sensing sampling, and enabling the cluster head node to collect sampling data and perform secondary compression so as to transmit the compressed data after the secondary compression to a sink node; and the aggregation node performs joint reconstruction on the compressed data to obtain corresponding reconstructed data. The two-stage compression mode is used, distributed compressed sensing is realized substantially, energy consumption of the wireless sensor network can be further reduced compared with the current compressed sensing technology, and the number of required measurement times in the period can be reduced. The distributed compressed sensing sampling can also improve the computing capacity and speed of the wireless sensor network, thereby saving a large amount of node energy and being beneficial to prolonging the life cycle of the whole wireless sensor network.
Description
Technical Field
The application relates to the technical field of the Internet of things, in particular to a distributed compressed sensing method of a wireless sensor network based on the Internet of things.
Background
With the continuous development of scientific technology, technologies such as the internet of things and big data have become popular concepts and are widely applied in various fields. Moreover, with the continuous development of the technology of the internet of things, a plurality of things are connected to form a larger network, so that the interconnection of everything is realized, and more information data is provided for the further development of human beings.
In the application of the internet of things, two key technologies are provided, namely a sensor technology and an embedded technology. The sensor is used as an important means for information and data acquisition, and forms three major pillars of an information technology together with a communication technology and a computer technology. In addition, with the development of wireless electronic communication technology, wireless sensor networks have become a new research hotspot. In particular, considering that a plurality of sensor nodes with low energy consumption, low cost and low storage capacity are formed in a multi-hop self-organizing mode, and information data required by a user is transmitted to the user through a wireless network.
The wireless sensor network is widely applied in many fields, and particularly, compared with the traditional sensor, the wireless sensor network can cooperatively work in a severe environment to collect required specific data, thereby providing more convenience for life or work of people. However, the wireless sensor network also has some disadvantages, such as low network reliability in a strong electromagnetic environment and large energy consumption of the sensor nodes during operation.
Compressive sensing is a technique for solving sparse solutions of an underdetermined linear system, which is commonly used to acquire and reconstruct sparse or compressible signals in signal processing of electrical engineering. Once the compressive sensing theory is put forward, it has attracted extensive attention in academia and industry, especially in the fields of information theory, image processing, earth science, optical/microwave imaging, fuzzy recognition, wireless engineering and biological engineering. In addition, in the sampling process, the sampling, conversion and compression of the signals can be combined into one, so that the information processing capacity is enhanced.
Therefore, the compressed sensing technology is applied to the wireless sensor network based on the Internet of things, the defects of the wireless sensor network can be overcome, the Internet of things technology can be further improved, the processing capacity of collected information is enhanced, the information data collecting capacity of the Internet of things is improved, and more information is provided for other layers of the Internet of things framework.
With the continuous development of wireless sensor networks, wireless sensor networks are widely applied in various fields, and help people to acquire accurate data information in monitoring work. In the internet of things, the wireless sensor network is more applied to node energy consumption.
Wang Jianning et al monitor CO in the atmosphere using a wireless sensor network 2 Concentration; the result shows that the wireless sensor network can transmit the acquired information data to an upper computer software platform, so that the greenhouse environment can be remotely monitored.
Liu Haoran et al propose a wireless sensor network fault-tolerant topology dynamic evolution model capable of balancing network energy consumption; through the comparative analysis of network performance with other classical models, the model is found to balance energy consumption, prolong the life cycle of the network, and have strong network fault tolerance and intrusion fault tolerance.
Zhang Xiaoshuan et al use wireless sensor network to monitor the original salt water well in the potash fertilizer production, the result shows that the detection system can operate effectively, can reflect the operating state and level of the salt well accurately, and has higher communication reliability.
The Luo Busheng applies a routing algorithm to energy consumption calculation of sensor nodes of the balanced Internet of things, cluster head nodes are reasonably selected, and node energy consumption can be well balanced.
Khan et al discovered that multiple applications could share the infrastructure of a wireless sensor network through the research on the virtualization of the wireless sensor network, and made a thorough review and thorough discussion of the latest technologies.
Zhan et al, using an unmanned aerial vehicle as a ground sensor node in a wireless sensor network, obtains a mobile data collector capable of effectively prolonging the service life of the network, and is an effective energy-saving technology.
Yang Hao and Wang Xiwei propose a data acquisition method based on regionalization compressed sensing to realize high-efficiency energy transmission; the result shows that the performance of compressed sensing is superior to that of direct transmission, distributed compressed sensing and a hybrid algorithm, and the feasibility of the design principle and the sampling stop principle of the regional measurement matrix is verified.
Li Yiqing et al, apply compressed sensing to wireless sensor networks, propose a block compressed sensing global reconstruction algorithm based on spatio-temporal correlation; the result shows that the reconstruction algorithm is superior to the traditional compressed sensing algorithm and the block reconstruction algorithm based on the space-time correlation.
In conclusion, the application of compressed sensing in the wireless sensor network of the internet of things is less, the internet of things is an important tool for information data acquisition at present, and ubiquitous links and online services are provided in various fields. Further development of the internet of things requires more accurate information data acquisition, larger storage capacity and lower energy consumption to acquire more information data. Therefore, applying the compressive sensing technology to the wireless sensor network is expected to make up for the deficiency of the wireless sensor network in terms of node storage, calculation, energy consumption and the like, so that the wireless sensor network in the internet of things has stronger information processing capability and better serves human beings. Based on this, the inventor of this application has studied on the aspect of signal transmission and the energy consumption control of thing networking wireless sensor.
Disclosure of Invention
An object of the embodiment of the application is to provide a distributed compressed sensing method for a wireless sensor network based on the internet of things, so that the energy consumption of the wireless sensor network is reduced under the condition that the data is real.
In order to achieve the above object, embodiments of the present application are implemented as follows:
in a first aspect, an embodiment of the present application provides a distributed compressed sensing method for a wireless sensor network based on the internet of things, where the wireless sensor network includes a sink node and a plurality of sensor network nodes, the method includes: dividing a plurality of sensor network nodes of the wireless sensor network into a plurality of clusters, wherein each cluster comprises a cluster head node; for each cluster head node, enabling the sensor network node in the cluster to perform compressed sensing sampling, and collecting sampling data and performing secondary compression on the cluster head node so as to transmit the compressed data after the secondary compression to a sink node; and the aggregation node performs joint reconstruction on the compressed data to obtain corresponding reconstructed data.
In the embodiment of the application, a plurality of sensor network nodes of a wireless sensor network are divided into a plurality of clusters (each cluster comprises a cluster head node); aiming at each cluster head node, enabling the sensor network node in the cluster where the cluster head node is located to perform compressed sensing sampling, and enabling the cluster head node to collect sampling data and perform secondary compression so as to transmit the compressed data after the secondary compression to a sink node; and the aggregation node performs joint reconstruction on the compressed data to obtain corresponding reconstructed data. The two-stage compression mode is utilized, distributed compression sensing is realized substantially, compared with the existing compression sensing technology, the energy consumption of a wireless sensor network can be further reduced, the measurement times required in the period can be reduced, the high occupied requirement in signal transmission is reduced, better service can be provided for the Internet of things in the aspect of sensing layer information acquisition, preparation is made for subsequent application, and wider application is realized. Moreover, the distributed compressed sensing sampling can improve the computing capacity and speed of the wireless sensor network, thereby saving a large amount of node energy and being beneficial to prolonging the life cycle of the whole wireless sensor network.
With reference to the first aspect, in a first possible implementation manner of the first aspect, dividing a plurality of sensor network nodes of the wireless sensor network into a plurality of clusters includes: acquiring one or more event source positions existing in the wireless sensor network in the same time period; judging whether the number of event sources exceeds a number threshold, wherein the number threshold is 3% -10% of the number of all sensor network nodes in the wireless sensor network; if the number of the event sources exceeds a number threshold, a first clustering mode is adopted to divide a plurality of sensor network nodes of the wireless sensor network into a plurality of clusters; and if the number of the event sources does not exceed the number threshold, dividing a plurality of sensor network nodes of the wireless sensor network into a plurality of clusters by adopting a second clustering mode.
In the implementation mode, the number of event source positions existing in the same time period is used as a judgment basis to determine which clustering mode is adopted to cluster the wireless sensor network. Therefore, clustering and actual operation of the wireless sensor network are more flexible, and the information transmission state to be faced by the wireless sensor network can be reflected according to the number of event sources (namely the number of the positions of the event sources), so that a plurality of sensor network nodes of the wireless sensor network are divided into a plurality of clusters by flexibly selecting a proper clustering mode. For example, when the number of event sources is large (exceeds a number threshold), the operating efficiency of the wireless sensor network may be prioritized, thereby reducing the resources occupied by the clusters and the resulting energy consumption. When the number of event sources is small (not exceeding the number threshold), another clustering mode can be adopted to solve the problem of energy consumption imbalance of the nodes of the sensor network in the wireless sensor network caused by the clustering mode adopted when the number of event sources is large, so that the life cycle of the wireless sensor network is effectively prolonged.
With reference to the first possible implementation manner of the first aspect, in a second possible implementation manner of the first aspect, dividing a plurality of sensor network nodes of the wireless sensor network into a plurality of clusters in a first clustering manner includes: for each event source position, determining at least one sensor network node closest to the event source position as a cluster head node from a plurality of sensor network nodes which are not determined as the cluster head nodes temporarily in the wireless sensor network; and clustering a plurality of sensor network nodes of the wireless sensor network by taking the determined plurality of cluster head nodes as a center, wherein each sensor network node belongs to a single cluster at the same time.
In the implementation mode, aiming at each event source position, at least one sensor network node closest to the event source position is determined to be a cluster head node from a plurality of sensor network nodes which are not determined to be cluster head nodes temporarily in the wireless sensor network; and clustering a plurality of sensor network nodes of the wireless sensor network by taking the determined plurality of cluster head nodes as a center. The cluster head nodes can be quickly and effectively determined in the clustering mode, so that the number of clusters can be determined, the rest of work is the division of other sensor network nodes, the clustering of the wireless sensor network can be quickly and effectively completed, the energy consumption caused by clustering is reduced, the information of the position of the event source can be well reported by the clustering, and the accurate positioning of the event source is facilitated.
With reference to the second possible implementation manner of the first aspect, in a third possible implementation manner of the first aspect, clustering a plurality of sensor network nodes of the wireless sensor network with the determined plurality of cluster head nodes as a center includes: determining the cluster of each sensor network node by adopting a first cluster model:
Epay(n，m)＝y(d(N n ，H m ))+z(d(H m ，Sink))，
wherein d is y_max ＝EX(max{d(N n ，H m )})，d z_max ＝max{d(H m ，Sink)}，d z_min ＝min{d(H m Sink) }, epay (N, m) denotes the sensor network node N n Joining cluster head node H m Energy cost of d (N) n ，H m ) Representing sensor network nodes to cluster head node H m Function y is used to realize sensor network node and cluster head node H m Energy costs of d (H) is minimized m Sink) represents a cluster head node H m Distance to Sink node Sink, function z for implementing cluster head node H m Energy consumption cost minimization with Sink node, and sensor network node N n Adding cluster head node H with minimum integral energy consumption cost Epay (n, m) m 。
In this implementation, by using such a node clustering model, the overall energy consumption of the wireless sensor network can be minimized in the case of determining the cluster head node, so as to reduce the energy consumption of the wireless sensor network as much as possible.
With reference to the first possible implementation manner of the first aspect, in a fourth possible implementation manner of the first aspect, dividing a plurality of sensor network nodes of the wireless sensor network into a plurality of clusters in a second clustering manner includes: performing cluster analysis on a plurality of sensor network nodes in the wireless sensor network by adopting a fuzzy clustering algorithm, so as to divide the wireless sensor network into L clusters based on the position information and the residual energy of the sensor network nodes, wherein L is more than or equal to 2 and less than or equal to N, and thus obtaining a target function:
wherein u represents a membership matrix, u li The ith parameter represents the membership degree matrix, the maximum correction value represents the membership degree of the ith sensor network node in the ith cluster, m is a weighting parameter and is more than 1,v l Denotes the cluster center of the ith cluster, A i Representing the ith sensor network node in the wireless sensor network.
In the implementation mode, a fuzzy clustering algorithm is adopted to perform clustering analysis on a plurality of sensor network nodes in the wireless sensor network, and clustering of the wireless sensor network is performed based on the position information and the residual energy of the sensor network nodes, so that clustering can be performed in a relatively balanced mode (the whole energy consumption of the wireless sensor network and the residual energy of the single sensor network node can be considered), and thus the energy consumption balance of the whole wireless sensor network is maintained, the whole energy consumption of the wireless sensor network is reduced, and the life cycle of the wireless sensor network is further prolonged.
With reference to the first aspect, in a fifth possible implementation manner of the first aspect, for each cluster head node, performing compressed sensing sampling on a sensor network node in a cluster in which the cluster head node is located, where the cluster head node collects sampling data and performs secondary compression, so as to transmit compressed data after the secondary compression to a sink node, the method includes: for each cluster head node, enabling N sensor network nodes in a cluster where the cluster head node is located to perform sparse transformation on a sensing information data N-dimensional signal x by using a sparse transformation matrix psi, wherein the sparse representation is x = psi theta; based on this, each cluster head node collects sampling data X with sparse representation of correlation from n sensor network nodes in the cluster where the cluster head node is located 1 ～X n (ii) a The cluster head nodes respectively utilize an observation matrix phi 1 ～Φ n For sampling data X 1 ～X n Respectively projecting to obtain measurement signals Y 1 ～Y n Realizing first-stage compression; using transmission channel to measure signal Y 1 ～Y n Rebuilding is carried out, redundant parts in the measuring signals are eliminated, and rebuilding signals Y 'are obtained' 1 ～Y’ n Realizing two-stage compression; will reconstruct signal Y' 1 ～Y’ n And transmitting the data to the sink node.
In the implementation mode, by closely combining the compressed sensing (primary compression) and the secondary compression, the distributed information processing technology and the compressed sensing technology can be effectively combined, so that the sensing data (namely, sensing information data and N-dimensional signals) in the wireless sensor network is related to time and space, the authenticity of the data can be ensured, and accurate information can be provided for subsequent work. The two-stage compression mode can reduce the measurement times, thereby reducing the high requirement for occupation during signal transmission, providing better service for the Internet of things in the aspect of induction layer information acquisition, preparing for subsequent application and realizing wider application. In addition, the mode of combining the distributed information processing technology and the compressed sensing technology does not carry out communication and coordination during data projection, and carries out reconstruction in a transmission channel, thereby improving the computing capacity and speed of the system and saving a large amount of node energy.
With reference to the fifth possible implementation manner of the first aspect, in a sixth possible implementation manner of the first aspect, the measurement signal Y is measured by using a transmission channel 1 ～Y n Rebuilding is carried out, redundant parts in the measuring signals are eliminated, and rebuilding signals Y 'are obtained' 1 ～Y’ n And realizing two-stage compression, comprising: measuring signal Y through transmission channel pair using Y = Φ Ψ θ 1 ～Y n And (4) carrying out reconstruction and integration to obtain:
setting column vectors of at least one column in the sparse transformation matrix Ψ as zero vectors, wherein the column vectors set as the zero vectors all belong to the highest frequency part in the sparse transformation matrix Ψ, and the number of the column vectors does not exceed 10% of the number of the whole column vectors in the sparse transformation matrix Ψ; from this, a reconstructed signal Y 'is obtained' 1 ～Y’ n Realizing two-stage compression; correspondingly, the sink node feeds the compressed dataAnd (3) performing line joint reconstruction to obtain corresponding reconstruction data, wherein the method comprises the following steps: aggregation node is based on reconstructed signal Y' 1 ～Y’ n Performing joint reconstruction to obtain sampling data X 1 ～X n One-to-one correspondence of reconstructed signals
In the implementation mode, the two-stage compression is realized by using the mode, the transmission energy consumption can be effectively reduced, and the sampling data X can be effectively reconstructed by using the joint reconstruction technology 1 ～X n One-to-one correspondence of reconstructed signals
Even if sensing data of the sensor network nodes are wrong in the transmission process in actual operation, the sink node can recover the wrong data (a joint reconstruction technology can be used), and certain fault tolerance is achieved.
With reference to the fifth possible implementation manner of the first aspect, in a seventh possible implementation manner of the first aspect, the signal Y 'is to be reconstructed' 1 ～Y’ n Before transmitting to the sink node, the method further comprises: acquiring a preset energy model, and calculating data transmission energy consumption; if the energy consumption of data transmission does not exceed the set threshold, the cluster head node executes the following steps: will reconstruct signal Y' 1 ～Y’ n And transmitting the data to the sink node.
In this implementation, signal Y 'will be reconstructed' 1 ～Y’ n Before the data are transmitted to the sink nodes, the preset energy model is obtained to calculate data transmission energy consumption, so that the node energy consumption in the wireless sensor network can be effectively controlled, and the life cycle of the wireless sensor network can be prolonged.
With reference to the seventh possible implementation manner of the first aspect, in an eighth possible implementation manner of the first aspect, the energy model is:
E rec (S J )＝w(S J )E elec ，
wherein E is sent (S J ) Indicating transmission energy consumption, E rec (S J ) Represents the reception power consumption, w (S) J ) Indicating the number of bits transmitted by the node, E elec Representing energy consumption per byte transmitted or received, d 0 Representing the distance between nodes, ξ a tunable parameter, E amp Representing the energy consumption per unit distance for transmitting a unit byte, r representing the signal attenuation index, d j，s Indicating the distance d that needs to be transmitted.
In the implementation mode, the energy consumption of the nodes in the wireless sensor network can be effectively controlled by using the energy model, and the setting of the adjustable parameter xi ensures that the energy model can keep high applicability when being applied to the wireless sensor network under different scenes and conditions.
In order to make the aforementioned objects, features and advantages of the present application more comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments of the present application will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and that those skilled in the art can also obtain other related drawings based on the drawings without inventive efforts.
Fig. 1 is a schematic diagram of a wireless sensor network based on the internet of things according to an embodiment of the present application.
FIG. 2 is a flow chart of compressed sensing.
Fig. 3 is a theoretical framework diagram of compressed sensing.
Fig. 4 is a flowchart of a distributed compressed sensing method for a wireless sensor network based on the internet of things according to an embodiment of the present application.
Fig. 5 is a schematic diagram illustrating a relationship between the number of nodes in the sensor network and the energy consumption of the nodes.
Fig. 6 is a schematic diagram of the effect of errors after reconstruction.
Icon: 100-a wireless sensor network; 110-a sensor network node; 120-cluster head node; 130-sink nodes; 200-a processing center.
Detailed Description
The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application.
Referring to fig. 1, fig. 1 is a schematic diagram of a wireless sensor network 100 based on the internet of things according to an embodiment of the present disclosure. In this embodiment, the wireless sensor network 100 may include a sink node 130 and a plurality of sensor network nodes 110. During the operation of the wireless sensor network 100, the wireless sensor network 100 needs to be clustered, and the clustering process needs to determine a plurality of cluster head nodes 120 from the plurality of sensor network nodes 110, where each cluster head node 120 corresponds to a cluster (a cluster usually includes a plurality of sensor network nodes 110). For each cluster, a plurality of sensor network nodes 110 in the cluster are used for data sampling, and the cluster head node 120 is used for collecting the sampled data of the sensor network nodes 110 in the cluster where the cluster head node is located and transmitting the data to the sink node 130. The sink node 130 is configured to process (e.g., reconstruct) the received data, and in addition, the sink node 130 may further transmit the data to the processing center 200 through a transmission network to implement data application, which is not limited herein.
In order to reduce the energy consumption of the wireless sensor network 100, the scheme adopts a concept of combining compressed sensing and distributed transmission, and provides a distributed compressed sensing method of the wireless sensor network 100 based on the internet of things, so that the energy consumption of the wireless sensor network 100 is reduced under the condition of ensuring real data.
To facilitate understanding of the present solution, the present embodiment will introduce knowledge of compressed sensing to facilitate understanding of the present solution.
Currently, various wireless sensor network operating systems have become an important component of wireless field platforms. Compressed Sensing (CS), also known as compressed sampling or sparse sampling, is an information acquisition and signal transmission scheme proposed by Candes, romberg and Tao et al. CS breaks the limitations of Shannon's sampling theory and recovers the signal even with a small number of samples. However, in practical applications, it is also found that the amount of information received by the terminal is greatly reduced, and the reliability and stability of the recovered data are affected to some extent. The flow of compressed sensing is shown in fig. 2, and the theoretical framework of compressed sensing is shown in fig. 3.
Therefore, sparse signals, uncorrelated observations and the ability to recover signals become critical in sensing technology. Compressed sensing can also be divided into traditional compressed sensing and distributed compressed sensing, but no matter traditional compressed sensing and distributed compressed sensing, sparsity of signals is a precondition for signal processing, and compressibility of the signals is determined. It is generally defined as:
||θ|| 0 ＝K，K＜＜N， (1)
wherein, theta is a sparse signal, and theta is belonged to R N 。
If sparsity is set to K, then there will be K basis vectors V i I is 1. Ltoreq. K and alpha i The position of the basis vector in the sparse signal is represented, and therefore, the sparse signal can be represented as:
normally, the other basis vectors are too small to count, and therefore, the signal can be considered sparse of order K. For K sparsity in a strict sense, the algorithm has a certain error, and an error threshold is given as epsilon, at this time, a K sparse signal can be expressed as:
in the internet of things, signals are generally considered to be not sparse, but can be considered to be sparse after being transformed. The sparse transform matrix for the N-dimensional signal x is Ψ, whose sparse representation is:
x＝Ψθ， (4)
in compressive sensing, non-correlated observations can reduce the dimensionality of signals and send the signals to other nodes, and finally energy conservation is achieved. In general, the original signal is assumed to be x, x ∈ R N And projecting through an observation matrix phi to finally obtain a measurement signal y, y epsilon R N The mathematical expression is as follows:
y＝Φx (5)
as can be derived from equations (4) and (5), Ψ and Φ are uncorrelated. In the description, in order to more conveniently express the observation effect, the product of Ψ and Φ is generally used as the sensing matrix a CS The requirements are as follows:
wherein, delta K To constrain constant, δ K ∈(0，1)。
When observing the original signal, the observation times of the compressed sensing technology are far less than the dimension. Therefore, when the signal is recovered and reconstructed, an underdetermined system of equations is required to solve this problem:
where l0 represents the minimum norm for solving the underdetermined equation.
In compressed sensing, a sparse signal is not directly measured generally, and the signal needs to be projected to a coefficient space to obtain a corresponding sparse coefficient, so that the compression operation of the signal is realized. The smaller the inner product value of the sparse basis of the observation matrix is, the smaller the correlation is; the larger the inner product, the greater the correlation.
If the matrix Φ can satisfy the K-order finite equidistant property:
thus, delta K The method ensures that different signals cannot be projected to the same frequency domain space, and takes all minimum values which enable the K-order sparse signals not to spread so as to ensure the accurate reconstruction of the signals.
In the optimization research of the wireless sensor network, a distributed information processing technology and a compressed sensing technology are combined. At this point, the amount of data required for signal measurement in the projection may be much smaller than that required for conventional signals. Furthermore, the low-dimensional data may be converted to high-dimensional data or the high-dimensional data may be converted to low-dimensional data with no distortion of the data. The sensory data in a wireless sensor network is both time and space dependent. Therefore, the combination of the two technologies can ensure the authenticity of data and provide accurate information for subsequent work; and the computing capacity and speed of the system (wireless sensor network) can be improved, and a large amount of node energy can be saved.
Based on the above, the inventor of the present application provides a distributed compressed sensing method for a wireless sensor network based on the internet of things. Referring to fig. 4, fig. 4 is a flowchart of a distributed compressed sensing method for a wireless sensor network based on the internet of things according to an embodiment of the present disclosure. In this embodiment, the distributed compressed sensing method for the wireless sensor network based on the internet of things may include step S10, step S20, and step S30.
In this embodiment, in order to implement the distributed compressed sensing of the wireless sensor network, the wireless sensor network may be initialized, and then step S10 is executed.
Step S10: and dividing a plurality of sensor network nodes of the wireless sensor network into a plurality of clusters, wherein each cluster comprises a cluster head node.
In this embodiment, the wireless sensor network may divide a plurality of sensor network nodes therein into a plurality of clusters, and each cluster includes a cluster head node (for maintaining communication with the sink node and other nodes).
For example, in order to balance the operation efficiency and the overall energy consumption of the wireless sensor network, in the embodiment, one or more event source locations existing in the wireless sensor network within the same time period (e.g., within 1 second, within 10 seconds, within 1 minute, etc.) may be obtained. Then, it may be determined whether the number of event sources (i.e., the number of event source locations existing in the time period) exceeds a number threshold (which may be set to 3% to 10% of the number of all sensor network nodes in the wireless sensor network, and preferably 5% to 8% of the number of all sensor network nodes in the wireless sensor network for the existing application scenarios, such as an ecological monitoring scenario, an agricultural monitoring scenario, a logistics monitoring scenario, an equipment monitoring scenario, and a production line monitoring scenario), so as to consider both the operating efficiency and the overall energy consumption balance of the wireless sensor network as much as possible (if the number of event source locations in the same time period is large, accuracy and timeliness of information acquisition and data transmission should be considered first, and if the number of event source locations in the same time period is small, energy consumption balance of sensor network nodes in the wireless sensor network should be considered first, so as to improve the life cycle of the wireless sensor network).
Therefore, if the number of event sources exceeds the number threshold, the wireless sensor network can divide a plurality of sensor network nodes in the wireless sensor network into a plurality of clusters by adopting a first clustering mode; if the number of event sources does not exceed the number threshold, a second clustering mode can be adopted to divide a plurality of sensor network nodes in the event sources into a plurality of clusters.
The number of event source positions existing in the same time period is used as a judgment basis, and the clustering mode is determined to be adopted for clustering the wireless sensor network. Therefore, clustering and actual operation of the wireless sensor network are more flexible, and the information transmission state to be faced by the wireless sensor network can be reflected according to the number of event sources (namely the number of the positions of the event sources), so that a plurality of sensor network nodes of the wireless sensor network are divided into a plurality of clusters by flexibly selecting a proper clustering mode. For example, when the number of event sources is large (exceeds a number threshold), the operating efficiency of the wireless sensor network may be prioritized, thereby reducing the resources occupied by the clusters and the resulting energy consumption. When the number of event sources is small (not exceeding the number threshold), another clustering mode can be adopted to overcome the problem of energy consumption imbalance of the sensor network nodes in the wireless sensor network caused by the clustering mode adopted when the number of event sources is large, so that the life cycle of the wireless sensor network is effectively prolonged.
For example, the first clustering manner may be:
for each event source position, at least one sensor network node closest to the event source position can be determined as a cluster head node from a plurality of sensor network nodes of the wireless sensor network which are not determined as the cluster head nodes temporarily, then the plurality of sensor network nodes of the wireless sensor network are clustered by taking the determined plurality of cluster head nodes as centers, and each sensor network node belongs to a single cluster at the same time.
At least one sensor network node closest to the source position of the event is determined to be a cluster head node, and the cluster head node can be flexibly selected according to actual conditions. For example, for a relatively important location area, in order to ensure accuracy, reliability, comprehensiveness, and the like of information, a plurality of (e.g., 3, 5, and the like) sensor network nodes closest to the event source location may be determined as cluster head nodes; for a location area that is relatively unimportant, the 1 sensor network node closest to the event source location may be determined to be a cluster head node, which is not limited herein.
The clustering mode can quickly and effectively determine the cluster head nodes so as to determine the number of clusters, and the rest of work is only the division of other sensor network nodes, so that the clustering of the wireless sensor network can be quickly and effectively finished, and the energy consumption brought by clustering is reduced. Moreover, the clustering can well report the information at the position of the event source, and is favorable for accurate positioning of the event source and continuous tracking of the event source.
Specifically, a first cluster model may be adopted to determine the cluster to which each sensor network node belongs:
Epay(n，m)＝y(d(N n ，H m ))+z(d(H m ，Sink))， (10)
wherein d is y_max ＝EX(max{d(N n ，H m )})，d z_max ＝max{d(H m ，Sink)}，d z_min ＝min{d(H m Sink) }, epay (N, m) denotes the sensor network node N n Joining cluster head node H m Energy cost of d (N) n ，H m ) Representing sensor network nodes to cluster head node H m Function y is used to realize sensor network node and cluster head node H m Energy costs of d (H) is minimized m Sink) represents a cluster head node H m Distance to Sink node Sink, function z for implementing cluster head node H m Energy consumption cost with Sink node is minimized, and sensor network node N n Adding cluster head node H with minimum integral energy consumption cost Epay (n, m) m 。
By adopting the node clustering model, the overall energy consumption of the wireless sensor network can be minimized under the condition of determining the cluster head nodes, so that the energy consumption of the wireless sensor network is reduced as much as possible.
For example, the second clustering manner may be:
clustering analysis is carried out on a plurality of sensor network nodes in the wireless sensor network by adopting a fuzzy clustering algorithm, so that the wireless sensor network is divided into L clusters based on the position information and the residual energy of the sensor network nodes, and L satisfies the condition that L is more than or equal to 2 and less than or equal to N, and thus an objective function is obtained:
wherein u represents a membership matrix, u li The ith parameter represents the membership degree matrix, the maximum correction value represents the membership degree of the ith sensor network node in the ith cluster, m is a weighting parameter and is more than 1,v l Denotes the cluster center of the ith cluster, A i Representing the ith sensor network node in the wireless sensor network.
The fuzzy clustering algorithm is adopted to perform clustering analysis on a plurality of sensor network nodes in the wireless sensor network, and clustering of the wireless sensor network is performed based on the position information and the residual energy of the sensor network nodes, so that clustering can be performed in a relatively balanced mode (the whole energy consumption of the wireless sensor network and the residual energy of the single sensor network node can be considered), the energy consumption balance of the whole wireless sensor network is kept, the whole energy consumption of the wireless sensor network is reduced, and the life cycle of the wireless sensor network is further prolonged.
Overall, the first clustering method focuses more on the operation efficiency (such as timeliness, reliability and comprehensiveness of message acquisition and data transmission) of the wireless sensor network in an emergency (a large number of event source locations), and thus a very simple and fast clustering method (determined by the distance from the event source location) which is very effective for message acquisition and data transmission is adopted, but this method has the disadvantage that the residual energy of the nodes is not considered, which is not favorable for the overall energy consumption balance of the wireless sensor network. And the second clustering mode is to perform clustering on the wireless sensor network by using the position information of the sensor network nodes and the residual energy under the condition of preferentially considering the overall energy consumption balance of the wireless sensor network under the normal condition (the number of the source positions of the event is small), so that the sensor network nodes with higher residual energy can be determined as cluster head nodes as far as possible under the condition of keeping the energy consumption of the whole wireless sensor network low (the cluster head nodes are responsible for data collection in the whole cluster and communication with other nodes, and the energy required to be consumed is higher), thereby being beneficial to keeping the overall energy consumption balance of the wireless sensor network and prolonging the life cycle of the whole wireless sensor network as far as possible. Therefore, the first clustering mode and the second clustering mode are matched for use, the running efficiency and the whole energy consumption balance of the wireless sensor network can be considered, and the life cycle of the whole wireless sensor network can be prolonged.
In addition, for the clustering of the wireless sensor network, the number of the sensor network nodes in each cluster is preferably 10-30 for the first clustering mode, and the number of the sensor network nodes in each cluster can be more for the second clustering mode, so that the number of the cluster head nodes is reduced, and the energy consumption is reduced.
After the clustering of the wireless sensor network is completed, step S20 may be executed.
Step S20: and aiming at each cluster head node, performing compressed sensing sampling on the sensor network node in the cluster where the cluster head node is positioned, and collecting sampling data and performing secondary compression on the sampling data by the cluster head node so as to transmit the compressed data subjected to the secondary compression to the sink node.
In this embodiment, for each cluster head node, N sensor network nodes in the cluster where the cluster head node is located may perform sparse transformation on the sensing information data N-dimensional signal x by using the sparse transformation matrix Ψ, where the sparse representation is x = Ψ θ. Namely, the n sensor network nodes in the cluster are enabled to collect data in a compressed sensing manner, and the cluster head node is used for collecting the data collected by each sensor network node.
Based on this, each cluster head node can collect sampling data X with sparse representation of correlation from n sensor network nodes in the cluster where the cluster head node is located 1 ～X n . Here, data X is sampled 1 ～X n It can be understood as an acquired signal that can be sparsely represented.
And the cluster head nodes can respectively utilize the observation matrix phi 1 ～Φ n For sampling data X 1 ～X n Respectively projecting to obtain measurement signals Y 1 ～Y n And realizing one-stage compression. Then, the transmission channel is used again to measure the signal Y 1 ～Y n Rebuilding is carried out, redundant parts in the measuring signals are eliminated, and rebuilding signals Y 'are obtained' 1 ～Y’ n Two-stage compression is implemented, so that signal Y 'is reconstructed' 1 ～Y’ n (i.e., the compressed data) is transmitted to the sink node.
By tightly combining the compressed sensing (primary compression) and the secondary compression, the distributed information processing technology and the compressed sensing technology can be effectively combined, so that the sensing data (namely, sensing information data and N-dimensional signals) in the wireless sensor network are related to time and space, the authenticity of the data can be ensured, and accurate information can be provided for subsequent work. The two-stage compression mode can reduce the measurement times, thereby reducing the high requirement for occupation during signal transmission, providing better service for the Internet of things in the aspect of induction layer information acquisition, preparing for subsequent application and realizing wider application. In addition, the mode of combining the distributed information processing technology and the compressed sensing technology does not carry out communication and coordination during data projection, and carries out reconstruction in a transmission channel, thereby improving the computing capacity and speed of the system and saving a large amount of node energy.
Illustratively, the measurement signal Y is measured by means of a transmission channel 1 ～Y n Rebuilding is carried out, redundant parts in the measuring signals are eliminated, and rebuilding signals Y 'are obtained' 1 ～Y’ n The specific way to realize the second-stage compression may be:
measuring signal Y through transmission channel pair using Y = Φ Ψ θ (see equation 7) 1 ～Y n And (3) reconstructing and integrating (namely sorting the data and further representing the data in a matrix form):
then, the column vectors of at least one column in the sparse transform matrix Ψ are set as zero vectors, wherein the column vectors set as zero vectors all belong to the highest frequency part in the sparse transform matrix Ψ and are not more than 10% of the total column vectors in the sparse transform matrix Ψ (of course, the values can be adjusted according to different situations, and are not specifically limited), thereby obtaining the reconstructed signal Y' 1 ～Y’ n To realize two-stage compression. Therefore, under the condition of ensuring the accuracy of data reconstruction as much as possible, the data can be further compressed, and the transmission energy consumption is reduced. Wherein, the form of the integrated sparse transform matrix can be understood as:
the examples are given for ease of understanding only and are not intended to be limiting.
The two-stage compression is realized by the method, the transmission energy consumption can be effectively reduced, and the sampling data X can be effectively reconstructed by utilizing the joint reconstruction technology 1 ～X n One-to-one correspondence of reconstructed signals
Even if sensing data of the sensor network nodes are wrong in the transmission process in actual operation, the sink node can recover the wrong data (a joint reconstruction technology can be used), and certain fault tolerance is achieved.
After signal Y 'is to be reconstructed' 1 ～Y’ n After transmission to the sink node, step S30 may be performed.
Step S30: and the aggregation node performs joint reconstruction on the compressed data to obtain corresponding reconstructed data.
In this embodiment, the sink node may perform joint reconstruction on the compressed data to obtain corresponding reconstructed data. For example, the sink node may be based on the reconstructed signal Y' 1 ～Y’ n Performing joint reconstruction to obtain sampling data X 1 ～X n One-to-one correspondence of reconstructed signals
In addition, in order to realize the energy consumption control of the signal, the signal Y 'is reconstructed' 1 ～Y’ n Before transmitting to the sink node, a preset energy model can be obtained, and data transmission energy consumption is calculated; if the energy consumption of data transmission does not exceed the set threshold, the cluster head node executes the following steps: will reconstruct signal Y' 1 ～Y’ n And transmitting the data to the sink node. Thus can haveThe energy consumption of the nodes in the wireless sensor network is effectively controlled, and the life cycle of the wireless sensor network is prolonged.
For example, the energy model may be designed as:
E rec (S J )＝w(S J )E elec ， (16)
wherein E is sent (S J ) Indicating transmission energy consumption, E rec (S J ) Represents the reception power consumption, w (S) J ) Indicating the number of bits transmitted by the node, E elec Representing energy consumption per byte transmitted or received, d 0 Representing the distance between nodes, xi represents an adjustable parameter (the value is more than or equal to 1), E amp Representing the energy consumption per unit distance for transmitting a unit byte, r representing the signal attenuation index, d j，S Indicating the distance d that needs to be transmitted.
By utilizing the energy model, the node energy consumption in the wireless sensor network can be effectively controlled, and the setting of the adjustable parameter xi ensures that the energy model can keep high applicability when being applied to the wireless sensor network under different scenes and conditions.
In this embodiment, according to different application scenarios of the wireless sensor network, the signal attenuation index r may be designed to be between 2 and 4 (for example, in an environment with good signal transmission conditions, the signal attenuation index r takes a value of 2, in an environment with poor signal transmission conditions, the signal attenuation index r takes a value of 4, and in general, an initial value of the signal attenuation index r may take a value of 2.5). In addition, the adjustable parameter xi is adjusted according to different scenes: for example, for an ecological environment monitoring scene or an agricultural monitoring scene, the environmental change is mild, the sampling frequency of the sensor network node is generally low, the working environment of the sensor network node is relatively mild, and the value of the adjustable parameter ξ can be smaller (ξ can be 1.0-1.1). And aiming at equipment monitoring scenes, production line monitoring scenes and the like, the environment changes violently, the sampling frequency of sensor network nodes is generally high, the working environment of the sensor network nodes is relatively extreme, the energy consumption is higher, and therefore the adjustable parameter xi can be higher (xi can be 1.1-1.3).
Referring to fig. 5 and 6, fig. 5 is a schematic diagram illustrating a relationship between the number of nodes in the sensor network and energy consumption of the nodes, and fig. 6 is a schematic diagram illustrating an error effect after reconstruction. As can be seen from the relationship curves in fig. 5 and fig. 6, the distributed compressive sensing technology combines the distributed acquisition and compressive sensing technology, which can save the node energy consumption to a greater extent, and the error after information reconstruction is small. And, as can be seen from fig. 5, the number threshold (3% to 10% of the number of all sensor network nodes in the wireless sensor network) for determining the clusters, the number of sensor network nodes (preferably 10 to 30) in each cluster in the first clustering manner, and the number of sensor network nodes (more than 30, preferably 40 to 50) in each cluster in the second clustering manner can all save energy consumption well.
In summary, an embodiment of the present application provides a distributed compressed sensing method for a wireless sensor network based on the internet of things, which divides a plurality of sensor network nodes of the wireless sensor network into a plurality of clusters (each cluster includes a cluster head node); aiming at each cluster head node, enabling the sensor network node in the cluster where the cluster head node is located to perform compressed sensing sampling, and enabling the cluster head node to collect sampling data and perform secondary compression so as to transmit the compressed data after the secondary compression to a sink node; and the aggregation node performs joint reconstruction on the compressed data to obtain corresponding reconstructed data. The two-stage compression mode is utilized, distributed compression sensing is realized substantially, compared with the existing compression sensing technology, the energy consumption of a wireless sensor network can be further reduced, the measurement times required in the period can be reduced, the high occupied requirement in signal transmission is reduced, better service can be provided for the Internet of things in the aspect of sensing layer information acquisition, preparation is made for subsequent application, and wider application is realized. Moreover, the distributed compressed sensing sampling can improve the computing capacity and speed of the wireless sensor network, thereby saving a large amount of node energy and being beneficial to prolonging the life cycle of the whole wireless sensor network.
In this document, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions.
The above description is only an example of the present application and is not intended to limit the scope of the present application, and various modifications and changes may be made to the present application by those skilled in the art. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present application shall be included in the protection scope of the present application.
Claims (7)
1. A distributed compressed sensing method of a wireless sensor network based on the Internet of things is characterized in that the wireless sensor network comprises a sink node and a plurality of sensor network nodes, and the method comprises the following steps:
dividing a plurality of sensor network nodes of the wireless sensor network into a plurality of clusters, wherein each cluster comprises a cluster head node;
for each cluster head node, enabling the sensor network node in the cluster to perform compressed sensing sampling, and collecting sampling data and performing secondary compression on the cluster head node so as to transmit the compressed data after the secondary compression to a sink node;
the aggregation node performs combined reconstruction on the compressed data to obtain corresponding reconstructed data;
wherein, divide a plurality of sensor network nodes of the wireless sensor network into a plurality of clusters, include:
acquiring one or more event source positions existing in the wireless sensor network in the same time period; judging whether the number of event sources exceeds a number threshold value, wherein the number threshold value is any value of 3% -10% of the number of all sensor network nodes in the wireless sensor network; if the number of the event sources exceeds a number threshold, a first clustering mode is adopted to divide a plurality of sensor network nodes of the wireless sensor network into a plurality of clusters; if the number of the event sources does not exceed the number threshold, dividing a plurality of sensor network nodes of the wireless sensor network into a plurality of clusters by adopting a second clustering mode;
aiming at each cluster head node, the sensor network node in the cluster where the cluster head node is located is enabled to carry out compressed sensing sampling, and the cluster head node collects sampling data and carries out secondary compression so as to transmit the compressed data after the secondary compression to the sink node, wherein the method comprises the following steps:
aiming at each cluster head node, enabling N sensor network nodes in a cluster where the cluster head node is located to perform sparse transformation on a sensing information data N-dimensional signal x by using a sparse transformation matrix psi, wherein the sparse representation is x = psi theta, and theta is a sparse signal; based on this, each cluster head node collects sampling data X with sparse representation of correlation from n sensor network nodes in the cluster where the cluster head node is located 1 ～X n (ii) a The cluster head nodes respectively utilize an observation matrix phi 1 ～Φ n For sampling data X 1 ～X n Respectively projecting to obtain measurement signals Y 1 ～Y n Realizing first-stage compression; using transmission channel to measure signal Y 1 ～Y n Rebuilding is carried out, redundant parts in the measuring signals are eliminated, and rebuilding signals Y 'are obtained' 1 ～Y’ n Realizing two-stage compression; the reconstructed signal Y 'is' 1 ～Y’ n Transmitting to a sink node;
using transmission channel to measure signal Y 1 ～Y n Rebuilding is carried out, redundant parts in the measuring signals are eliminated, and rebuilding signals Y 'are obtained' 1 ～Y’ n And realizing two-stage compression, comprising:
measuring signal Y through transmission channel pair using Y = Φ Ψ θ 1 ～Y n And (4) carrying out reconstruction and integration to obtain:
wherein, y 1 ，y n Respectively representing the measurement signal of line 1 and the measurement signal of line n, [ theta ] 1 ，θ n Respectively representing the sparse signal of the 1 st row and the sparse signal of the n-th row;
setting column vectors of at least one column in the sparse transformation matrix psi as zero vectors, wherein the column vectors set as the zero vectors all belong to the highest frequency part in the sparse transformation matrix psi and the number of the column vectors does not exceed 10% of the number of the whole column vectors in the sparse transformation matrix psi; from this, a reconstructed signal Y 'is obtained' 1 ～Y’ n And realizing two-stage compression.
2. The internet of things-based distributed compressed sensing method for the wireless sensor network according to claim 1, wherein the step of dividing a plurality of sensor network nodes of the wireless sensor network into a plurality of clusters in a first clustering manner comprises the steps of:
for each event source position, determining at least one sensor network node closest to the event source position as a cluster head node from a plurality of sensor network nodes which are not determined as the cluster head nodes temporarily in the wireless sensor network;
clustering a plurality of sensor network nodes of the wireless sensor network by taking the determined plurality of cluster head nodes as a center, wherein each sensor network node belongs to a single cluster at the same time.
3. The internet of things-based distributed compressed sensing method for the wireless sensor network, according to claim 2, wherein the clustering of the plurality of sensor network nodes of the wireless sensor network is performed by taking the determined plurality of cluster head nodes as a center, and includes:
determining the cluster of each sensor network node by adopting a first cluster model:
Epay(n，m)＝y(d(N n ，H m ))+z(d(H m ，Sink))，
wherein d is y_max ＝EX(max{d(N n ，H m )})，d z_max ＝max{d(H m ，Sink)}，d z_min ＝min{d(H m Sink) }, epay (N, m) denotes the sensor network node N n Joining cluster head node H m Energy cost of, EX () represents expectation, d (N) n ，H m ) Representing sensor network nodes to cluster head node H m Function y is used to realize sensor network node and cluster head node H m Energy costs of d (H) m Sink) represents a cluster head node H m Distance to Sink node Sink, function z for implementing cluster head node H m Energy consumption cost minimization with Sink node, and sensor network node N n Adding cluster head node H with minimum integral energy consumption cost Epay (n, m) m 。
4. The internet of things-based distributed compressed sensing method for the wireless sensor network according to claim 1, wherein a second clustering mode is adopted to divide a plurality of sensor network nodes of the wireless sensor network into a plurality of clusters, and the method comprises the following steps:
performing cluster analysis on a plurality of sensor network nodes in the wireless sensor network by adopting a fuzzy clustering algorithm, so as to divide the wireless sensor network into L clusters based on the position information and the residual energy of the sensor network nodes, wherein L is more than or equal to 2 and less than or equal to N, thereby obtaining a target function:
wherein u represents a membership matrix, u li The ith parameter represents a membership matrix, the maximum correction value of the ith parameter represents the membership of the ith sensor network node in the ith cluster, m is a weighting parameter and is more than 1,v l Is shown asCluster center of l clusters, A i Representing the ith sensor network node in the wireless sensor network.
5. The internet of things-based distributed compressed sensing method for the wireless sensor network, according to claim 1, wherein the joint reconstruction of the compressed data by the aggregation node to obtain corresponding reconstructed data comprises:
aggregation node is based on reconstructed signal Y' 1 ～Y’ n Performing joint reconstruction to obtain sampling data X 1 ～X n One-to-one correspondence of reconstructed signals
6. The distributed compressed sensing method for the wireless sensor network based on the Internet of things of claim 1, wherein the signal Y 'is reconstructed' 1 ～Y’ n Before transmitting to the sink node, the method further comprises:
acquiring a preset energy model, and calculating data transmission energy consumption;
if the energy consumption of data transmission does not exceed the set threshold, the cluster head node executes the following steps: will reconstruct signal Y' 1 ～Y’ n And transmitting the data to the sink node.
7. The method for distributed compressed sensing of the wireless sensor network based on the internet of things according to claim 6, wherein the energy model is as follows:
E rec (S J )＝w(S J )E elec ，
wherein E is sent (S J ) Represents transmission power consumption, E rec (S J ) Represents the reception power consumption, w (S) J ) Representing nodesNumber of bits transmitted, E elec Representing energy consumption per byte transmitted or received, ξ representing an adjustable parameter, E amp Representing the energy consumption per unit distance for transmitting a unit byte, r representing the signal attenuation index, d j，S Indicating the distance d that needs to be transmitted.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110965561.5A CN113660614B (en) | 2021-08-23 | 2021-08-23 | Distributed compressed sensing method of wireless sensor network based on Internet of things |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110965561.5A CN113660614B (en) | 2021-08-23 | 2021-08-23 | Distributed compressed sensing method of wireless sensor network based on Internet of things |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113660614A CN113660614A (en) | 2021-11-16 |
| CN113660614B true CN113660614B (en) | 2022-10-04 |
Family
ID=78491928
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110965561.5A Active CN113660614B (en) | 2021-08-23 | 2021-08-23 | Distributed compressed sensing method of wireless sensor network based on Internet of things |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113660614B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115514789B (en) * | 2022-11-01 | 2023-05-02 | 国网天津市电力公司电力科学研究院 | Compression-sensing vehicle network interaction data lightweight security convergence transmission method and system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103269525A (en) * | 2013-05-11 | 2013-08-28 | 嘉兴市宏联电子科技有限公司 | Multi-sensor data distributed compression method and multi-sensor data distributed compression device for internet of things system |
| CN104618947A (en) * | 2015-02-03 | 2015-05-13 | 中国人民解放军信息工程大学 | Compressive sensing based dynamic clustering wireless sensor network data collecting method and device |
| CN106559799A (en) * | 2016-11-21 | 2017-04-05 | 重庆大学 | Refuse landfill intelligent monitor system and method based on wireless sensor network |
| CN108174411A (en) * | 2017-12-12 | 2018-06-15 | 扬州大学 | Wireless sensor network routing method based on fuzzy logic |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9086499B2 (en) * | 2008-09-23 | 2015-07-21 | Purdue Research Foundation | Clustering protocol for directional sensor networks |
| CN112367675B (en) * | 2020-11-11 | 2022-04-08 | 内蒙古大学 | Wireless sensor network data fusion method and network system based on self-encoder |
-
2021
- 2021-08-23 CN CN202110965561.5A patent/CN113660614B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103269525A (en) * | 2013-05-11 | 2013-08-28 | 嘉兴市宏联电子科技有限公司 | Multi-sensor data distributed compression method and multi-sensor data distributed compression device for internet of things system |
| CN104618947A (en) * | 2015-02-03 | 2015-05-13 | 中国人民解放军信息工程大学 | Compressive sensing based dynamic clustering wireless sensor network data collecting method and device |
| CN106559799A (en) * | 2016-11-21 | 2017-04-05 | 重庆大学 | Refuse landfill intelligent monitor system and method based on wireless sensor network |
| CN108174411A (en) * | 2017-12-12 | 2018-06-15 | 扬州大学 | Wireless sensor network routing method based on fuzzy logic |
Non-Patent Citations (1)
| Title |
|---|
| 压缩感知在无线传感器网络数据融合中的应用研究;张瑞;《中国优秀博硕士学位论文全文数据库(硕士)信息科技辑》;20131215;第三章 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113660614A (en) | 2021-11-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Qin et al. | Federated learning and wireless communications | |
| Atia et al. | Dynamic online-calibrated radio maps for indoor positioning in wireless local area networks | |
| CN102638889B (en) | Indoor wireless terminal positioning method based on Bayes compression sensing | |
| CN104618947B (en) | Dynamic clustering wireless sense network method of data capture and device based on compressed sensing | |
| Zhou et al. | Novel energy-efficient data gathering scheme exploiting spatial-temporal correlation for wireless sensor networks | |
| CN113660614B (en) | Distributed compressed sensing method of wireless sensor network based on Internet of things | |
| Zhang et al. | Medical data fusion algorithm based on internet of things | |
| Al-Qurabat et al. | Important extrema points extraction-based data aggregation approach for elongating the WSN lifetime | |
| CN116599602B (en) | Low-energy-consumption long-distance sonar communication system for ocean monitoring | |
| Chamberland et al. | The impact of fading on decentralized detection in power constrained wireless sensor networks | |
| Luo et al. | Decentralized estimation in an inhomogeneous sensing environment | |
| Wei et al. | The compressed sensing of wireless sensor networks based on Internet of Things | |
| CN101237357B (en) | Online failure detection method for industrial wireless sensor network | |
| CN104038993B (en) | A kind of wireless sensor network node system of selection of blind source separating | |
| Wang et al. | Collaborative event-region and boundary-region detections in wireless sensor networks | |
| Kumar et al. | Localization estimation using artificial intelligence technique in wireless sensor networks | |
| Yu et al. | Multi-attribute missing data reconstruction based on adaptive weighted nuclear norm minimization in IoT | |
| Razzak et al. | Collaborative image compression in wireless sensor networks | |
| Li et al. | Multi-sensor complex network data fusion under the condition of uncertainty of coupling occurrence probability | |
| Chen et al. | The importance of data compression for energy efficiency in sensor networks | |
| Zhai et al. | Data reconstructing algorithm in unreliable links based on matrix completion for heterogeneous wireless sensor networks | |
| Yu et al. | Data anomaly detection and data fusion based on incremental principal component analysis in fog computing | |
| Wajgi et al. | Localization in wireless sensor networks and wireless multimedia sensor networks using clustering techniques | |
| Ngo et al. | Deep learning-based adaptive beamforming for mmWave wireless body area network | |
| CN107787002B (en) | Method for rapidly evaluating information transmission rate in wireless power supply communication |
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 |