Compressed data collection method based on wireless sensor network

Abstract

The invention relates to a compressed data collection method based on a wireless sensor network, and the method comprises the steps: obtaining utility functions of sensor nodes, taking the sensor node with the highest utility function as a cluster head, and transmitting a seed code through a cluster head node, thereby dividing a cluster; taking each non-cluster head node in the cluster as a forward ant, selecting a next node according to a probability selection rule until the next node reaches the cluster head node, returning to the original node, updating pheromones of a round-trip path, and obtaining the round-trip path with the highest pheromones as an optimal path; and obtaining the reading of each sensor, generating # imgabs0 # according to the seed code, sending the product of the reading of each sensor and # imgabs1 # to the corresponding cluster head node according to the optimal path, obtaining an observation value, tracking the observation value, and obtaining an original value. The optimal transmission path in the cluster is searched through the efficient ant colony algorithm, and the method is suitable for a cluster communication mode of a compressed sensing data collection mode.

Description

Technical Field

[0001] The present invention relates to the field of wireless transmission technology, and in particular to a compressed data collection method based on a wireless sensor network. Background Art

[0002] Wireless sensor networks (WSNs) are an important part of the Internet of Things (IoT) and play a vital role in applications such as environmental monitoring, smart homes, and industrial automation. For resource-limited wireless sensor networks, effective data collection methods can significantly reduce energy consumption and extend network lifetime.

[0003] Wireless sensor networks consist of interconnected sensor nodes with flexibility and wide deployment capabilities, enabling them to collect and transmit data in a variety of environments. Such sensor nodes are usually powered by batteries with limited energy, and the main energy consumption occurs during data processing and transmission, among which the communication process accounts for a large part of the total energy consumption of the network, which directly affects the overall efficiency and duration of network operation.

[0004] Optimization of existing wireless sensor network routing methods is a key strategy to extend the lifetime of the network. Clustering strategies involving hierarchical network structures have been proposed to improve energy efficiency. In these strategies, nodes do not communicate directly with the base station, balancing the energy consumption of the entire network and extending its lifetime. This structure also enhances the scalability of the network, manages nodes effectively, and improves the stability of the network. However, clustering strategies bring a key problem: cluster head nodes, which are responsible for data aggregation and transmitting large amounts of data, consume more energy than ordinary nodes. This uneven distribution of energy consumption may cause premature node failure, thereby affecting the overall network performance.

[0005] The existing Low Energy Adaptive Clustering Hierarchy (LEACH) introduces a low-power, random cluster head rotation method. However, it also has obvious disadvantages, such as the uncertainty of the number of cluster heads due to random probability selection, resulting in unreasonable cluster size and distribution, and not considering the residual energy of cluster heads. After that, the LEACH algorithm was improved. REAC-IN solved the problem of isolated nodes caused by unreasonable cluster head allocation, and the network life cycle was increased by 50% compared with the traditional LEACH algorithm. The R-LEACH algorithm considers the ratio between the residual energy and the initial energy of the node, and the cluster head selection is more reasonable, which increases the network life cycle by 66% and the residual energy by 64%. In addition, Singh uses the PSO algorithm to select cluster heads, and uses a semi-distributed method to locate the best cluster head position in the cluster based on residual energy, intra-cluster distance, node degree and possible number of cluster heads in a low-cost manner. Haleem Farman used the multi-criteria decision-making tool of ANP, which comprehensively considered multiple parameters such as energy, distance between adjacent nodes, distance to the regional center, and number of times to reach the cluster head, thereby improving the stability and life of the network. These clustering algorithms take into account the selection of cluster heads as much as possible to optimize the transmission energy consumption in the network, but such methods are difficult to fundamentally reduce the transmission volume of the entire network. Therefore, a compressed data collection method based on wireless sensor networks is urgently needed. Summary of the invention

[0006] In order to solve this problem, the present invention provides a compressed data collection method based on wireless sensor networks. Under the data acquisition framework of compressed sensing, the network structure is redesigned, and an efficient ant colony algorithm is studied to find the optimal transmission path within the cluster, which is suitable for the cluster communication mode of compressed sensing data collection.

[0007] To achieve the above object, the present invention provides the following solutions:

[0008] A compressed data collection method based on a wireless sensor network, comprising:

[0009] Obtaining the utility functions of the sensor nodes, taking the sensor node with the highest utility function as the cluster head node, and using the cluster head node to send a seed code, thereby dividing the cluster;

[0010] Each non-cluster head node in the cluster is used as a forward ant, and the next node is selected according to a probability selection rule until the cluster head node is reached, and the ant goes back and forth to the original node, and the pheromone of the round-trip path is updated, and the round-trip path with the highest pheromone is obtained as the optimal path;

[0011] Get each sensor's own reading x i , and generates itself according to the seed code in, is a floating point number, j represents the cluster number, i represents the sensor number; based on all nodes Together they form the observation matrix Φ∈R m×n Each node compares its own reading with its own The product of is sent to the corresponding cluster head node according to the optimal path to obtain the observation value y j , track the observed value and solve to obtain the original reading x i .

[0012] Optionally, the utility function of the sensor node is:

[0013]

[0014] Among them, UE i represents the utility value of the i-th node in the cluster, i represents a node in cluster k, C k represents the set of all nodes in cluster k, and represents the distance between the farthest and nearest nodes in cluster k from the sink, d i represents the distance from node i to the sink, E i represents the residual energy of node i, E0 represents the initial energy of the node, and w represents the weight of the control index.

[0015] Optionally, dividing the clusters includes: using the cluster head node to send a seed code, and using the non-cluster head nodes to select and join the closest cluster according to the strength of received signals.

[0016] Optionally, acquiring the optimal path includes:

[0017] Step 1: Set pheromone on each forwarding path corresponding to each non-cluster head node;

[0018] Step 2: Use each non-cluster head node in the cluster to send information packets as a forward ant, select the next node according to the probability selection rule, and when reaching the next node, record the current node information and put it into its own information packet until reaching the cluster head node or the forwarding life cycle ends. When the forward ant reaches the cluster head node, calculate the global pheromone, introduce a penalty term to discard inefficient paths, and further send backward ants;

[0019] Step 3: Use the backward ant to carry the global pheromone and perform reverse propagation along the original forward path until it reaches the starting node; when the backward ant reaches the next node, the local pheromone is calculated according to the current node information, and each node updates the pheromone of its own path according to the local pheromone;

[0020] Step 4: Repeat steps 1-3 until the number of iterations is reached, and obtain the forwarding path with the highest pheromone as the optimal path.

[0021] Optionally, selecting the next node according to the probability selection rule includes:

[0022]

[0023] Among them, p k (i, j) represents the next node selected, η(i, j) represents the heuristic information on the path from node i to node j, τ(i, j) represents the pheromone amount on the path from node i to node j, α and β represent the weights between the control heuristic information and the pheromone amount, J k (i) represents the neighbor set of node i, τ(i,u) represents the pheromone amount from node i to node u, u is the neighbor node of i, and in the set J k (i).

[0024] Optionally, the current node information includes: the number of hops, the distance between the current node and the cluster head, and the transmission distance.

[0025] Optionally, calculating the global pheromone includes:

[0026]

[0027] Among them, Δτ total represents the global pheromone, E avg 、E min Represents the average residual energy and minimum residual energy of all nodes on this path, E init represents the initial energy of the node, μ represents the parameter between the two, and d k Indicates the transmission distance, h k represents the number of hops, g(k) represents, represents the distance between the ith node and the cluster head, In the path set g(k) of ant k.

[0028] Optionally, introducing the penalty term includes:

[0029]

[0030] Among them, Δτ′ total represents the global pheromone after the penalty term is introduced, and e represents the base of the natural logarithm.

[0031] Optionally, calculating the local pheromone includes:

[0032]

[0033] Among them, Δτ kIndicates local pheromone, E i Indicates the remaining energy of the current sensor i, E init represents the initial energy of the sensor node when it is deployed, Δτ′ total Represents the global pheromone after the penalty term is introduced.

[0034] Optionally, updating the pheromone of the own path includes:

[0035] τ(i,j)=(1-ρ)·τ(i,j)+Δτ k ;

[0036] Among them, τ(i,j) represents the pheromone of the path from node i to node j, ρ∈(0,1) represents the pheromone evaporation value, Δτ k Indicates local pheromone.

[0037] The beneficial effects of the present invention are:

[0038] The present invention discloses a compressed data collection method based on a wireless sensor network, which greatly reduces the amount of data collected by the cluster and reduces the energy consumption of the cluster head. Secondly, the method adopts an efficient ant colony algorithm to find the route, and finds the optimal transmission path within the cluster according to the characteristics of compressed data transmission, which greatly reduces the communication overhead within the cluster. Thirdly, the ant colony algorithm of the method is a distributed algorithm, which does not require large-scale communication to exchange information, and reduces the communication overhead. Finally, the method can be started at any time or in the interval of data transmission rounds using the efficient ant colony algorithm, so that it can better adapt to the dynamic network that may change and has high robustness. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0040] Figure 1 A schematic diagram of a compressed data collection method based on a wireless sensor network according to an embodiment of the present invention;

[0041] Figure 2 A schematic diagram of a simulated deployment of non-uniformly distributed nodes according to an embodiment of the present invention;

[0042] Figure 3 A schematic diagram of the number of active nodes in an embodiment of the present invention;

[0043] Figure 4 A schematic diagram of total energy comparison of an embodiment of the present invention;

[0044] Figure 5 A dead node comparison schematic diagram of an embodiment of the present invention;

[0045] Figure 6 A schematic diagram of the number of data packets transmitted to a receiver according to an embodiment of the present invention;

[0046] Figure 7 It is a schematic diagram of control overhead and energy saving comparison of an embodiment of the present invention. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0049] Wireless sensor networks (WSNs) are an important part of the Internet of Things (IoT) and play an important role in applications such as environmental monitoring, smart homes, and industrial automation. For resource-constrained wireless sensor networks, effective data collection methods can significantly reduce energy consumption and extend the network lifetime. Compressed sensing technology provides an efficient sampling method that significantly reduces the amount of data transmitted in the network and allows the recovery of original data under a certain noise level.

[0050] like Figure 1 As shown, a compressed data collection method based on a wireless sensor network includes: obtaining the utility function of the sensor node, taking the sensor node with the highest utility function as the cluster head node, using the cluster head node to send a seed code, thereby dividing the cluster; taking each non-cluster head node in the cluster as a forward ant, selecting the next node according to a probability selection rule, until reaching the cluster head node, and going back and forth to the original node, updating the pheromone of the round-trip path, and obtaining the round-trip path with the highest pheromone as the optimal path; obtaining the self-reading of each sensor, and generating its own according to the seed code Combine self-reading with self The product of is sent to the corresponding cluster head node according to the optimal path to obtain the observed value, which is then tracked and processed to obtain the original value.

[0051] Furthermore, dividing the clusters includes: using the cluster head node to send a seed code, and using the non-cluster head nodes to select and join the closest cluster according to the strength of the received signal.

[0052] Specifically:

[0053] Step 1: Cluster division: The sink node obtains the deployment area of ​​the sensor node through a broadcast to determine the size of the entire sensing area, and calculates the number of cluster divisions and the cluster size based on the number of sensors and the transmission distance of the sensors. The sink node calculates the utility function UE of each node and selects the highest value in the cluster as the cluster head:

[0054]

[0055] Among them, UE i represents the utility value of the i-th node in the cluster, where i belongs to a node in cluster k, C k is the set of all nodes in cluster k, and represents the distance between the farthest and nearest nodes in cluster k from the sink, d i represents the distance from node i to the sink, E i represents the residual energy of node i, E0 represents the initial energy of the node, and w is used to control the weights of the two indicators.

[0056] The selected cluster head node sends its cluster identification and seed code (used to generate the same measurement matrix) through broadcasting, and the remaining sensor nodes choose to join the closest cluster according to the strength of the received signal, thus completing the entire cluster division.

[0057] Furthermore, obtaining the optimal path includes: step 1, setting pheromones for each forwarding path corresponding to each non-cluster head node; step 2, using each non-cluster head node in the cluster to send information packets as a forward ant, selecting the next node according to a probability selection rule, and when arriving at the next node, recording the current node information and putting it into its own information packet until reaching the cluster head node or the forwarding life cycle ends. When the forward ant reaches the cluster head node, the global pheromone is calculated, and a penalty item is introduced to abandon the inefficient path, and further send the backward ant; the current node information includes: the number of hops, the distance between the current node and the cluster head, and the transmission distance; step 3, using the backward ant to carry the global pheromone to perform reverse propagation according to the original forward path until reaching the starting node; when the backward ant reaches the next node, the local pheromone is calculated according to the current node information, and each node updates the pheromone of its own path according to the local pheromone; step 4, repeating steps 1-3 until the number of iterations is reached, and obtaining the forwarding path with the highest pheromone as the optimal path.

[0058] Specifically:

[0059] Step 2: Optimal transmission path construction:

[0060] The construction of the path relies on the efficient ant colony algorithm designed in this embodiment. The specific steps are as follows:

[0061] 1. Each non-cluster head node sets the pheromone τ(i,j) to the initial value C for each of its forwarding paths:

[0062]

[0063] 2. Each non-cluster head node sends a message packet, which is called a transmitting ant and has a sequence number of k.

[0064] 3. The forward ant selects the next node according to the probability selection rule:

[0065]

[0066] Among them, p k (i,j) is calculated at each node based on the heuristic information and pheromone of each path under the node. Among them, η(i,j) represents the heuristic information on the path from node i to node j, τ(i,j) represents the amount of pheromone on the path from node i to node j, α and β control the weight between pheromone and heuristic information, and J k (i) is the set of neighbors of node i. The heuristic information in this algorithm is the inverse of the path, i.e.

[0067] 4. After the forward ant reaches each node, it records the current node information and puts it into its own information packet, and then continues to forward it until the forward ant reaches the cluster head or the forwarding life cycle ends. The recorded information is as follows:

[0068] (1) Hop count h k :h k =h k +1;

[0069] (2) Record L i (distance between current node and cluster head);

[0070] (3) Transmission distance: d k =d k +d ij ; where d ij Represents the path distance from node i to node j.

[0071] 5. After the forward ant reaches the cluster head node, the data is used to calculate the global pheromone Δτ total :

[0072]

[0073] Among them, E avg and E min Represents the average residual energy and minimum residual energy of all nodes on this path, E initRepresents the initial energy of the node, μ is used to control the energy consumption factor and path factors The weight between the two.

[0074] In order to speed up the convergence, a penalty term is introduced here to discard inefficient paths:

[0075]

[0076] 6. Every time a forward ant cluster head arrives, it sends a backward ant, which carries the forward path and the global pheromone Δτ total , and verify the forward path backpropagation.

[0077] 7. When the ant reaches the next node, it calculates the local pheromone based on the current node information:

[0078]

[0079] 8. Each node updates the pheromone of its own path according to the local pheromone given by the backward ant:

[0080] τ(i,j)=(1-c)·τ(i,j)+Δτ k (i,j);

[0081] Where ρ represents the evaporation rate of pheromone.

[0082] 9. The backward ants continue forwarding back to the final starting node.

[0083] 10. After receiving the message, the starting node sends a new forward ant after a period of time and repeats steps 1-9 to complete the iteration.

[0084] Step 3: Data transfer and recovery:

[0085] In each sending round, each sensor in the cluster sends its own reading x i , and generates itself according to the seed code in Usually a floating point number, j represents the cluster number, i represents the sensor number, and all nodes obtain Together they form the observation matrix Φ∈R m×n Each node compares its own reading with its own The product of is selected according to the highest pheromone value in the second step until it is transmitted to the cluster head node to obtain the observation value y i .

[0086]

[0087] Where y∈R mIndicates that a set of observed data has the observed value y j The dimension is m. In this network, m represents the number of clusters. Generally speaking, m<<n. i represents the data finally collected by the i-th cluster, x∈R n Represents the set of all readings.

[0088] Finally, the sink node obtains a set of observation values ​​y and uses the common basis pursuit (BP) algorithm and orthogonal matching pursuit (OMP) algorithm to solve the following problem to obtain the original value x:

[0089] min x ||x||1subject to Φx=y;

[0090] Among them, min x ||x||1 represents the minimization of a normal form of the original value x, which is obtained by solving the condition Φx=y.

[0091] The experimental simulation parameters are shown in Table 1. This embodiment simulates the deployment of 100 non-uniformly distributed nodes in an environment of 100*100 meters, and the BS is located in the center of the network. Figure 2 .

[0092] Table 1

[0093]

[0094]

[0095] In order to demonstrate the overall performance of the protocol of this embodiment, the EARCS proposed in this embodiment is compared with the R-leach algorithm, the CEEC algorithm, the DEC algorithm, and the D-wsn algorithm. The R-leach algorithm is a distributed clustering protocol that uses multiple parameters to control cluster head selection. The CEEC algorithm is a centralized clustering protocol that integrates a balanced clustering algorithm, a low-cost maintenance mechanism, and an energy-aware solution. DEC and D-wsn also introduce compressed sensing into clustering algorithms.

[0096] Figure 3 The comparison of the network life cycle of the four protocols is shown. The results show that due to the limitations of the compressed sensing method in cluster head selection and intra-cluster transmission during round change, the DEC algorithm has the earliest death time of the first node, about 200 rounds, and cannot achieve the most balanced selection. The EARCS method overcomes this limitation and maintains a similar balance to the R-leach and CEEC algorithms, indicating that the ant colony optimization of this embodiment focuses on load balancing.

[0097] The last node death of the EARCS algorithm occurs at 2300 rounds, which is 2.55 times better than R-leach, 2.09 times better than CEEC, and 1.21 times and 1.18 times better than DEC and D-wsn, respectively, both of which use compressed sensing.

[0098] Figure 4 The total energy of the network is shown. It can be seen from the figure that the energy of the EARCS algorithm decreases the slowest, followed by D-wsn and DEC, and CEEC and R-leach are the last. This shows that the algorithm of this embodiment has the highest overall energy efficiency.

[0099] Figure 5 The network throughput during the first node death, half of the node death, and all the node death periods is compared. EARCS has the highest throughput in all three stages, and its total data volume advantage is significant, reflecting the efficiency of the algorithm's transmission method.

[0100] Figure 6 The number of packets arriving at the aggregation node compared with ERACS and other algorithms is shown. It can be observed that the algorithm of this embodiment significantly reduces the amount of data transmitted in the network. In the simulation experiment, the compression ratio (CR) in the early stable transmission stage is 14.9%, which is related to the node density. As most nodes die and the remaining nodes strive to form a cluster in the later stage, the compression ratio will increase.

[0101] In the protocol proposed in this embodiment, ants must be started to explore at the beginning of cluster construction. According to the size of the cluster, this embodiment allows each node to deploy 5 ants before data transmission begins, and includes an ant in each data packet after the transmission begins. This embodiment regards the generation cost of these ants as the control overhead of the ant colony algorithm, and calculates it according to formulas (1) and (2) based on the data packet size and transmission volume of the ants. This embodiment also experimentally obtains the energy saved by the ant colony algorithm optimized path relative to the original single-hop routing within the cluster. This embodiment takes the average value through 5 simulations and compares these two data segments. Figure 7 As shown, it can be concluded that although the introduction of the ant colony algorithm increases the control overhead, it is almost negligible compared with the energy saved by the optimized path, which also shows the advantage of introducing the ant colony algorithm in this embodiment.

[0102] In summary, due to the smaller transmission volume, the algorithm of this embodiment is more energy-efficient and has a longer lifespan than the R-leach and CEEC algorithms. Compared with the DEC algorithm, the method of this embodiment uses the characteristics of compressed sensing to optimize routing, slowing down the death of nodes far away from the cluster head, thereby making the network load more balanced and delaying the time of the first node death.

[0103] This embodiment combines the principles of compressed sensing and ant colony algorithm to propose a compressed data collection method based on wireless sensor networks. This method re-imagines data fusion in clustered networks through compressed sensing and changes the traditional intra-cluster communication mode. A new intra-cluster routing optimization method is proposed, and the ant colony algorithm is redesigned. Simulation results show that compared with traditional methods, this embodiment significantly reduces the overall network energy consumption and significantly extends the network life.

[0104] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A compressed data collection method based on a wireless sensor network, characterized in that: include: Obtaining the utility functions of the sensor nodes, taking the sensor node with the highest utility function as the cluster head node, and using the cluster head node to send a seed code, thereby dividing the cluster; Each non-cluster head node in the cluster is used as a forward ant, and the next node is selected according to a probability selection rule until the cluster head node is reached, and the ant goes back and forth to the original node, and the pheromone of the round-trip path is updated, and the round-trip path with the highest pheromone is obtained as the optimal path; Get each sensor's own reading x i , and generates itself according to the seed code in, is a floating point number, j represents the cluster number, i represents the sensor number; based on all nodes Together they form the observation matrix Φ∈R m×n Each node compares its own reading with its own The product of is sent to the corresponding cluster head node according to the optimal path to obtain the observation value y j , track the observed value and solve to obtain the original reading x i .

2. The compressed data collection method based on wireless sensor network according to claim 1 is characterized in that: The utility function of the sensor node is: Among them, UE i represents the utility value of the i-th node in the cluster, i represents a node in cluster k, C k represents the set of all nodes in cluster k, and represents the distance between the farthest and nearest nodes in cluster k from the sink, d i represents the distance from node i to the sink, E i represents the residual energy of node i, E0 represents the initial energy of the node, and w represents the weight of the control index.

3. The compressed data collection method based on wireless sensor network according to claim 1 is characterized in that: Dividing the clusters includes: using the cluster head node to send a seed code, and using the non-cluster head nodes to select and join the closest cluster according to the strength of received signals.

4. The compressed data collection method based on wireless sensor network according to claim 1 is characterized in that: Acquiring the optimal path includes: Step 1: Set pheromone on each forwarding path corresponding to each non-cluster head node; Step 2: Use each non-cluster head node in the cluster to send information packets as a forward ant, select the next node according to the probability selection rule, and when reaching the next node, record the current node information and put it into its own information packet until reaching the cluster head node or the forwarding life cycle ends. When the forward ant reaches the cluster head node, calculate the global pheromone, introduce a penalty term to discard inefficient paths, and further send backward ants; Step 3: Use the backward ant to carry the global pheromone and perform reverse propagation along the original forward path until it reaches the starting node; when the backward ant reaches the next node, the local pheromone is calculated according to the current node information, and each node updates the pheromone of its own path according to the local pheromone; Step 4: Repeat steps 1-3 until the number of iterations is reached, and obtain the forwarding path with the highest pheromone as the optimal path.

5. The compressed data collection method based on wireless sensor network according to claim 4 is characterized in that: The next node is selected according to the probability selection rule including: Among them, p k (i, j) represents the next node selected, η(i, j) represents the heuristic information on the path from node i to node j, τ(i, j) represents the amount of pheromone on the path from node i to node j, α and β represent the weights between the control heuristic information and the amount of pheromone, J k (i) represents the neighbor set of node i, τ(i,u) represents the pheromone amount from node i to node u, u is the neighbor node of i, and in the set J k (i).

6. The compressed data collection method based on wireless sensor network according to claim 4 is characterized in that: The current node information includes: the number of hops, the distance between the current node and the cluster head, and the transmission distance.

7. The compressed data collection method based on wireless sensor network according to claim 4 is characterized in that: Calculating the global pheromone includes: Among them, Δτ total represents the global pheromone, E avg 、E min Represents the average residual energy and minimum residual energy of all nodes on this path, E init represents the initial energy of the node, μ represents the parameter between the two, and d k Indicates the transmission distance, h k represents the number of hops, g(k) represents, represents the distance between the ith node and the cluster head, In the path set g(k) of ant k.

8. The compressed data collection method based on wireless sensor network according to claim 7 is characterized in that: The penalty items introduced include: Among them, Δτ′ total represents the global pheromone after the penalty term is introduced, and e represents the base of the natural logarithm.

9. The compressed data collection method based on wireless sensor network according to claim 4, characterized in that: Calculating the local pheromone includes: Among them, Δτ k Indicates local pheromone, E i Indicates the remaining energy of the current sensor i, E init represents the initial energy of the sensor node when it is deployed, Δτ′ total Represents the global pheromone after the penalty term is introduced.

10. The compressed data collection method based on wireless sensor network according to claim 4, characterized in that: Updating the pheromone of the self-path includes: τ(i,j)=(1-ρ)·τ(i,j)+Δτ k ; Among them, τ(i,j) represents the pheromone of the path from node i to node j, ρ∈(0,1) represents the pheromone evaporation value, Δτ k Indicates local pheromone.

Images