CN115967631A - Internet of things topology optimization method based on breadth learning and application thereof - Google Patents
Internet of things topology optimization method based on breadth learning and application thereof Download PDFInfo
- Publication number
- CN115967631A CN115967631A CN202211632251.2A CN202211632251A CN115967631A CN 115967631 A CN115967631 A CN 115967631A CN 202211632251 A CN202211632251 A CN 202211632251A CN 115967631 A CN115967631 A CN 115967631A
- Authority
- CN
- China
- Prior art keywords
- topology
- internet
- things
- nodes
- data
- 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.)
- Pending
Links
- 238000005457 optimization Methods 0.000 title claims abstract description 33
- 238000000034 method Methods 0.000 title claims abstract description 18
- 239000011159 matrix material Substances 0.000 claims abstract description 38
- 238000012360 testing method Methods 0.000 claims abstract description 23
- 238000012549 training Methods 0.000 claims abstract description 18
- 238000004891 communication Methods 0.000 claims description 24
- 238000013507 mapping Methods 0.000 claims description 16
- 238000004364 calculation method Methods 0.000 claims description 6
- 239000013598 vector Substances 0.000 claims description 4
- 230000004913 activation Effects 0.000 claims description 3
- 230000006870 function Effects 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims description 2
- 230000008569 process Effects 0.000 claims description 2
- 238000013500 data storage Methods 0.000 abstract description 3
- 238000005516 engineering process Methods 0.000 description 5
- 238000005265 energy consumption Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 230000002776 aggregation Effects 0.000 description 2
- 238000004220 aggregation Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 230000002068 genetic effect Effects 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000003213 activating effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000009194 climbing Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 230000002028 premature Effects 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000002922 simulated annealing Methods 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
Images
Classifications
-
- 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
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention discloses a topology optimization method based on width learning and application thereof, comprising the following steps: step 1: initializing the Internet of things in a scale-free network mode to generate an initial topology; and 2, step: optimizing the topology of the Internet of things by a multi-population evolution algorithm to generate a target topology; and step 3: the initial topology and the target topology are subjected to de-duplication to obtain a topology pair data set; and 4, step 4: coding each topology pair in the topology pair data set through the node degree to obtain a topology adjacency matrix; and 5: testing the width learning model by splitting the topological adjacency matrix into a training set and a testing set; step 6: calculating the tested width learning model by the following formula to obtain an optimized width learning model; the method ensures that the width learning can fully learn the topological characteristics before and after optimization by adjusting the data storage format of the nodes in the topology, and improves the robustness of the topology of the Internet of things with extremely low time overhead.
Description
Technical Field
The invention belongs to the technical field of Internet of things and topology optimization, and particularly relates to a width learning-based Internet of things topology optimization method and application thereof.
Background
The internet of things is a comprehensive system integrating the technologies of mechanical control, wireless sensing, data acquisition and sharing and the like. With the development of 5G technology, the Internet of things is increasingly used. For a part of application scenarios of the specific internet of things, ultra-dense networking technology has been proposed and used for improving the peak value of network traffic. The centralized arrangement of the sensors can improve the power and spectrum utilization rate of the network, but can also increase the energy consumption of the nodes. In addition, due to unbalanced communication loads among a plurality of aggregation nodes, energy consumption difference of each area can be caused, so that energy of part of nodes is exhausted, and service quality of the network is reduced. Network attacks or offline damages also increase the failure probability of nodes, and reduce the connection performance of the network, so that reasonable network layout is the key for improving the network efficiency. Network topology refers to the connection and communication relationship between nodes in a network. The reasonable network topology structure plays a crucial role in the stable operation of the whole network and is also an important factor influencing the safety and reliability of the ad hoc network of the sensing layer of the internet of things. However, the traditional network topology structure is directly used in a large-scale internet of things, and the problems of stability, energy consumption, time delay and the like of a network system cannot be well solved. In addition, in the dense internet of things, due to the fact that the data volume is large, the performance of some aggregation nodes is prone to be reduced. The reason is that the multi-hop forwarding mechanism in the data routing makes the data volume and the required energy consumption of the nodes in the area higher than those of other nodes. Meanwhile, the burden task is heavy, and the communication performance of the whole topology is reduced due to the fact that the communication is easily and maliciously attacked and the communication cannot work. Therefore, in order to ensure the network communication capacity, the topology is optimized by adjusting the connection relationship among the nodes, the degree of topology failure caused by the fault of the sink node can be reduced, and the service life of the network is prolonged.
In addition, when the 5G technology is developed, the user's demand for network service quality is higher and higher, the network communication service quality is guaranteed through topology robustness optimization, and it will become a necessary trend for topology research to reduce the computation time overhead of topology optimization. The traditional heuristic algorithm such as hill climbing algorithm and simulated annealing is a method for searching the optimal solution based on random edge changing. The method can quickly find a solution meeting the requirement, but has the condition of easy falling into local optimum and higher time overhead. The genetic algorithm is also an effective means for solving the optimization problem, but premature convergence is easy, so that the cultural genetic algorithm and the multi-population evolution algorithm solve the defect, but the defects of no consideration of the communication distance characteristic of the Internet of things and long convergence time exist respectively.
Disclosure of Invention
In order to solve the problems, the invention provides a width learning-based topology optimization method of the internet of things, which ensures that the width learning can fully learn the topological characteristics before and after optimization by adjusting the data storage format of nodes in the topology and improves the robustness of the topology of the internet of things with extremely low time overhead; according to the invention, related technologies such as breadth learning, internet of things and multi-population evolution are combined, and the robustness of the Internet of things topology is improved on the premise of ensuring that the overall degree distribution in the initial topology is not changed.
The invention solves the practical problem by adopting the following technical scheme:
a topology optimization method based on width learning comprises the following steps:
101, initializing the Internet of things in a scale-free network mode to generate an initial topology;
102, optimizing the topology of the Internet of things to generate a target topology through a multi-population evolution algorithm;
103, repeating the initial topology and the target topology to obtain a topology pair data set;
step 104, coding each topology pair in the topology pair data set through the node degree to obtain a topology adjacency matrix;
105, splitting the topological adjacency matrix into a training set and a test set to test a width learning model;
step 106, calculating the tested width learning model through the following formula to obtain an optimized width learning model:
wherein: n is the number of nodes in the topology, and MCS (N) represents that after the nodes with the highest degree of the current topology are attacked maliciously for N times, the maximum connected subgraph of the topology comprises the number of the nodes; m is the data size of the training set or test set, R i A robustness value representing the ith optimized topology,
and representing the topology robustness value of the ith initial topology after width learning optimization.
The invention can also be implemented by adopting the following technical scheme:
a topology optimization method based on breadth learning is applied to the Internet of things, and comprises the following steps:
step 201, initializing the internet of things in a scale-free network mode to generate an initial internet of things topology
202, optimizing the topology of the Internet of things through a multi-population evolution algorithm to generate a target topology of the Internet of things;
step 203, repeating the initial internet of things topology and the target internet of things topology to obtain a topology pair data set;
step 204, coding each topology pair in the topology pair data set through the node degree to obtain an internet of things topology adjacency matrix;
step 205, testing the Internet of things width learning model by splitting the topological adjacency matrix into a training set and a testing set;
wherein: n is the number of nodes in the topology, and MCS (N) represents that after the nodes with the highest degree of the current topology are attacked maliciously for N times, the maximum connected subgraph of the topology comprises the number of the nodes; m is the data size of the training set or test set, R i A robustness value representing the ith optimized topology,
representing a topology robustness value of the ith initial topology after width learning optimization;
and step 207, executing steps 201, 203 and 204 in the Internet of things optimization width learning model to output the optimized Internet of things topology.
Further, the step 3 of obtaining the topology adjacency matrix by encoding each topology pair in the topology pair dataset through the node degree includes the following steps:
401. numbering nodes in each topological pair in the topological pair data set according to the sequence of the degrees of the nodes from large to small to obtain the topological number data information of the Internet of things;
402. establishing an Internet of things topology number matrix by adopting an adjacency matrix to store topology network number data information;
403. and (3) dividing the topology number matrix information into three types according to the interrelation of the two nodes:
the first type is that two nodes are connected, and the matrix value is 1;
the second type is that there can be no connection between two nodes, i.e. each other is out of communication range, and the matrix value is-1.
The third type is that two nodes are not connected but within communication range with each other, and the matrix value d ij The calculation is as follows:
wherein: p represents a node position, and r represents a node communication range; the fraction represents the ratio of the distance between the two nodes and the communication range, and the value obtained after the integral calculation is inversely related to the distance between the two nodes and is normalized to (0,1);
404. and expanding the upper triangular matrix of the divided topological matrix to obtain vector representation.
Further, the process of obtaining the internet of things optimized topology model through splitting the training set and the testing set of the topology adjacency matrix and testing the width learning model in the step 5 comprises the following steps:
501. mapping the data of the training set and the test set to a feature mapping layer to obtain a first data feature;
502. the sparse encoder restricts the extraction of the first data features through KL divergence to obtain second data features;
503. the enhancement node layer performs secondary mapping on the second data characteristics to obtain third data characteristics;
504. calculating the second data characteristic and the third data characteristic through an activation function to obtain output layer node data;
505. and establishing an Internet of things optimized topology model by calculating the weight of the nodes in the Internet of things node data through pseudo-inverse.
Has the advantages that:
1. the invention introduces an efficient machine learning means, namely width learning, to optimize the network topology, thereby greatly reducing the time overhead of the prior topology optimization.
2. The invention designs a unique data storage content, the data design considers the consistency of topology information, and the position, the communication range and the like of the topology are stored in the data together, thereby preventing gradient disappearance and improving the performance of the model.
Description of the drawings:
FIG. 1 is an overall framework diagram of the topology optimization of the Internet of things of the present invention;
FIG. 2 is a schematic diagram of a network topology and data vector conversion;
FIG. 3 is a schematic diagram of a width learning mechanism
Detailed Description
The invention is described in detail below with reference to the accompanying drawings:
as shown in fig. 1, the present invention provides a topology optimization method based on width learning, which includes the following steps:
step 1: initializing the Internet of things in a scale-free network mode to generate an initial Internet of things topology;
step 2: optimizing the topology of the Internet of things by a multi-population evolution algorithm to generate a target topology of the Internet of things;
and step 3: the initial internet of things topology and the target internet of things topology are subjected to de-duplication to obtain a topology pair data set;
the steps 1 to 3 are the improvement of the initial topological structure, namely: determining the size of a node deployment area to be 500 x 500m 2 And the total number of nodes is 100. The communication radius of the nodes is 200m, and all the sensor nodes are scattered randomly in a monitoring area to perform self-organized connection according to a scale-free network generation mode. And then optimizing the collected scale-free initial topology by utilizing a multi-population evolution algorithm to obtain an optimized topology. And carrying out repeatability detection on the topology pairs consisting of each pair of initial topology and corresponding optimized topology, if repeated topology pairs appear, rejecting the repeated topology pairs, and finally keeping the topology pairs unique.
And 4, step 4: coding each topology pair in the topology pair data set through the node degree to obtain an internet of things topology adjacency matrix; and processing the collected topological data. The invention introduces a unique topological data information representation form, as shown in fig. 2:
firstly, nodes in the network topology are renumbered, the nodes are endowed with numbers 1,2.. 100 according to the sequence of the degrees of the nodes from large to small, and the nodes are renumbered, so that the node degree distribution of the initial scale-free topology can present a rule. The purpose of doing so is to enable all high-degree nodes to have smaller node numbers in the whole network topology, and the low-degree nodes to have larger node numbers, so that the collected topological nodes all have the inverse correlation relationship between the degrees and the numbers;
then, an adjacency matrix storage topology is established, and the storage content is divided into three types. The first type indicates that a connection exists between two nodes, and the matrix value is 1; the second type indicates that connection cannot exist between two nodes, namely the two nodes are out of the communication range, and the matrix value is-1; the third type indicates that two nodes are not connected but are in communication range with each other, and the matrix value is
The numerator represents the distance between two nodes in the numerator and the denominator represents the communication distance, which is regarded as a representation form of connection probability because two nodes close to each other are theoretically easier to generate connection, namely:
wherein: p represents the node position, and r represents the node communication range; the fraction represents the ratio of the distance between the two nodes and the communication range, and the value obtained after the integral calculation is inversely related to the distance between the two nodes and is normalized to (0,1); .
Finally, because the topology is an undirected graph, the adjacency matrix may represent all the information with an upper triangular matrix, which is then expanded into vectors of dimension 5050 for subsequent use as a data set. The series of topology data processing is beneficial to the convergence of width learning, simultaneously, the unique attributes of the Internet of things, such as position coordinates, communication distances and the like are considered, the connection relation of the topology is reflected, the storage cost is reduced, and the operation efficiency is improved.
And 5: the method comprises the steps that a training set and a test set of the topology adjacency matrix of the Internet of things are split to check a width learning model to obtain an Internet of things optimization topology model;
inputting the initial topological data into a feature mapping layer for mapping, and then extracting the most essential features of the data through a sparse automatic encoder, so that the data feature representation is more compact. The sparsity is constrained by using KL divergence, and punishment is carried out if the actual sparsity deviates from the set sparsity; and outputting the obtained hidden layer to an enhanced node layer for secondary mapping, activating by an activation function tanh (), and then using the output of the characteristic mapping layer and the output of the enhanced node layer as the input of an output node. And calculating the weight of the output node through the pseudo-inverse to obtain a topology optimization model. If the value of formula (3) in the training set is too large, the performance of the representative model is too strong, i.e. overfitting. A width learning structure with moderate model performance degree needs to be selected, and the final model structure is 25 feature mapping windows, 25 feature mapping windows in size and 850 enhanced node number.
Wherein: and modifying the feature mapping layer structure of width learning, and increasing the learning capability of the model on linear features by expanding the capacity of a feature mapping window.
Wherein: and modifying the feature mapping layer structure of the width learning, and increasing the learning capacity of the model to the linear features by increasing the number of feature mapping windows.
Wherein: and modifying the enhanced node layer structure of the width learning, and increasing the learning capacity of the model on the nonlinear features by increasing the number of enhanced nodes.
Step 6: judging the robustness of the Internet of things optimization topology model by the following formula to output an optimal topology Internet of things model:
wherein: n is the number of nodes in the topology, and MCS (N) represents that after malicious attack is carried out on the node with the highest degree of the current topology for N times, the maximum connected subgraph of the topology contains the number of the nodes; m is the data size of the training set or test set, R i A robustness value representing the ith optimized topology,
and the topology robustness value of the ith initial topology after width learning optimization is represented. Wherein: the degree of topological robustness is evaluated according to the formula (2), and the model is evaluated according to the formula (3).
And (2) a robustness R value calculation formula, wherein N is the number of nodes in the topology, and MCS (N) represents that after the node with the highest degree of the current topology is attacked maliciously for N times, the maximum connected subgraph of the topology contains the number of the nodes. The network robustness is measured by calculating the number of nodes in the maximum connected subgraph after each attack, and the value range is 0 to 0.5. M in equation (3) is the data size of the training set or test set, R i A robustness value representing the ith optimized topology,
and representing the topology robustness value of the ith initial topology after width learning optimization.
And 6: judging the robustness of the Internet of things optimization topology model by the following formula to output an optimal topology Internet of things model:
wherein: n is the number of nodes in the topology, and MCS (N) represents that after malicious attack is carried out on the node with the highest degree of the current topology for N times, the maximum connected subgraph of the topology contains the number of the nodes; m is the data size of the training set or test set, R i A robustness value representing the ith optimized topology,
and the topology robustness value of the ith initial topology after width learning optimization is represented.
The present invention is not limited to the embodiments described above. The foregoing description of the specific embodiments is intended to describe and illustrate the technical solutions of the present invention, and the specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make various changes in form and details without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (4)
1. A topology optimization method based on width learning is characterized by comprising the following steps:
101, initializing the Internet of things in a scale-free network mode to generate an initial topology;
102, optimizing the topology of the Internet of things to generate a target topology through a multi-population evolution algorithm;
103, repeating the initial topology and the target topology to obtain a topology pair data set;
step 104, coding each topology pair in the topology pair data set through the node degree to obtain a topology adjacency matrix;
105, splitting the topological adjacency matrix into a training set and a test set to test the width learning model;
step 106, calculating the tested width learning model through the following formula to obtain an optimized width learning model:
wherein: n is the number of nodes in the topology, and MCS (N) represents that after malicious attack is carried out on the node with the highest degree of the current topology for N times, the maximum connected subgraph of the topology contains the number of the nodes; m is the data size of the training set or test set, R i A robustness value representing the ith optimized topology,
and representing the topology robustness value of the ith initial topology after width learning optimization.
2. A topology optimization method based on breadth learning is applied to the Internet of things, and is characterized by comprising the following steps:
step 201, initializing the internet of things in a scale-free network mode to generate an initial internet of things topology
202, optimizing the topology of the Internet of things through a multi-population evolution algorithm to generate a target topology of the Internet of things;
step 203, repeating the initial internet of things topology and the target internet of things topology to obtain a topology pair data set;
step 204, coding each topology pair in the topology pair data set through the node degree to obtain an internet of things topology adjacency matrix;
step 205, testing the Internet of things width learning model by splitting the topological adjacency matrix into a training set and a testing set;
step 206, calculating the tested width learning model through the following formula to obtain an internet of things optimized width learning model:
wherein: n is the number of nodes in the topology, and MCS (N) represents that after the nodes with the highest degree of the current topology are attacked maliciously for N times, the maximum connected subgraph of the topology comprises the number of the nodes; m is the data size of the training set or test set, R i A robustness value representing the ith optimized topology,
representing a topology robustness value of the ith initial topology after width learning optimization;
and step 207, executing steps 201, 203 and 204 in the internet of things optimization width learning model to output the optimized internet of things topology.
3. The topology optimization method based on width learning according to claim 1, wherein the step 3 of obtaining the topology adjacency matrix by encoding each topology pair in the topology pair data set by node degree includes the following steps:
401. numbering the nodes in each topological pair in the data set according to the sequence of the degrees of the nodes from large to small to obtain the topological number data information of the Internet of things;
402. adopting an adjacency matrix to store topology serial number data information to establish an Internet of things topology serial number matrix;
403. and (3) carrying out three types of division on the topology number matrix information according to the interrelation of the two nodes:
the first type is that two nodes are connected, and the matrix value is 1;
the second type is that there cannot be a connection between two nodes, i.e. each other is out of communication range, and the matrix value is-1.
The third type is that two nodes are not connected but within communication range with each other, and the matrix value d ij The calculation is as follows:
wherein: p represents a node position, and r represents a node communication range; the fraction represents the ratio of the distance between the two nodes and the communication range, and the value obtained after the integral calculation is inversely related to the distance between the two nodes and is normalized to (0,1);
404. and expanding the upper triangular matrix of the divided topological matrix to obtain vector representation.
4. The topology optimization method based on breadth learning according to claim 1, wherein the process of obtaining the internet of things optimized topology model through splitting the training set and the testing set of the topological adjacency matrix to check the breadth learning model in the step 5 includes the following steps:
501. mapping the data of the training set and the test set to a feature mapping layer to obtain a first data feature;
502. the sparse encoder restricts the extraction of the first data features through KL divergence to obtain second data features;
503. the enhancement node layer performs secondary mapping on the second data characteristics to obtain third data characteristics;
504. calculating the second data characteristic and the third data characteristic through an activation function to obtain output layer node data;
505. and establishing an Internet of things optimized topology model by calculating the weight of the nodes in the Internet of things node data through pseudo-inverse.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211632251.2A CN115967631A (en) | 2022-12-19 | 2022-12-19 | Internet of things topology optimization method based on breadth learning and application thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211632251.2A CN115967631A (en) | 2022-12-19 | 2022-12-19 | Internet of things topology optimization method based on breadth learning and application thereof |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115967631A true CN115967631A (en) | 2023-04-14 |
Family
ID=87354226
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211632251.2A Pending CN115967631A (en) | 2022-12-19 | 2022-12-19 | Internet of things topology optimization method based on breadth learning and application thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115967631A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108734301A (en) * | 2017-06-29 | 2018-11-02 | 澳门大学 | A kind of machine learning method and machine learning device |
| CN110209825A (en) * | 2019-06-17 | 2019-09-06 | 大连海事大学 | A kind of fast network representative learning algorithm based on width learning system |
| CN111935724A (en) * | 2020-07-06 | 2020-11-13 | 天津大学 | Wireless sensor network topology optimization method based on asynchronous deep reinforcement learning |
| CN113111913A (en) * | 2021-03-12 | 2021-07-13 | 浙江工业大学 | Time series multi-classification method and system based on width learning |
| CN113159310A (en) * | 2020-12-21 | 2021-07-23 | 江西理工大学 | Intrusion detection method based on residual error sparse width learning system |
| CN113422695A (en) * | 2021-06-17 | 2021-09-21 | 天津大学 | Optimization method for improving robustness of topological structure of Internet of things |
-
2022
- 2022-12-19 CN CN202211632251.2A patent/CN115967631A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108734301A (en) * | 2017-06-29 | 2018-11-02 | 澳门大学 | A kind of machine learning method and machine learning device |
| CN110209825A (en) * | 2019-06-17 | 2019-09-06 | 大连海事大学 | A kind of fast network representative learning algorithm based on width learning system |
| CN111935724A (en) * | 2020-07-06 | 2020-11-13 | 天津大学 | Wireless sensor network topology optimization method based on asynchronous deep reinforcement learning |
| CN113159310A (en) * | 2020-12-21 | 2021-07-23 | 江西理工大学 | Intrusion detection method based on residual error sparse width learning system |
| CN113111913A (en) * | 2021-03-12 | 2021-07-13 | 浙江工业大学 | Time series multi-classification method and system based on width learning |
| CN113422695A (en) * | 2021-06-17 | 2021-09-21 | 天津大学 | Optimization method for improving robustness of topological structure of Internet of things |
Non-Patent Citations (3)
| Title |
|---|
| C. L. PHILIP CHEN，ZHULIN LIU: "Broad Learning System An Effective and Efficient Incremental Learning System Without the Need for Deep Architecture", IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, 21 July 2017 (2017-07-21), pages 10 - 24 * |
| NING CHEN, TIE QIU, XIAOBO ZHOU, KEQIU LI, AND MOHAMMED ATIQUZZAMAN: "An Intelligent Robust Networking Mechanism for the Internet of Things", IEEE COMMUNICATIONS MAGAZINE, 21 November 2019 (2019-11-21), pages 91 - 95, XP011757285, DOI: 10.1109/MCOM.001.1900094 * |
| TIE QIU, ET AL: "Robustness Optimization Scheme With Multi-Population Co-Evolution for Scale-Free Wireless Sensor Networks", IEEE/ACM TRANSACTIONS ON NETWORKING, 12 April 2019 (2019-04-12), pages 1028 - 1042 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111935724B (en) | Wireless sensor network topology optimization method based on asynchronous deep reinforcement learning | |
| CN108737191B (en) | Topology control method for ultra-dense wireless sensor network based on unsupervised learning | |
| CN113297429B (en) | Social network link prediction method based on neural network architecture search | |
| CN113422695B (en) | Optimization method for improving robustness of topological structure of Internet of things | |
| CN113411766B (en) | Intelligent Internet of things comprehensive sensing system and method | |
| CN103249179A (en) | Multi-objective mother foraging algorithm based optimization method for relay node deployment in wireless sensor network | |
| Diao et al. | Energy-efficient boundary detection of continuous objects in IoT sensing networks | |
| CN107578136A (en) | The overlapping community discovery method extended based on random walk with seed | |
| CN115967631A (en) | Internet of things topology optimization method based on breadth learning and application thereof | |
| CN111464327A (en) | Spatial information network survivability evaluation method based on graph convolution network | |
| CN114329867B (en) | Scale-free network robustness measuring method based on die body | |
| CN112101608A (en) | Offshore wind farm site selection method and device | |
| CN110365006B (en) | Power grid partitioning method based on nwbbo algorithm | |
| CN110048945B (en) | Node mobility clustering method and system | |
| CN105722173A (en) | Large-scale WSNs topology control method based on two-way attraction mechanism | |
| Yue et al. | Critical nodes identification: A non-cooperative method for unknown topology information in ad hoc networks | |
| CN114051217B (en) | Safe routing method and system for sensor network of comprehensive energy Internet of things | |
| Wei et al. | A multi-objective optimized OLSR routing protocol | |
| CN111818043B (en) | Dual-network coupling topological structure design method based on memetic algorithm | |
| CN114943430B (en) | Smart grid key node identification method based on deep reinforcement learning | |
| CN116032727B (en) | Electric power internet of things sensing layer self-repairing method based on regional collaboration | |
| CN117729606B (en) | Power grid route determining method and device, electronic equipment and storage medium | |
| CN107509234A (en) | Flight MANET key node detection method and system based on limited routing information | |
| Liu et al. | Energy Allocation Strategy of Earthquake Monitoring Wireless Sensor Network Based on Longitude and Latitude Coding and Differential Evolution Algorithm | |
| Sahu et al. | Energy Efficient Multipath Routing in Zone-based Mobile Ad-hoc Networks: Mathematical Formulation |
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 |