CN116915746A - Network addressing method - Google Patents
Network addressing method Download PDFInfo
- Publication number
- CN116915746A CN116915746A CN202311181075.XA CN202311181075A CN116915746A CN 116915746 A CN116915746 A CN 116915746A CN 202311181075 A CN202311181075 A CN 202311181075A CN 116915746 A CN116915746 A CN 116915746A
- Authority
- CN
- China
- Prior art keywords
- matrix
- network
- solution
- solutions
- ipv6
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 35
- 239000013598 vector Substances 0.000 claims abstract description 63
- 239000011159 matrix material Substances 0.000 claims abstract description 55
- 238000005192 partition Methods 0.000 claims abstract description 22
- 238000005457 optimization Methods 0.000 claims abstract description 21
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 16
- 230000002068 genetic effect Effects 0.000 claims abstract description 13
- 230000006872 improvement Effects 0.000 claims abstract description 12
- 238000013507 mapping Methods 0.000 claims abstract description 12
- 238000013135 deep learning Methods 0.000 claims abstract description 9
- 230000006870 function Effects 0.000 claims description 35
- 238000009826 distribution Methods 0.000 claims description 13
- 238000004458 analytical method Methods 0.000 claims description 5
- 239000012530 fluid Substances 0.000 claims description 3
- 238000002360 preparation method Methods 0.000 claims description 3
- 108090000623 proteins and genes Proteins 0.000 claims description 3
- 230000001846 repelling effect Effects 0.000 claims description 3
- 238000007621 cluster analysis Methods 0.000 abstract description 6
- 238000007726 management method Methods 0.000 description 13
- 230000008859 change Effects 0.000 description 10
- 238000004364 calculation method Methods 0.000 description 9
- 238000005516 engineering process Methods 0.000 description 8
- 230000008569 process Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 238000004590 computer program Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000012986 modification Methods 0.000 description 4
- 230000004048 modification Effects 0.000 description 4
- 238000007405 data analysis Methods 0.000 description 3
- 238000005265 energy consumption Methods 0.000 description 3
- 238000005304 joining Methods 0.000 description 3
- 230000009467 reduction Effects 0.000 description 3
- 238000005070 sampling Methods 0.000 description 3
- 239000002699 waste material Substances 0.000 description 3
- 241000899793 Hypsophrys nicaraguensis Species 0.000 description 2
- 230000004075 alteration Effects 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013506 data mapping Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 239000002360 explosive Substances 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000003064 k means clustering Methods 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The application relates to the technical field of IPv6 addressing, in particular to a network addressing method. Collecting characteristic state vectors of the Internet of things equipment, deducing a similarity matrix, constructing a Laplace matrix of the graph, analyzing the Laplace matrix, constructing a new matrix, updating a clustering center, and putting the similar equipment in the same group to realize intelligent network partition of the Internet of things equipment; in each network partition, mapping the characteristic state vector of the equipment to an IPv6 address through deep learning, constructing an IPv6 addressing model, regarding an IPv6 address space as a multi-dimensional wind field, and optimizing through a multi-target genetic algorithm to obtain an optimal IPv6 address allocation scheme. The method solves the problems that the prior art can not effectively cope with dynamic changes in the environment of the Internet of things, and the prior art lacks of carrying out cluster analysis and address dynamic optimization allocation on equipment, so that the network management efficiency and the improvement of the addressing efficiency are directly influenced.
Description
Technical Field
The application relates to the technical field of IPv6 addressing, in particular to a network addressing method.
Background
The internet of things (Internet of Things, ioT) refers to a variety of physical devices and objects that connect and communicate over the internet, which devices can connect to the internet, share data and perform specific tasks without human intervention. As technology continues to develop, a wide variety of physical devices and sensors are connected to the internet. These devices include smart home devices, industrial machines, smart city infrastructure, and the like. This results in an explosive increase in the number of devices. IPv4 is the most widely used version of the protocol for the internet, but its address space is limited, with only about 40 more than a hundred million available addresses. As the number of devices in the internet of things increases, the IPv4 address pool is quickly exhausted, which causes a problem that an IP address cannot be allocated to a new device.
IPv6 (Internet Protocol version 6) is the next generation protocol of the internet for assigning unique IP addresses to devices connected to the internet. IPv6 expands the IP address space, providing an almost unlimited number of IP addresses, allowing for unique global IP addresses to be provided for internet of things devices, no longer subject to address shortages. The large address space of the IPv6 means that more Internet of things equipment can be accommodated, so that more applications and innovations are supported, and the large address space comprises various fields of intelligent home, intelligent medical treatment, intelligent transportation systems and the like. The IPv6 design also considers better routing and address allocation methods, simplifying the routing table.
Chinese patent application number: CN201610709778.9, publication date: 2019.06.18 discloses a geographical location and application information based addressing method for IPv6, comprising: and acquiring the position information and the application information of the equipment of the Internet of things: the method comprises the steps that an Internet of things device sends a broadcast request; the sink node receives the broadcast request and sends the measured state of the Internet of things equipment to an address configuration server; mapping of IPv6 address with location information and application information: the address configuration server maps the obtained position information and application information into a unique IPv6 address containing geographic position and application information; IPv6 address configuration of an Internet of things device: and completing communication and configuration mechanisms with sink nodes and the Internet of things equipment on an address configuration server. According to the method, binding and configuration mechanisms of IPv6 and IoT device position information and application information are completed aiming at the characteristics of the internet of things device and the internet of things data acquisition, a mapping algorithm is researched, and the above-described models are realized. And a loop of the Internet of things taking the information as the center is initially established, so that the use efficiency of the Internet of things is improved.
The above technology has at least the following technical problems: the prior art generally fails to effectively cope with dynamic changes in the environment of the internet of things, such as frequent changes of network topology caused by dynamic joining and off-line of equipment, and availability, stability and the like of the network are affected by equipment faults; in addition, the prior art also lacks of performing cluster analysis and address dynamic optimization allocation on the devices, and the deficiency directly affects the improvement of network management efficiency and addressing efficiency.
Disclosure of Invention
The embodiment of the application solves the problem that the prior art generally cannot effectively cope with dynamic changes in the environment of the Internet of things by providing the network addressing method, such as frequent changes of network topology caused by dynamic joining and off-line of equipment, and the availability, stability and the like of the network are influenced by the failure of the equipment; in addition, the problem that the prior art lacks in performing cluster analysis and address dynamic optimization allocation on equipment, so that the network management efficiency and the improvement of the addressing efficiency are directly influenced is solved. The application realizes intelligent equipment clustering and intelligent management equipment, can better utilize address resources, reduce address conflict, improve the efficiency of address allocation, reduce delay and improve the bandwidth utilization rate.
The application provides a network addressing method, which specifically comprises the following technical scheme:
a network addressing method comprising the steps of:
s1, collecting characteristic state vectors of each piece of Internet of things equipment, deducing a similarity matrix, constructing a Laplace matrix of a graph, carrying out characteristic vector analysis on the Laplace matrix, constructing a new matrix, updating a clustering center, and putting similar equipment in the same group to realize intelligent network partition of the pieces of Internet of things equipment;
s2, mapping the characteristic state vector of the equipment to an IPv6 address through deep learning in each network partition, constructing an IPv6 addressing model, regarding an IPv6 address space as a multi-dimensional wind field, converting the IPv6 address allocation problem into a multi-dimensional optimization problem, and optimizing through a multi-objective genetic algorithm to obtain an optimal IPv6 address allocation scheme.
Preferably, the step S1 specifically includes:
based on the characteristic state vector, calculating the similarity between each device and each initial cluster centerAnd generating a similarity matrix, and constructing a Laplacian matrix of the standardized graph based on the similarity matrix:
，
wherein ,representing the identity matrix, the D-representation matrix being a diagonal matrix with diagonal elements +.>Is the degree of node i, +.>，/>The method comprises the steps of carrying out a first treatment on the surface of the And calculating the eigenvectors corresponding to the minimum K nonzero eigenvalues of the normalized Laplace matrix L.
Preferably, the step S1 further includes:
the minimum K non-zero eigenvalues of the Laplace matrix are used for corresponding eigenvectors to form a matrix，/>Is +.>Clustering each row of the matrix U; taking the mean value of the characteristic state vectors of the group internal devices of each initial cluster as a new cluster center; intelligent network implemented by clusteringThe network partitions are automatically divided according to the characteristics of the devices, and similar devices are placed in the same group.
Preferably, the step S2 specifically includes:
in each network partition, encoding and decoding the characteristic state vector of the device, then mapping the characteristic state vector of the device to an IPv6 address, allocating a locally unique IPv6 address to each device, and introducing randomness to adapt to continuously changing network conditions so that each input is mapped to at least one potential output; constructing a variational automatic encoder model, and mapping characteristic state vectors to IPv6 addresses by combining an encoder, a decoder and probability distributions; the IPv6 address is optimized by generating new data points against the network.
Preferably, the step S2 further includes:
constructing an IPv6 addressing model through a multi-target genetic algorithm, and regarding an IPv6 address space as a multi-dimensional wind field, wherein each IPv6 address is regarded as a fluid element in the wind field and has the properties of position, speed and pressure; the Internet of things equipment is regarded as an attraction source or a rejection source in a wind field; randomly generating a set of initial solutions, each solution representing an IPv6 address allocation scheme, each address allocation scheme being considered as a state in a multi-dimensional wind park, said solutions constituting a population; the location, speed and pressure of each address element will be the genes of the individual, while the attracting or repelling device will be considered part of the environment.
Preferably, the step S2 further includes:
for each solution in the population, calculating each objective function, the objective function comprising: minimizing address conflict objective functionMinimizing address allocation distance->And maximizing resource utilization +.>The method comprises the steps of carrying out a first treatment on the surface of the Based on objective functionNumber, defining fitness function:
，
wherein ,weights are represented for balancing the importance of the respective targets.
Preferably, the step S2 further includes:
the specific steps of non-dominant ordering include:
a. initializing a non-dominant solution set:
creating an empty non-dominant solution set to store solutions in Pareto fronts;
b. calculating a dominance relation:
for each pair of solutions, solutions A and B, determining a dominance relationship between them; one solution a dominates the other solution B, then solution a is at least as good as solution B on all objective functions and better than solution B on at least one objective function; the dominance relationship is calculated by:
for all objective functions j, aj < = Bj and at least one j is present such that Aj < Bj, then solution A dominates solution B;
c. updating the non-dominant solution set:
adding solutions that are not subject to solutions other than self to a non-subject solution set, and when no solution other than self subjects itself, adding a solution itself to a non-subject solution set;
d. removing the dominant solution from the population:
removing solutions from the population that are dominated by solutions other than themselves in preparation for generating a next generation population;
e. repeating steps b-d:
repeating steps b-d until no more solutions are added to the non-dominant solution set;
f. forming Pareto front:
finally, the solutions in the non-dominant solution set form a Pareto front, which contains a set of optimal solutions, and the optimal solutions have no obvious improvement space under at least two objective functions, and at the same time represent the optimal IPv6 address allocation scheme.
The beneficial effects are that:
the technical schemes provided by the embodiment of the application have at least the following technical effects or advantages:
1. according to the application, the management efficiency, performance and maintainability of the Internet of things equipment network are improved through data analysis and automation technology, and meanwhile, the network partition can be dynamically adjusted according to the actual change of the equipment state, so that the Internet of things environment is more suitable for the change and different requirements are better met. Clustering the devices into different groups by computing the similarity between the devices helps to organize the devices in order, thereby making management and maintenance easier. After clustering the devices, the center of each cluster may be the new cluster center. The intelligent network partition can be automatically divided according to the characteristics of the equipment, and the maintainability and the performance of the network can be improved. By clustering, similar devices are placed within the same group, thereby reducing redundancy and confusion in the network;
2. the characteristic state vector of the equipment is mapped to the IPv6 address through the deep learning technology, so that the address can be more accurately distributed to the equipment, and the utilization efficiency of address resources is improved; the IPv6 address allocation problem is converted into a multidimensional optimization problem, so that a plurality of optimization targets, such as address conflict reduction, network performance improvement and the like, are allowed to be considered simultaneously, and the network is more comprehensively optimized; the multi-objective genetic algorithm allows finding a group of balanced solutions, which is helpful for meeting different network requirements and improving the flexibility of the network; through intelligent dynamic optimization allocation, IPv6 address resources can be managed more effectively, waste and excessive allocation of resources are reduced, address conflicts and faults are reduced, and therefore stability and reliability of the network are improved.
3. The technical scheme of the application solves the problem that the prior art generally cannot effectively cope with dynamic changes in the environment of the Internet of things, such as frequent changes of network topology caused by dynamic addition and off-line of equipment, and the availability, stability and the like of the network are affected by the faults of the equipment; in addition, the problem that the prior art lacks of carrying out cluster analysis and address dynamic optimization allocation on equipment is solved, so that the network management efficiency and the improvement of addressing efficiency are directly influenced, the intelligent equipment clustering and intelligent management equipment are realized, address resources can be better utilized, address conflicts are reduced, the efficiency of address allocation is improved, delay is reduced, and the bandwidth utilization rate is improved.
Drawings
FIG. 1 is a flow chart of a network addressing method according to the present application;
Detailed Description
The embodiment of the application solves the problem that the prior art generally cannot effectively cope with dynamic changes in the environment of the Internet of things by providing the network addressing method, such as frequent changes of network topology caused by dynamic joining and off-line of equipment, and the availability, stability and the like of the network are influenced by the failure of the equipment; in addition, the problem that the prior art lacks in performing cluster analysis and address dynamic optimization allocation on equipment, so that the network management efficiency and the improvement of the addressing efficiency are directly influenced is solved.
The technical scheme in the embodiment of the application aims to solve the problems, and the overall thought is as follows:
firstly, acquiring characteristic state vectors of each Internet of things device, calculating the similarity between each device and each initial clustering center, constructing a Laplace matrix of a graph, carrying out feature vector analysis on the Laplace matrix, forming a new matrix by the feature vectors, forming a new clustering center, and carrying out network partition on the Internet of things device; the management efficiency, performance and maintainability of the Internet of things equipment network are improved through data analysis and automation technology, and meanwhile, the network partition can be dynamically adjusted according to the actual change of the equipment state, so that the Internet of things environment is more suitable for the change and different requirements are better met. Then, in each network partition, mapping the characteristic state vector of the equipment to an IPv6 address through deep learning, constructing an IPv6 addressing model, regarding an IPv6 address space as a multi-dimensional wind field, thereby converting the IPv6 address allocation problem into a multi-dimensional optimization problem, and optimizing through a multi-objective genetic algorithm to realize dynamic optimization allocation of the IPv6 address. The characteristic state vector of the equipment is mapped to the IPv6 address through the deep learning technology, so that the address can be more accurately distributed to the equipment, and the utilization efficiency of address resources is improved; the IPv6 address allocation problem is converted into a multidimensional optimization problem, so that a plurality of optimization targets, such as address conflict reduction, network performance improvement and the like, are allowed to be considered simultaneously, and the network is more comprehensively optimized; the multi-objective genetic algorithm allows finding a group of balanced solutions, which is helpful for meeting different network requirements and improving the flexibility of the network; through intelligent dynamic optimization allocation, IPv6 address resources can be managed more effectively, waste and excessive allocation of resources are reduced, address conflicts and faults are reduced, and therefore stability and reliability of the network are improved.
In order to better understand the above technical solutions, the following detailed description will refer to the accompanying drawings and specific embodiments.
Referring to fig. 1, a network addressing method according to the present application includes the following steps:
s1, collecting characteristic state vectors of each piece of Internet of things equipment, deducing a similarity matrix, constructing a Laplace matrix of a graph, carrying out characteristic vector analysis on the Laplace matrix, constructing a new matrix, updating a clustering center, and putting similar equipment in the same group to realize intelligent network partition of the pieces of Internet of things equipment;
and collecting a characteristic state vector of each piece of equipment of the Internet of things, wherein the characteristic state vector is a comprehensive equipment characteristic state vector which consists of static characteristics such as sensor data, hardware configuration, running state and the like of the equipment and dynamic network states such as networking state, network quality, energy consumption and the like of the equipment.
Calculating the similarity between each device and each initial clustering center based on the characteristic state vector, generating a similarity matrix, constructing a Laplacian matrix of a graph based on the similarity matrix, carrying out feature vector analysis on the Laplacian matrix, forming a new matrix by the feature vector, dividing the feature vector into different clusters by using a clustering algorithm such as K-means clustering, and dividing the center of each cluster into a new clustering center so as to carry out network partition on the Internet of things device;
as a specific embodiment, N characteristic state vectors are set for the Internet of things equipment, so that the characteristic state vectors of the Internet of things equipmentComprising the functions, location, energy consumption, processing capacity, memory capacity of the device, for example, wherein the functions of the device are +.>The device position is->The energy consumption of the seed and the equipment is->The device processing capability is +.>The device memory capacity is->A kind of module is assembled in the module and the module is assembled in the module. The characteristic state vector of the device is represented by a binary vector, which is marked 1 if the device has the current characteristic, and 0 otherwise. All vectors can be spliced into one +.>Characteristic state vector of dimension, denoted +.>。
Randomly selecting K devices as initial clustering centers, and recording asEach of which is->Is a kind of having->Device for maintaining a characteristic state vector, < >>，/>Representing the device characteristic state vector dimensions,. For each device, calculating the distance between the current device and each cluster center based on the characteristic state vector of the device, and introducing a weight coefficient when calculating the distance>The importance of each characteristic state vector of the corresponding equipment is calculated according to the following formula:
，
wherein ,a state vector representing the u-th characteristic of the device, < >>Representing the kth initial cluster center device +.>Is the u-th characteristic state vector of +.>Representing the importance of the u-th characteristic state vector,/->Represents dot product->Representing the transpose of the vector.
According to the similarity between the equipment and the distance calculation equipment of each initial clustering center, the calculation formula is as follows:
，
wherein ,representing the similarity between the ith device and the jth device,/and->Represents the distance of the ith device from each cluster center,/->A u-th characteristic state vector representing an i-th device,>representing the variance of the characteristic state vector.
The similarity matrix can be obtained by the calculation formula of the similarity between the calculation devices, the Laplacian matrix of the graph is constructed based on the similarity matrix, and the degree matrix D is a diagonal matrix with diagonal elementsIs the degree of node i, i.e
，
Then, a normalized Laplace matrix is constructed:
，
wherein ,representing an identity matrix>。
The feature vectors corresponding to the minimum K non-zero feature values of the standardized Laplace matrix L are calculated, and the steps are as follows:
(1) Randomly initializing a non-zero vectorAnd normalize it, i.e. +.>, wherein />Representation->2-norms of (2);
(2) For each iteration i=1, 2,3, … …, the following steps are performed:
calculating a new vector:
，
orthogonalizationTo remove the effects of known feature vectors:
,/>，
calculation of2-norm of (2), noted +.>；
Updating estimated feature vectors：
，
Calculating an inverse estimate of the eigenvalue:
，
(3) Repeating step (2) until a stop condition (the variation of the characteristic value estimation is smaller than a preset threshold value) is satisfiedOr up to a maximum number of iterations);
(4) Repeating steps (2) to (3) until K different minimum eigenvalue estimates and their corresponding eigenvectors are obtained.
The minimum K non-zero eigenvalues of the Laplace matrix are used for corresponding eigenvectors to form a matrix，/>Is +.>Each row of the matrix U is clustered. The mean value calculation formula of the characteristic state vector of the group internal device of each initial cluster is as follows:
，
wherein ,representing the number of devices within the current cluster group. Taking the mean value as a new cluster center, and ending the algorithm if the change of all cluster centers is smaller than a certain preset threshold value or the preset maximum iteration number is carried out. The change of all the cluster centers can be measured by calculating the sum of Euclidean distances of the new cluster center and the old cluster center, and the calculation formula is as follows:
，
wherein ,representing the magnitude of the change in cluster center, +.>Representing the changed kth cluster center,/, for example>Representing the kth cluster center before the change.
The intelligent network partition according to the method can be automatically divided according to the characteristics of the equipment, and the maintainability and the performance of the network can be improved. By clustering, similar devices are placed within the same group, thereby reducing redundancy and confusion in the network.
S2, mapping the characteristic state vector of the equipment to an IPv6 address through deep learning in each network partition, constructing an IPv6 addressing model, regarding an IPv6 address space as a multi-dimensional wind field, converting the IPv6 address allocation problem into a multi-dimensional optimization problem, and optimizing through a multi-objective genetic algorithm to obtain an optimal IPv6 address allocation scheme.
In each network partition, the device's characteristic state vector is encoded and decoded, then mapped to an IPv6 address, each device is assigned a locally unique IPv6 address, and randomness is introduced to accommodate changing network conditions so that each input can be mapped to multiple possible outputs. A deep learning based generation model, a variational automatic encoder (Variational Autoencoder, VAE) model, is constructed that is used to learn potential representations of data and can be used to generate new data points. The automatic variable encoder is realized by combining the automatic encoder and probability distribution, and the specific VAE realization steps are as follows:
(1) An encoder:
the encoder gaussian distribution (normal distribution) of the VAE will input dataMapping to probability distributions in potential space. The encoder consists of two parts: mean->And standard deviation->. Encoder calculation process:
，
，
wherein , and />Are neural network layers, which are derived from the input +.>The mean and the logarithmic variance are extracted to define a normal distribution in the potential space.
(2) Sampling:
sampling a point from a normal distribution to obtain potential vectorsThis is an encoded representation. The sampling process is as follows:
，
wherein ,is distributed from standard normal (+)>N (0, 1)).
(3) A decoder:
the decoder uses the potential vector through the neural networkMapping back to the original data space for generating and inputting data +.>Similar data.
Decoder calculation process:
，
(4) Loss function:
a loss function is defined that includes the reconstruction loss and the KL divergence loss. Reconstruction loss is used to measure inputAnd output->The difference between them is given by:
，
KL divergence loss is used to penalize potential vectorsThe difference from the standard normal distribution to ensure that the vector distribution in the potential space is close to the normal distribution is calculated as follows:
，
the total loss function is a weighted sum of the reconstruction loss and the KL divergence loss:
，
wherein ,is the weight of the KL divergence penalty for balancing the two penalty terms.
Generating a countermeasure network (Generative Adversarial Networks, GAN) can be used to generate new data points, using GAN to generate and optimize IPv6 addresses, with the following implementation steps:
a generator:，
a discriminator:，
wherein ,representing the fake IPv6 address generated by the generator, < +.>Indicating the authenticity of the arbiter to the IPv6 address,/-> and />Representing an activation function-> and />Respectively representing the weight and bias of the generator, +.> and />The weights and biases of the discriminators are represented, respectively. By alternately training the generator and the arbiter, performance is gradually improved. Once training is complete, a new device characteristic state vector is generated by a generator>And maps it to an IPv6 address.
In the IPv6 addressing process based on the Internet of things, an IPv6 addressing model is built through a Multi-objective genetic algorithm (Multi-Objective Genetic Algorithm, MOGA), and the MOGA can find a group of solutions which are optimal or near optimal on a plurality of objectives, so that the optimal trade-off in the aspects of network performance, address utilization rate, address uniqueness and the like is ensured.
The specific implementation process of the IPv6 addressing model is as follows:
(1) Modeling an address space:
the IPv6 address space is considered a multi-dimensional wind farm, where each IPv6 address is considered a fluid element in the wind farm, with the following attributes:
position: representing the position of an IPv6 address in an address space (wind farm), using multidimensional coordinates;
speed of: representing the motion speed of an IPv6 address in an address space (wind field) for simulating the movement of elements;
pressure: representing the density or distribution of IPv6 addresses in an address space (wind farm) that may be varied by the influence of the internet of things devices.
(2) Modeling equipment:
the internet of things devices are considered as attractive or repulsive sources in the wind field, and have the following properties:
position: the location of the device may affect the wind field flow in the address space;
attractive or repulsive forces: each device may generate an attractive or repulsive force affecting the motion and distribution of nearby address elements.
(3) Multi-objective genetic algorithm:
a set of initial solutions are randomly generated, each representing an IPv6 address allocation scheme, each of which can be viewed as a state in a multi-dimensional wind farm. These solutions constitute a population. The location, speed and pressure of each address element will be the genes of the individual, while the attracting or repelling device will be considered part of the environment.
For each solution in the population, each objective function is calculated, including:
there are N devices, each of which needs to be assigned an IPv6 address, denoted as。
Objective function 1: address collision is minimized, and each device is assigned as much as possible a unique IPv6 address. When all devices are assigned unique IPv6 addresses, the value of the objective function will be zero, indicating no address collision. The minimized address collision objective function is defined as follows:
，
wherein ,indicating the IPv6 address assigned by device i, +.>Indicating that the address +.>Is a number of times (1).
Objective function 2: minimizing the address allocation distance can reduce the delay of communication and improve the network performance by minimizing the address allocation distance between devices. The minimum address allocation distance objective function is defined as follows:
，
wherein , and />Indicating the IPv6 address assigned by device i and device j, respectively,>representing address->And address->Euclidean distance between them.
Objective function 3: maximizing resource utilization, by maximizing the ratio of the number of addresses that have been allocated to the total number of available addresses, achieves maximum utilization of resources. The maximum resource utilization objective function is defined as follows:
let the total number of IPv6 addresses be N, where M addresses have been assigned to the internet of things device.
，
The fitness function is defined as follows:
，
weights are represented for balancing the importance of the respective targets.
(4) Non-dominant ordering:
a. initializing a non-dominant solution set:
an empty non-dominant solution set (as a list or set) is created to store solutions in Pareto fronts.
b. Calculating a dominance relation:
for each pair of solutions (solutions a and B), the dominant relationship between them is determined. One solution a dominates the other solution B, meaning that solution a is at least as good as solution B on all objective functions and better than solution B on at least one objective function. The dominance relationship may be calculated by:
solution a dominates solution B if for all objective functions j, a [ j ] < = B [ j ] and there is at least one j such that a [ j ] < B [ j ].
c. Updating the non-dominant solution set:
solutions that are not subject to other solutions are added to the non-subject solution set, and if there are no other solutions subject to solution a, then solution a is added to the non-subject solution set.
d. Removing the dominant solution from the population:
solutions that are dominated by other solutions are removed from the population in preparation for generating a next generation population.
e. Repeating steps b-d:
steps b-d are repeated until no more solutions are added to the non-dominant solution set.
f. Forming Pareto front:
finally, solutions in the non-dominant solution set form a Pareto front, which contains a set of optimal solutions, none of which has significant room for improvement under multiple objective functions, and at the same time, represent the optimal IPv6 address allocation scheme under multiple objectives.
By the mode, the model can adapt to the behavior dynamic state of the equipment in an efficient mode, and dynamically performs self-optimization, so that IPv6 addressing of the equipment of the Internet of things is realized.
In summary, the network addressing method of the present application is completed.
The technical scheme provided by the embodiment of the application at least has the following technical effects or advantages:
1. according to the application, the management efficiency, performance and maintainability of the Internet of things equipment network are improved by utilizing data analysis and automation technology, and meanwhile, the network partition can be dynamically adjusted according to the actual change of the equipment state, so that the Internet of things environment is more suitable for the change and different requirements are better met. Clustering the devices into different groups by computing the similarity between the devices helps to organize the devices in order, thereby making management and maintenance easier. After clustering the devices, the center of each cluster may be the new cluster center. The intelligent network partition can be automatically divided according to the characteristics of the equipment, and the maintainability and the performance of the network can be improved. By clustering, similar devices are placed within the same group, thereby reducing redundancy and confusion in the network;
2. the characteristic state vector of the equipment is mapped to the IPv6 address through the deep learning technology, so that the address can be more accurately distributed to the equipment, and the utilization efficiency of address resources is improved; the IPv6 address allocation problem is converted into a multidimensional optimization problem, so that a plurality of optimization targets, such as address conflict reduction, network performance improvement and the like, are allowed to be considered simultaneously, and the network is more comprehensively optimized; the multi-objective genetic algorithm allows finding a group of balanced solutions, which is helpful for meeting different network requirements and improving the flexibility of the network; through intelligent dynamic optimization allocation, IPv6 address resources can be managed more effectively, waste and excessive allocation of resources are reduced, address conflicts and faults are reduced, and therefore stability and reliability of the network are improved.
Effect investigation:
the technical scheme of the application solves the problem that the prior art generally cannot effectively cope with dynamic changes in the environment of the Internet of things, such as frequent changes of network topology caused by dynamic addition and off-line of equipment, and the availability, stability and the like of the network are affected by the faults of the equipment; in addition, the problem that the prior art lacks of carrying out cluster analysis and address dynamic optimization allocation on equipment is solved, so that the network management efficiency and the improvement of addressing efficiency are directly influenced, the intelligent equipment clustering and intelligent management equipment are realized, address resources can be better utilized, address conflicts are reduced, the efficiency of address allocation is improved, delay is reduced, and the bandwidth utilization rate is improved.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present application have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiments and all such alterations and modifications as fall within the scope of the application.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present application without departing from the spirit or scope of the application. Thus, it is intended that the present application also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.
Claims (7)
1. A network addressing method comprising the steps of:
s1, collecting characteristic state vectors of each piece of Internet of things equipment, deducing a similarity matrix, constructing a Laplace matrix of a graph, carrying out characteristic vector analysis on the Laplace matrix, constructing a new matrix, updating a clustering center, and putting similar equipment in the same group to realize intelligent network partition of the pieces of Internet of things equipment;
s2, mapping the characteristic state vector of the equipment to an IPv6 address through deep learning in each network partition, constructing an IPv6 addressing model, regarding an IPv6 address space as a multi-dimensional wind field, converting the IPv6 address allocation problem into a multi-dimensional optimization problem, and optimizing through a multi-objective genetic algorithm to obtain an optimal IPv6 address allocation scheme.
2. The network addressing method according to claim 1, wherein the step S1 specifically comprises:
based on the characteristic state vector, calculating the similarity between each device and each initial cluster centerAnd generating a similarity matrix, and constructing a Laplacian matrix of the standardized graph based on the similarity matrix:
，
wherein ,representing the identity matrix, the D-representation matrix being a diagonal matrix with diagonal elements +.>Is the degree of the node i and,，/>the method comprises the steps of carrying out a first treatment on the surface of the And calculating the eigenvectors corresponding to the minimum K nonzero eigenvalues of the normalized Laplace matrix L.
3. A network addressing method according to claim 2, characterized in that said step S1 further comprises:
the minimum K non-zero eigenvalues of the Laplace matrix are used for corresponding eigenvectors to form a matrix，/>Is +.>Clustering each row of the matrix U; taking the mean value of the characteristic state vectors of the group internal devices of each initial cluster as a new cluster center; the intelligent network partition realized by clustering is automatically divided according to the characteristics of the devices, and similar devices are placed in the same group.
4. The network addressing method according to claim 1, wherein the step S2 specifically comprises:
in each network partition, encoding and decoding the characteristic state vector of the device, then mapping the characteristic state vector of the device to an IPv6 address, allocating a locally unique IPv6 address to each device, and introducing randomness to adapt to continuously changing network conditions so that each input is mapped to at least one potential output; constructing a variational automatic encoder model, and mapping characteristic state vectors to IPv6 addresses by combining an encoder, a decoder and probability distributions; the IPv6 address is optimized by generating new data points against the network.
5. The network addressing method according to claim 1, wherein said step S2 further comprises:
constructing an IPv6 addressing model through a multi-target genetic algorithm, and regarding an IPv6 address space as a multi-dimensional wind field, wherein each IPv6 address is regarded as a fluid element in the wind field and has the properties of position, speed and pressure; the Internet of things equipment is regarded as an attraction source or a rejection source in a wind field; randomly generating a set of initial solutions, each solution representing an IPv6 address allocation scheme, each address allocation scheme being considered as a state in a multi-dimensional wind park, said solutions constituting a population; the location, speed and pressure of each address element will be the genes of the individual, while the attracting or repelling device will be considered part of the environment.
6. The network addressing method of claim 5, wherein said step S2 further comprises:
for each solution in the population, calculating each objective function, the objective function comprising: minimizing address conflict objective functionMinimizing address allocation distance->And maximizing resource utilization +.>The method comprises the steps of carrying out a first treatment on the surface of the Based on the objective function, an fitness function is defined:
，
wherein ,weights are represented for balancing the importance of the respective targets.
7. The network addressing method of claim 6, wherein said step S2 further comprises:
the specific steps of non-dominant ordering include:
a. initializing a non-dominant solution set:
creating an empty non-dominant solution set to store solutions in Pareto fronts;
b. calculating a dominance relation:
for each pair of solutions, solutions A and B, determining a dominance relationship between them; one solution a dominates the other solution B, then solution a is at least as good as solution B on all objective functions and better than solution B on at least one objective function; the dominance relationship is calculated by:
for all objective functions j, aj < = Bj and at least one j is present such that Aj < Bj, then solution A dominates solution B;
c. updating the non-dominant solution set:
adding solutions that are not subject to solutions other than self to a non-subject solution set, and when no solution other than self subjects itself, adding a solution itself to a non-subject solution set;
d. removing the dominant solution from the population:
removing solutions from the population that are dominated by solutions other than themselves in preparation for generating a next generation population;
e. repeating steps b-d:
repeating steps b-d until no more solutions are added to the non-dominant solution set;
f. forming Pareto front:
finally, the solutions in the non-dominant solution set form a Pareto front, which contains a set of optimal solutions, and the optimal solutions have no obvious improvement space under at least two objective functions, and at the same time represent the optimal IPv6 address allocation scheme.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311181075.XA CN116915746B (en) | 2023-09-14 | 2023-09-14 | IPv6 addressing method based on Internet of things |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311181075.XA CN116915746B (en) | 2023-09-14 | 2023-09-14 | IPv6 addressing method based on Internet of things |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116915746A true CN116915746A (en) | 2023-10-20 |
| CN116915746B CN116915746B (en) | 2023-11-21 |
Family
ID=88367268
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311181075.XA Active CN116915746B (en) | 2023-09-14 | 2023-09-14 | IPv6 addressing method based on Internet of things |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116915746B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117354063A (en) * | 2023-12-04 | 2024-01-05 | 明阳产业技术研究院(沈阳)有限公司 | IPv 6-based intelligent internet terminal management method, system, medium and equipment |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108363740A (en) * | 2018-01-22 | 2018-08-03 | 中国平安人寿保险股份有限公司 | Analysis method, device, storage medium and the terminal of IP address |
| CN109871972A (en) * | 2017-12-01 | 2019-06-11 | 北京普华亿能风电技术有限公司 | A kind of optimization placement method and system of wind field inner blower |
| CN111065103A (en) * | 2019-12-11 | 2020-04-24 | 哈尔滨工程大学 | Multi-objective optimization wireless sensor network node deployment method |
| US20210012199A1 (en) * | 2019-07-04 | 2021-01-14 | Zhejiang University | Address information feature extraction method based on deep neural network model |
| CN114662201A (en) * | 2022-03-31 | 2022-06-24 | 中国石油大学(华东) | Optimizing method for intelligent regulation and control of natural ventilation |
| CN116628903A (en) * | 2023-05-22 | 2023-08-22 | 哈尔滨工业大学（深圳） | Optimal arrangement method for urban wind field environment monitoring sensors |
-
2023
- 2023-09-14 CN CN202311181075.XA patent/CN116915746B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109871972A (en) * | 2017-12-01 | 2019-06-11 | 北京普华亿能风电技术有限公司 | A kind of optimization placement method and system of wind field inner blower |
| CN108363740A (en) * | 2018-01-22 | 2018-08-03 | 中国平安人寿保险股份有限公司 | Analysis method, device, storage medium and the terminal of IP address |
| US20210012199A1 (en) * | 2019-07-04 | 2021-01-14 | Zhejiang University | Address information feature extraction method based on deep neural network model |
| CN111065103A (en) * | 2019-12-11 | 2020-04-24 | 哈尔滨工程大学 | Multi-objective optimization wireless sensor network node deployment method |
| CN114662201A (en) * | 2022-03-31 | 2022-06-24 | 中国石油大学(华东) | Optimizing method for intelligent regulation and control of natural ventilation |
| CN116628903A (en) * | 2023-05-22 | 2023-08-22 | 哈尔滨工业大学（深圳） | Optimal arrangement method for urban wind field environment monitoring sensors |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117354063A (en) * | 2023-12-04 | 2024-01-05 | 明阳产业技术研究院(沈阳)有限公司 | IPv 6-based intelligent internet terminal management method, system, medium and equipment |
| CN117354063B (en) * | 2023-12-04 | 2024-04-02 | 明阳产业技术研究院(沈阳)有限公司 | IPv 6-based intelligent internet terminal management method, system, medium and equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116915746B (en) | 2023-11-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Balakrishnan et al. | Diverse client selection for federated learning via submodular maximization | |
| CN107172166B (en) | Cloud and mist computing system for industrial intelligent service | |
| CN110688219B (en) | Adaptive weight load balancing algorithm based on reverse chaotic cuckoo search | |
| CN116915746B (en) | IPv6 addressing method based on Internet of things | |
| CN110765320B (en) | Data processing method, device, storage medium and computer equipment | |
| Gagné et al. | Distributed beagle: An environment for parallel and distributed evolutionary computations | |
| EP4350572A1 (en) | Method, apparatus and system for generating neural network model, devices, medium and program product | |
| CN114332984B (en) | Training data processing method, device and storage medium | |
| CN115358487A (en) | Federal learning aggregation optimization system and method for power data sharing | |
| WO2021008675A1 (en) | Dynamic network configuration | |
| Chen et al. | Deep-broad learning system for traffic flow prediction toward 5G cellular wireless network | |
| CN110809066A (en) | IPv6 address generation model creation method, device and address generation method | |
| Zhu et al. | FedOVA: one-vs-all training method for federated learning with non-IID data | |
| Li et al. | Data analytics for fog computing by distributed online learning with asynchronous update | |
| CN114781688A (en) | Method, device, equipment and storage medium for identifying abnormal data of business expansion project | |
| Gogineni et al. | Personalized online federated learning for IoT/CPS: Challenges and future directions | |
| CN104618480A (en) | Cloud system source distributing method driven on basis of network link utilization rates | |
| CN108595272B (en) | Request distribution method and device and computer readable storage medium | |
| CN111415265A (en) | Social relationship data generation method of generative confrontation network | |
| CN116187383A (en) | Method for determining 3D NoC optimal test plan based on improved whale optimization algorithm | |
| CN114531665A (en) | Wireless sensor network node clustering method and system based on Laiwei flight | |
| CN114329127A (en) | Characteristic box dividing method, device and storage medium | |
| CN115618532A (en) | Network system simulation method and related device | |
| CN117424813B (en) | Node expansion method for block chain | |
| CN111369374B (en) | Social network time sequence link prediction method and device based on probability generation formula |
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 |