CN114666254A - Network performance testing method and system for whole-house router system - Google Patents
Network performance testing method and system for whole-house router system Download PDFInfo
- Publication number
- CN114666254A CN114666254A CN202210235630.1A CN202210235630A CN114666254A CN 114666254 A CN114666254 A CN 114666254A CN 202210235630 A CN202210235630 A CN 202210235630A CN 114666254 A CN114666254 A CN 114666254A
- Authority
- CN
- China
- Prior art keywords
- matrix
- feature
- router
- classification
- signal strength
- 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.)
- Withdrawn
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/50—Testing arrangements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Environmental & Geological Engineering (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The field of network performance testing of routers of the application specifically discloses a network performance testing method and system for a whole-house router system, which extracts topological features among a plurality of sub-routers based on a convolutional neural network model, obtains global strength correlation information of each sub-router by using a context coding model comprising an embedded layer, further extracts correlation information of a data sample due to feature information and irregular topological structure information by using a graph neural network, and performs orthogonal dimension decomposition on signal features, so that weak correlation information of the data sample is ignored, and only strong correlation information is retained, and the accuracy of the obtained classification result is higher. In this way, the performance of the router system can be better checked to ensure that the basic communication needs are met at all locations indoors.
Description
Technical Field
The present invention relates to the field of network performance testing of routers, and more particularly, to a network performance testing method and system for a whole-house router system.
Background
At present, with the improvement of the living quality of people, houses are larger and larger, the attention on Wi-Fi signal coverage is higher and higher, the traditional router cannot well meet the requirement of a user on network coverage, and a Wi-Fi System is produced. The whole-house routing can meet the requirements of larger and larger home area and wider network coverage of users, and on the other hand, the whole-house Wi-Fi is a cornerstone for all hardware devices of the smart home to access into the network. The whole-house router is a router system formed by a parent router (GW) and a plurality of child Routers (RE) through special links, and the problem of coverage of the traditional router is solved.
After the router system is constructed, the performance of the router system needs to be tested at various locations in the room to ensure that the basic communication requirements can be met at various locations indoors. However, when testing the performance of the router system, the existing testing method only considers the received signal strength of the terminal device at a specific position, which may make the measured result inaccurate. Therefore, a network performance testing method for a whole-house router system is desired in order to better test the performance of the router system to ensure that the basic communication requirements can be met at various locations indoors.
Disclosure of Invention
The present application is proposed to solve the above-mentioned technical problems. The embodiment of the application provides a network performance testing method and a system thereof for a whole house router system, which are used for extracting topological features among a plurality of sub-routers based on a convolutional neural network model, simultaneously obtaining global strength associated information of each sub-router by utilizing a context coding model comprising an embedded layer, further using a graph neural network to extract associated information of a data sample due to feature information and irregular topological structure information, and further performing orthogonal dimension decomposition on signal features, thereby ignoring weak associated information thereof and only keeping strong associated information, and enabling the accuracy of the obtained classification result to be higher. In this way, the performance of the router system can be better checked to ensure that the basic communication needs are met at all locations indoors.
According to one aspect of the present application, there is provided a network performance testing method for a whole house router system, comprising:
acquiring a topology matrix among a plurality of sub routers connected with a parent router, wherein the characteristic value of each position on the off-diagonal position of the topology matrix is the distance between two corresponding sub routers, and the characteristic value of each position on the diagonal position of the topology matrix is the distance between a sub router and the parent router;
processing the topological matrix using a convolutional neural network as a feature extractor to obtain a topological feature matrix;
obtaining the signal strength of each terminal device connected with each sub-router and the signal strength of the test device at the first position;
inputting the signal strength of each terminal device and the signal strength of the test device into a context coding model containing an embedded layer to obtain a strength feature vector of each terminal device and a test feature vector of the test device;
calculating a cross entropy value between the test feature vector and each of the intensity feature vectors;
calculating the weighting of each intensity feature vector by taking the cross entropy value as the weight of each intensity feature vector, and cascading the weighted intensity feature vectors to obtain feature expression vectors of each sub-router;
adjusting the feature expression vectors of the sub-routers to be uniform in length, and then performing two-dimensional splicing to obtain a feature expression matrix;
passing the feature representation matrix and the topological feature matrix through a graph neural network to obtain a global feature matrix;
performing eigenvalue decomposition on the global eigenvalue matrix to obtain n eigenvalues, wherein the n eigenvalues are used for representing essential characteristics of the global eigenvalue matrix in n mutually orthogonal dimensions of a high-dimensional space;
obtaining n characteristic values of the test equipment from the second position to the Yth position;
constructing n eigenvalues of the test equipment at Y positions into a classification matrix; and
and passing the classification matrix through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the network performance of the router system meets a preset requirement or not.
According to another aspect of the present application, there is provided a network performance testing system for a whole house router system, comprising:
a topology matrix obtaining unit, configured to obtain a topology matrix among a plurality of child routers connected to a parent router, where a feature value of each position on an off-diagonal position of the topology matrix is a distance between two corresponding child routers, and a feature value of each position on a diagonal position of the topology matrix is a distance between a child router and the parent router;
a feature extraction unit, configured to process the topology matrix obtained by the topology matrix obtaining unit using a convolutional neural network as a feature extractor to obtain a topology feature matrix;
a signal strength obtaining unit, configured to obtain a signal strength of each terminal device connected to each sub-router and a signal strength of the test device at the first location;
an encoding unit, configured to input the signal strength of the terminal device obtained by each signal strength obtaining unit and the signal strength of the test device obtained by the signal strength obtaining unit into a context coding model including an embedded layer to obtain a strength feature vector of each terminal device and a test feature vector of the test device;
a cross entropy value calculation unit, configured to calculate a cross entropy value between the test feature vector obtained by the encoding unit and the intensity feature vector obtained by each encoding unit;
a feature expression vector generation unit, configured to calculate a weight of each of the intensity feature vectors by using the cross entropy values obtained by the cross entropy value calculation unit as weights of the intensity feature vectors obtained by each of the encoding units, and cascade the weighted intensity feature vectors to obtain feature expression vectors of each of the sub routers;
the two-dimensional splicing unit is used for adjusting the feature expression vectors of the sub-routers obtained by the feature expression vector generation units to be uniform in length and then performing two-dimensional splicing to obtain a feature expression matrix;
the graph neural network processing unit is used for enabling the feature representation matrix obtained by the two-dimensional splicing unit and the topological feature matrix obtained by the feature extraction unit to pass through a graph neural network so as to obtain a global feature matrix;
an eigenvalue decomposition unit, configured to perform eigenvalue decomposition on the global eigenvalue matrix obtained by the graph neural network processing unit to obtain n eigenvalues, where the n eigenvalues are used to represent essential characteristics of the global eigenvalue matrix in n mutually orthogonal dimensions of a high-dimensional space;
the characteristic value acquisition unit is used for acquiring n characteristic values of the test equipment from a second position to a Y-th position;
a classification matrix constructing unit, configured to construct n eigenvalues of the test device at Y positions, which are obtained by the eigenvalue obtaining unit, into a classification matrix;
and the classification unit is used for enabling the classification matrix obtained by the classification matrix constructing unit to pass through a classifier so as to obtain a classification result, and the classification result is used for indicating whether the network performance of the router system meets the preset requirement or not.
Compared with the prior art, the network performance testing method and system for the whole house router system extract the topological features among the plurality of sub-routers based on the convolutional neural network model, obtain the global strength associated information of each sub-router by using the context coding model comprising the embedded layer, further use the graph neural network to extract the associated information of the data sample due to the feature information and the irregular topological structure information, and perform orthogonal dimension decomposition on the signal features, thereby neglecting the weak associated information and only keeping the strong associated information, so that the obtained classification result is higher in accuracy. In this way, the performance of the router system can be better checked to ensure that the basic communication needs are met at all locations indoors.
Drawings
The above and other objects, features and advantages of the present application will become more apparent by describing in more detail embodiments of the present application with reference to the attached drawings. The accompanying drawings are included to provide a further understanding of the embodiments of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the principles of the application. In the drawings, like reference numbers generally represent like parts or steps.
Fig. 1 is an application scenario diagram of a network performance testing method for a whole-house router system according to an embodiment of the present application;
fig. 2 is a flowchart of a network performance testing method for a whole house router system according to an embodiment of the present application;
fig. 3 is a schematic system architecture diagram of a network performance testing method for a whole-house router system according to an embodiment of the present application;
FIG. 4 is a block diagram of a network performance testing system for a whole house router system according to an embodiment of the application;
fig. 5 is a block diagram of an electronic device according to an embodiment of the application.
Detailed Description
Hereinafter, example embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be understood that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and that the present application is not limited by the example embodiments described herein.
Overview of a scene
As mentioned above, after the router system is constructed, the performance of the router system needs to be tested at various locations in the room to ensure that the basic communication requirements can be met at various locations in the room. However, when testing the performance of the router system, the existing testing method only considers the received signal strength of the terminal device at a specific position, which may make the measured result inaccurate. Therefore, a network performance testing method for a whole-house router system is desired in order to better test the performance of the router system to ensure that the basic communication requirements can be met at various locations indoors.
Based on this, in the technical scheme of the application, a topology matrix among a plurality of sub-routers is obtained, the corresponding position of the matrix is the distance among the sub-routers, and the diagonal position is the distance between the sub-router and the parent router, and the topology matrix is input to a convolutional neural network to obtain the topology characteristic matrix.
Then, the signal strength of the terminal device connected to each sub-router and the signal strength of the test device at a predetermined position are obtained, and a context coding model including an embedded layer is input to obtain a strength feature vector of each sub-router and a test feature vector of the test device. Then, the cross entropy value of the test feature vector and each strength feature vector is calculated to be used as the weight of the strength feature vector, and the strength feature vectors are weighted and then cascaded to obtain the feature expression vector of each sub-router.
And after the feature expression vectors of the plurality of sub-routers are linearly converted into the same length, performing two-dimensional splicing to obtain a feature expression matrix, and then passing through a graph neural network together with the topological feature matrix to obtain a global feature matrix.
Here, it is considered that although the global feature matrix includes rich information including the topology information of the correlation between the sub-routers and the signal feature information of the terminal devices of each sub-router based on the correlation of the test device, the correlation between the information has strong correlation and weak correlation, and the information expression degree is different when the feature expression vector is subjected to linear transformation, so that there is a lot of noise correlation information between the feature values of each position of the global feature matrix, and if the classification is directly performed by the classifier including the full connection layer, the classification accuracy is adversely affected.
Therefore, for the n × n diagonal matrix characteristic of the topological feature matrix forming the global feature matrix, the global feature matrix is subjected to eigenvalue decomposition to obtain n eigenvalues, so that the essential features of the global feature matrix in n mutually orthogonal dimensions of the high-dimensional space are represented by the n eigenvalues, which can also be understood as performing orthogonal dimension decomposition on the signal features, so as to ignore weak correlation information thereof and only retain strong correlation information.
Therefore, after each position in the plurality of positions is tested by using the testing equipment, the n characteristic values corresponding to each position are combined into the classification characteristic vector in a two-dimensional splicing mode to obtain a classification matrix, and then the classification matrix is input into a classifier, so that whether the network performance meets the requirement or not can be obtained.
Based on this, the present application provides a network performance testing method for a whole-house router system, which includes: acquiring a topology matrix among a plurality of sub routers connected with a parent router, wherein the characteristic value of each position on the off-diagonal position of the topology matrix is the distance between two corresponding sub routers, and the characteristic value of each position on the diagonal position of the topology matrix is the distance between the sub router and the parent router; processing the topological matrix by using a convolutional neural network as a feature extractor to obtain a topological feature matrix; obtaining the signal strength of each terminal device connected with each sub-router and the signal strength of the test device at the first position; inputting the signal strength of each terminal device and the signal strength of the test device into a context coding model containing an embedded layer to obtain a strength feature vector of each terminal device and a test feature vector of the test device; calculating a cross entropy value between the test feature vector and each of the intensity feature vectors; calculating the weighting of each intensity feature vector by taking the cross entropy value as the weight of each intensity feature vector, and cascading the weighted intensity feature vectors to obtain feature expression vectors of each sub-router; adjusting the feature expression vectors of the sub-routers to be uniform in length, and then performing two-dimensional splicing to obtain a feature expression matrix; passing the feature representation matrix and the topological feature matrix through a graph neural network to obtain a global feature matrix; performing eigenvalue decomposition on the global eigenvalue matrix to obtain n eigenvalues, wherein the n eigenvalues are used for representing essential characteristics of the global eigenvalue matrix in n mutually orthogonal dimensions of a high-dimensional space; obtaining n characteristic values of the test equipment from the second position to the Yth position; constructing n eigenvalues of the test equipment at Y positions into a classification matrix; and enabling the classification matrix to pass through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the network performance of the router system meets a preset requirement or not.
Fig. 1 is a diagram illustrating an application scenario of a network performance testing method for a whole-house router system according to an embodiment of the present application. As shown in fig. 1, in this application scenario, first, a topology matrix between a plurality of child routers (e.g., R as illustrated in fig. 1) connected to a parent router (e.g., P as illustrated in fig. 1) is acquired, wherein the eigenvalue of each position on the off-diagonal positions of the topology matrix is the distance between the corresponding two sub-routers, the characteristic value of each position on the diagonal position of the topology matrix is the distance between the child router and the parent router, and, the signal strength of each terminal device (for example, T as illustrated in fig. 1) connected to each of the sub-routers and the signal strength of the test device (for example, H as illustrated in fig. 1) at the first to nth positions are obtained, the terminal device includes, but is not limited to, a smart phone, a computer, a tablet, and the like, and the test device includes, but is not limited to, a signal tester, and the like. Then, the obtained topology matrix, the signal strength of each terminal device, and the signal strength of the test device are input into a server (e.g., S as illustrated in fig. 1) deployed with a network performance test algorithm for the whole-house router system, where the server can process the topology matrix, the signal strength of each terminal device, and the signal strength of the test device with the network performance test algorithm for the whole-house router system to generate a classification result indicating whether the network performance of the router system meets a preset requirement.
Having described the general principles of the present application, various non-limiting embodiments of the present application will now be described with reference to the accompanying drawings.
Exemplary method
Fig. 2 illustrates a flow chart of a network performance testing method for a whole house router system. As shown in fig. 2, a network performance testing method for a whole house router system according to an embodiment of the present application includes: s110, acquiring a topology matrix among a plurality of sub routers connected with a parent router, wherein the characteristic value of each position on the off-diagonal position of the topology matrix is the distance between two corresponding sub routers, and the characteristic value of each position on the diagonal position of the topology matrix is the distance between the sub router and the parent router; s120, processing the topological matrix by using a convolutional neural network as a feature extractor to obtain a topological feature matrix; s130, obtaining the signal intensity of each terminal device connected with each sub-router and the signal intensity of the test device at the first position; s140, inputting the signal strength of each terminal device and the signal strength of the testing device into a context coding model including an embedded layer to obtain a strength feature vector of each terminal device and a testing feature vector of the testing device; s150, calculating cross entropy values between the test feature vectors and the intensity feature vectors; s160, calculating the weight of each intensity characteristic vector by taking the cross entropy value as the weight of each intensity characteristic vector, and cascading the weighted intensity characteristic vectors to obtain the characteristic representation vector of each sub-router; s170, adjusting the feature expression vectors of the sub-routers to be uniform in length, and then performing two-dimensional splicing to obtain a feature expression matrix; s180, passing the feature representation matrix and the topological feature matrix through a neural network to obtain a global feature matrix; s190, performing eigenvalue decomposition on the global eigenvalue matrix to obtain n eigenvalues, wherein the n eigenvalues are used for representing essential characteristics of the global eigenvalue matrix on n mutually orthogonal dimensions of a high-dimensional space; s200, acquiring n characteristic values of the test equipment from the second position to the Y position; s210, constructing n characteristic values of the test equipment at Y positions into a classification matrix; and S220, enabling the classification matrix to pass through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the network performance of the router system meets preset requirements or not.
Fig. 3 is a schematic diagram illustrating an architecture of a network performance testing method for a whole house router system according to an embodiment of the present application. As shown in fig. 3, in the network architecture of the network performance test method for the whole house router system, first, the obtained topology matrix (e.g., M1 as illustrated in fig. 3) is processed using a convolutional neural network (e.g., CNN as illustrated in fig. 3) as a feature extractor to obtain a topology feature matrix (e.g., MF1 as illustrated in fig. 3); then, inputting the obtained signal strength of each of the terminal devices (e.g., P1 as illustrated in fig. 3) and the signal strength of the test device (e.g., P2 as illustrated in fig. 3) into a context coding model (e.g., E as illustrated in fig. 3) including an embedding layer to obtain a strength feature vector (e.g., VF1 as illustrated in fig. 3) of each of the terminal devices and a test feature vector of the test device (e.g., VF2 as illustrated in fig. 3); then, cross-entropy values (e.g., CEV as illustrated in fig. 3) between the test feature vectors and the respective intensity feature vectors are calculated; then, calculating the weight of each intensity feature vector by taking the cross entropy value as the weight of each intensity feature vector and cascading each weighted intensity feature vector (for example, VF3 as illustrated in fig. 3) to obtain a feature representation vector (for example, VF4 as illustrated in fig. 3) of each sub-router; then, adjusting the feature representation vectors of the sub-routers to be of uniform length, and performing two-dimensional splicing to obtain a feature representation matrix (for example, MF2 as illustrated in fig. 3); then, passing the feature representation matrix and the topological feature matrix through a graph neural network (e.g., GNN as illustrated in fig. 3) to obtain a global feature matrix (e.g., MF as illustrated in fig. 3); then, performing eigenvalue decomposition on the global eigenvalue matrix to obtain n eigenvalues (e.g., CV1 as illustrated in fig. 3); next, n characteristic values (e.g., CVs 2-CVn as illustrated in fig. 3) of the test device at the second position through the Y-th position are obtained; then, constructing n eigenvalues of the test device at Y positions as a classification matrix (e.g., M as illustrated in fig. 3); and, finally, passing the classification matrix through a classifier (e.g., a classifier as illustrated in fig. 3) to obtain a classification result, which is used to indicate whether the network performance of the router system meets the preset requirement.
In steps S110 and S120, a topology matrix between a plurality of sub-routers connected to a parent router is obtained, where a feature value of each position on an off-diagonal position of the topology matrix is a distance between two corresponding sub-routers, and a feature value of each position on a diagonal position of the topology matrix is a distance between a sub-router and the parent router, and the topology matrix is processed using a convolutional neural network as a feature extractor to obtain the topology feature matrix. As described above, in consideration of the fact that the existing test method only considers the received signal strength of the terminal device at a specific location when testing the performance of the router system, in the technical solution of the present application, it is desirable to perform the network performance test by integrating the various elements based on the influence of other routers on the communication of the target router connected to the terminal device, wherein the number of terminal devices connected to the other routers needs to be considered when evaluating the influence of the other routers on the target router, and the interference of the other terminal devices connected to the target router on the signal strength of the terminal device to be evaluated needs to be considered.
Accordingly, in the technical solution of the present application, a topology matrix between a plurality of child routers connected to a parent router needs to be obtained first, where a feature value of each position on an off-diagonal position of the topology matrix is a distance between two corresponding child routers, and a feature value of each position on a diagonal position of the topology matrix is a distance between a child router and a parent router. Then, the topological matrix is processed in a convolutional neural network serving as a feature extractor to extract high-dimensional associated features of each position in the topological matrix, so that a topological feature matrix is obtained.
Specifically, in the embodiment of the present application, the process of inputting the topology matrix into a convolutional neural network to obtain a topology feature matrix includes: firstly, carrying out convolution processing, pooling processing and activation processing on input data in the forward transmission process of layers by each layer except the last layer of the convolution neural network to obtain a topological characteristic diagram; then, the last layer of the convolutional neural network performs global mean pooling along the channel dimension on the topological feature map to obtain the topological feature matrix. It should be appreciated that by performing global mean pooling on the topological feature map, the number of parameters may be reduced to increase the speed of training, thereby normalizing the entire network structure to prevent overfitting.
In steps S130 and S140, the signal strength of each terminal device connected to each sub-router and the signal strength of the testing device at the first location are obtained, and the signal strength of each terminal device and the signal strength of the testing device are input into a context coding model including an embedded layer to obtain a strength feature vector of each terminal device and a testing feature vector of the testing device. That is, the signal strength of each terminal device connected to each sub-router, including but not limited to a smart phone, a computer, a tablet, etc., and the signal strength of the testing device at the first location, including but not limited to a signal tester, etc., are acquired. Then, the signal strength of each terminal device and the signal strength of the test device are further input into a context coding model containing an embedded layer for coding processing, so as to obtain a strength feature vector of each terminal device and a test feature vector of the test device. It should be appreciated that since the context-based encoder model is capable of encoding the input vector based on context, the resulting feature vector has global signal strength correlation information.
Specifically, in this embodiment of the present application, the process of inputting the signal strength of each terminal device and the signal strength of the test device into a context coding model including an embedded layer to obtain a strength feature vector of each terminal device and a test feature vector of the test device includes: firstly, converting the signal intensity of each terminal device and the signal intensity of the test device into input vectors by using an embedded layer of the encoder model so as to obtain a sequence of the input vectors; then, the sequence of input vectors is passed through a converter of the encoder model to obtain an intensity feature vector of each of the terminal devices and a test feature vector of the test device.
In steps S150 and S160, cross entropy values between the test feature vectors and the intensity feature vectors are calculated, weights of the intensity feature vectors are calculated by using the cross entropy values as weights of the intensity feature vectors, and the weighted intensity feature vectors are concatenated to obtain feature representation vectors of the sub-routers. That is, in the technical solution of the present application, after obtaining the test feature vector and each of the intensity feature vectors, further calculating a cross entropy value between the test feature vector and each of the intensity feature vectors, so as to obtain a degree of consistency between each of the intensity feature vectors and the test feature vector. Then, the cross entropy value is used as the weight of each intensity feature vector to weight each intensity feature vector, so as to obtain each weighted intensity feature vector having communication interference influence. Then, cascading the weighted intensity characteristic vectors to obtain the characteristic representation vector of each sub-router. In this way, the weighted intensity feature information can be integrated to obtain the feature expression vector of each sub-router, thereby facilitating the subsequent calculation processing.
In step S170 and step S180, the feature representation vectors of the sub-routers are adjusted to have a uniform length, two-dimensional splicing is performed to obtain a feature representation matrix, and then the feature representation matrix and the topological feature matrix are passed through a graph neural network to obtain a global feature matrix. That is, the feature expression vectors of the sub-routers are further adjusted to be of uniform length and then two-dimensional splicing is performed to obtain a feature expression matrix. Accordingly, in a specific example, the feature representation vectors of the sub-routers may be adjusted to a uniform length by linear transformation and then two-dimensionally spliced to obtain a feature representation matrix. Then, the feature representation matrix and the topological feature matrix pass through a graph neural network to obtain a global feature matrix. It should be understood that the graph neural network can be used for processing graph data in irregular non-euclidean air, so as to extract the associated information of the data samples due to the feature information and the irregular topological structure information, and therefore, the obtained global feature matrix can improve the accuracy of classification compared with a feature representation matrix obtained by directly splicing.
In steps S190 and S200, performing eigenvalue decomposition on the global eigenvalue matrix to obtain n eigenvalues, where the n eigenvalues are used to represent essential characteristics of the global eigenvalue matrix in n mutually orthogonal dimensions of a high-dimensional space, and obtaining n eigenvalues of the test equipment at the second position to the Y position. It should be understood that, although the global feature matrix includes rich information including the topology information of the correlation between the sub-routers and the signal feature information of the terminal devices of each sub-router based on the correlation of the testing device, the correlation between these information has strong correlation and weak correlation, and the information expression degree is different when the feature expression vector is subjected to linear transformation, so that there is a lot of noise correlation information between the feature values at the positions of the global feature matrix, and if the classifier including the full connection layer is used for classification directly, the classification accuracy is adversely affected. Therefore, in the technical solution of the present application, for the n × n diagonal matrix characteristic of the topological feature matrix forming the global feature matrix, the global feature matrix is subjected to eigenvalue decomposition to obtain n eigenvalues, so that the n eigenvalues represent essential features of the global feature matrix in n mutually orthogonal dimensions of a high-dimensional space, which may also be understood as performing orthogonal dimension decomposition on the signal features, thereby ignoring weak correlation information thereof and only retaining strong correlation information. Then, after further performing a test using a test apparatus at each of a plurality of positions, the n feature values corresponding to said each position, that is, the n feature values of said test apparatus at the second position to the Y-th position, are obtained.
Specifically, in this embodiment of the present application, a process of performing eigenvalue decomposition on the global eigenvalue matrix to obtain n eigenvalues includes: performing eigenvalue decomposition on the global eigenvalue matrix to obtain the n eigenvalues; wherein the formula is: c is A Λ ATWherein Λ ═ diag (λ)1,λ2,…,λN) And A ═ A1,A2,…,AN]Is a matrix of eigenvectors containing the corresponding eigenvectors as columns, where λ1,λ2,…,λNThe n characteristic values are obtained.
In steps S210 and S220, the n eigenvalues of the test device at the Y positions are constructed as a classification matrix, and the classification matrix is passed through a classifier to obtain a classification result, where the classification result is used to indicate whether the network performance of the router system meets a preset requirement. That is, after n eigenvalues corresponding to each position are obtained, the n eigenvalues corresponding to each position are further grouped into a classification eigenvector, the classification eigenvector of each position is two-dimensionally spliced to obtain a classification matrix, and then the classification matrix is input to a classifier, so that a classification result indicating whether the network performance of the router system meets preset requirements can be obtained.
Specifically, in the embodiment of the present application, the process of passing the classification matrix through a classifier to obtain a classification result includes: first, the classification matrix is full-join coded using a plurality of full-join layers of the classifier to obtain a classification feature vector. Then, the classification feature vector is input into a Softmax classification function of the classifier to obtain a first probability that network performance of the classification feature vector belonging to the router system meets a preset requirement and a second probability that the network performance of the router system does not meet the preset requirement. Finally, the classification result is generated based on a comparison between the first probability and the second probability. Specifically, when the first probability is greater than the second probability, the classification result is that the network performance of the router system meets a preset requirement; when the first probability is smaller than the second probability, the classification result is that the network performance of the router system does not meet preset requirements.
In particular, in a specific example, the process of passing the classification matrix through a classifier to obtain a classification result further includes: the classifier processes the classification matrix by the following formula to obtain the classification result; wherein the formula is:
softmax{(Wn,Bn):…:(W1,B1)|Project(F)}
wherein project (F) represents projecting the classification matrix as a vector; w1To WnIs a weight matrix, B1To BnIs a bias vector.
In summary, the network performance testing method for the whole house router system according to the embodiment of the present application is illustrated, which extracts topological features between a plurality of sub-routers based on a convolutional neural network model, and obtains global strength correlation information of each sub-router by using a context coding model including an embedded layer, further uses a graph neural network to extract correlation information existing in a data sample due to feature information and irregular topological structure information, and further performs orthogonal dimension decomposition on the signal features, so as to ignore weak correlation information thereof and only retain strong correlation information, thereby enabling the obtained classification result to be more accurate. In this way, the performance of the router system can be better checked to ensure that the basic communication needs are met at all locations indoors.
Exemplary System
Fig. 4 illustrates a block diagram of a network performance testing system for a whole house router system according to an embodiment of the application. As shown in fig. 4, a network performance testing system 400 for a whole house router system according to an embodiment of the present application includes: a topology matrix obtaining unit 410, configured to obtain a topology matrix among a plurality of child routers connected to a parent router, where a feature value of each position on an off-diagonal position of the topology matrix is a distance between two corresponding child routers, and a feature value of each position on a diagonal position of the topology matrix is a distance between a child router and the parent router; a feature extraction unit 420, configured to process the topology matrix obtained by the topology matrix obtaining unit 410 using a convolutional neural network as a feature extractor to obtain a topology feature matrix; a signal strength obtaining unit 430, configured to obtain a signal strength of each terminal device connected to each sub-router and a signal strength of a testing device at a first location; an encoding unit 440, configured to input the signal strength of the terminal device obtained by each of the signal strength obtaining units 430 and the signal strength of the test device obtained by the signal strength obtaining unit 430 into a context coding model including an embedded layer to obtain a strength feature vector of each of the terminal devices and a test feature vector of the test device; a cross entropy value calculating unit 450, configured to calculate a cross entropy value between the test feature vector obtained by the encoding unit 440 and the intensity feature vector obtained by each of the encoding units 440; a feature representation vector generating unit 460, configured to calculate a weighting of each of the intensity feature vectors by using the cross entropy values obtained by the cross entropy value calculating unit 450 as weights of the intensity feature vectors obtained by each of the encoding units 440, and concatenate the weighted intensity feature vectors to obtain feature representation vectors of each of the sub-routers; a two-dimensional splicing unit 470, configured to perform two-dimensional splicing after adjusting the feature representation vectors of the sub routers obtained by each feature representation vector generation unit 460 to a uniform length to obtain a feature representation matrix; a graph neural network processing unit 480, configured to pass the feature representation matrix obtained by the two-dimensional stitching unit 470 and the topological feature matrix obtained by the feature extraction unit 420 through a graph neural network to obtain a global feature matrix; an eigenvalue decomposition unit 490, configured to perform eigenvalue decomposition on the global eigenvalue matrix obtained by the graph neural network processing unit 480 to obtain n eigenvalues, where the n eigenvalues are used to represent essential characteristics of the global eigenvalue matrix in n mutually orthogonal dimensions of a high-dimensional space; a characteristic value obtaining unit 500, configured to obtain n characteristic values of the test device at a second position to a Y-th position; a classification matrix constructing unit 510, configured to construct n eigenvalues of the test equipment at Y positions, which are obtained by the eigenvalue obtaining unit 500, into a classification matrix; a classifying unit 520, configured to pass the classification matrix obtained by the classification matrix constructing unit 510 through a classifier to obtain a classification result, where the classification result is used to indicate whether the network performance of the router system meets a preset requirement.
In an example, in the network performance testing system 400 for a whole house router system, the feature extraction unit 420 is further configured to: all layers except the last layer of the convolutional neural network carry out convolution processing, pooling processing and activation processing on input data in the forward transmission process of the layers so as to obtain a topological characteristic diagram; and the last layer of the convolutional neural network performs global mean pooling along the channel dimension on the topological feature map to obtain the topological feature matrix.
In an example, in the network performance testing system 400 for a whole house router system, the encoding unit 440 is further configured to: converting the signal strength of each terminal device and the signal strength of the test device into input vectors by using an embedded layer of the encoder model to obtain a sequence of input vectors; and passing the sequence of input vectors through a converter of the encoder model to obtain an intensity feature vector for each of the terminal devices and a test feature vector for the test device.
In one example, in the network performance testing system 400 for a whole house router system described above, the two-dimensional stitching unit 470 is further configured to: and adjusting the feature expression vectors of the sub-routers into a uniform length through linear transformation, and then performing two-dimensional splicing to obtain a feature expression matrix.
In an example, in the network performance testing system 400 for a whole house router system, the eigenvalue decomposition unit 490 is further configured to: performing eigenvalue decomposition on the global eigenvalue matrix to obtain the n eigenvalues; wherein the formula is: a Λ a ═ CTWherein Λ ═ diag (λ)1,λ2,…,λN) And A ═ A1,A2,…,AN]Is a matrix of eigenvectors containing the corresponding eigenvectors as columns, where1,λ2,…,λNThe n characteristic values are obtained.
In an example, in the network performance testing system 400 for a whole house router system, the classifying unit 520 is further configured to: fully concatenating the classification matrix using a plurality of fully concatenated layers of the classifier to obtain a classification feature vector; inputting the classification feature vector into a Softmax classification function of the classifier to obtain a first probability that network performance of the classification feature vector belonging to a router system meets a preset requirement and a second probability that the network performance of the router system does not meet the preset requirement; and generating the classification result based on a comparison between the first probability and the second probability.
In an example, in the network performance testing system 400 for a whole house router system, the classifying unit 520 is further configured to: the classifier processes the classification matrix according to the following formula to obtain the classification result; wherein the formula is:
softmax{(Wn,Bn):…:(W1,B1)|Project(F)}
wherein project (F) represents projecting the classification matrix as a vector; w1To WnIs a weight matrix, B1To BnIs a bias vector.
Here, it will be understood by those skilled in the art that the specific functions and operations of the respective units and modules in the network performance testing system 400 for the whole house router system described above have been described in detail in the description of the network performance testing method for the whole house router system with reference to fig. 1 to 3, and thus, a repetitive description thereof will be omitted.
As described above, the network performance test system 400 for the whole house router system according to the embodiment of the present application may be implemented in various terminal devices, such as a server of a network performance test algorithm for the whole house router system. In one example, the network performance testing system 400 for a whole house router system according to embodiments of the present application may be integrated into a terminal device as a software module and/or a hardware module. For example, the network performance testing system 400 for a whole house router system may be a software module in the operating system of the terminal device, or may be an application developed for the terminal device; of course, the network performance testing system 400 for a whole house router system may also be one of many hardware modules of the terminal device.
Alternatively, in another example, the network performance testing system 400 for the whole house router system and the terminal device may be separate devices, and the network performance testing system 400 for the whole house router system may be connected to the terminal device through a wired and/or wireless network and transmit the interaction information according to an agreed data format.
The basic principles of the present application have been described above with reference to specific embodiments, but it should be noted that advantages, effects, etc. mentioned in the present application are only examples and are not limiting, and the advantages, effects, etc. must not be considered to be possessed by various embodiments of the present application. Furthermore, the foregoing disclosure of specific details is for the purpose of illustration and description and is not intended to be limiting, since the foregoing disclosure is not intended to be exhaustive or to limit the disclosure to the precise details disclosed.
The block diagrams of devices, apparatuses, systems referred to in this application are only given as illustrative examples and are not intended to require or imply that the connections, arrangements, configurations, etc. must be made in the manner shown in the block diagrams. These devices, apparatuses, devices, systems may be connected, arranged, configured in any manner, as will be appreciated by those skilled in the art. Words such as "including," "comprising," "having," and the like are open-ended words that mean "including, but not limited to," and are used interchangeably therewith. The words "or" and "as used herein mean, and are used interchangeably with, the word" and/or, "unless the context clearly dictates otherwise. The word "such as" is used herein to mean, and is used interchangeably with, the phrase "such as but not limited to".
It should also be noted that in the devices, apparatuses, and methods of the present application, the components or steps may be decomposed and/or recombined. These decompositions and/or recombinations are to be considered as equivalents of the present application.
The previous description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other aspects without departing from the scope of the application. Thus, the present application is not intended to be limited to the aspects shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
The foregoing description has been presented for purposes of illustration and description. Furthermore, the description is not intended to limit embodiments of the application to the form disclosed herein. While a number of example aspects and embodiments have been discussed above, those of skill in the art will recognize certain variations, modifications, alterations, additions and sub-combinations thereof.
Claims (10)
1. A network performance testing method for a whole-house router system is characterized by comprising the following steps:
acquiring a topology matrix among a plurality of sub routers connected with a parent router, wherein the characteristic value of each position on the off-diagonal position of the topology matrix is the distance between two corresponding sub routers, and the characteristic value of each position on the diagonal position of the topology matrix is the distance between the sub router and the parent router;
processing the topological matrix by using a convolutional neural network as a feature extractor to obtain a topological feature matrix;
obtaining the signal strength of each terminal device connected with each sub-router and the signal strength of the test device at the first position;
inputting the signal strength of each terminal device and the signal strength of the test device into a context coding model containing an embedded layer to obtain a strength feature vector of each terminal device and a test feature vector of the test device;
calculating cross entropy values between the test feature vectors and the intensity feature vectors;
calculating the weight of each intensity feature vector by taking the cross entropy value as the weight of each intensity feature vector, and cascading the weighted intensity feature vectors to obtain feature expression vectors of each sub-router;
adjusting the feature expression vectors of the sub-routers to be uniform in length, and then performing two-dimensional splicing to obtain a feature expression matrix;
passing the feature representation matrix and the topological feature matrix through a graph neural network to obtain a global feature matrix;
performing eigenvalue decomposition on the global eigenvalue matrix to obtain n eigenvalues, wherein the n eigenvalues are used for representing essential characteristics of the global eigenvalue matrix in n mutually orthogonal dimensions of a high-dimensional space;
obtaining n characteristic values of the test equipment from the second position to the Yth position;
constructing n eigenvalues of the test equipment at Y positions into a classification matrix; and
and passing the classification matrix through a classifier to obtain a classification result, wherein the classification result is used for indicating whether the network performance of the router system meets a preset requirement or not.
2. The network performance testing method for a whole house router system as claimed in claim 1, wherein inputting the topology matrix into a convolutional neural network to obtain a topology feature matrix comprises:
all layers except the last layer of the convolutional neural network carry out convolution processing, pooling processing and activation processing on input data in the forward transmission process of the layers so as to obtain a topological characteristic diagram; and
and the last layer of the convolutional neural network performs global mean pooling along channel dimensions on the topological feature map to obtain the topological feature matrix.
3. The network performance testing method for a whole house router system of claim 2, wherein inputting the signal strength of each of the terminal devices and the signal strength of the testing device into a context coding model containing an embedding layer to obtain the strength feature vector of each of the terminal devices and the testing feature vector of the testing device comprises:
converting the signal strength of each terminal device and the signal strength of the test device into input vectors by using an embedded layer of the encoder model to obtain a sequence of input vectors; and
and passing the sequence of input vectors through a converter of the encoder model to obtain an intensity feature vector of each terminal device and a test feature vector of the test device.
4. The network performance testing method for the whole house router system as claimed in claim 3, wherein the adjusting of the eigen expression vectors of each sub-router to a uniform length and then the two-dimensional splicing to obtain the eigen expression matrix comprises:
and adjusting the feature expression vectors of the sub-routers into a uniform length through linear transformation, and then performing two-dimensional splicing to obtain a feature expression matrix.
5. The network performance testing method for a whole house router system of claim 4, wherein performing eigenvalue decomposition on the global eigenvalue matrix to obtain n eigenvalues comprises:
performing eigenvalue decomposition on the global eigenvalue matrix to obtain the n eigenvalues;
wherein the formula is:
C＝AΛATwherein Λ ═ diag (λ)1,λ2,…,λN) And A ═ A1,A2,…,AN]Is a matrix of eigenvectors containing the corresponding eigenvectors as columns, where1,λ2,…,λNThe n characteristic values are obtained.
6. The network performance testing method for the whole house router system as claimed in claim 5, wherein passing the classification matrix through a classifier to obtain a classification result comprises:
fully concatenating the classification matrix using a plurality of fully concatenated layers of the classifier to obtain a classification feature vector;
inputting the classification feature vector into a Softmax classification function of the classifier to obtain a first probability that network performance of the classification feature vector belonging to a router system meets a preset requirement and a second probability that the network performance of the router system does not meet the preset requirement; and
generating the classification result based on a comparison between the first probability and the second probability.
7. The network performance testing method for the whole house router system as claimed in claim 5, wherein passing the classification matrix through a classifier to obtain a classification result comprises:
the classifier processes the classification matrix according to the following formula to obtain the classification result;
wherein the formula is:
softmax{(Wn,Bn):…:(W1,B1)|Project(F)}
wherein project (F) represents projecting the classification matrix as a vector; w is a group of1To WnIs a weight matrix, B1To BnIs a bias vector.
8. A network performance testing system for a full house router system, comprising:
a topology matrix obtaining unit, configured to obtain a topology matrix among a plurality of child routers connected to a parent router, where a feature value of each position on an off-diagonal position of the topology matrix is a distance between two corresponding child routers, and a feature value of each position on a diagonal position of the topology matrix is a distance between a child router and the parent router;
a feature extraction unit, configured to process the topology matrix obtained by the topology matrix obtaining unit using a convolutional neural network as a feature extractor to obtain a topology feature matrix;
a signal strength obtaining unit, configured to obtain a signal strength of each terminal device connected to each sub-router and a signal strength of the test device at the first location;
the encoding unit is used for inputting the signal strength of the terminal equipment obtained by each signal strength obtaining unit and the signal strength of the test equipment obtained by the signal strength obtaining unit into a context encoding model containing an embedded layer so as to obtain a strength characteristic vector of each terminal equipment and a test characteristic vector of the test equipment;
a cross entropy value calculation unit, configured to calculate a cross entropy value between the test feature vector obtained by the encoding unit and the intensity feature vector obtained by each encoding unit;
a feature expression vector generation unit, configured to calculate, by using the cross entropy value obtained by the cross entropy value calculation unit as a weight of the intensity feature vector obtained by each encoding unit, a weight of each intensity feature vector, and concatenate the weighted intensity feature vectors to obtain a feature expression vector of each sub-router;
the two-dimensional splicing unit is used for adjusting the feature expression vectors of the sub routers, which are obtained by the feature expression vector generating units, into a uniform length and then performing two-dimensional splicing to obtain a feature expression matrix;
the graph neural network processing unit is used for enabling the feature representation matrix obtained by the two-dimensional splicing unit and the topological feature matrix obtained by the feature extraction unit to pass through a graph neural network so as to obtain a global feature matrix;
an eigenvalue decomposition unit, configured to perform eigenvalue decomposition on the global eigenvalue matrix obtained by the graph neural network processing unit to obtain n eigenvalues, where the n eigenvalues are used to represent essential characteristics of the global eigenvalue matrix in n mutually orthogonal dimensions of a high-dimensional space;
the characteristic value acquisition unit is used for acquiring n characteristic values of the test equipment from a second position to a Y-th position;
a classification matrix constructing unit, configured to construct n eigenvalues of the test equipment at Y positions, which are obtained by the eigenvalue obtaining unit, as a classification matrix;
and the classification unit is used for enabling the classification matrix obtained by the classification matrix constructing unit to pass through a classifier so as to obtain a classification result, and the classification result is used for indicating whether the network performance of the router system meets the preset requirement or not.
9. The network performance testing system for a whole house router system of claim 8, wherein the feature extraction unit is further configured to:
all layers except the last layer of the convolutional neural network carry out convolution processing, pooling processing and activation processing on input data in the forward transmission process of the layers so as to obtain a topological characteristic diagram; and the last layer of the convolutional neural network performs global mean pooling along channel dimensions on the topological feature map to obtain the topological feature matrix.
10. The network performance testing system for a whole house router system of claim 8, wherein the encoding unit is further configured to:
converting the signal strength of each terminal device and the signal strength of the test device into input vectors by using an embedded layer of the encoder model to obtain a sequence of input vectors; and passing the sequence of input vectors through a converter of the encoder model to obtain an intensity feature vector of each of the terminal devices and a test feature vector of the test device.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210235630.1A CN114666254A (en) | 2022-03-11 | 2022-03-11 | Network performance testing method and system for whole-house router system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210235630.1A CN114666254A (en) | 2022-03-11 | 2022-03-11 | Network performance testing method and system for whole-house router system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114666254A true CN114666254A (en) | 2022-06-24 |
Family
ID=82029744
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210235630.1A Withdrawn CN114666254A (en) | 2022-03-11 | 2022-03-11 | Network performance testing method and system for whole-house router system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114666254A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115617636A (en) * | 2022-12-17 | 2023-01-17 | 华测国软技术服务南京有限公司 | Distributed performance test system |
| CN115955410A (en) * | 2023-03-13 | 2023-04-11 | 国网智联电商有限公司 | Data visualization modeling and matching method based on cloud computing environment |
-
2022
- 2022-03-11 CN CN202210235630.1A patent/CN114666254A/en not_active Withdrawn
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115617636A (en) * | 2022-12-17 | 2023-01-17 | 华测国软技术服务南京有限公司 | Distributed performance test system |
| CN115955410A (en) * | 2023-03-13 | 2023-04-11 | 国网智联电商有限公司 | Data visualization modeling and matching method based on cloud computing environment |
| CN115955410B (en) * | 2023-03-13 | 2023-08-18 | 国网智联电商有限公司 | Data visualization modeling and matching method based on cloud computing environment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107743103B (en) | Multi-node access detection and channel estimation method of MMTC (multimedia messaging and control) system based on deep learning | |
| CN114666254A (en) | Network performance testing method and system for whole-house router system | |
| CN110210330B (en) | Electromagnetic signal identification method and device for constructing graph convolution network based on implicit knowledge | |
| CN113313156A (en) | Internet of things equipment identification method and system based on time sequence load flow fingerprints | |
| CN110636445B (en) | WIFI-based indoor positioning method, device, equipment and medium | |
| CN115512166B (en) | Intelligent preparation method and system of lens | |
| CN115982736B (en) | Data encryption method and system for computer network information | |
| CN113095370A (en) | Image recognition method and device, electronic equipment and storage medium | |
| CN113111543A (en) | Internet of things service system | |
| CN115577248A (en) | Fault diagnosis system and method for wind generating set | |
| CN115235612A (en) | Intelligent fault diagnosis system and method for servo motor | |
| Geng et al. | Multipoint channel charting with multiple-input multiple-output convolutional autoencoder | |
| CN114205846A (en) | Router test management method and system with adjustable antenna and electronic equipment | |
| CN114821169A (en) | Method-level non-intrusive call link tracking method under micro-service architecture | |
| CN115219845A (en) | Power grid fault diagnosis system and method | |
| CN114866427A (en) | Whole house route networking method and system thereof | |
| CN114757975A (en) | Pedestrian trajectory prediction method based on transformer and graph convolution network | |
| CN112464712B (en) | Rotating machine fault diagnosis method based on blind extraction algorithm | |
| CN116320459B (en) | Computer network communication data processing method and system based on artificial intelligence | |
| CN112801268A (en) | Positioning method based on graph convolution and multilayer perceptron hybrid network | |
| CN115775055B (en) | Method, device, equipment and medium for predicting personnel evacuation time of multi-storey building | |
| CN115617636B (en) | Distributed performance test system | |
| Kukolj et al. | Indoor location fingerprinting based on data reduction | |
| CN117034123A (en) | Fault monitoring system and method for fitness equipment | |
| Sonny et al. | Fingerprint Image-Based Multi-Building 3D Indoor Wi-Fi Localization Using Convolutional Neural Networks |
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 | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20220624 |
|
| WW01 | Invention patent application withdrawn after publication |