CN114615147A - Bandwidth control method and system for router and electronic equipment - Google Patents
Bandwidth control method and system for router and electronic equipment Download PDFInfo
- Publication number
- CN114615147A CN114615147A CN202210244423.2A CN202210244423A CN114615147A CN 114615147 A CN114615147 A CN 114615147A CN 202210244423 A CN202210244423 A CN 202210244423A CN 114615147 A CN114615147 A CN 114615147A
- Authority
- CN
- China
- Prior art keywords
- bandwidth
- matrix
- feature map
- feature
- terminal device
- 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- Withdrawn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
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/08—Configuration management of networks or network elements
- H04L41/0896—Bandwidth or capacity management, i.e. automatically increasing or decreasing capacities
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The application relates to the field of router bandwidth allocation, and particularly discloses a bandwidth control method, a system and an electronic device for a router. The method adopts a deep learning technology based on a convolutional neural network to fully excavate correlation characteristics between bandwidth information and signal strength information of each terminal device, uses uplink parameters to fuse the bandwidth information and the signal strength information, and then corrects the obtained characteristic graph based on the principle of classical multi-dimensional scale transformation so as to enable the percentage of the bandwidth distributed by each terminal device to be more accurate. In this way, the bandwidth resources allocated to each terminal device can be allocated reasonably.
Description
Technical Field
The present invention relates to the field of router bandwidth allocation, and more particularly, to a bandwidth control method, system and electronic device for a router.
Background
A router may also be referred to as a gateway device. In network communication, the router has the functions of judging network addresses and selecting IP paths, a flexible link system can be constructed in a plurality of network environments, and each subnet is linked through different data packets and medium access modes.
Currently, in an existing router, bandwidth resources of each terminal device in the router are generally controlled based on a time-sharing control mode, that is, first, signal strength of each terminal device is obtained, bandwidth resources corresponding to each terminal device are determined based on the signal strength, and then, the calculated corresponding bandwidth resources are allocated to each terminal device.
However, this way of allocating bandwidth does not take into account the actual situation of each terminal device and the overall situation of each terminal device. This may result in insufficient bandwidth resources being allocated by each terminal device. Therefore, to achieve reasonable bandwidth allocation, a bandwidth control scheme for routers is desired.
Disclosure of Invention
The present application is proposed to solve the above-mentioned technical problems. The embodiment of the application provides a bandwidth control method, a system and electronic equipment for a router, wherein deep learning technology based on a convolutional neural network is adopted to fully excavate correlation characteristics between bandwidth information and signal strength information of each terminal device, uplink parameters are used for fusing the bandwidth information and the signal strength information, and then the obtained characteristic diagram is corrected based on the principle of classical multi-dimensional scale transformation, so that the obtained percentage of the bandwidth allocated to each terminal device is more accurate. In this way, the bandwidth resources allocated to each terminal device can be allocated reasonably.
According to an aspect of the present application, there is provided a bandwidth control method for a router, including:
acquiring bandwidth data of each terminal device communicated with a router at a series of preset time points, and constructing the bandwidth data of the series of preset time points into a bandwidth matrix according to device sample dimensions and time dimensions;
acquiring signal intensity data of each terminal device at a same series of preset time points, and constructing the signal intensity data of the series of preset time points into a signal matrix according to the device sample dimension and the time dimension;
obtaining a first feature map and a second feature map from the bandwidth matrix and the signal matrix using a first convolutional neural network and a second convolutional neural network as feature extractors;
fusing the first characteristic diagram and the second characteristic diagram to obtain a third characteristic diagram based on uplink communication parameters of each terminal device to the router, wherein the uplink communication parameters are determined based on small-scale attenuation power components of the terminal devices, small-scale attenuation effects of the terminal devices, additive white Gaussian noise power and interference in the selected channel;
scaling the third feature map using classical multi-dimensional scaling to obtain a modified fused feature map based on a distance between corresponding locations in the first and second feature maps;
acquiring the current bandwidth allocation data of each terminal device and arranging the bandwidth allocation data of each terminal into bandwidth allocation vectors;
encoding the bandwidth allocation vector using an encoder consisting of a plurality of fully-connected layers to obtain a bandwidth allocation eigenvector;
taking the bandwidth allocation characteristic vector as a query vector to perform matrix multiplication with the corrected fusion characteristic diagram so as to map the bandwidth allocation characteristic vector into a characteristic space of the corrected fusion characteristic diagram to obtain an allocation characteristic vector; and
and calculating a Softmax function value of each position of the allocation feature vector as a bandwidth allocation percentage of the terminal device corresponding to each position, the Softmax function value being a weighted sum of a natural exponent function value raised to the power of the feature value of each position divided by the natural exponent function value raised to the power of the feature value of each position.
According to another aspect of the present application, there is provided a bandwidth control system for a router, including:
the bandwidth matrix construction unit is used for acquiring bandwidth data of each terminal device communicated with the router at a series of preset time points and constructing the bandwidth data of the series of preset time points into a bandwidth matrix according to the device sample dimension and the time dimension;
the signal matrix constructing unit is used for acquiring signal intensity data of each terminal device at a same series of preset time points and constructing the signal intensity data of the series of preset time points into a signal matrix according to the device sample dimension and the time dimension;
a feature extraction unit configured to obtain a first feature map and a second feature map from the bandwidth matrix obtained by the bandwidth matrix construction unit and the signal matrix obtained by the signal matrix construction unit using a first convolutional neural network and a second convolutional neural network as a feature extractor;
a fusion unit configured to fuse the first feature map obtained by the feature extraction unit and the second feature map obtained by the feature extraction unit to obtain a third feature map based on uplink communication parameters from each of the terminal devices to the router, the uplink communication parameters being determined based on a small-scale attenuation power component of the terminal device, a small-scale attenuation effect of the terminal device, an additive white gaussian noise power, and interference in the selected channel;
a scaling unit configured to scale the third feature map obtained by the fusion unit using classical multidimensional scaling based on a distance between the first feature map obtained by the feature extraction unit and a corresponding position in the second feature map obtained by the feature extraction unit to obtain a modified fusion feature map;
a bandwidth allocation vector generating unit, configured to obtain current bandwidth allocation data of each terminal device and arrange the bandwidth allocation data of each terminal as a bandwidth allocation vector;
an encoding unit configured to encode the bandwidth allocation vector obtained by the bandwidth allocation vector generation unit using an encoder configured by a plurality of fully-connected layers to obtain a bandwidth allocation feature vector;
a mapping unit, configured to perform matrix multiplication on the bandwidth allocation feature vector obtained by the encoding unit as a query vector and the modified fusion feature map obtained by the scale transformation unit to map the bandwidth allocation feature vector into a feature space of the modified fusion feature map to obtain an allocation feature vector; and
a result calculation unit configured to calculate, as a bandwidth allocation percentage of the terminal device corresponding to each location, a Softmax function value of each location of the allocation feature vector obtained by the mapping unit, the Softmax function value being a weighted sum of a natural exponent function value raised to a power of the feature value of each location divided by the natural exponent function value raised to a power of the feature value of each location.
According to yet another aspect of the present application, there is provided an electronic device including: a processor; and a memory in which are stored computer program instructions which, when executed by the processor, cause the processor to perform a bandwidth control method for a router as described above.
According to yet another aspect of the present application, there is provided a computer readable medium having stored thereon computer program instructions which, when executed by a processor, cause the processor to perform the bandwidth control method for a router as described above.
Compared with the prior art, the bandwidth control method, the system and the electronic device for the router provided by the application adopt a deep learning technology based on a convolutional neural network to fully dig out the correlation characteristics between the bandwidth information and the signal strength information of each terminal device, and use the uplink parameters to fuse the bandwidth information and the signal strength information, and then modify the obtained characteristic diagram based on the principle of classical multidimensional scaling so as to enable the percentage of the bandwidth allocated to each terminal device to be more accurate. In this way, the bandwidth resources allocated to each terminal device can be allocated reasonably.
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 bandwidth control method for a router according to an embodiment of the present application;
fig. 2 is a flowchart of a bandwidth control method for a router according to an embodiment of the present application;
fig. 3 is a system architecture diagram of a bandwidth control method for a router according to an embodiment of the present application;
fig. 4 is a flowchart of scaling the third feature map using classical multidimensional scaling to obtain a modified fused feature map based on a distance between corresponding positions in the first feature map and the second feature map in the bandwidth control method for a router according to the embodiment of the present application;
FIG. 5 is a block diagram of a bandwidth control system for a router according to an embodiment of the present application;
FIG. 6 is a block diagram of a scaling unit in a bandwidth control system for a router according to an embodiment of the present application;
fig. 7 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 scenes
As described above, in the conventional router, the bandwidth resources of each terminal device in the router are generally controlled based on the time-sharing control mode, that is, the signal strength of each terminal device is first obtained, the bandwidth resources corresponding to each terminal device are determined based on the signal strength, and then the calculated corresponding bandwidth resources are allocated to each terminal device.
However, this way of allocating bandwidth does not take into account the actual situation of each terminal device and the overall situation of each terminal device. This may result in insufficient bandwidth resources being allocated by the respective terminal devices. Therefore, to achieve reasonable bandwidth allocation, a bandwidth control scheme for routers is desired.
Based on this, the applicant of the present application expects to sufficiently express the bandwidth information and the signal condition information of each terminal device and the correlation therebetween by the deep learning technique based on the convolutional neural network, thereby realizing reasonable bandwidth allocation.
Specifically, bandwidth data of each terminal device at a series of predetermined time points is obtained to obtain a bandwidth matrix, signal strength data of each terminal device at the same series of predetermined time points is also obtained to obtain a signal matrix, and a first characteristic diagram and a second characteristic diagram are obtained through a first convolutional neural network and a second convolutional neural network respectively.
Here, since the first characteristic diagram and the second characteristic diagram respectively represent bandwidth information and signal strength information, and the bandwidth information here is actually an upstream bandwidth from the terminal device to the router, the applicant of the present application uses an upstream parameter to fuse the bandwidth information and the signal strength information to obtain a third characteristic diagram, which is represented as:
wherein f is1、f2And f3Is the eigenvalue of each position of the first, second and third profiles, respectively, h represents the fractional attenuation power component in the sample dimension, i.e. the sample dimension of the terminal device, which is frequency dependent and can be assumed to be an exponential distribution of unit means, a represents the fractional attenuation effect in the sample dimension, including path loss and occlusion loss. In addition, σ2Represents the power of additive white Gaussian noise and IcIs interference in a selected channel
After the third feature map is obtained, the third feature map is corrected based on the principle of classical multi-dimensional scaling in consideration of the scaling problem thereof with respect to the original data. Specifically, first, the distance between corresponding positions in the first feature map and the second feature map is calculated to obtain a fourth feature map, and then, for each of a third matrix and a fourth matrix of the third feature map and the fourth feature map in the channel dimension, a correction matrix is calculated, which is expressed as:
wherein M is3Is a third matrix, M4Is a fourth matrix, M4 ⊙2Indicating that the respective positions of the fourth matrix are squared. And finally, arranging the correction matrix M' according to the channel dimension to obtain a corrected fusion characteristic diagram.
In this way, when acquiring the current bandwidth allocation data of each terminal device, the encoder may first convert the current bandwidth allocation data into the high-dimensional feature space to obtain a query vector, multiply the query vector by the modified fused feature map to obtain an allocation feature vector, and then calculate the Softmax function value of each position of the allocation feature vector, that is, the percentage of the bandwidth allocated to each terminal device.
Based on this, the present application proposes a bandwidth control method for a router, which includes: acquiring bandwidth data of each terminal device communicated with a router at a series of preset time points, and constructing the bandwidth data of the series of preset time points into a bandwidth matrix according to device sample dimensions and time dimensions; acquiring signal intensity data of each terminal device at a same series of preset time points, and constructing the signal intensity data of the series of preset time points into a signal matrix according to the device sample dimension and the time dimension; obtaining a first feature map and a second feature map from the bandwidth matrix and the signal matrix using a first convolutional neural network and a second convolutional neural network as feature extractors; fusing the first characteristic diagram and the second characteristic diagram to obtain a third characteristic diagram based on uplink communication parameters of each terminal device to the router, wherein the uplink communication parameters are determined based on small-scale attenuation power components of the terminal devices, small-scale attenuation effects of the terminal devices, additive white Gaussian noise power and interference in the selected channel; scaling the third feature map using classical multi-dimensional scaling to obtain a modified fused feature map based on a distance between corresponding locations in the first and second feature maps; acquiring the current bandwidth allocation data of each terminal device and arranging the bandwidth allocation data of each terminal into bandwidth allocation vectors; encoding the bandwidth allocation vector using an encoder consisting of a plurality of fully-connected layers to obtain a bandwidth allocation eigenvector; taking the bandwidth allocation characteristic vector as a query vector to perform matrix multiplication with the corrected fusion characteristic diagram so as to map the bandwidth allocation characteristic vector into a characteristic space of the corrected fusion characteristic diagram to obtain an allocation characteristic vector; and calculating a Softmax function value of each position of the allocation feature vector as a bandwidth allocation percentage of the terminal device corresponding to each position, the Softmax function value being a weighted sum of a natural exponent function value raised to the power of the feature value of each position divided by the natural exponent function value raised to the power of the feature value of each position.
Fig. 1 illustrates an application scenario diagram of a bandwidth control method for a router according to an embodiment of the present application. As shown in fig. 1, in this application scenario, first, bandwidth data of each terminal device (e.g., T as illustrated in fig. 1) communicating with a router (e.g., R as illustrated in fig. 1) at a series of predetermined time points is acquired, and simultaneously signal strength data of each terminal device at the same series of predetermined time points is acquired. And obtaining the current bandwidth allocation data of each terminal device in the same way. Then, the obtained bandwidth data and signal strength data of each terminal device at a series of predetermined time points and the current bandwidth allocation data are input into a server (e.g., S as illustrated in fig. 1) deployed with a bandwidth control algorithm for a router, wherein the server is capable of processing the bandwidth data and signal strength data of each terminal device at a series of predetermined time points and the current bandwidth allocation data with the bandwidth control algorithm for the router to generate a Softmax function value representing a bandwidth allocation percentage of each terminal device. And then, the bandwidth resource to be allocated to each terminal device is obtained by multiplying the total value of the bandwidth allocation by the bandwidth allocation percentage of each terminal device, so that reasonable bandwidth allocation can be realized for each terminal device.
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 bandwidth control method for a router. As shown in fig. 2, a bandwidth control method for a router according to an embodiment of the present application includes: s110, acquiring bandwidth data of each terminal device communicated with the router at a series of preset time points, and constructing the bandwidth data of the series of preset time points into a bandwidth matrix according to device sample dimensions and time dimensions; s120, acquiring signal intensity data of each terminal device at a series of same preset time points, and constructing the signal intensity data of the series of preset time points into a signal matrix according to the device sample dimension and the time dimension; s130, using a first convolutional neural network and a second convolutional neural network as feature extractors to obtain a first feature map and a second feature map from the bandwidth matrix and the signal matrix; s140, fusing the first characteristic diagram and the second characteristic diagram to obtain a third characteristic diagram based on uplink communication parameters from each terminal device to the router, wherein the uplink communication parameters are determined based on the small-scale attenuation power component of the terminal device, the small-scale attenuation effect of the terminal device, the additive white Gaussian noise power and the interference in the selected channel; s150, carrying out scale transformation on the third feature map by using classical multi-dimensional scale transformation based on the distance between corresponding positions in the first feature map and the second feature map so as to obtain a modified fusion feature map; s160, obtaining the current bandwidth allocation data of each terminal device and arranging the bandwidth allocation data of each terminal into bandwidth allocation vectors; s170, encoding the bandwidth allocation vector by using an encoder composed of a plurality of fully-connected layers to obtain a bandwidth allocation characteristic vector; s180, taking the bandwidth allocation characteristic vector as a query vector to perform matrix multiplication with the corrected fusion characteristic diagram so as to map the bandwidth allocation characteristic vector to a characteristic space of the corrected fusion characteristic diagram to obtain an allocation characteristic vector; and S190, calculating Softmax function values of the respective positions of the allocation feature vector as bandwidth allocation percentages of the terminal devices corresponding to each position, the Softmax function values being weighted sums of natural exponent function values raised to powers of the feature values of the respective positions divided by natural exponent function values raised to powers of the feature values of the respective positions.
Fig. 3 illustrates an architecture diagram of a bandwidth control method for a router according to an embodiment of the present application. As shown in fig. 3, in the network architecture of the bandwidth control method for a router, first, the obtained bandwidth data (e.g., P1 as illustrated in fig. 3) of each of the terminal devices communicating with the router at a series of predetermined time points is constructed as a bandwidth matrix (e.g., M1 as illustrated in fig. 3) according to a device sample dimension and a time dimension; then, constructing the obtained signal strength data (for example, P2 as illustrated in fig. 3) of each terminal device at the same series of predetermined time points into a signal matrix (for example, M2 as illustrated in fig. 3) according to the device sample dimension and the time dimension; then, a first eigenmap (e.g., F1 as illustrated in fig. 3) and a second eigenmap (e.g., F2 as illustrated in fig. 3) are obtained from the bandwidth matrix and the signal matrix using a first convolutional neural network (e.g., CNN1 as illustrated in fig. 3) and a second convolutional neural network (e.g., CNN2 as illustrated in fig. 3) as eigen extractors; then, fusing the first characteristic diagram and the second characteristic diagram based on uplink communication parameters (for example, U as illustrated in fig. 3) of each terminal device to the router to obtain a third characteristic diagram (for example, F3 as illustrated in fig. 3); then, scaling the third feature map using classical multi-dimensional scaling to obtain a feature map (e.g., FC as illustrated in fig. 3) based on a distance between corresponding locations in the first and second feature maps; then, arranging the obtained current bandwidth allocation data (e.g., Q as illustrated in fig. 3) of each of the terminals into a bandwidth allocation vector (e.g., V as illustrated in fig. 3); then, encoding the bandwidth allocation vector using an encoder (e.g., E as illustrated in fig. 3) composed of a plurality of fully connected layers to obtain a bandwidth allocation feature vector (e.g., VF1 as illustrated in fig. 3); then, taking the bandwidth allocation eigenvector as a query vector to perform matrix multiplication with the modified fused feature map to map the bandwidth allocation eigenvector into an eigenspace of the modified fused feature map to obtain an allocation eigenvector (e.g., VF2 as illustrated in fig. 3); and finally, calculating a Softmax function value (for example, SF as illustrated in fig. 3) of each location of the allocation feature vector as a bandwidth allocation percentage of the terminal device corresponding to each location, the Softmax function value being a weighted sum of a natural exponent function value raised to the power of the feature value of each location divided by the natural exponent function value raised to the power of the feature value of each location.
In step S110 and step S120, acquiring bandwidth data of each terminal device communicating with the router at a series of predetermined time points and constructing the bandwidth data of the series of predetermined time points into a bandwidth matrix according to a device sample dimension and a time dimension; and acquiring signal intensity data of each terminal device at a series of same preset time points and constructing the signal intensity data of the series of preset time points into a signal matrix according to the device sample dimension and the time dimension. As described above, the bandwidth allocation of the existing bandwidth allocation method for routers may be insufficient because the actual signal condition of each terminal device and the overall situation of each terminal device are not considered. Therefore, in the technical solution of the present application, the bandwidth information and the signal condition information of each terminal device and the association therebetween are fully expressed by using a deep learning technique based on a convolutional neural network, so as to implement reasonable bandwidth allocation. Before this, first, bandwidth data of each of the terminal devices communicating with the router at a series of predetermined time points and signal strength data of each of the terminal devices at the same series of predetermined time points need to be acquired. Then, the bandwidth data of the series of predetermined time points are constructed as a bandwidth matrix according to the device sample dimension and the time dimension, and in the same way, the signal strength data of the series of predetermined time points are also constructed as a signal matrix according to the device sample dimension and the time dimension.
In step S130, a first feature map and a second feature map are obtained from the bandwidth matrix and the signal matrix using a first convolutional neural network and a second convolutional neural network as feature extractors. That is, first, the bandwidth matrix is processed by a first convolutional neural network serving as a feature extractor to extract high-dimensional correlation features between bandwidth information of the terminal devices at a series of predetermined time points, thereby obtaining a first feature map. Then, the signal matrix is processed by a second convolution neural network serving as a feature extractor to extract high-dimensional correlation features among signal strength information of each terminal device at a series of preset time points, so that a second feature map is obtained.
Specifically, in the embodiment of the present application, the first convolutional neural network and the second convolutional neural network have the same network structure; wherein the first convolutional neural network and the second convolutional neural network process the bandwidth matrix and the signal matrix with the following formula to obtain the first feature map and the second feature map;
the formula is:
fi＝active(Ni×fi-1+Bi)
wherein f isi-1Is the input of the i-th convolutional neural network, fiIs the output of the ith convolutional neural network, NiIs the convolution kernel of the ith convolutional neural network, and BiActive represents a nonlinear activation function for the bias vector of the ith layer of convolutional neural network.
In step S140, the first feature map and the second feature map are fused based on uplink communication parameters from the respective terminal devices to the router to obtain a third feature map. It should be understood that, since the first characteristic diagram and the second characteristic diagram respectively represent bandwidth information and signal strength information of each terminal device at a series of predetermined time points, and the bandwidth information here is actually an upstream bandwidth of each terminal device to the router. Therefore, in order to effectively fuse the bandwidth information and the signal strength information, in the technical solution of the present application, the first characteristic diagram and the second characteristic diagram are fused based on uplink communication parameters from each terminal device to the router, so as to obtain a third characteristic diagram. In one specific example, the uplink communication parameter is determined based on a small-scale fading power component of the terminal device, a small-scale fading effect of the terminal device, an additive white gaussian noise power, and interference in the selected channel
Specifically, in this embodiment of the present application, a process of fusing the first feature map and the second feature map to obtain a third feature map based on uplink communication parameters from each terminal device to the router includes: fusing the first characteristic diagram and the second characteristic diagram to obtain a third characteristic diagram according to the following formula based on the uplink communication parameters from the terminal equipment to the router;
the formula is:
wherein f is1、f2And f3Is an eigenvalue of each position of said first, second and third profiles, respectively, h represents a small-scale attenuation power component in the sample dimension of the terminal device, which is frequency dependent and can be assumed to be an exponential distribution of unity mean, a represents small-scale attenuation effects in the sample dimension of the terminal device, including path loss and occlusion loss, σ2Represents the power of additive white Gaussian noise and IcIs interference in the selected channel.
In step S150, the third feature map is scaled using classical multi-dimensional scaling to obtain a modified fused feature map based on the distance between the corresponding positions in the first and second feature maps. It should be understood that, after the third feature map is obtained, considering the problem of scale transformation of the third feature map with respect to the original data, in the technical solution of the present application, based on the distance between the corresponding positions in the first feature map and the second feature map, the third feature map is modified by using the principle of classical multidimensional scale transformation, so as to obtain a modified fused feature map.
Specifically, in this embodiment of the present application, a process of scaling the third feature map using classical multidimensional scaling to obtain a modified fused feature map based on a distance between corresponding positions in the first feature map and the second feature map includes: first, the distance between the corresponding positions in the first feature map and the second feature map is calculated to obtain a fourth feature map. In a specific example, the L1 distance or the L2 distance between corresponding positions in the first profile and the second profile may be calculated to obtain the fourth profile. It should be appreciated that the L1 distance function, also referred to as the minimum absolute deviation (LAD), is the sum of the absolute differences of the target value and the estimated value. By calculating the L1 distance, the feature difference degree between the feature vector of each word in the sentence vector sequence and the query vector can be reflected in a numerical dimension. The L2 distance function, also known as the Least Squares Error (LSE), is the sum of the squares of the differences between the target and estimated values, also called the euclidean distance. The L2 distance function has a stable solution compared to the L1 distance function.
Then, for each of a third matrix and a fourth matrix of the third feature map and the fourth feature map in a channel dimension, a correction matrix between the third matrix and the fourth matrix is calculated. In a specific example, for each of the third matrix and the fourth matrix of the third feature map and the fourth feature map in the channel dimension, calculating a modification matrix between the third matrix and the fourth matrix in the following formula; the formula is:
wherein M is3Is a third matrix, M4Is a fourth matrix, M4 ⊙2Indicating that the respective positions of the fourth matrix are squared.
And finally, arranging the correction matrix M' into a corrected fusion characteristic diagram according to the channel dimension.
Fig. 4 illustrates a flowchart of scaling the third feature map using classical multidimensional scaling to obtain a modified fused feature map based on a distance between corresponding positions in the first feature map and the second feature map in a bandwidth control method for a router according to an embodiment of the present application. As shown in fig. 4, in the embodiment of the present application, scaling the third feature map by using classical multidimensional scaling to obtain a modified fused feature map based on a distance between corresponding positions in the first feature map and the second feature map includes: s210, calculating the distance between corresponding positions in the first feature map and the second feature map to obtain a fourth feature map; s220, calculating a correction matrix between a third matrix and a fourth matrix for each of the third matrix and the fourth matrix of the third feature map and the fourth feature map in the channel dimension; and S230, arranging the correction matrixes into the corrected fusion characteristic diagram according to channel dimensions.
In steps S160 and S170, bandwidth allocation data of each of the current terminal devices is obtained and arranged as a bandwidth allocation vector, and the bandwidth allocation vector is encoded using an encoder composed of a plurality of fully-connected layers to obtain a bandwidth allocation feature vector. That is, when bandwidth allocation data of each current terminal device is obtained, first, the bandwidth allocation data of each terminal device is arranged as a bandwidth allocation vector; then, the bandwidth allocation vector is encoded by an encoder composed of a plurality of fully connected layers to convert the bandwidth allocation vector into a high-dimensional feature space for subsequent fusion processing, so as to obtain a bandwidth allocation feature vector, i.e., a query vector.
In step S180, the bandwidth allocation eigenvector is used as a query vector to perform matrix multiplication with the modified fused feature map to map the bandwidth allocation eigenvector into the eigenspace of the modified fused feature map to obtain an allocation eigenvector. That is, in order to fuse the bandwidth allocation eigenvector with the information in the modified fused feature map to map the bandwidth allocation eigenvector into the feature space of the modified fused feature map, in a specific example, the bandwidth allocation eigenvector may be matrix-multiplied with the modified fused feature map as a query vector to obtain an allocation eigenvector.
In step S190, Softmax function values of the respective positions of the allocation feature vector, which are weighted sums of natural exponent function values raised to powers of the feature values of the respective positions divided by natural exponent function values raised to powers of the feature values of the respective positions, are calculated as bandwidth allocation percentages of the terminal devices corresponding to each position. Specifically, in this embodiment of the present application, the process of calculating the Softmax function value of each position of the allocation feature vector as the bandwidth allocation percentage of the terminal device corresponding to each position includes: calculating a Softmax function value of each position of the allocation feature vector as a bandwidth allocation percentage of the terminal device corresponding to each position according to the following formula: exp (pi)/Sigmaiexp (pi), where pi represents the eigenvalue of each position in the assigned eigenvector.
Further, the bandwidth resource to be allocated to each terminal device is obtained by multiplying the total value of the bandwidth allocation by the bandwidth allocation percentage of each terminal device, so that reasonable bandwidth allocation can be realized for each terminal device.
In summary, the bandwidth control method for the router according to the embodiment of the present application is elucidated, which adopts a deep learning technique based on a convolutional neural network to fully find out correlation characteristics between bandwidth information and signal strength information of each terminal device, and uses an uplink parameter to fuse the bandwidth information and the signal strength information, and then corrects the obtained characteristic diagram based on the principle of classical multidimensional scaling, so that the percentage of the bandwidth allocated to each terminal device is more accurate. In this way, the bandwidth resources allocated to each terminal device can be allocated reasonably.
Exemplary System
Fig. 5 illustrates a block diagram of a bandwidth control system for a router according to an embodiment of the application. As shown in fig. 5, a bandwidth control system 500 for a router according to an embodiment of the present application includes: a bandwidth matrix constructing unit 510, configured to acquire bandwidth data of each terminal device that is communicated with the router at a series of predetermined time points and construct the bandwidth data of the series of predetermined time points into a bandwidth matrix according to a device sample dimension and a time dimension; a signal matrix constructing unit 520, configured to acquire signal intensity data of each terminal device at a same series of predetermined time points and construct the signal intensity data of the series of predetermined time points into a signal matrix according to a device sample dimension and a time dimension; a feature extraction unit 530 for obtaining a first feature map and a second feature map from the bandwidth matrix obtained by the bandwidth matrix construction unit 510 and the signal matrix obtained by the signal matrix construction unit 520 using a first convolutional neural network and a second convolutional neural network as feature extractors; a merging unit 540, configured to merge the first feature map obtained by the feature extraction unit 530 and the second feature map obtained by the feature extraction unit 530 to obtain a third feature map based on uplink communication parameters from each of the terminal devices to the router, where the uplink communication parameters are determined based on the small-scale attenuation power component of the terminal device, the small-scale attenuation effect of the terminal device, the additive white gaussian noise power, and the interference in the selected channel; a scaling unit 550, configured to scale the third feature map obtained by the fusion unit by using classical multidimensional scaling to obtain a modified fusion feature map based on a distance between the first feature map obtained by the feature extraction unit 530 and a corresponding position in the second feature map obtained by the feature extraction unit 540; a bandwidth allocation vector generating unit 560, configured to obtain current bandwidth allocation data of each terminal device and arrange the bandwidth allocation data of each terminal device as a bandwidth allocation vector; an encoding unit 570 configured to encode the bandwidth allocation vector obtained by the bandwidth allocation vector generation unit 560 using an encoder configured by a plurality of fully-connected layers to obtain a bandwidth allocation feature vector; a mapping unit 580, configured to perform matrix multiplication on the bandwidth allocation feature vector obtained by the encoding unit 570 as a query vector and the modified fused feature map obtained by the scale transforming unit 550 to map the bandwidth allocation feature vector into a feature space of the modified fused feature map to obtain an allocation feature vector; and a result calculation unit 590 configured to calculate, as a bandwidth allocation percentage of the terminal device corresponding to each location, a Softmax function value of each location of the allocation feature vector obtained by the mapping unit 580, the Softmax function value being a weighted sum of a natural exponent function value raised to the power of the feature value of each location divided by a natural exponent function value raised to the power of the feature value of each location.
In one example, in the bandwidth control system 500 for a router described above, the first convolutional neural network and the second convolutional neural network have the same network structure; wherein the first convolutional neural network and the second convolutional neural network process the bandwidth matrix and the signal matrix with the following formula to obtain the first feature map and the second feature map;
the formula is:
fi＝active(Ni×fi-1+Bi)
wherein f isi-1Is the input of the i-th convolutional neural network, fiIs the output of the ith convolutional neural network, NiConvolution for i-th layer convolution neural networkA core, and BiActive represents a nonlinear activation function for the bias vector of the ith layer of convolutional neural network.
In an example, in the bandwidth control system 500 for a router, the merging unit 540 is further configured to: fusing the first characteristic diagram and the second characteristic diagram to obtain a third characteristic diagram according to the following formula based on the uplink communication parameters from the terminal equipment to the router;
the formula is:
wherein f is1、f2And f3Is a feature value of each position of said first, second and third feature maps, respectively, h represents a small-scale attenuation power component in a sample dimension of the terminal device, a represents a small-scale attenuation effect in the sample dimension of the terminal device, σ represents a maximum value of the attenuation power component in the sample dimension of the terminal device, and2represents the power of additive white Gaussian noise and IcIs interference in the selected channel.
In an example, in the bandwidth control system 500 for a router, as shown in fig. 6, the scaling unit 550 includes: a fourth feature map generation subunit 551, configured to calculate distances between corresponding positions in the first feature map and the second feature map to obtain a fourth feature map; a modification matrix generation subunit 552 configured to calculate, for each of a third matrix and a fourth matrix of the fourth feature map in the channel dimension obtained by the third feature map generation subunit 551, a modification matrix between the third matrix and the fourth matrix; and a modified arrangement subunit 553, configured to arrange the modified matrix obtained by the modified matrix generation subunit 552 into the modified fusion feature map according to the channel dimension.
In one example, in the bandwidth control system 500 for a router described above, the fourth feature map generating subunit 551 is further configured to: and calculating Euclidean distances between corresponding positions in the first feature map and the second feature map to obtain the fourth feature map.
In an example, in the bandwidth control system 500 for a router, the modification matrix generating subunit 552 is further configured to: for each third matrix and fourth matrix of the third feature map and the fourth feature map in the channel dimension, calculating a correction matrix between the third matrix and the fourth matrix according to the following formula;
the formula is:
wherein M is3Is a third matrix, M4Is a fourth matrix, M4 ⊙2Indicating that the respective positions of the fourth matrix are squared.
In an example, in the bandwidth control system 500 for a router described above, the result calculating unit 590 is further configured to: calculating a Softmax function value of each position of the allocation feature vector as a bandwidth allocation percentage of the terminal device corresponding to each position according to the following formula: exp (pi)/Sigmaiexp (pi), where pi represents the eigenvalue of each position in the assigned eigenvector.
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 above-described bandwidth control system 500 for a router have been described in detail in the above description of the bandwidth control method for a router with reference to fig. 1 to 4, and thus, a repetitive description thereof will be omitted.
As described above, the bandwidth control system 500 for a router according to the embodiment of the present application may be implemented in various terminal devices, such as a server for a bandwidth control algorithm of a router, and the like. In one example, the bandwidth control system 500 for a router according to the embodiment of the present application may be integrated into a terminal device as one software module and/or hardware module. For example, the bandwidth control system 500 for a router 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 bandwidth control system 500 for a router may also be one of many hardware modules of the terminal device.
Alternatively, in another example, the bandwidth control system for router 500 and the terminal device may also be separate devices, and the bandwidth control system for router 500 may be connected to the terminal device through a wired and/or wireless network and transmit the interaction information according to the agreed data format.
Exemplary electronic device
Next, an electronic apparatus according to an embodiment of the present application is described with reference to fig. 7. As shown in fig. 7, the electronic device 10 includes one or more processors 11 and a memory 12. The processor 11 may be a Central Processing Unit (CPU) or other form of processing unit having data processing capabilities and/or instruction execution capabilities, and may control other components in the electronic device 10 to perform desired functions.
In one example, the electronic device 10 may further include: an input system 13 and an output system 14, which are interconnected by a bus system and/or other form of connection mechanism (not shown).
The input system 13 may comprise, for example, a keyboard, a mouse, etc.
The output system 14 may output various information to the outside including the percentage of bandwidth allocation, etc. The output system 14 may include, for example, a display, speakers, a printer, and a communication network and its connected remote output devices, among others.
Of course, for simplicity, only some of the components of the electronic device 10 relevant to the present application are shown in fig. 7, and components such as buses, input/output interfaces, and the like are omitted. In addition, the electronic device 10 may include any other suitable components depending on the particular application.
Exemplary computer program product and computer-readable storage Medium
In addition to the above-described methods and apparatus, embodiments of the present application may also be a computer program product comprising computer program instructions that, when executed by a processor, cause the processor to perform the steps in the functions in the bandwidth control method for a router according to the various embodiments of the present application described in the "exemplary methods" section of this specification above.
The computer program product may be written with program code for performing the operations of embodiments of the present application in any combination of one or more programming languages, including an object oriented programming language such as Java, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device, or entirely on the remote computing device or server.
Furthermore, embodiments of the present application may also be a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, cause the processor to perform the steps in the bandwidth control method for a router described in the "exemplary methods" section above in this specification.
The computer-readable storage medium may take any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may include, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
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, devices, systems referred to in this application are only used as illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured 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 bandwidth control method for a router, comprising:
acquiring bandwidth data of each terminal device communicated with the router at a series of preset time points, and constructing the bandwidth data of the series of preset time points into a bandwidth matrix according to the device sample dimension and the time dimension;
acquiring signal intensity data of each terminal device at a same series of preset time points, and constructing the signal intensity data of the series of preset time points into a signal matrix according to the device sample dimension and the time dimension;
obtaining a first feature map and a second feature map from the bandwidth matrix and the signal matrix using a first convolutional neural network and a second convolutional neural network as feature extractors;
fusing the first characteristic diagram and the second characteristic diagram to obtain a third characteristic diagram based on uplink communication parameters from each terminal device to the router, wherein the uplink communication parameters are determined based on small-scale attenuation power components of the terminal devices, small-scale attenuation effects of the terminal devices, additive white gaussian noise power and interference in the selected channel;
scaling the third feature map using classical multi-dimensional scaling to obtain a modified fused feature map based on a distance between corresponding locations in the first and second feature maps;
acquiring the current bandwidth allocation data of each terminal device and arranging the bandwidth allocation data of each terminal into bandwidth allocation vectors;
encoding the bandwidth allocation vector using an encoder consisting of a plurality of fully-connected layers to obtain a bandwidth allocation eigenvector;
taking the bandwidth allocation characteristic vector as a query vector to perform matrix multiplication with the corrected fusion characteristic diagram so as to map the bandwidth allocation characteristic vector into a characteristic space of the corrected fusion characteristic diagram to obtain an allocation characteristic vector; and
and calculating a Softmax function value of each position of the allocation feature vector as a bandwidth allocation percentage of the terminal device corresponding to each position, the Softmax function value being a weighted sum of a natural exponent function value raised to the power of the feature value of each position divided by the natural exponent function value raised to the power of the feature value of each position.
2. The bandwidth control method for a router according to claim 1, wherein the first convolutional neural network and the second convolutional neural network have the same network structure;
wherein the first convolutional neural network and the second convolutional neural network process the bandwidth matrix and the signal matrix with the following formula to obtain the first feature map and the second feature map;
the formula is:
fi＝active(Ni×fi-1+Bi)
wherein, fi-1Is the ith layerInput of convolutional neural network, fiIs the output of the ith convolutional neural network, NiIs the convolution kernel of the ith convolutional neural network, and BiActive represents a nonlinear activation function for the bias vector of the ith layer of convolutional neural network.
3. The bandwidth control method for a router according to claim 2, wherein fusing the first feature map and the second feature map based on uplink communication parameters from the respective terminal devices to the router to obtain a third feature map comprises:
fusing the first characteristic diagram and the second characteristic diagram to obtain a third characteristic diagram according to the following formula based on the uplink communication parameters from the terminal equipment to the router;
the formula is:
wherein f is1、f2And f3Is an eigenvalue of each position of said first, second and third profiles, respectively, h represents a small-scale attenuation power component in the sample dimension of the terminal device, a represents a small-scale attenuation effect in the sample dimension of the terminal device, σ2Represents the power of additive white Gaussian noise and IcIs interference in the selected channel.
4. The bandwidth control method for a router of claim 3, wherein scaling the third feature map using classical multi-dimensional scaling to obtain a modified fused feature map based on a distance between corresponding locations in the first and second feature maps comprises:
calculating the distance between corresponding positions in the first feature map and the second feature map to obtain a fourth feature map;
for each third matrix and fourth matrix of the third feature map and the fourth feature map in the channel dimension, calculating a correction matrix between the third matrix and the fourth matrix; and
and arranging the correction matrix into the corrected fusion characteristic diagram according to the channel dimension.
5. The bandwidth control method for a router of claim 4, wherein calculating a distance between corresponding positions in the first and second feature maps to obtain a fourth feature map comprises:
and calculating Euclidean distances between corresponding positions in the first feature map and the second feature map to obtain the fourth feature map.
6. The bandwidth control method for a router of claim 5, wherein for each of a third matrix and a fourth matrix of the third feature map and the fourth feature map in a channel dimension, calculating a modification matrix between the third matrix and the fourth matrix comprises:
for each third matrix and fourth matrix of the third feature map and the fourth feature map in the channel dimension, calculating a correction matrix between the third matrix and the fourth matrix according to the following formula;
the formula is:
wherein M is3Is a third matrix, M4Is a fourth matrix, M4 ⊙2Indicating that the respective positions of the fourth matrix are squared.
7. The bandwidth control method for a router according to claim 1, wherein calculating the Softmax function value of each position of the allocation feature vector as a bandwidth allocation percentage of the terminal device corresponding to each position includes:
calculating each of the assigned feature vectors by the following formulaAnd taking the Softmax function value of the position as the bandwidth allocation percentage of the terminal device corresponding to each position, wherein the formula is as follows: exp (pi)/Sigmaiexp (pi), where pi represents the eigenvalue of each position in the assigned eigenvector.
8. A bandwidth control system for a router, comprising:
the bandwidth matrix construction unit is used for acquiring bandwidth data of each terminal device communicated with the router at a series of preset time points and constructing the bandwidth data of the series of preset time points into a bandwidth matrix according to the device sample dimension and the time dimension;
the signal matrix constructing unit is used for acquiring signal intensity data of each terminal device at a same series of preset time points and constructing the signal intensity data of the series of preset time points into a signal matrix according to the device sample dimension and the time dimension;
a feature extraction unit configured to obtain a first feature map and a second feature map from the bandwidth matrix obtained by the bandwidth matrix construction unit and the signal matrix obtained by the signal matrix construction unit using a first convolutional neural network and a second convolutional neural network as a feature extractor;
a fusion unit configured to fuse the first feature map obtained by the feature extraction unit and the second feature map obtained by the feature extraction unit to obtain a third feature map based on uplink communication parameters from each of the terminal devices to the router, the uplink communication parameters being determined based on a small-scale attenuation power component of the terminal device, a small-scale attenuation effect of the terminal device, an additive white gaussian noise power, and interference in the selected channel;
a scaling unit configured to scale the third feature map obtained by the fusion unit using classical multidimensional scaling based on a distance between the first feature map obtained by the feature extraction unit and a corresponding position in the second feature map obtained by the feature extraction unit to obtain a modified fusion feature map;
a bandwidth allocation vector generating unit, configured to obtain current bandwidth allocation data of each terminal device and arrange the bandwidth allocation data of each terminal as a bandwidth allocation vector;
an encoding unit configured to encode the bandwidth allocation vector obtained by the bandwidth allocation vector generation unit using an encoder configured by a plurality of fully-connected layers to obtain a bandwidth allocation feature vector;
a mapping unit, configured to perform matrix multiplication on the bandwidth allocation feature vector obtained by the encoding unit as a query vector and the modified fusion feature map obtained by the scale transformation unit to map the bandwidth allocation feature vector into a feature space of the modified fusion feature map to obtain an allocation feature vector; and
a result calculation unit configured to calculate, as a bandwidth allocation percentage of the terminal device corresponding to each location, a Softmax function value of each location of the allocation feature vector obtained by the mapping unit, the Softmax function value being a weighted sum of a natural exponent function value raised to a power of the feature value of each location divided by the natural exponent function value raised to a power of the feature value of each location.
9. The bandwidth control system for a router of claim 8, wherein the first convolutional neural network and the second convolutional neural network have the same network structure;
wherein the first convolutional neural network and the second convolutional neural network process the bandwidth matrix and the signal matrix with the following formula to obtain the first characteristic diagram and the second characteristic diagram;
the formula is:
fi＝active(Ni×fi-1+Bi)
wherein f isi-1Is the input of the i-th convolutional neural network, fiIs the output of the ith convolutional neural network, NiIs the convolution kernel of the ith convolutional neural network, and BiActive represents not for bias vector of i-th layer convolutional neural networkA linear activation function.
10. An electronic device, comprising:
a processor; and
memory having stored therein computer program instructions which, when executed by the processor, cause the processor to carry out the method of bandwidth control for a router of any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210244423.2A CN114615147A (en) | 2022-03-14 | 2022-03-14 | Bandwidth control method and system for router and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210244423.2A CN114615147A (en) | 2022-03-14 | 2022-03-14 | Bandwidth control method and system for router and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114615147A true CN114615147A (en) | 2022-06-10 |
Family
ID=81863521
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210244423.2A Withdrawn CN114615147A (en) | 2022-03-14 | 2022-03-14 | Bandwidth control method and system for router and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114615147A (en) |
-
2022
- 2022-03-14 CN CN202210244423.2A patent/CN114615147A/en not_active Withdrawn
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111414987B (en) | Training method and training device of neural network and electronic equipment | |
| Zheng | Gradient descent algorithms for quantile regression with smooth approximation | |
| JP2021517316A (en) | Neural network display generation method and its equipment, computer programs and equipment | |
| CN113111543B (en) | Internet of things service system | |
| CN115373813A (en) | Scheduling method and system based on GPU virtualization in cloud computing environment and electronic equipment | |
| CN115982736B (en) | Data encryption method and system for computer network information | |
| CN116579618B (en) | Data processing method, device, equipment and storage medium based on risk management | |
| CN114884772A (en) | Bare computer vxlan deployment method, system and electronic equipment | |
| CN115878319A (en) | Load balancing method and system and electronic equipment | |
| CN114423030A (en) | Full-house dynamic automatic networking method, system and electronic equipment | |
| CN115718846A (en) | Big data mining method and system for intelligent interactive network | |
| CN114821169A (en) | Method-level non-intrusive call link tracking method under micro-service architecture | |
| CN115617636B (en) | Distributed performance test system | |
| CN114610488A (en) | Data acquisition method and system based on load balancing and electronic equipment | |
| JP2021033994A (en) | Text processing method, apparatus, device and computer readable storage medium | |
| CN114615147A (en) | Bandwidth control method and system for router and electronic equipment | |
| CN113626646A (en) | Intelligent electric meter data compression method and device and electronic equipment | |
| CN115526320A (en) | Neural network model inference acceleration method, apparatus, electronic device and medium | |
| CN115409617A (en) | Foreign currency exchange method and system based on block chain | |
| CN111222722B (en) | Method, neural network model and device for business prediction for business object | |
| CN114610445A (en) | Virtual machine-based resource scheduling method and system for cloud computing | |
| CN114221830A (en) | Network distribution method and system of intelligent electronic equipment and electronic equipment | |
| Tran et al. | Family‐wise error rate control in Gaussian graphical model selection via distributionally robust optimization | |
| CN116794975B (en) | Intelligent control method and system for electric butterfly valve | |
| CN114585056A (en) | Proxy AP (Access Point) selection method, system and electronic equipment |
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: 20220610 |
|
| WW01 | Invention patent application withdrawn after publication |