CN117215766A - Multimedia intelligent dispatcher and method thereof - Google Patents
Multimedia intelligent dispatcher and method thereof Download PDFInfo
- Publication number
- CN117215766A CN117215766A CN202310756083.6A CN202310756083A CN117215766A CN 117215766 A CN117215766 A CN 117215766A CN 202310756083 A CN202310756083 A CN 202310756083A CN 117215766 A CN117215766 A CN 117215766A
- Authority
- CN
- China
- Prior art keywords
- feature vector
- running state
- time sequence
- parameter
- multimedia
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 26
- 239000013598 vector Substances 0.000 claims abstract description 141
- 238000003860 storage Methods 0.000 claims abstract description 42
- 238000012098 association analyses Methods 0.000 claims abstract description 9
- 239000011159 matrix material Substances 0.000 claims description 62
- 238000001514 detection method Methods 0.000 claims description 16
- 230000011218 segmentation Effects 0.000 claims description 14
- 238000000605 extraction Methods 0.000 claims description 13
- 238000012545 processing Methods 0.000 claims description 13
- 238000009826 distribution Methods 0.000 claims description 12
- 238000005728 strengthening Methods 0.000 claims description 11
- 238000005457 optimization Methods 0.000 claims description 9
- 238000011156 evaluation Methods 0.000 claims description 5
- 230000006870 function Effects 0.000 claims description 5
- 230000004927 fusion Effects 0.000 claims description 5
- 230000002787 reinforcement Effects 0.000 claims description 3
- 238000010219 correlation analysis Methods 0.000 claims description 2
- 239000012634 fragment Substances 0.000 claims description 2
- 238000004458 analytical method Methods 0.000 abstract description 9
- 238000012544 monitoring process Methods 0.000 description 13
- 238000010586 diagram Methods 0.000 description 10
- 238000012549 training Methods 0.000 description 7
- 238000011176 pooling Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 230000002159 abnormal effect Effects 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 5
- 230000008859 change Effects 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 238000007726 management method Methods 0.000 description 4
- 230000004913 activation Effects 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 3
- 230000015556 catabolic process Effects 0.000 description 3
- 230000002596 correlated effect Effects 0.000 description 3
- 230000000875 corresponding effect Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000012706 support-vector machine Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000003066 decision tree Methods 0.000 description 2
- 230000001419 dependent effect Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000007637 random forest analysis Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000005094 computer simulation Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000013468 resource allocation Methods 0.000 description 1
- 238000012731 temporal analysis Methods 0.000 description 1
- 238000000700 time series analysis Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The application discloses a multimedia intelligent dispatcher and a method thereof, which acquire CPU occupation ratios, memory occupation ratios, disk storage amounts and network bandwidth values at a plurality of preset time points in a preset time period; performing time sequence association analysis on CPU occupancy rate, memory occupancy rate, disk storage amount and network bandwidth value of the plurality of preset time points to obtain an operation state feature vector; and determining whether the running state of the multimedia intelligent dispatcher is normal or not based on the running state feature vector. Therefore, the service condition of the multimedia intelligent dispatcher can be detected through comprehensive analysis, so that whether the running state of the multimedia intelligent dispatcher is normal or not is judged, the normal work of the multimedia intelligent dispatcher is ensured, and the stability and the performance of a multimedia system are improved.
Description
Technical Field
The application relates to the technical field of intelligent scheduling, in particular to a multimedia intelligent scheduling machine and a method thereof.
Background
With the continuous development of multimedia technology, multimedia systems play an increasingly important role in our daily lives. For a large multimedia system, how to perform efficient resource scheduling and monitoring is a very critical issue.
At present, the traditional manual management mode is mainly to monitor and schedule various resources in a multimedia system in a manual mode. However, the manual monitoring and scheduling methods are prone to errors, thereby affecting the normal operation of the entire multimedia system. Meanwhile, in a large-scale multimedia system, the use condition of various resources is difficult to be mastered completely in time and accurately by manual monitoring, and faults and even breakdown caused by monitoring loopholes exist. In addition, the multimedia system has a plurality of resource types needing to be scheduled, has wide involved areas, and has low manual scheduling efficiency. The traditional manual management mode can not realize real-time monitoring and dynamic scheduling, so that the system has no quick response capability.
Therefore, an optimized multimedia intelligent scheduler is desired.
Disclosure of Invention
The embodiment of the application provides a multimedia intelligent scheduler and a method thereof, which acquire CPU occupancy, memory occupancy, disk storage and network bandwidth values at a plurality of preset time points in a preset time period; performing time sequence association analysis on CPU occupancy rate, memory occupancy rate, disk storage amount and network bandwidth value of the plurality of preset time points to obtain an operation state feature vector; and determining whether the running state of the multimedia intelligent dispatcher is normal or not based on the running state feature vector. Therefore, the service condition of the multimedia intelligent dispatcher can be detected through comprehensive analysis, so that whether the running state of the multimedia intelligent dispatcher is normal or not is judged, the normal work of the multimedia intelligent dispatcher is ensured, and the stability and the performance of a multimedia system are improved.
The embodiment of the application also provides a multimedia intelligent dispatcher, which comprises:
the data acquisition module is used for acquiring CPU occupancy rate, memory occupancy rate, disk storage amount and network bandwidth value of a plurality of preset time points in a preset time period;
the data processing module is used for carrying out time sequence association analysis on the CPU occupancy rate, the memory occupancy rate, the disk storage amount and the network bandwidth value at a plurality of preset time points so as to obtain an operation state feature vector; and
and the running state detection module is used for determining whether the running state of the multimedia intelligent dispatcher is normal or not based on the running state feature vector.
The embodiment of the application also provides a multimedia intelligent scheduling method, which comprises the following steps:
acquiring CPU occupancy rate, memory occupancy rate, disk storage amount and network bandwidth value of a plurality of preset time points in a preset time period;
performing time sequence association analysis on CPU occupancy rate, memory occupancy rate, disk storage amount and network bandwidth value of the plurality of preset time points to obtain an operation state feature vector; and
and determining whether the running state of the multimedia intelligent dispatcher is normal or not based on the running state feature vector.
Drawings
In order to more clearly illustrate the embodiments of the application or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art. In the drawings:
fig. 1 is a block diagram of a multimedia intelligent scheduler provided in an embodiment of the present application.
Fig. 2 is a block diagram of a data processing module in a multimedia intelligent scheduler according to an embodiment of the present application.
Fig. 3 is a block diagram of the running state detection module in the multimedia intelligent scheduler according to an embodiment of the present application.
Fig. 4 is a flowchart of a multimedia intelligent scheduling method provided in an embodiment of the present application.
Fig. 5 is a schematic diagram of a system architecture of a multimedia intelligent scheduling method according to an embodiment of the present application.
Fig. 6 is an application scenario diagram of a multimedia intelligent scheduler provided in an embodiment of the present application.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present application more apparent, the embodiments of the present application will be described in further detail with reference to the accompanying drawings. The exemplary embodiments of the present application and their descriptions herein are for the purpose of explaining the present application, but are not to be construed as limiting the application.
Unless defined otherwise, all technical and scientific terms used in the embodiments of the application have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in the present application is for the purpose of describing particular embodiments only and is not intended to limit the scope of the present application.
In describing embodiments of the present application, unless otherwise indicated and limited thereto, the term "connected" should be construed broadly, for example, it may be an electrical connection, or may be a communication between two elements, or may be a direct connection, or may be an indirect connection via an intermediate medium, and it will be understood by those skilled in the art that the specific meaning of the term may be interpreted according to circumstances.
It should be noted that, the term "first\second\third" related to the embodiment of the present application is merely to distinguish similar objects, and does not represent a specific order for the objects, it is to be understood that "first\second\third" may interchange a specific order or sequence where allowed. It is to be understood that the "first\second\third" distinguishing objects may be interchanged where appropriate such that embodiments of the application described herein may be practiced in sequences other than those illustrated or described herein.
In one embodiment of the present application, fig. 1 is a block diagram of a multimedia intelligent scheduler provided in an embodiment of the present application. As shown in fig. 1, a multimedia intelligent scheduler 100 according to an embodiment of the present application includes: a data acquisition module 110, configured to acquire CPU occupancy, memory occupancy, disk storage, and network bandwidth values at a plurality of predetermined time points within a predetermined period of time; the data processing module 120 is configured to perform time-sequence association analysis on the CPU occupancy, the memory occupancy, the disk storage, and the network bandwidth values at the plurality of predetermined time points to obtain an operation state feature vector; and an operation state detection module 130, configured to determine whether the operation state of the multimedia intelligent scheduler is normal based on the operation state feature vector.
The data collection module 110 can monitor and analyze the operation state of the multimedia intelligent dispatcher by collecting the data, so as to discover and solve the problem in time, and can collect a large amount of data quickly and transmit the data to the data processing module. The data processing module 120 is capable of rapidly processing large amounts of data and extracting useful information. By analyzing the data, the running state of the multimedia intelligent dispatcher can be known, and the problems can be found and solved in time, so that the reliability and stability of the system are improved. The running state detection module 130 can find and solve problems in time, and can ensure the normal running of the multimedia intelligent dispatcher and improve the reliability and stability of the system by detecting the running state.
The multimedia intelligent scheduler 100 monitors and analyzes the CPU occupancy, memory occupancy, disk storage and network bandwidth values at a plurality of predetermined time points through the above three modules, thereby finding and solving problems in time and improving the reliability and stability of the system.
Specifically, the data acquisition module 110 is configured to acquire the CPU occupancy, the memory occupancy, the disk storage, and the network bandwidth value at a plurality of predetermined time points within a predetermined period of time. With the continuous development of multimedia technology, multimedia systems play an increasingly important role in our daily lives. For a large multimedia system, how to perform efficient resource scheduling and monitoring is a very critical issue.
At present, the traditional manual management mode is mainly to monitor and schedule various resources in a multimedia system in a manual mode. However, the manual monitoring and scheduling methods are prone to errors, thereby affecting the normal operation of the entire multimedia system. Meanwhile, in a large-scale multimedia system, the use condition of various resources is difficult to be mastered completely in time and accurately by manual monitoring, and faults and even breakdown caused by monitoring loopholes exist. In addition, the multimedia system has a plurality of resource types needing to be scheduled, has wide involved areas, and has low manual scheduling efficiency. The traditional manual management mode can not realize real-time monitoring and dynamic scheduling, so that the system has no quick response capability. Therefore, an optimized multimedia intelligent scheduler is desired.
The intelligent multimedia dispatcher is one system for managing and dispatching various media resource, and may be used in managing and dispatching different kinds of media resource, such as audio, video, image, etc. Also, the system may collect media assets from multiple sources and store them in a central media library. It can then automatically select and allocate appropriate media resources to meet various needs, such as advertising, information distribution, entertainment, etc., according to the needs of the user. Thus, various resources are efficiently managed through resource allocation and scheduling, and system efficiency and resource utilization are improved.
Furthermore, the multimedia intelligent dispatcher can also provide real-time monitoring and reporting functions so that an administrator can know the running condition and the resource use condition of the system. It should be understood that the system resource utilization refers to the use of various resources (e.g., CPU, memory, disk, network bandwidth, etc.) in the system. If the utilization of some resources is too high, bottlenecks may occur in the system, requiring adjustment.
Based on the above, in the technical scheme of the application, the use condition of the multimedia intelligent scheduler is expected to be detected based on comprehensive analysis of various resources in the utilization rate of system resources, such as CPU, memory, disk and network bandwidth, so as to judge whether the running state of the multimedia intelligent scheduler is normal, thereby ensuring the normal operation of the multimedia intelligent scheduler and improving the stability and performance of a multimedia system.
Specifically, in the technical scheme of the present application, first, the CPU occupancy, the memory occupancy, the disk storage and the network bandwidth value at a plurality of predetermined time points in a predetermined period are obtained. It should be understood that the system resource utilization refers to the use of various resources (e.g., CPU, memory, disk, network bandwidth, etc.) in the system. If the utilization of some resources is too high, bottlenecks may occur in the system, requiring adjustment. Therefore, it is desirable to determine the operating load and operating state of the system based on time series analysis of CPU occupancy, memory occupancy, disk storage, and network bandwidth values. That is, when the system running state is unstable, the current resource usage of the system can be better evaluated by collecting various resource indexes in the system, so as to determine whether adaptive adjustment is needed to optimize the system performance. For example, when high CPU occupation tasks such as video encoding and decoding are performed, the intelligent dispatcher can be helped to judge whether the current system is bottleneck or not and take corresponding measures to avoid system breakdown caused by overlarge load by monitoring CPU occupation rate change; meanwhile, the intelligent dispatcher can be helped to estimate the running time by monitoring data such as disk storage capacity, network bandwidth value and the like.
Specifically, the data processing module 120 is configured to perform timing correlation analysis on the CPU occupancy, the memory occupancy, the disk storage, and the network bandwidth value at the multiple predetermined time points to obtain an operation state feature vector. Fig. 2 is a block diagram of a data processing module in a multimedia intelligent scheduler according to an embodiment of the present application, as shown in fig. 2, the data processing module 120 includes: a data timing arrangement unit 121, configured to arrange the CPU occupancy, the memory occupancy, the disk storage, and the network bandwidth values at the plurality of predetermined time points into a resource multiparameter timing matrix according to a time dimension and a sample dimension; a context resource multi-parameter time sequence associated feature extraction unit 122, configured to perform feature extraction on the resource multi-parameter time sequence matrix to obtain a plurality of context resource multi-parameter time sequence associated feature vectors; and a local association strengthening unit 123, configured to perform local association feature strengthening on the multiple context resource multi-parameter time sequence association feature vectors to obtain the running state feature vector. In the data timing arrangement unit 121, considering that the CPU occupancy rate, the memory occupancy rate, the disk storage amount, and the network bandwidth value all have a dynamic change rule of time sequence in a time dimension, the CPU occupancy rate, the memory occupancy rate, the disk storage amount, and the network bandwidth value at the plurality of predetermined time points are further arranged into a resource multi-parameter time sequence matrix according to the time dimension and the sample dimension, so as to integrate the time sequence distribution information of the CPU occupancy rate, the memory occupancy rate, the disk storage amount, and the network bandwidth value in the time dimension, respectively, to facilitate the subsequent time sequence change correlation feature analysis of various resources in the system and the running state detection of the multimedia intelligent scheduler.
In the resource multi-parameter timing matrix, the time dimension refers to a column of data formed by arranging CPU occupancy, memory occupancy, disk storage and network bandwidth values at a plurality of preset time points according to time sequence. For example, if the data is scheduled to be collected once per minute, the time dimension is the data collected per minute, forming a time series. The arrangement of the time dimension facilitates analysis of the time sequence variation of various resources in the system.
The sample dimension refers to a line of data formed by arranging the CPU occupancy, the memory occupancy, the disk storage and the network bandwidth values at a plurality of predetermined time points in the sample order. For example, if the data of multiple servers is collected, the sample dimension is the data collected by each server, forming a sample sequence. The arrangement of sample dimensions facilitates analysis of the correlation characteristics between different resources.
In one embodiment of the application, 1. CPU occupancy, memory occupancy, disk storage, and network bandwidth values at a plurality of predetermined points in time are arranged in a time dimension to form a time series. 2. And arranging CPU occupancy, memory occupancy, disk storage and network bandwidth values at a plurality of preset time points according to sample dimensions to form a sample sequence. 3. The time series and the sample series are combined to form a resource multi-parameter time sequence matrix.
In another embodiment of the present application, 1. CPU occupancy, memory occupancy, disk storage, and network bandwidth values at a plurality of predetermined points in time are arranged in a time dimension to form a time series. 2. And arranging CPU occupancy, memory occupancy, disk storage and network bandwidth values at a plurality of preset time points according to sample dimensions to form a sample sequence. 3. And combining the time sequence and the sample sequence according to the resource types respectively to form a time sequence matrix of a plurality of resource types. 4. And combining the time sequence matrixes of the plurality of resource types to form a resource multi-parameter time sequence matrix.
The CPU occupancy rate, the memory occupancy rate, the disk storage amount and the network bandwidth value at a plurality of preset time points are arranged into a resource multi-parameter time sequence matrix according to the time dimension and the sample dimension, so that time sequence distribution information of various resources in the time dimension can be conveniently integrated, and time sequence change association characteristic analysis of various resources in the system and running state detection of the multimedia intelligent scheduler are carried out.
Further, in the context resource multi-parameter timing-related feature extraction unit 122, it includes: the matrix dividing subunit is used for carrying out matrix segmentation on the resource multi-parameter time sequence matrix to obtain a plurality of resource multi-parameter local time sequence matrices; and the global associated feature extraction subunit is used for enabling the plurality of resource multi-parameter local time sequence matrixes to pass through a converter module comprising an embedded layer to obtain a plurality of context resource multi-parameter time sequence associated feature vectors.
Wherein, considering that the time sequence characteristic distribution information about various resource parameters exists in the resource multi-parameter time sequence matrix, the time sequence associated characteristic extraction of the various resource parameters is performed by using a convolution neural network model with excellent performance in the aspect of local implicit associated characteristic extraction. In particular, considering that due to the inherent limitations of convolution operations, the pure CNN method can only capture local implicit correlation features, has limited receptive fields, and is difficult to learn explicit global and remote semantic information interactions.
Therefore, in order to capture the time sequence global associated characteristic information of the various resource parameters, the resource multi-parameter time sequence matrix is further subjected to matrix segmentation to obtain a plurality of resource multi-parameter local time sequence matrices. It should be understood that, by performing matrix segmentation on the involved resource multi-parameter timing matrix, the original data can be segmented into a plurality of small local matrices according to a certain rule and method. Therefore, the characteristics of independent and fine granularity of various resource parameters can be better extracted, and complexity and excessive noise interference caused by calculation on the whole matrix are avoided.
In the application, the matrix segmentation is a process of dividing a large matrix into a plurality of small local matrices, so that large-scale calculation tasks can be parallelized, thereby improving the calculation efficiency. The matrix segmentation includes: two modes, namely horizontal segmentation and vertical segmentation.
A horizontal split is a split of rows of a matrix into a plurality of small partial matrices, each partial matrix containing a portion of the rows. For example, an m n matrix is split horizontally into k partial matrices, each containing m/k rows. This manner of segmentation is suitable for computational tasks that are independent of each other between rows.
Vertical slicing is the slicing of the columns of a matrix into a plurality of small partial matrices, each partial matrix containing a portion of the columns. For example, an m×n matrix is vertically split into k partial matrices, each partial matrix containing
n/k columns. This manner of segmentation is suitable for computational tasks that are independent of each other between columns.
Besides horizontal segmentation and vertical segmentation, the method also comprises segmentation according to a block mode, namely, a matrix is segmented into a plurality of block matrixes with equal size, and the segmentation mode is suitable for tasks needing to perform parallel calculation on the whole matrix.
And then, the multiple resource multi-parameter local time sequence matrixes are encoded in a converter module comprising an embedded layer to obtain multiple context resource multi-parameter time sequence associated feature vectors. It should be noted that, here, the converter module including the embedded layer can extract the local time sequence associated features of the various resource parameters based on the global context resource multi-parameter time sequence associated feature information, so as to aggregate the local implicit time sequence associated features of the various resource parameters, so as to facilitate the subsequent running state detection of the multimedia intelligent scheduler.
Specifically, the converter module including an embedded layer includes: and the embedding layer maps each element in each local time sequence matrix into a vector space to obtain an embedding matrix. And the encoder takes the embedded matrix as input, and processes the embedded matrix through a plurality of encoding layers to obtain a context feature vector. And the decoder takes the context feature vector as input, and processes the context feature vector through a plurality of decoding layers to obtain a plurality of context resource multi-parameter time sequence associated feature vectors.
In the encoder, each encoding layer includes a self-attention layer and a feedforward neural network layer. The self-attention layer is used to calculate the degree of association between each element in the input matrix and other elements, thereby capturing context information. In the decoder, each decoding layer also contains a self-attention layer and a feedforward neural network layer. In contrast, the self-attention layer of the decoder considers not only the degree of association between each element in the input matrix and other elements, but also the degree of association between the context feature vector output by the encoder and each element in the input matrix, thereby better capturing the context information.
Through the converter module comprising the embedded layer, the multi-parameter local time sequence matrix of the plurality of resources can be encoded into a plurality of context resource multi-parameter time sequence associated feature vectors, so that the analysis of the time sequence change associated feature of the resources and the detection of the running state of the multimedia intelligent scheduling machine can be better carried out.
In one embodiment of the application, first, an embedding layer is defined in the converter module, and each element in each local timing matrix is mapped into a vector space to obtain an embedding matrix. Then, each element in each local time sequence matrix is mapped into a vector space through an embedding layer to obtain an embedded local time sequence matrix. The embedded local timing matrix is then encoded by an encoder module to obtain a context vector that contains the context information of the local timing matrix. Finally, a plurality of context resource multi-parameter time sequence associated feature vectors are obtained: and sequentially performing embedding and coding operations on the plurality of local time sequence matrixes to obtain a plurality of context vectors, wherein the vectors form a plurality of resource multi-parameter time sequence associated feature vectors.
Further, in the local association strengthening unit 123, it is used to: and after the context resource multi-parameter time sequence associated feature vectors are arranged into one-dimensional feature vectors, the one-dimensional feature vectors are passed through a local associated feature strengthening module comprising a first convolution layer and a second convolution layer to obtain running state feature vectors.
Then, it is also considered that in the actual running state detection process of the multimedia intelligent scheduler, the various resource parameters, such as CPU occupancy, memory occupancy, disk storage and network bandwidth values, have different correlation characteristics under different time period spans and different resource parameter sample spans, and these different local cooperative correlation characteristic information plays an important role in running state detection of the multimedia intelligent scheduler. However, the converter module has excellent performance for long-range dependent correlated feature extraction, and has poor capability for medium-short range dependent correlated time series correlated feature extraction.
Therefore, in the technical scheme of the application, the context resource multi-parameter time sequence associated feature vectors are further arranged into one-dimensional feature vectors and then the one-dimensional feature vectors are processed by the local associated feature strengthening module comprising the first convolution layer and the second convolution layer to obtain the running state feature vectors. In particular, the first convolution layer and the second convolution layer use one-dimensional convolution kernels with different scales to extract multi-scale local time sequence collaborative correlation characteristic information of the various resource parameters at different times and different sample spans. It should be appreciated that features of multiple degrees of abstraction can be progressively extracted through convolution kernels and pooling operations of different sizes, thereby capturing various details and differences in the system operating state. In this way, the dependency relationship in the sample set can be learned, and the local information is reserved and enhanced while the global feature is extracted, so that the running state feature vector with more sufficient feature information is obtained.
Wherein the first convolution layer and the second convolution layer each comprise: and the convolution layer carries out convolution operation by using one-dimensional convolution kernels with different scales to obtain a plurality of convolution feature matrixes with different scales. Pooling layer: and the convolution feature matrix is subjected to pooling operation, and the dimension of the feature matrix is reduced and the calculated amount of the model is reduced by using maximum pooling or average pooling. And a nonlinear activation layer for performing nonlinear transformation on the pooled feature matrix, wherein a ReLU activation function or a variant thereof is generally used for enhancing the nonlinear expression capability of the model.
It should be noted that the input to the second convolution layer is typically the output of the first convolution layer, so the convolution kernel size and number of the second convolution layer may be different from the first convolution layer to extract higher level features. Meanwhile, the pooling and nonlinear activation operations of the second convolution layer can also be different from those of the first convolution layer, so as to further improve the expression capability of the model.
In one embodiment of the present application, a plurality of context resource multi-parameter timing-related feature vectors are arranged as one-dimensional feature vectors, comprising: 1. and arranging the multiple context resource multi-parameter time sequence associated feature vectors according to the time dimension to obtain a time sequence matrix. 2. And arranging the time sequence matrix according to the dimension of the sample to obtain a one-dimensional feature vector. 3. And inputting the one-dimensional feature vector into a local associated feature strengthening module comprising a first convolution layer and a second convolution layer to obtain an operation state feature vector.
More specifically, for a plurality of context resource multi-parameter time sequence associated feature vectors, assuming that the dimension of each feature vector is d, the time span is T, and the number of samples is N, arranging them according to the time dimension to obtain a time sequence matrix with the shape of (T, N, d); arranging the time sequence matrix according to the dimension of the sample to obtain a one-dimensional feature vector with the shape of (N x T, d); and inputting the one-dimensional feature vector into a local associated feature strengthening module comprising a first convolution layer and a second convolution layer to obtain an operation state feature vector. Specifically, the one-dimensional feature vector may be convolved, pooled, and nonlinear activated according to the steps described above to obtain an operational state feature vector with a dimension k.
Specifically, the running state detection module 130 is configured to determine whether the running state of the multimedia intelligent scheduler is normal based on the running state feature vector. Fig. 3 is a block diagram of the running state detection module in the multimedia intelligent scheduler according to an embodiment of the present application, as shown in fig. 3, the running state detection module 130 includes: a feature optimization unit 131, configured to perform feature distribution optimization on the running state feature vector to obtain an optimized running state feature vector; the scheduler operation state evaluation unit 132 is configured to pass the optimized operation state feature vector through a classifier to obtain a classification result, where the classification result is used to indicate whether the operation state of the multimedia intelligent scheduler is normal.
In particular, in the technical solution of the present application, when the plurality of resource multi-parameter local timing matrices are passed through the converter module including the embedded layer to obtain a plurality of context resource multi-parameter timing correlation feature vectors, the plurality of context resource multi-parameter timing correlation feature vectors express context-correlated multi-parameter sample-timing cross correlation features of each resource multi-parameter local timing matrix, and after the plurality of context resource multi-parameter timing correlation feature vectors are arranged into one-dimensional feature vectors and passed through the local correlation feature reinforcement module including the first convolution layer and the second convolution layer, local feature distribution can be further reinforced, but in order to better utilize the context-correlated sample-timing cross correlation features of resource parameters, the running state feature vector is preferably optimized by fusing the one-dimensional feature vectors and the running state feature vectors.
Here, considering that the one-dimensional feature vector is obtained by cascading the plurality of context resource multi-parameter time-sequence association feature vectors, and the running state feature vector is obtained by further extracting local association features under a local association scale on the basis of the one-dimensional feature vector, the one-dimensional feature vector and the running state feature vector both have serialized distribution attributes based on local fragment semantics.
Therefore, in order to promote the fusion effect of the one-dimensional feature vector and the running state feature vector, the applicant of the present application marks the one-dimensional feature vector as, for example, V 1 And the operating state feature vector, e.g. denoted as V 2 Performing segment enrichment fusion of the local sequence semantics to obtain an optimized running state feature vector, for example denoted as V 2 ' specifically expressed as: performing partial sequence semantic segment enrichment fusion on the one-dimensional feature vector and the running state feature vector by using the following optimization formula to obtain the optimized running stateA state feature vector; wherein, the optimization formula is:
wherein V is 1 Is the one-dimensional feature vector, V 2 Is the operating state feature vector, D (V 1 ,V 2 ) For the distance matrix between the one-dimensional feature vector and the operation state feature vector, V 1 And V 2 Are column vectors, and alpha is a weight super parameter,representing vector multiplication, ++>Representing vector addition, V 2 ' is the optimized operational state feature vector.
Here, the partial sequence semantic segment enrichment fuses the coding effect of the sequence-based segment feature distribution on the directional semantics in the predetermined distribution direction of the sequence to embed similarity between sequence segments as a re-weighting factor for inter-sequence association, thereby capturing the similarity between sequences based on the feature representation (feature appearance) at each segment level, realizing the one-dimensional feature vector V 1 And the operating state feature vector V 2 Is fused in a enrichment way, thereby improving the optimized running state feature vector V 2 ' for the one-dimensional feature vector V 1 And the operating state feature vector V 2 To promote the optimized running state feature vector V' 2 Is characterized by the expression of (3). Therefore, the running state of the multimedia intelligent dispatcher can be accurately detected in real time, so that the normal work of the multimedia intelligent dispatcher is ensured, and the stability and performance of a multimedia system are improved.
And further, classifying the optimized running state feature vector in a classifier to obtain a classification result, wherein the classification result is used for indicating whether the running state of the multimedia intelligent scheduler is normal or not. That is, in the technical solution of the present application, the labels of the classifier include a normal running state (first label) of the multimedia intelligent scheduler and an abnormal running state (second label) of the multimedia intelligent scheduler, where the classifier determines to which classification label the running state feature vector belongs through a soft maximum function.
It should be noted that the first tag p1 and the second tag p2 do not include a concept set by human, and in fact, during the training process, the computer model does not have a concept of "whether the operation state of the multimedia intelligent dispatcher is normal", which is simply that there are two kinds of classification tags and the probability that the output feature is under the two kinds of classification tags, that is, the sum of p1 and p2 is one. Therefore, the classification result of whether the running state of the multimedia intelligent dispatcher is normal is actually converted into the classification probability distribution conforming to the natural rule through classifying the labels, and the physical meaning of the natural probability distribution of the labels is essentially used instead of the language text meaning of whether the running state of the multimedia intelligent dispatcher is normal.
It should be understood that, in the technical scheme of the present application, the classification label of the classifier is a detection evaluation label for whether the operation state of the multimedia intelligent scheduler is normal, so that after the classification result is obtained, the operation state of the multimedia intelligent scheduler can be detected based on the classification result to ensure the normal operation of the multimedia intelligent scheduler.
A classifier is a machine learning model that maps input data into predefined categories. In this case, the input optimized operation state feature vector is mapped to one of two categories, indicating whether the operation state of the multimedia intelligent scheduler is normal.
The structure of a classifier is generally composed of three main components: an input layer, a hidden layer, and an output layer. The input layer receives the optimized running state feature vector as input, the hidden layer is a set of neurons and is used for processing and extracting features of input data, and the output layer maps the input data into a required category according to the result of the hidden layer. Common classifiers include Support Vector Machines (SVMs), decision trees, random forests, neural networks, and the like. Among them, neural networks are a powerful classifier because it can automatically learn the features in the input data and optimize model parameters during training to improve classification performance.
In one embodiment of the present application, the scheduler operation state evaluation unit 132 includes: the full-connection coding subunit is used for carrying out full-connection coding on the optimized running state feature vector by using a plurality of full-connection layers of the classifier so as to obtain a coding classification feature vector; and the classification subunit is used for passing the coding classification feature vector through a Softmax classification function of the classifier to obtain the classification result.
In another embodiment of the present application, passing the optimized operation state feature vector through a classifier to obtain a classification result may further include: data preparation, wherein the optimized running state feature vector and the corresponding label (normal or abnormal) form a training set and a testing set. Feature extraction, namely carrying out feature extraction and sequence modeling on each time segment by using a segment type enrichment fusion technology of local sequence semantics to obtain richer local sequence semantic information, and fusing the local sequence semantic features of all the time segments to obtain a complete optimized running state feature vector. Classifier training, training the classifier using the training set. Common classifiers include support vector machines, decision trees, random forests, neural networks, and the like. During the training process, the classifier updates the model parameters through a back propagation algorithm to minimize classification errors and improve classification performance. And testing the classifier by using the test set to evaluate the performance of the classifier, inputting the feature vector of the optimized running state into the classifier, and outputting a corresponding classification result by the classifier. And judging the running state, and judging whether the running state of the multimedia intelligent dispatcher is normal or not according to the classification result. If the classification result is normal, the operation state of the multimedia intelligent dispatcher is normal; if the classification result is abnormal, the abnormal running state of the multimedia intelligent dispatcher is indicated.
The running state of the intelligent multimedia dispatcher can be monitored and judged in real time by passing the optimized running state feature vector through the classifier to obtain a classification result, abnormal conditions can be processed in time, and the running efficiency and stability of the intelligent multimedia dispatcher are improved.
In summary, the multimedia intelligent scheduler 100 according to the embodiment of the present application is illustrated, and detects the use condition of the multimedia intelligent scheduler through comprehensive analysis, so as to determine whether the operation state of the multimedia intelligent scheduler is normal, thereby ensuring the normal operation of the multimedia intelligent scheduler and improving the stability and performance of the multimedia system.
In one embodiment of the present application, fig. 4 is a flowchart of a multimedia intelligent scheduling method provided in the embodiment of the present application. Fig. 5 is a schematic diagram of a system architecture of a multimedia intelligent scheduling method according to an embodiment of the present application. As shown in fig. 4 and 5, a multimedia intelligent scheduling method includes: 210, acquiring CPU occupancy rate, memory occupancy rate, disk storage amount and network bandwidth value of a plurality of preset time points in a preset time period; 220, performing time sequence association analysis on CPU occupancy rate, memory occupancy rate, disk storage amount and network bandwidth value of the plurality of preset time points to obtain an operation state feature vector; and 230, determining whether the running state of the multimedia intelligent scheduler is normal or not based on the running state feature vector.
In a specific example of the present application, in the above-mentioned multimedia intelligent scheduling method, performing time-series association analysis on CPU occupancy, memory occupancy, disk storage, and network bandwidth values at the plurality of predetermined time points to obtain an operation state feature vector includes: arranging the CPU occupancy rate, the memory occupancy rate, the disk storage amount and the network bandwidth value of the plurality of preset time points into a resource multi-parameter time sequence matrix according to the time dimension and the sample dimension; extracting features of the resource multi-parameter time sequence matrix to obtain a plurality of context resource multi-parameter time sequence associated feature vectors; and carrying out local association feature reinforcement on the context resource multi-parameter time sequence association feature vectors to obtain the running state feature vector.
It will be appreciated by those skilled in the art that the specific operation of each step in the above-described multimedia intelligent scheduling method has been described in detail in the above description of the multimedia intelligent scheduler with reference to fig. 1 to 3, and thus, repeated descriptions thereof will be omitted.
Fig. 6 is an application scenario diagram of a multimedia intelligent scheduler provided in an embodiment of the present application. As shown in fig. 6, in the application scenario, first, a CPU occupancy (e.g., C1 as illustrated in fig. 6), a memory occupancy (e.g., C2 as illustrated in fig. 6), a disk storage (e.g., C3 as illustrated in fig. 6), and a network bandwidth value (e.g., C4 as illustrated in fig. 6) of a multimedia intelligent scheduler (e.g., M as illustrated in fig. 6) of a plurality of predetermined time points within a predetermined period of time are acquired, and in a specific example, the model of the multimedia intelligent scheduler is PPHO-236DST-V1; then, the acquired CPU occupancy, memory occupancy, disk storage, and network bandwidth value are input to a server (e.g., S as illustrated in fig. 6) where a multimedia intelligent scheduling algorithm is deployed, wherein the server is capable of processing the CPU occupancy, memory occupancy, disk storage, and network bandwidth value based on the multimedia intelligent scheduling algorithm to determine whether the operation state of the multimedia intelligent scheduler is normal.
The foregoing description of the embodiments has been provided for the purpose of illustrating the general principles of the application, and is not meant to limit the scope of the application, but to limit the application to the particular embodiments, and any modifications, equivalents, improvements, etc. that fall within the spirit and principles of the application are intended to be included within the scope of the application.
Claims (10)
1. A multimedia intelligent scheduler, comprising:
the data acquisition module is used for acquiring CPU occupancy rate, memory occupancy rate, disk storage amount and network bandwidth value of a plurality of preset time points in a preset time period;
the data processing module is used for carrying out time sequence association analysis on the CPU occupancy rate, the memory occupancy rate, the disk storage amount and the network bandwidth value at a plurality of preset time points so as to obtain an operation state feature vector; and
and the running state detection module is used for determining whether the running state of the multimedia intelligent dispatcher is normal or not based on the running state feature vector.
2. The multimedia intelligent scheduler of claim 1, wherein the data processing module comprises:
the data time sequence arrangement unit is used for arranging the CPU occupancy rate, the memory occupancy rate, the disk storage amount and the network bandwidth value of the plurality of preset time points into a resource multi-parameter time sequence matrix according to the time dimension and the sample dimension;
the context resource multi-parameter time sequence associated feature extraction unit is used for extracting features of the resource multi-parameter time sequence matrix to obtain a plurality of context resource multi-parameter time sequence associated feature vectors; and
and the local association strengthening unit is used for carrying out local association feature strengthening on the multiple context resource multi-parameter time sequence association feature vectors so as to obtain the running state feature vector.
3. The multimedia intelligent scheduler of claim 2, wherein the context resource multi-parameter timing-related feature extraction unit comprises:
the matrix dividing subunit is used for carrying out matrix segmentation on the resource multi-parameter time sequence matrix to obtain a plurality of resource multi-parameter local time sequence matrices;
and the global associated feature extraction subunit is used for enabling the plurality of resource multi-parameter local time sequence matrixes to pass through a converter module comprising an embedded layer to obtain a plurality of context resource multi-parameter time sequence associated feature vectors.
4. A multimedia intelligent scheduler according to claim 3, wherein the local association strengthening unit is configured to: and after the context resource multi-parameter time sequence associated feature vectors are arranged into one-dimensional feature vectors, the one-dimensional feature vectors are passed through a local associated feature strengthening module comprising a first convolution layer and a second convolution layer to obtain running state feature vectors.
5. The intelligent multimedia scheduler of claim 4, wherein the first convolution layer and the second convolution layer of the local correlation feature enhancement module each employ one-dimensional convolution kernels of different scales.
6. The multimedia intelligent scheduler of claim 5, wherein the run state detection module comprises:
the feature optimization unit is used for performing feature distribution optimization on the running state feature vector to obtain an optimized running state feature vector;
the scheduler operation state evaluation unit is used for enabling the optimized operation state feature vector to pass through the classifier to obtain a classification result, and the classification result is used for indicating whether the operation state of the multimedia intelligent scheduler is normal or not.
7. The intelligent multimedia scheduler of claim 6, wherein the feature optimization unit is configured to:
carrying out partial sequence semantic fragment enrichment fusion on the one-dimensional feature vector and the running state feature vector by using the following optimization formula to obtain the optimized running state feature vector;
wherein, the optimization formula is:
wherein V is 1 Is the one-dimensional feature vector, V 2 Is the operating state feature vector, D (V 1 ,V 2 ) For the one-dimensional feature vector and the operating state feature vectorDistance matrix between V 1 And V 2 Are column vectors, and alpha is a weight super parameter,representing vector multiplication, ++>Representing vector addition, V 2 ' is the optimized operational state feature vector.
8. The multimedia intelligent scheduler of claim 7, wherein the scheduler operating state evaluation unit comprises:
the full-connection coding subunit is used for carrying out full-connection coding on the optimized running state feature vector by using a plurality of full-connection layers of the classifier so as to obtain a coding classification feature vector; and
and the classification subunit is used for passing the coding classification feature vector through a Softmax classification function of the classifier to obtain the classification result.
9. The intelligent multimedia scheduling method is characterized by comprising the following steps of:
acquiring CPU occupancy rate, memory occupancy rate, disk storage amount and network bandwidth value of a plurality of preset time points in a preset time period;
performing time sequence association analysis on CPU occupancy rate, memory occupancy rate, disk storage amount and network bandwidth value of the plurality of preset time points to obtain an operation state feature vector; and
and determining whether the running state of the multimedia intelligent dispatcher is normal or not based on the running state feature vector.
10. The method of claim 9, wherein performing timing correlation analysis on the CPU occupancy, memory occupancy, disk storage, and network bandwidth values at the plurality of predetermined time points to obtain the running state feature vector comprises:
arranging the CPU occupancy rate, the memory occupancy rate, the disk storage amount and the network bandwidth value of the plurality of preset time points into a resource multi-parameter time sequence matrix according to the time dimension and the sample dimension;
extracting features of the resource multi-parameter time sequence matrix to obtain a plurality of context resource multi-parameter time sequence associated feature vectors; and
and carrying out local correlation characteristic reinforcement on the multiple context resource multi-parameter time sequence correlation characteristic vectors to obtain the running state characteristic vectors.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310756083.6A CN117215766A (en) | 2023-06-26 | 2023-06-26 | Multimedia intelligent dispatcher and method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310756083.6A CN117215766A (en) | 2023-06-26 | 2023-06-26 | Multimedia intelligent dispatcher and method thereof |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117215766A true CN117215766A (en) | 2023-12-12 |
Family
ID=89049894
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310756083.6A Pending CN117215766A (en) | 2023-06-26 | 2023-06-26 | Multimedia intelligent dispatcher and method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117215766A (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111368980A (en) * | 2020-03-06 | 2020-07-03 | 京东数字科技控股有限公司 | State detection method, device, equipment and storage medium |
| CN113379033A (en) * | 2021-06-11 | 2021-09-10 | 国网湖北省电力有限公司黄石供电公司 | Cable health state intelligent early warning method based on time-space network enhanced deep learning |
| CN115373813A (en) * | 2022-04-03 | 2022-11-22 | 福建福清核电有限公司 | Scheduling method and system based on GPU virtualization in cloud computing environment and electronic equipment |
| CN115564203A (en) * | 2022-09-23 | 2023-01-03 | 杭州国辰智企科技有限公司 | Equipment real-time performance evaluation system and method based on multi-dimensional data cooperation |
| CN115878319A (en) * | 2022-12-13 | 2023-03-31 | 中国计量大学 | Load balancing method and system and electronic equipment |
| CN116257406A (en) * | 2023-03-15 | 2023-06-13 | 湖南华彩伟业网络科技有限公司 | Gateway data management method and system for smart city |
| CN116319599A (en) * | 2023-02-20 | 2023-06-23 | 养哇(南京)科技有限公司 | Channel distribution system for bearing VOIP service and method thereof |
-
2023
- 2023-06-26 CN CN202310756083.6A patent/CN117215766A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111368980A (en) * | 2020-03-06 | 2020-07-03 | 京东数字科技控股有限公司 | State detection method, device, equipment and storage medium |
| CN113379033A (en) * | 2021-06-11 | 2021-09-10 | 国网湖北省电力有限公司黄石供电公司 | Cable health state intelligent early warning method based on time-space network enhanced deep learning |
| CN115373813A (en) * | 2022-04-03 | 2022-11-22 | 福建福清核电有限公司 | Scheduling method and system based on GPU virtualization in cloud computing environment and electronic equipment |
| CN115564203A (en) * | 2022-09-23 | 2023-01-03 | 杭州国辰智企科技有限公司 | Equipment real-time performance evaluation system and method based on multi-dimensional data cooperation |
| CN115878319A (en) * | 2022-12-13 | 2023-03-31 | 中国计量大学 | Load balancing method and system and electronic equipment |
| CN116319599A (en) * | 2023-02-20 | 2023-06-23 | 养哇(南京)科技有限公司 | Channel distribution system for bearing VOIP service and method thereof |
| CN116257406A (en) * | 2023-03-15 | 2023-06-13 | 湖南华彩伟业网络科技有限公司 | Gateway data management method and system for smart city |
Non-Patent Citations (1)
| Title |
|---|
| 曲星宇;曾鹏;李俊鹏;: "基于RNN-LSTM的磨矿系统故障诊断技术", 信息与控制, no. 02, 15 April 2019 (2019-04-15) * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103299324B (en) | Potential son is used to mark the mark learnt for video annotation | |
| CN106651057B (en) | Mobile terminal user age prediction method based on installation package sequence list | |
| CN108563739B (en) | Weather data acquisition method and device, computer device and readable storage medium | |
| CN110598620B (en) | Deep neural network model-based recommendation method and device | |
| CN111310041B (en) | Image-text publishing method, model training method and device and storage medium | |
| CN111930526B (en) | Load prediction method, load prediction device, computer equipment and storage medium | |
| CN110990121B (en) | Kubernetes scheduling strategy based on application portraits | |
| CN110751191A (en) | Image classification method and system | |
| CN111931809A (en) | Data processing method and device, storage medium and electronic equipment | |
| CN115905528A (en) | Event multi-label classification method and device with time sequence characteristics and electronic equipment | |
| CN115294397A (en) | Classification task post-processing method, device, equipment and storage medium | |
| CN111339052A (en) | Unstructured log data processing method and device | |
| CN117197722B (en) | User perception and analysis system based on mobile internet video | |
| CN113569955A (en) | Model training method, user portrait generation method, device and equipment | |
| CN115982646B (en) | Management method and system for multisource test data based on cloud platform | |
| CN112949778A (en) | Intelligent contract classification method and system based on locality sensitive hashing and electronic equipment | |
| CN111950623A (en) | Data stability monitoring method and device, computer equipment and medium | |
| CN117215766A (en) | Multimedia intelligent dispatcher and method thereof | |
| CN115438658A (en) | Entity recognition method, recognition model training method and related device | |
| CN112183752B (en) | End-to-end multi-example learning method based on automatic example selection | |
| CN115168609A (en) | Text matching method and device, computer equipment and storage medium | |
| CN114399407A (en) | Power dispatching monitoring data anomaly detection method based on dynamic and static selection integration | |
| CN117724858B (en) | Geospatial analysis model scheduling method, device, equipment and medium | |
| CN110837894A (en) | Feature processing method and device and storage medium | |
| CN116450485B (en) | Detection method and system for application performance interference |
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 | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20240207 Address after: Room 415, 4th Floor, Building B, Yanxiang Technology Building, No. 333 Jianghong Road, Changhe Street, Binjiang District, Hangzhou City, Zhejiang Province, 310000 (self declared) Applicant after: Hangzhou Jiujiu Kechuang Technology Co.,Ltd. Country or region after: China Address before: 311100, 3rd Floor, Building 1, No. 163 Wuchang Avenue, Wuchang Street, Yuhang District, Hangzhou City, Zhejiang Province Applicant before: HANGZHOU XIBEI TECHNOLOGY Co.,Ltd. Country or region before: China |
|
| TA01 | Transfer of patent application right |