CN-122027630-A - Multi-platform interactive data transmission system

CN122027630ACN 122027630 ACN122027630 ACN 122027630ACN-122027630-A

Abstract

The invention provides a multi-platform interactive data transmission system, which relates to the technical field of data transmission and comprises edge node pools which are arranged at corresponding platforms one by one, wherein each edge node pool consists of a plurality of edge nodes with coding and decoding functions, an edge node sequencing module for sequencing the priorities of the edge nodes, a transmission route construction module for constructing transmission routes based on the priority sequencing, an anomaly monitoring model for monitoring coding and decoding processes, and a transmission route adjustment module for changing the edge nodes when the coding and decoding are abnormal.

Inventors

Hu Guangtie
Pang Tuzhou
YAO ZIZHONG
HU XIAOJUN
WEI ZHENDONG

Assignees

深圳市爱租机科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260210

Claims (8)

1. A multi-platform interactive data transmission system, comprising: The data processing module consists of a plurality of edge node pools which are arranged at the corresponding platform one by one, the edge node pools consist of a plurality of edge nodes, and encoders for converting original platform data into standard platform data and decoders for converting the standard platform data into new original platform data are arranged on the edge nodes; The intermediate server is used for realizing data transmission among different edge node pools; The edge node ordering module is used for determining a data sending platform and a data receiving platform according to the interactive instruction, calculating a coding error rate and a decoding error rate according to state parameters of the edge nodes in a current time period, further selecting a first edge node group for the data sending platform and a second edge node group for the data receiving platform, and respectively carrying out priority ordering on the edge nodes in the first edge node group and the second edge node group by combining the state parameters, the coding error rate and the decoding error rate, wherein the state parameters comprise CPU temperature, CPU use rate, memory size, network bandwidth and adapter voltage of the edge nodes; The transmission route construction module is used for constructing a transmission route by taking the edge node with the highest priority in the first edge node group, the intermediate server and the edge node with the highest priority in the second edge node group as intermediate nodes so as to realize data transmission between the data sending platform and the data receiving platform; The data transmission system comprises an anomaly monitoring model, a data transmission module and a data decoding module, wherein the anomaly monitoring model is used for monitoring a data transmission process based on a transmission route so as to determine a data transmission state, and the data transmission state comprises a normal state, data coding anomalies and data decoding anomalies; and the transmission route adjustment module is used for replacing the edge node with the highest priority in the first edge node group with the next-level edge node when the data coding is abnormal, and replacing the edge node with the highest priority in the second edge node group with the next-level edge node when the data decoding is abnormal.
2. The multi-platform interactive data transmission system according to claim 1, wherein the encoder converts the original platform data into standard platform data as follows: 1, 1) carrying out format conversion on received original platform data so as to convert the format of the original platform data into a JSON format; 1.2 Preprocessing the original platform data after format conversion, wherein the preprocessing comprises standardization processing, abnormal value elimination and missing value supplement; 1.3 Compressing the preprocessed data based on a Gzip compression algorithm to obtain standard platform data.
3. The multi-platform interactive data transmission system according to claim 1, wherein the decoder converts standard platform data into new original platform data as follows: 2.1 Decompression is carried out on the standard platform data based on the Gzip compression algorithm, and anti-standardization processing is carried out on the decompressed standard platform data; 2.2 Format conversion is carried out on the standard platform data after the inverse standardization processing to obtain new original platform data, and the new original platform data meets the data format requirement of the platform corresponding to the edge node where the encoder is located.
4. The multi-platform interactive data transmission system according to claim 1, wherein the method for obtaining the coding error rate is as follows: 3.1 Acquiring a plurality of pre-coding data points and corresponding coded data points of known labels to form a first sample set, wherein the labels are correct codes or errors, constructing a coding abnormality detection model which is input into the pre-coding data points and the coded data points and is output into the correct codes or the errors based on a support vector machine, and training the coding abnormality detection model based on the first sample set; 3.2 Acquiring state parameter time sequence data of the encoder in each historical time period, wherein the state parameter time sequence data comprises CPU temperature time sequence data, CPU utilization rate time sequence data, memory size time sequence data, network bandwidth time sequence data and adapter voltage time sequence data, acquiring pre-coding data points and post-coding data points of the encoder in the corresponding historical time period, inputting the pre-coding data points and the post-coding data points into a coding abnormality detection model to obtain the total number of coding errors of the encoder in the corresponding historical time period, comparing the total number of coding errors with the total number of pre-coding data points to obtain the coding error rate of the encoder in the corresponding historical time period, and marking each coding error rate on the state parameter time sequence data in the corresponding historical time period to form a second sample set; 3.3 Constructing time sequence data with state parameters based on a long-period memory network, outputting a coding error rate prediction model with a coding error rate, and training the coding error rate prediction model based on a second sample set; 3.4 Acquiring state parameter time sequence data of each edge node corresponding to the data sending platform in the current time period, and inputting the state parameter time sequence data into the coding error rate prediction model after training is completed to obtain the coding error rate of each edge node corresponding to the data sending platform.
5. The multi-platform interactive data transmission system according to claim 4, wherein the decoding error rate obtaining method specifically comprises the following steps: 4.1 Acquiring a plurality of pre-decoding data points and corresponding data points after decoding of a known label to form a third sample set, wherein the label is correct in decoding or incorrect in decoding, constructing a decoding anomaly detection model which is input as the pre-decoding data points and the post-decoding data points and is output as correct in decoding or incorrect in decoding based on a support vector machine, and training the decoding anomaly detection model based on the third sample set; 4.2 Acquiring state parameter time sequence data of the decoder in each historical time period, wherein the state parameter time sequence data comprises CPU temperature time sequence data, CPU utilization rate time sequence data, memory size time sequence data, network bandwidth time sequence data and adapter voltage time sequence data, acquiring pre-decoding data points and post-decoding data points of the decoder in the corresponding historical time period, inputting the pre-decoding data points and the post-decoding data points into a decoding abnormality detection model to obtain the total number of decoding errors of the decoder in the corresponding historical time period, comparing the total number of decoding errors with the total number of pre-decoding data points to obtain the decoding error rate of the decoder in the corresponding historical time period, and marking each decoding error rate on the state parameter time sequence data in the corresponding historical time period to form a fourth sample set; 4.3 Constructing time sequence data with state parameters based on a long-period memory network, outputting a decoding error rate prediction model with decoding error rate, and training the decoding error rate prediction model based on a fourth sample set; 4.4 Acquiring state parameter time sequence data of each edge node corresponding to the data receiving platform in the current time period, and inputting the state parameter time sequence data into a decoding error rate prediction model after training is completed to obtain the decoding error rate of each edge node corresponding to the data sending platform.
6. The multi-platform interactive data transmission system according to claim 5, wherein the anomaly monitoring model is composed of a coding anomaly detection model and a decoding anomaly detection model.
7. The multi-platform interactive data transmission system according to claim 1, wherein the frame selection logic of the first edge node group is that the statistics data send-out platform corresponds to the coding error rate of each edge node in the edge node pool, and the edge nodes with the coding error rate smaller than the coding error threshold are summarized to form the first edge node group; the frame selection logic of the second edge node group is used for counting the decoding error rate of each edge node in the edge node pool corresponding to the data receiving platform and summarizing the edge nodes with the decoding error rate smaller than the decoding error threshold value to form the second edge node group.
8. The multi-platform interactive data transmission system of claim 1, wherein the specific logic for prioritizing edge nodes in the first edge node group and the second edge node group is as follows: 5.1 Data processing is carried out on the state parameters acquired by each edge node in the current time period to obtain the characteristic parameters of each edge node in the current time period, wherein the characteristic parameters comprise a CPU temperature average value, a CPU temperature standard deviation, a CPU temperature maximum value, a CPU temperature minimum value, a CPU utilization rate current value, a memory size current value, a network bandwidth average value, a network bandwidth standard deviation, an adapter voltage average value, an adapter voltage standard deviation, an adapter voltage maximum value and an adapter voltage minimum value of each edge node in the current time period; 5.2 Based on the CPU temperature mean value, the CPU temperature standard deviation and the CPU temperature maximum value, generating a temperature evaluation parameter for evaluating the temperature condition of the edge node, wherein the calculation formula is as follows: In the formula, Respectively represent the first CPU maximum allowable temperature, CPU minimum allowable temperature, CPU proper working temperature of each edge node, An index of edge nodes within the first edge node group or the second edge node group, Represent the first The CPU temperature maximum of the individual edge nodes, Represent the first The high Wen Fengxian coefficients of the individual edge nodes, Represent the first The CPU temperature minimum of the individual edge nodes, Represent the first The low temperature risk coefficient of the individual edge nodes, 、 Respectively represent the first The CPU temperature mean and CPU temperature standard deviation of the individual edge nodes, Is the first The temperature evaluation parameters of the individual edge nodes during the current time period, 、、 Are all preset weights, and And (2) and ; 5.3 Based on the adapter voltage mean, adapter voltage standard deviation, adapter voltage maximum, and adapter voltage minimum, generating a voltage evaluation parameter for evaluating the edge node voltage condition, the calculation formula is as follows: In the formula, 、 Respectively the first The upper voltage input limit and the lower voltage input limit of each edge node, Is the first The adapter voltage average of the individual edge nodes over the current time period, Is the first The voltage deviation value of the individual edge nodes at the current time period, 、、 Respectively the first The adapter voltage standard deviation, the adapter voltage maximum value, the adapter voltage minimum value of each edge node within the current time period, Is the first Voltage evaluation parameters of the edge nodes in the current time period; 5.4 Based on the temperature evaluation parameter, the voltage evaluation parameter, the current value of CPU usage, the current value of memory size, the average value of network bandwidth, the standard deviation of network bandwidth, the coding error rate and the decoding error rate, the coding suitability parameter and the decoding suitability parameter are generated, the coding suitability parameter is used for evaluating the suitability of the edge node for coding work, the decoding suitability parameter is used for evaluating the suitability of the edge node for decoding work, and the calculation formula of the coding suitability parameter is as follows: In the formula, 、、、、 Respectively represent the first The CPU utilization current value, the memory size current value, the network bandwidth mean value, the network bandwidth standard deviation and the coding error rate of each edge node, Represent the first The coding of the individual edge nodes is adapted to the parameters, Are all preset weights, and The specific value of (2) is determined by an analytic hierarchy process; the calculation formula of the decoding suitable parameters is as follows: In the formula, Respectively represent the first The decoding error rate of the individual edge nodes, Represent the first The decoding suitability parameters of the individual edge nodes, Are all preset weights, and The specific value of (2) is determined by an analytic hierarchy process; 5.5 The priority of the edge nodes in the first edge node group is ordered according to the order of the coding suitability parameters from the big to the small, the earlier the ordering is, the higher the priority of the decoding work based on the edge node is, and sequencing the priority of the edge nodes in the second edge node group according to the sequence of the decoding proper parameters from big to small, wherein the higher the sequencing is, the higher the priority of the decoding work based on the edge node is.

Description

Multi-platform interactive data transmission system Technical Field The invention relates to the technical field of data transmission, in particular to a multi-platform interactive data transmission system. Background With the rapid development of information technology, data plays an increasingly important role in various industries, namely finance, medical treatment, education and electronic commerce, and collection, storage, processing and transmission of data are important tasks of enterprises and institutions. In this context, multi-platform data interactions are becoming more and more common, but the data problems that follow are also becoming more and more serious. In the prior art, the multi-platform data interaction safety supervision system (main classification number G06F) based on the artificial intelligence with the publication number of CN119227101A comprises a monitoring center, wherein the monitoring center is in communication connection with a data acquisition and integration module, an artificial intelligence analysis and detection module, a data encryption and protection module, an identity verification and data interaction module and a real-time monitoring and early warning module, acquires interaction data of a plurality of platforms and visitor identity data, performs preprocessing operation on the interaction data to obtain data to be interacted, builds a data analysis and detection standard model, performs analysis and detection operation on the data to be interacted to obtain interactable data, performs data encryption operation on the interactable data to obtain encrypted interaction data, performs identity verification operation on visitor identity data through identity verification, performs multi-platform data interaction operation on the encrypted interaction data, evaluates the safety state, and sends early warning and notification according to the corresponding safety state, thereby greatly improving the data safety. However, the prior art still has a great defect, such as inconsistent data formats among different platforms, so that in the process of multi-platform data interaction, an encoder and a decoder are often arranged at a transmission node to perform data format conversion, but the transmission node may have the problem of encoding errors and decoding errors in the encoding and decoding processes, the technology lacks effective monitoring on the encoding and decoding processes, cannot effectively early warn when encoding or decoding errors occur, has the problem of data distortion but is unknown, and adopts single fixed path transmission when data transmission occurs, and shutdown maintenance is required when encoding or decoding errors occur, so that the multi-platform data interaction efficiency is greatly reduced. The above information disclosed in the background section is only for enhancement of understanding of the background of the disclosure and therefore it may include information that does not form the prior art that is already known to a person of ordinary skill in the art. Disclosure of Invention The present invention is directed to a multi-platform interactive data transmission system, so as to solve the problems set forth in the background art. In order to achieve the above purpose, the present invention provides the following technical solutions: A multi-platform interactive data transmission system, comprising: The data processing module consists of a plurality of edge node pools which are arranged at the corresponding platform one by one, the edge node pools consist of a plurality of edge nodes, and encoders for converting original platform data into standard platform data and decoders for converting the standard platform data into new original platform data are arranged on the edge nodes; The intermediate server is used for realizing data transmission among different edge node pools; The edge node ordering module is used for determining a data sending platform and a data receiving platform according to the interactive instruction, calculating a coding error rate and a decoding error rate according to state parameters of the edge nodes in a current time period, further selecting a first edge node group for the data sending platform and a second edge node group for the data receiving platform, and respectively carrying out priority ordering on the edge nodes in the first edge node group and the second edge node group by combining the state parameters, the coding error rate and the decoding error rate, wherein the state parameters comprise CPU temperature, CPU use rate, memory size, network bandwidth and adapter voltage of the edge nodes; The transmission route construction module is used for constructing a transmission route by taking the edge node with the highest priority in the first edge node group, the intermediate server and the edge node with the highest priority in the second edge node group as intermediate nodes so as to realize data transmis