Data loss state analysis system for network transceiver ends

Abstract

The application relates to a data loss state analysis system for network transmitting and receiving ends, which comprises a customized learning device for establishing an AI prediction model for predicting the loss state of a file segment during the transmission process from a set network transmitting end to a set network receiving end; and a prediction judgment mechanism for taking a latest received file at the set network receiving end as a target file, intercepting a file segment with a length equal to a set character length from the target file as a reference file segment, and predicting the loss degree of the reference file segment by using the AI prediction model. The data loss state analysis system for network transmitting and receiving ends is simple to operate and compact in structure. Since the AI prediction model for predicting the loss state of the file segment during the network transmission process is established between the fixed network transmitting and receiving ends by using the targeted learning mechanism, the overall content of multiple loads transmitted by multiple data packets can be analyzed and judged in a complete state.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of network, in particular to a data loss state analysis system for network transmitting and receiving ends. BACKGROUND

[0002] A computer can only communicate normally with other computers on the network to achieve data transmission at the network transmitting and receiving ends on the premise of complying with network protocols. Network protocols are usually divided into several layers, each layer completes its own function. The two parties in communication can only contact each other at the common level. Common protocols include TCP / IP protocol, IPX / SPX protocol, NetBEUI protocol, etc. IPX / SPX is more commonly used in LAN. If a user accesses the Internet, TCP / IP protocol must be added to the network protocol.

[0003] At the network receiving end, data needs to be extracted from each received TCP data packet, restored in the order before sending, and checked. If errors are found, TCP will require retransmission. For ordinary users, they do not need to understand the entire structure of the network protocol, and only need to understand the IP address format to communicate with the world through the network.

[0004] File data composed of multiple characters is one of the main transmission contents between the network transmitting and receiving ends. However, in the actual network transmission process, TCP / IP protocol can only judge the transmission error, i.e. the transmission error rate, of the short file content transmitted in each data packet, and cannot perform integrity judgment on the file fragments composed of multiple loads corresponding to multiple data packets received in succession, resulting in that the network receiving end can only determine whether the received characters in each data packet are error-free, and cannot determine whether the entire file fragment is complete and the degree of character loss of the file fragment when it is incomplete. The existing technology with publication number CN115883540A provides a file transmission method and system based on a distributed file storage system, which includes: a first client configures an integrity check code for a pre-read file directory and a corresponding initial file to generate a target file; a second client receives the target file through a distributed file storage system and stores the downloaded target file in a file directory to be downloaded; and the second client checks the integrity of the downloaded target file according to the integrity check code. This method has the technical effect of improving the integrity and stability of file transmission to a certain extent. SUMMARY

[0005] In order to overcome the technical defects in the prior art, the application provides a data loss state analysis system for network receiving and transmitting ends, which can adopt a targeted learning mechanism to establish an AI prediction model for predicting the loss state of file segments in the network transmission process between the fixed set network receiving end and the set network transmitting end, thereby analyzing and judging the complete state of the overall content of multiple loads transmitted by multiple data packets, and avoiding falling into a complex and tedious load comparison information processing process of each data packet.

[0006] The application provides a data loss state analysis system for network receiving and transmitting ends, which comprises:

[0007] Batch extraction device, used for extracting file segments of a set character length in batches at the set network receiving end, the extraction of each batch obtains a corresponding single file segment, and the number of batches is positively correlated with the value of the set character length, each file segment is from the set network transmitting end and has a set character length which is obtained by statistics at the set network receiving end;

[0008] State acquisition device, connected with the batch extraction device, used for acquiring state data of each extracted file segment, wherein the state data comprises the complete identification of the corresponding single file segment and the number of missing characters;

[0009] Custom learning device, connected with the state acquisition device and the batch extraction device respectively, used for establishing an AI prediction model for predicting the loss state of file segments in the transmission process from the set network transmitting end to the set network receiving end, wherein the AI prediction model is a BP neural network after a set number of learning, and the value of the set number is equal to the value of the number of batches;

[0010] Prediction and judgment mechanism, connected with the custom learning device, used for taking a file with uncertain state data newly received from the set network transmitting end at the set network receiving end as a target file, intercepting a file segment with a length equal to the set character length from the target file as a reference file segment, and using the AI prediction model to predict the state data of the reference file segment based on the respective ASCII code values of each constituent character of the reference file segment, the actual geographical distance from the set network transmitting end to the set network receiving end, the transmission bandwidth and the maximum operation amount per unit time of the set network transmitting end, and the transmission bandwidth and the maximum operation amount per unit time of the set network receiving end;

[0011] Loss warning mechanism, connected with the prediction and judgment mechanism, used for performing corresponding loss warning actions when the complete identification in the predicted state data of the reference file segment indicates that the reference file segment is incomplete;

[0012] The state data of each extracted file segment includes the complete identification of the corresponding single file segment and the number of missing characters, and the number of missing characters of the corresponding single file segment is 0 when the complete identification of the corresponding single file segment indicates that the corresponding single file segment is complete.

[0013] The data loss state analysis system for network transmitting and receiving ends of the application is simple to operate and compact in structure. The AI prediction model for predicting the loss state of file segments during network transmission is established between fixed network transmitting and receiving ends by using a targeted learning mechanism, so that the overall content of multiple loads transmitted by multiple data packets can be analyzed and judged for complete state. BRIEF DESCRIPTION OF DRAWINGS

[0014] The embodiments of the application will be described below with reference to the accompanying drawings.

[0015] Figure 1 The internal structure diagram of the data loss state analysis system for network transmitting and receiving ends according to the primary embodiment of the application is shown.

[0016] Figure 2 The internal structure diagram of the data loss state analysis system for network transmitting and receiving ends according to the secondary embodiment of the application is shown.

[0017] Figure 3 The internal structure diagram of the data loss state analysis system for network transmitting and receiving ends according to the secondary embodiment of the application is shown. DETAILED DESCRIPTION

[0018] The embodiments of the data loss state analysis system for network transmitting and receiving ends of the application will be described in detail below with reference to the accompanying drawings.

[0019] Figure 1 The internal structure diagram of the data loss state analysis system for network transmitting and receiving ends according to the primary embodiment of the application is shown, which includes:

[0020] The batch extraction device is used to extract each file segment of a set character length at a set network receiving end, and each batch of extraction actions obtains a corresponding single file segment. The number of batches is positively correlated with the value of the set character length. Each file segment comes from a set network transmitting end and has a set character length which is statistically obtained at the set network receiving end.

[0021] Specifically, the file types of each file segment are the same, which can be parts of network novels or electronic books.

[0022] For example, the positive correlation between the batch number of the extracted files and the set character length value includes: when the set character length value is 5000 characters, the batch number of the extracted files is 500; when the set character length value is 10000 characters, the batch number of the extracted files is 600; when the set character length value is 20000 characters, the batch number of the extracted files is 700; and so on.

[0023] The state acquisition device is connected with the batch-by-batch extraction device, and is configured to acquire state data of each extracted file segment, wherein the state data includes a complete identifier of the corresponding single file segment and a number of missing characters.

[0024] The customized learning device is connected with the state acquisition device and the batch-by-batch extraction device, and is configured to establish an AI prediction model for predicting a loss state of a file segment during a transmission process from a set network transmission end to a set network receiving end, wherein the AI prediction model is a BP neural network after a set number of learning times, and the set number of times is equal to the number of the batch number of the extracted files.

[0025] For example, the establishment of the AI prediction model for predicting the loss state of the file segment during the transmission process from the set network transmission end to the set network receiving end includes: the AI prediction model can be established by using a numerical simulation mode.

[0026] The prediction judgment mechanism is connected with the customized learning device, and is configured to take a file with uncertain state data newly received at the set network receiving end from the set network transmission end as a target file, intercept a file segment with a length equal to a set character length from the target file as a reference file segment, and use the AI prediction model to predict state data of the reference file segment based on respective ASCII code values of respective constituent characters of the reference file segment, an actual geographical distance between the set network transmission end and the set network receiving end, a transmission bandwidth and a maximum operation amount per unit time of the set network transmission end, and a transmission bandwidth and a maximum operation amount per unit time of the set network receiving end.

[0027] The loss warning mechanism is connected with the prediction judgment mechanism, and is configured to perform a corresponding loss warning action when the complete identifier in the predicted state data of the reference file segment indicates that the reference file segment is incomplete.

[0028] Specifically, the status data of each extracted file segment is obtained. The status data includes the complete identifier of the corresponding single file segment and the number of missing characters. Specifically, when the complete identifier of the corresponding single file segment indicates that the corresponding single file segment is complete, the number of missing characters is 0.

[0029] Among them, an AI prediction model is established to predict the loss state of executable file fragments during transmission from a set network transmitting end to a set network receiving end. The AI ​​prediction model is a BP neural network after a set number of learning iterations. The set number is equal to the number of batches extracted, including: the multiple learning iterations of the BP neural network correspond to each file fragment.

[0030] The multiple learning iterations of the BP neural network corresponding to each file segment include: in each learning iteration of the BP neural network, the ASCII code values ​​corresponding to each constituent character of the single file segment corresponding to this learning iteration, the actual geographical distance between the network transmitter and the network receiver, the transmission bandwidth and maximum computational load per unit time of the network transmitter, and the transmission bandwidth and maximum computational load per unit time of the network receiver are used as the input content of the BP neural network, and the known state data of the single file segment corresponding to this learning iteration are used as the output content of the BP neural network to complete this learning iteration.

[0031] Figure 2 This is a schematic diagram of the internal structure of a data loss status analysis system for both ends of a network, as shown in a secondary embodiment of the present invention.

[0032] like Figure 2 As shown, with Figure 1 different, Figure 2 The data loss status analysis system for both ends of the network can also include the following components:

[0033] The on-site display mechanism, connected to the prediction and judgment mechanism, is used to receive and display the status data of the reference document fragment in real time;

[0034] For example, the on-site display mechanism can be selected as a liquid crystal display device, an LED display array, or an LCD display array;

[0035] The status data for receiving and displaying reference file fragments in real time includes: the complete identifier of the received reference file fragment and the number of missing characters.

[0036] Figure 3 This is a schematic diagram of the internal structure of a data loss status analysis system for both ends of a network, as shown in a further embodiment of the present invention.

[0037] likeFigure 3 As shown, Figure 1 different, Figure 3 The data loss state analysis system for network transmitting and receiving ends can further include the following components:

[0038] The parameter storage mechanism is connected with the customized learning device and is used to receive and store the AI prediction model.

[0039] The receiving and storing of the AI prediction model includes receiving and storing each model parameter of the AI prediction model.

[0040] Next, the specific structure of the data loss state analysis system for network transmitting and receiving ends of the present application will be further described.

[0041] In the data loss state analysis system for network transmitting and receiving ends according to various embodiments of the present application:

[0042] The state data of each extracted file segment includes the complete identification of the corresponding single file segment and the number of missing characters, and the number of missing characters of the corresponding single file segment is the total number of characters actually lost in the transmission process from the set network transmitting end to the set network receiving end when the complete identification of the corresponding single file segment indicates that the corresponding single file segment is incomplete.

[0043] In the data loss state analysis system for network transmitting and receiving ends according to various embodiments of the present application:

[0044] The state data of each extracted file segment includes the complete identification of the corresponding single file segment and the number of missing characters, and the number of missing characters of the corresponding single file segment is the total number of characters actually lost in the transmission process from the set network transmitting end to the set network receiving end when the complete identification of the corresponding single file segment indicates that the corresponding single file segment is incomplete.

[0045] In the data loss state analysis system for network transmitting and receiving ends according to various embodiments of the present application:

[0046] The loss warning mechanism is further used to temporarily suspend the corresponding loss warning action of the reference file segment when the complete identification in the state data of the predicted reference file segment indicates that the reference file segment is complete.

[0047] In the data loss state analysis system for network transmitting and receiving ends according to various embodiments of the present application:

[0048] The network receiving end extracts each file segment of the set character length in batches, and each batch of extraction actions obtains a corresponding single file segment, and the number of extracted batches is positively correlated with the value of the set character length, including: each file segment is composed of characters;

[0049] The network receiving end extracts each file segment of the set character length in batches, and each batch of extraction actions obtains a corresponding single file segment, and the number of extracted batches is positively correlated with the value of the set character length, including: each file segment is composed of characters;

[0050] The network receiving end extracts each file segment of the set character length in batches, and each batch of extraction actions obtains a corresponding single file segment, and the number of extracted batches is positively correlated with the value of the set character length, including: each file segment is composed of characters;

[0051] The network receiving end extracts each file segment of the set character length in batches, and each batch of extraction actions obtains a corresponding single file segment, and the number of extracted batches is positively correlated with the value of the set character length, including: each file segment is composed of characters;

[0052] In addition, in the data loss state analysis system for the network transmitting and receiving ends:

[0053] The network receiving end extracts each file segment of the set character length in batches, and each batch of extraction actions obtains a corresponding single file segment, and the number of extracted batches is positively correlated with the value of the set character length, including: each file segment is composed of characters;

[0054] The AI prediction model is used to predict the state data of the reference file segment based on the respective ASCII code values corresponding to respective constituent characters of the reference file segment, the actual geographical distance from the set network transmitting end to the set network receiving end, the transmission bandwidth and maximum operation amount per unit time of the set network transmitting end, and the transmission bandwidth and maximum operation amount per unit time of the set network receiving end, and the state data of the reference file segment output by the AI prediction model is obtained.

[0055] According to the above-mentioned embodiments, the innovation of the technical scheme of the present application is embodied in:

[0056] Firstly, an AI prediction model for predicting the loss state of a file segment in a network transmission process is established between the network receiving end and the network transmitting end of the set network, and the following establishment mechanism is performed on the AI prediction model to ensure the reliability of the prediction result: the AI prediction model is a BP neural network after a set number of learning times, and the number of learning times is positively correlated with the character length of each file segment.

[0057] Secondly, in each learning, the respective ASCII code values corresponding to respective constituent characters of a single file segment corresponding to this learning, the actual geographical distance from the set network transmitting end to the set network receiving end, the transmission bandwidth and maximum operation amount per unit time of the set network transmitting end, and the transmission bandwidth and maximum operation amount per unit time of the set network receiving end are used as the respective input contents of the BP neural network, and the known state data of the single file segment corresponding to this learning is used as the output content of the BP neural network, thereby completing this learning and ensuring the effect of each learning.

[0058] Thirdly, a file with uncertain state data newly received at the set network receiving end from the set network transmitting end is taken as a target file, a file segment with a length equal to the set character length is taken from the target file as a reference file segment, and the AI prediction model is used to predict the state data of the reference file segment based on the respective ASCII code values corresponding to respective constituent characters of the reference file segment, the actual geographical distance from the set network transmitting end to the set network receiving end, the transmission bandwidth and maximum operation amount per unit time of the set network transmitting end, and the transmission bandwidth and maximum operation amount per unit time of the set network receiving end, and the state data includes the complete identification of the reference file segment and the number of missing characters, thereby realizing effective judgment of the integrity of the received file segment and the degree of missing data.

[0059] Those skilled in the art will understand that the embodiments of the application described above and shown in the drawings are merely illustrative and that numerous other modifications and configurations can be devised without departing from the principles of the present application. The scope of the application is best defined by the appended claims.

Claims

1. A data loss status analysis system for both ends of a network, characterized in that, The system includes: The batch extraction device is used to extract file segments of a set character length in batches at a set network receiving end. Each batch of extraction action obtains a corresponding single file segment. The number of batches extracted is positively correlated with the value of the set character length. Each file segment comes from a set network transmitting end, and its set character length is statistically obtained at the set network receiving end. A status acquisition device, connected to the batch extraction device, is used to acquire status data for each extracted file segment. The status data includes the complete identifier of the corresponding single file segment and the number of missing characters. A custom learning device is connected to the state acquisition device and the batch extraction device respectively, and is used to establish an AI prediction model for predicting the loss of state of executable file fragments during transmission from a set network transmission end to a set network receiving end. The AI ​​prediction model is a BP neural network after a set number of learning iterations, and the set number is equal to the number of batches extracted. The prediction and judgment mechanism, connected to the customized learning device, is used to take the file with uncertain state data that is latest received from the set network transmitter at the set network receiver as the target file, extract a file segment with a length equal to the set character length from the target file as a reference file segment, and use the AI ​​prediction model to predict the state data of the reference file segment based on the ASCII code values ​​corresponding to each constituent character of the reference file segment, the actual geographical distance between the set network transmitter and the set network receiver, the transmission bandwidth and maximum computational load per unit time of the set network transmitter, and the transmission bandwidth and maximum computational load per unit time of the set network receiver. The missing information warning mechanism, connected to the prediction and judgment mechanism, is used to execute a corresponding missing information warning action when the complete identifier in the predicted state data of the reference file segment indicates that the reference file segment is incomplete. Specifically, the status data of each extracted file segment is obtained. The status data includes the complete identifier of the corresponding single file segment and the number of missing characters. Specifically, when the complete identifier of the corresponding single file segment indicates that the corresponding single file segment is complete, the number of missing characters is 0.

2. The data loss status analysis system for both ends of a network as described in claim 1, characterized in that: An AI prediction model is established to predict the loss state of executable file fragments during transmission from a designated network transmitter to a designated network receiver. The AI ​​prediction model is a BP neural network that has undergone a set number of learning iterations. The set number is equal to the number of batches extracted, where each of the multiple learning iterations of the BP neural network corresponds to a different file fragment. The multiple learning iterations of the BP neural network corresponding to each file segment include: in each learning iteration of the BP neural network, the ASCII code values ​​corresponding to each constituent character of the single file segment corresponding to this learning iteration, the actual geographical distance between the network transmitter and the network receiver, the transmission bandwidth and maximum computational load per unit time of the network transmitter, and the transmission bandwidth and maximum computational load per unit time of the network receiver are used as the input content of the BP neural network, and the known state data of the single file segment corresponding to this learning iteration are used as the output content of the BP neural network to complete this learning iteration.

3. The data loss status analysis system for both ends of a network as described in claim 2, characterized in that, The system also includes: The on-site display mechanism, connected to the prediction and judgment mechanism, is used to receive and display the status data of the reference document fragment in real time; The status data for receiving and displaying reference file fragments in real time includes: the complete identifier of the received reference file fragment and the number of missing characters.

4. The data loss status analysis system for both ends of a network as described in claim 2, characterized in that, The system also includes: A parameter storage mechanism, connected to the customized learning device, is used to receive and store the AI ​​prediction model; The process of receiving and storing the AI ​​prediction model includes receiving and storing the AI ​​prediction model by receiving and storing various model parameters of the AI ​​prediction model.

5. The data loss status analysis system for both ends of a network as described in any one of claims 2-4, characterized in that: Obtain the status data of each extracted file segment. The status data includes the complete identifier of the corresponding single file segment and the number of missing characters. When the complete identifier of the corresponding single file segment indicates that the corresponding single file segment is incomplete, the number of missing characters is the total number of characters actually lost by the corresponding single file segment during the transmission from the set network transmitting end to the set network receiving end.

6. The data loss status analysis system for both ends of a network as described in any one of claims 2-4, characterized in that: Obtain the status data of each extracted file segment. The status data includes the complete identifier of the corresponding single file segment and the number of missing characters. The data includes: using 0B00 as the complete identifier of the corresponding single file segment to indicate that the corresponding single file segment is complete, and using 0B01 as the complete identifier of the corresponding single file segment to indicate that the corresponding single file segment is incomplete.

7. The data loss status analysis system for both ends of a network as described in any one of claims 2-4, characterized in that: The missing information alert mechanism is also used to temporarily suspend the execution of the corresponding missing information alert action when the complete identifier in the predicted state data of the reference file fragment indicates that the reference file fragment is complete.

8. The data loss status analysis system for both ends of a network as described in any one of claims 2-4, characterized in that: At the network receiving end, file segments of a set character length are extracted in batches. Each batch of extractions yields a single file segment. The number of batches extracted is positively correlated with the set character length, including the fact that each file segment consists of individual characters. The process of extracting file segments of a set character length in batches at the network receiving end, with each batch of extraction obtaining a corresponding single file segment, and the positive correlation between the number of extraction batches and the set character length, also includes: using a numerical correspondence formula to represent the numerical correspondence between the number of extraction batches and the set character length.

9. The data loss status analysis system for both ends of a network as described in claim 8, characterized in that: The numerical correspondence formula used to express the positive correlation between the number of extracted batches and the set character length includes: in the numerical correspondence formula, the number of extracted batches is the output information of the numerical correspondence formula; The numerical correspondence between the extracted batch number and the set character length, expressed by the numerical correspondence formula, includes the following: in the numerical correspondence formula, the set character length is the input information of the numerical correspondence formula.

Images