Communication data processing method and device

Abstract

The embodiment of the invention provides a communication data processing method and device. The method comprises the steps that a first communication end obtains communication data to be transmitted; determining a compression threshold value of the communication data according to the current network condition; compressing the communication data by adopting an encoder of a data decompression model, and adjusting the compression ratio of the communication data according to a compression threshold in the compression process; wherein the compression threshold value is in negative correlation with the compression rate; sending the compressed communication data to a second communication end; wherein the second communication end is provided with a decoder of a data decompression model, and the second communication end is used for decompressing the compressed communication data by adopting the decoder of the data decompression model. Through the embodiment of the invention, the communication data is compressed and decompressed through the encoder and the decoder of the data decompression model, the compression rate is improved, the compression rate of the communication data is controlled by combining the current network condition, and network resources are effectively utilized.

Description

Technical Field

[0001] The present invention relates to the field of communication technologies, and particularly to a method and device for processing communication data. Background Art

[0002] In the scenario of mobile network communication, there are significant differences in the performance of different terminal devices, and the data transmission process has become an important bottleneck for the large-scale promotion and development of the data economy. Among them, the scale and format of data transmission have changed greatly compared with the prior data transmission, and data transmission is developing towards characteristics such as massive data, rich-format data, and tidal data.

[0003] In the prior art, a method of compressing and then transmitting data is proposed to improve the stability and rate of transmission by reducing the amount of data. However, the existing compression technologies have disadvantages such as low compression ratio and failure to effectively utilize network resources, and it is difficult to meet the requirements. Summary of the Invention

[0004] In view of the above problems, a method and device for processing communication data are proposed to overcome or at least partially solve the above problems, including:

[0005] A method for processing communication data, applied to a first communication end, where the first communication end is provided with an encoder of a data decompression model, and the method includes:

[0006] Obtain the communication data to be transmitted;

[0007] Determine the compression threshold of the communication data according to the current network situation;

[0008] Use the encoder of the data decompression model to compress the communication data, and during the compression process, adjust the compression ratio of the communication data according to the compression threshold; wherein, the compression threshold is negatively correlated with the compression ratio;

[0009] Send the compressed communication data to a second communication end; wherein, the second communication end is provided with a decoder of the data decompression model, and the second communication end is used to decompress the compressed communication data by using the decoder of the data decompression model.

[0010] Optionally, the determining the compression threshold of the communication data according to the current network situation includes:

[0011] Determine the current network attribute information according to the current network situation;

[0012] Obtain the preset size of the compressed data buffer;

[0013] Determine the compression threshold of the communication data according to the current network attribute information and the size of the compressed data buffer.

[0014] Optionally, the current network attribute information includes: current network bandwidth, current network latency.

[0015] Optionally, the encoder and decoder of the data decompression model adopt a symmetric structure.

[0016] Optionally, the encoder and decoder of the data decompression model are provided with a convolutional layer and an attention layer based on the attention mechanism.

[0017] Optionally, before determining the compression threshold of the communication data according to the current network situation, it further includes:

[0018] Obtain the data header identifier of the communication data;

[0019] When the data header identifier indicates that the communication data is video stream data, execute determining the compression threshold of the communication data according to the current network situation.

[0020] Optionally, before determining the compression threshold of the communication data according to the current network situation, it further includes:

[0021] Standardize the communication data into two-dimensional data according to the time dimension.

[0022] A method for processing communication data, applied to a second communication end, where the second communication end is provided with a decoder of a data decompression model, and the method includes:

[0023] Obtain the compressed communication data sent by the first communication end; wherein, the compressed communication data is obtained by the first communication end using the encoder of the data decompression model to compress the communication data, and during the compression process, adjust the compression ratio of the communication data according to the compression threshold; the compression threshold is determined according to the current network situation, and the compression threshold is negatively correlated with the compression ratio;

[0024] Use the decoder of the data decompression model to decompress the compressed communication data.

[0025] A device for processing communication data, applied to a first communication end, where the first communication end is provided with an encoder of a data decompression model, and the device is used for:

[0026] Obtain the communication data to be transmitted;

[0027] Determine the compression threshold of the communication data according to the current network situation;

[0028] The encoder of the data decompression model is used to compress the communication data, and during the compression process, the compression rate of the communication data is adjusted according to the compression threshold; wherein, the compression threshold is negatively correlated with the compression rate;

[0029] The compressed communication data is sent to the second communication end; wherein, the second communication end is provided with a decoder of the data decompression model, and the second communication end is used to decompress the compressed communication data by using the decoder of the data decompression model.

[0030] A device for processing communication data is applied to the second communication end, and the second communication end is provided with a decoder of the data decompression model. The device is used for:

[0031] Obtain the compressed communication data sent by the first communication end; wherein, the compressed communication data is obtained by the first communication end using the encoder of the data decompression model to compress the communication data, and during the compression process, the compression rate of the communication data is adjusted according to the compression threshold; the compression threshold is determined according to the current network condition, and the compression threshold is negatively correlated with the compression rate;

[0032] Use the decoder of the data decompression model to decompress the compressed communication data.

[0033] The embodiments of the present invention have the following advantages:

[0034] In the embodiments of the present invention, the first communication end obtains the communication data to be transmitted, determines the compression threshold of the communication data according to the current network condition, uses the encoder of the data decompression model to compress the communication data, and during the compression process, adjusts the compression rate of the communication data according to the compression threshold, and sends the compressed communication data to the second communication end. The second communication end uses the decoder of the data decompression model to decompress the compressed communication data, realizes the compression and decompression of communication data through the encoder and decoder of the data decompression model, improves the compression rate, and controls the compression rate of communication data by combining the current network condition, effectively utilizing network resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions of the present invention, the drawings required for the description of the present invention will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0036] Figure 1 It is a flowchart of the steps of a method for processing communication data provided by some embodiments of the present invention;

[0037] Figure 2 It is a schematic diagram of a model structure provided by some embodiments of the present invention;

[0038] Figure 3 It is a flowchart of the steps of a method for processing communication data provided by some embodiments of the present invention;

[0039] Figure 4 It is a flowchart of the steps of another method for processing communication data provided by some embodiments of the present invention. Detailed implementation manners

[0040] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific implementation manners. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0041] In mobile network communication, the transmission of video stream data accounts for a large proportion of daily data transmission. The transmission scenario of video stream data is different from that of traditional communication data transmission, with characteristics such as a large data transmission scale, a large occupied bandwidth, and high requirements for network performance.

[0042] In the related art, the data transmission method is to directly form an Ethernet communication packet with the data to be transmitted using the corresponding transmission protocol on the mobile network card of the terminal device and perform the transmission. However, this data transmission method will cause untimely data transmission and a large network delay under the condition of scarce network resources, reducing the network bandwidth utilization rate.

[0043] In view of the characteristics of video stream data transmission, a method of compressing the data first and then transmitting it is proposed to improve the transmission stability and rate by reducing the data volume. For example, the compression technologies include dictionary compression method, Huffman coding method, and predictive coding method. However, the compression ratios of these compression technologies are relatively low, and it is difficult to maintain a stable data throughput when the network resources are at a low ebb. Moreover, these compression technologies do not fully consider the phased fluctuations of network resources, adopt the same compression strategy during the peak and trough periods of network resources, and fail to effectively utilize network resources, resulting in a decrease in data throughput.

[0044] In order to improve the transmission rate and stability of video stream data in a mobile communication network and to adapt to the performance differences of different terminal devices, the embodiments of the present invention use an attention mechanism in combination with deep learning to propose an stacked symmetric autoencoder. Through the attention mechanism, data compression and decompression are realized, spatial transformation of input data during the compression and decompression processes is prevented, the compression ratio is increased, and the reconstruction loss is minimized, that is, the difference between the original data and the decompressed data is reduced.

[0045] Moreover, in order to meet the applicability of dynamic network resources, the embodiments of the present invention propose a method for adaptively adjusting the compression ratio, which can automatically adjust the compression ratio according to network performance, so as to maximize the utilization of network resources under different network conditions, adapt to different network conditions, make full use of network bandwidth, increase data throughput, reduce data transmission delay, and improve transmission rate and stability.

[0046] The following provides an exemplary description of the present invention with reference to the accompanying drawings:

[0047] Refer to Figure 1 , which shows a flowchart of the steps of a method for processing communication data provided by some embodiments of the present invention, applied to a first communication end. The first communication end communicates with a second communication end. The first communication end can be a data sending end, and the second communication end is a data receiving end. In some examples, the first communication end and the second communication end can be mobile terminal devices.

[0048] Among them, an encoder of a data decompression model is provided at the first communication end, and a decoder of the data decompression model is provided at the second communication end. In some examples, the trained encoder of the data decompression model is used as a compression module and embedded in the network card of the mobile terminal device. The data to be transmitted is first processed by the compression module. The compression module adjusts the compression threshold in real time according to the network status of the terminal device and transmits the compressed data. A decoder of the corresponding data decompression model is deployed at the receiving end as a decompression module to decompress the compressed data for subsequent processing.

[0049] Specifically, a compression plugin and a decompression plugin can be deployed at the data sending end and the receiving end respectively. The compression plugin is deployed at the data sending end to monitor the data sending port and compress the data. The decompression plugin is deployed at the data receiving end to monitor the data reporting port and decompress the reported data.

[0050] Such as Figure 2, the encoder and decoder of the data decompression model adopt a symmetric structure, that is, a stacked symmetric autoencoder is implemented, and the encoder and decoder of the data decompression model are provided with a convolutional layer and an attention layer based on the attention mechanism. Through the attention mechanism, data compression and decompression are realized, preventing the spatial transformation of the input data during the compression and decompression processes, improving the compression rate, and minimizing the reconstruction loss, that is, reducing the difference between the original data and the decompressed data.

[0051] Among them, the data decompression model includes two parts: data compression and data decompression.

[0052] In the data compression part, a multi-layer encoder is constructed. The number of neurons in the first layer matches the dimension of the input vector. Through the attention layer and the convolutional layer, the dimension of the input data is gradually halved, and the high-dimensional data is sequentially mapped to the low-dimensional space.

[0053] In the data decompression part, a structure symmetric to the compression encoder is adopted. The number of neurons in the input layer of the decoder is the number of neurons in the output layer of the encoder, and it doubles layer by layer starting from the second layer.

[0054] In the embodiment of the present invention, by proposing a stacked symmetric autoencoder, the compression rate of data is improved, and combined with the attention mechanism, the reconstruction loss of data during the compression and decompression processes is reduced, which can improve the compression rate and enable the decompressed data to retain the feature information of the original data to a great extent, and make full use of the network bandwidth to improve the data transmission rate.

[0055] In some examples, the loss function of the data decompression model is shown in the following formula:

[0056]

[0057] Among them, X i represents the input data, represents the reconstructed data, n is the total amount of data, Z e is the sum of the distances from the input data to the cluster centers in the hidden space in the encoder, Z d is the sum of the distances from the data to the cluster centers in the hidden space in the decoder. The difference between Z e and Z d in the loss function suppresses the offset of the spatial position of the data points to the cluster centers during the compression and decompression processes, thereby reducing the reconstruction loss. In some examples, the Adam optimizer is used to train the decompression model during the training process.

[0058] Specifically, it may include the following steps:

[0059] Step 101, obtain the communication data to be transmitted.

[0060] In practical applications, the first communication end can obtain communication data to be transmitted from a data source or an application. As an example, the communication data can be video stream data. Of course, it can also be audio data, text data, or any other data type that needs to be transmitted over a network.

[0061] Step 102: Determine the compression threshold of the communication data according to the current network conditions.

[0062] In practical applications, an appropriate compression threshold can be comprehensively evaluated and determined based on network conditions such as the current network bandwidth and the current network latency, as well as the preset size of the compressed data buffer. This compression threshold will serve as an important basis for subsequent adjustment of the communication data compression ratio.

[0063] In some embodiments of the present invention, the determining the compression threshold of the communication data according to the current network conditions includes: determining the current network attribute information according to the current network conditions; obtaining the preset size of the compressed data buffer; and determining the compression threshold of the communication data according to the current network attribute information and the size of the compressed data buffer.

[0064] In some embodiments of the present invention, the current network attribute information includes:

[0065] The current network bandwidth and the current network latency.

[0066] In practical applications, an adaptive compression ratio setting method is proposed, which automatically adjusts the compression ratio according to the dynamic changes of network resources and can be defined by the following formula:

[0067] L = D·N / T

[0068] Where N represents the network bandwidth, T represents the network latency, D represents the preset size of the compressed data buffer, which is default to the buffer size of the network card. L represents the data compression threshold, where N / T represents the current network data transmission rate, and D·N / T is the compression threshold, which also represents the maximum data transmission amount that the current network node can bear.

[0069] In practical applications, when the network bandwidth is larger or the latency is lower, the compression threshold is larger, more data can be directly transmitted, and the data compression ratio is lower, indicating that when network resources are sufficient, the compression ratio is reduced to make full use of network resources to improve data throughput. On the contrary, when the network bandwidth is smaller or the latency is higher, the compression threshold is smaller, and the compression ratio is increased to adapt to the poor network situation.

[0070] Such as Figure 3, a compression threshold L is set, which is dynamically calculated based on the current network bandwidth N, network latency T, and buffer size D. When the amount of data to be compressed is less than the threshold L, the data will not be compressed; when the amount of data exceeds the threshold L, the data within the threshold remains uncompressed, and the excess part is compressed.

[0071] In the embodiments of the present invention, in order to adapt to the characteristics of tidal data transmission and different network conditions, a method of adaptively adjusting the compression ratio is designed, which can automatically adjust the data compression rate under different network conditions and data transmission scales, maintain the stability of data transmission latency, improve the throughput of data and the utilization rate of network bandwidth, and meet the requirements of different performance terminal devices to provide a stable data transmission process in the scenario of tidal data transmission.

[0072] In some embodiments of the present invention, before determining the compression threshold of the communication data according to the current network situation, it further includes: obtaining the data header identifier of the communication data; when the data header identifier indicates that the communication data is video stream data, performing the determination of the compression threshold of the communication data according to the current network situation.

[0073] In practical applications, it is possible to determine whether it is video stream data through packet content analysis, that is, obtain the packets of communication data, check the data header identifier, and determine whether the packet is of the video stream data type according to the identifier. If it is determined to be video stream data, subsequent data compression processes can be performed.

[0074] In some embodiments of the present invention, before determining the compression threshold of the communication data according to the current network situation, it further includes:

[0075] Standardize the communication data into two-dimensional data according to the time dimension.

[0076] In practical applications, standard two-dimensional processing can be performed on the communication data to be transmitted. Specifically, taking the timestamp as the division basis, the communication data to be transmitted can be sorted through the time dimension to form standardized two-dimensional data.

[0077] For example, convert the network communication data into a two-dimensional data format with time as the dimension. The timestamp in each row is the collected communication data, and a socket is formed.

[0078] By normalizing the data into two-dimensional data, the structuring and standardization of the data are achieved, making the subsequent data compression process more efficient and stable. In the two-dimensional data format, the data compression strategy can be analyzed and adjusted more intuitively to meet the data transmission requirements under different network conditions. In addition, the two-dimensional data format helps to improve the speed and accuracy of data processing, providing a strong guarantee for the real-time transmission of video stream data. In the embodiment of the present invention, by proposing a standardized two-dimensional data processing method and combining an adaptive compression rate adjustment mechanism, the efficient and stable transmission of video stream data is achieved, improving the transmission performance and user experience of the mobile communication network.

[0079] Step 103: Use the encoder of the data decompression model to compress the communication data, and during the compression process, adjust the compression rate of the communication data according to the compression threshold; wherein, the compression threshold is negatively correlated with the compression rate.

[0080] In practical applications, the encoder of the data decompression model receives the input communication data, focuses on the data center through convolutional layer downsampling and attention mechanism, and after multiple layers of convolution and attention, maps the high-dimensional input data to a low-dimensional hidden space, thereby obtaining the compressed communication data.

[0081] Among them, the attention mechanism dynamically adjusts the weights of each data point during the compression process by calculating the correlation and importance between different data points, so that more key information can be retained and redundant information can be ignored during the compression process, thereby improving the compression efficiency and data quality.

[0082] Among them, the convolutional layer performs a sliding window operation on the input data through a convolutional kernel, extracts local features, and maps these features into the hidden space. By stacking multiple convolutional layers, higher-level features can be gradually extracted, thereby achieving effective compression of the data.

[0083] Moreover, during the compression process, the encoder dynamically adjusts the compression rate according to a preset compression threshold. When the data volume is small and lower than the compression threshold, the encoder can choose not to compress the data or use a lower compression rate to reduce data loss and maintain data integrity. When the data volume is large and exceeds the compression threshold, the encoder will increase the compression rate to reduce the data volume and meet the requirements of network transmission. This way of dynamically adjusting the compression rate can ensure the stability and efficiency of data transmission under different network conditions and transmission data scales.

[0084] Step 104: Send the compressed communication data to the second communication end; wherein, the second communication end is provided with a decoder of the data decompression model, and the second communication end is used to decompress the compressed communication data by using the decoder of the data decompression model.

[0085] After obtaining the compressed communication data, the compressed communication data can be sent to the second communication terminal. After receiving the data, if the received data is uncompressed original data, the second communication terminal directly performs subsequent data forwarding or receiving processing. If the received data is compressed data, the second communication terminal can use a decoder to decompress the data before performing subsequent processing.

[0086] In practical applications, the decoder can perform the reverse operation of the encoder, through multiple convolution upsampling and reverse processing of the attention mechanism, and finally output the actual decoded data.

[0087] In some examples, the decoder can use a multi-layer decoder structure to gradually restore data from a low-dimensional hidden space to the original high-dimensional space. The number of neurons in each layer of the decoder gradually increases, corresponding to the structure of the encoder, to ensure that the data can be accurately restored from the compressed state to the original state. In this way, the decoder can efficiently restore the original data, so that the decompressed data can retain the original feature information as much as possible while meeting the needs of subsequent data processing or application.

[0088] In an embodiment of the present invention, the first communication end obtains communication data to be transmitted, determines a compression threshold of the communication data according to the current network conditions, uses an encoder of a data decompression model to compress the communication data, and during the compression process, adjusts the compression rate of the communication data according to the compression threshold, and sends the compressed communication data to the second communication end. The second communication end uses a decoder of the data decompression model to decompress the compressed communication data, thereby realizing compression and decompression of the communication data through the encoder and decoder of the data decompression model, thereby improving the compression rate, and controlling the compression rate of the communication data in combination with the current network conditions, thereby effectively utilizing network resources.

[0089] Reference Figure 4 , shows a flowchart of another method for processing communication data provided by some embodiments of the present invention, which is applied to a second communication terminal, a first communication terminal and a second communication terminal communicate, the first communication terminal may be a data sending terminal, and the second communication terminal may be a data receiving terminal. In some examples, the first communication terminal and the second communication terminal may be mobile terminal devices.

[0090] Among them, an encoder of a data decompression model is provided at the first communication end, and a decoder of the data decompression model is provided at the second communication end. In some examples, the encoder of the trained data decompression model is embedded as a compression module into the network card of the mobile terminal device. The data to be transmitted is first processed by the compression module. The compression module adjusts the compression threshold in real time according to the network status of the terminal device, and transmits the compressed data. At the receiving end, a decoder of the corresponding data decompression model is deployed as a decompression module to decompress the compressed data for subsequent processing.

[0091] Specifically, a compression plug-in and a decompression plug-in can be deployed at the data sending end and the receiving end respectively. The compression plug-in is deployed at the data sending end to monitor the data sending port and compress the data. The decompression plug-in is deployed at the data receiving end to monitor the data reporting port and decompress the reported data.

[0092] Such as Figure 2 , the encoder and decoder of the data decompression model adopt a symmetric structure, that is, a stacked symmetric autoencoder is implemented, and the encoder and decoder of the data decompression model are provided with a convolutional layer and an attention layer based on the attention mechanism. The compression and decompression of data are realized through the attention mechanism, preventing the spatial transformation of the input data during the compression and decompression processes, improving the compression rate, and minimizing the reconstruction loss, that is, reducing the difference between the original data and the decompressed data.

[0093] Among them, the data decompression model includes two parts: data compression and data decompression.

[0094] In the data compression part, a multi-layer encoder is constructed. The number of neurons in the first layer matches the dimension of the input vector. Through the attention layer and the convolutional layer, the dimension of the input data is gradually halved, and the high-dimensional data is mapped to the low-dimensional space in turn.

[0095] In the data decompression part, a structure symmetric to the compression encoder is adopted. The number of neurons in the input layer of the decoder is the number of neurons in the output layer of the encoder, and it doubles layer by layer starting from the second layer.

[0096] In the embodiment of the present invention, by proposing a stacked symmetric autoencoder, the compression rate of data is improved, and combined with the attention mechanism, the reconstruction loss of data during the compression and decompression processes is reduced. It can improve the compression rate and enable the decompressed data to retain the characteristic information of the original data to a great extent, and make full use of the network bandwidth to improve the data transmission rate.

[0097] In some examples, the loss function of the data decompression model is shown in the following formula:

[0098]

[0099] Among them, Xi Represents the input data, Represents the reconstructed data, where n is the total amount of data, and Z e Is the sum of the distances from the input data in the encoder to the clustering centers in the hidden space, and Z d Is the sum of the distances from the data in the decoder to the clustering centers in the hidden space. In the loss function, Z e And Z d The difference in... suppresses the offset of the data points to the spatial positions of the clustering centers during the compression and decompression processes, thereby reducing the reconstruction loss. In some examples, the Adam optimizer is used to train the decompression model during the training process.

[0100] Specifically, it may include the following steps:

[0101] Step 401, obtain the compressed communication data sent by the first communication end;

[0102] Among them, the compressed communication data is obtained by the first communication end using the encoder of the data decompression model to compress the communication data, and during the compression process, the compression rate of the communication data is adjusted according to the compression threshold; the compression threshold is determined according to the current network situation, and the compression threshold is negatively correlated with the compression rate.

[0103] In practical applications, the first communication end can obtain the communication data to be transmitted from the data source or the application program. As an example, the communication data can be video stream data. Of course, it can also be audio data, text data, or any other data type that needs to be transmitted through the network.

[0104] For the communication data to be transmitted, a suitable compression threshold can be comprehensively evaluated and determined according to the network situation such as the current network bandwidth and the current network delay, as well as the preset size of the compressed data buffer. This compression threshold will be an important basis for adjusting the compression rate of the communication data subsequently.

[0105] In practical applications, the encoder of the data decompression model receives the input communication data, focuses on the data center through convolutional layer downsampling and the attention mechanism, and after multiple convolutional and attention operations, maps the high-dimensional input data to a low-dimensional hidden space, thereby obtaining the compressed communication data.

[0106] Among them, the attention mechanism dynamically adjusts the weights of each data point during the compression process by calculating the correlation and importance between different data points, so that more key information can be retained and redundant information can be ignored during the compression process, thereby improving the compression efficiency and data quality.

[0107] The convolution layer uses the convolution kernel to perform a sliding window operation on the input data, extract local features, and map these features to the latent space. By stacking multiple convolution layers, higher-level features can be gradually extracted, thereby achieving effective data compression.

[0108] Moreover, during the compression process, the encoder will dynamically adjust the compression rate according to the preset compression threshold. When the amount of data is small and below the compression threshold, the encoder can choose not to compress the data or use a lower compression rate to reduce data loss and maintain data integrity. When the amount of data is large and exceeds the compression threshold, the encoder will increase the compression rate to reduce the amount of data and adapt to the needs of network transmission. This method of dynamically adjusting the compression rate can ensure that the stability and efficiency of data transmission can be maintained under different network conditions and transmission data scales.

[0109] Step 402: Decompress the compressed communication data using the decoder of the data decompression model.

[0110] After obtaining the compressed communication data, the compressed communication data can be sent to the second communication terminal. After receiving the data, if the received data is uncompressed original data, the second communication terminal directly performs subsequent data forwarding or receiving processing. If the received data is compressed data, the second communication terminal can use a decoder to decompress the data before performing subsequent processing.

[0111] In practical applications, the decoder can perform the reverse operation of the encoder, through multiple convolution upsampling and reverse processing of the attention mechanism, and finally output the actual decoded data.

[0112] In some examples, the decoder can use a multi-layer decoder structure to gradually restore data from a low-dimensional hidden space to the original high-dimensional space. The number of neurons in each layer of the decoder gradually increases, corresponding to the structure of the encoder, to ensure that the data can be accurately restored from the compressed state to the original state. In this way, the decoder can efficiently restore the original data, so that the decompressed data can retain the original feature information as much as possible while meeting the needs of subsequent data processing or application.

[0113] In an embodiment of the present invention, the first communication end obtains communication data to be transmitted, determines a compression threshold of the communication data according to the current network condition, uses an encoder of a data decompression model to compress the communication data, and adjusts the compression rate of the communication data according to the compression threshold during the compression process, and sends the compressed communication data to the second communication end. The second communication end uses a decoder of the data decompression model to decompress the compressed communication data, realizing the compression and decompression of communication data through the encoder and decoder of the data decompression model, improving the compression rate, and effectively utilizing network resources by controlling the compression rate of communication data in combination with the current network condition.

[0114] It should be noted that, for method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that the embodiments of the present invention are not limited by the described action sequence, because according to the embodiments of the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily essential for the embodiments of the present invention.

[0115] Some embodiments of the present invention also provide a device for processing communication data, which is applied to the first communication end. The first communication end is provided with an encoder of a data decompression model. The device is used for:

[0116] Obtain communication data to be transmitted;

[0117] Determine the compression threshold of the communication data according to the current network condition;

[0118] Use the encoder of the data decompression model to compress the communication data, and adjust the compression rate of the communication data according to the compression threshold during the compression process; wherein, the compression threshold is negatively correlated with the compression rate;

[0119] Send the compressed communication data to the second communication end; wherein, the second communication end is provided with a decoder of the data decompression model, and the second communication end is used to use the decoder of the data decompression model to decompress the compressed communication data.

[0120] Optionally, the determining the compression threshold of the communication data according to the current network condition includes:

[0121] Determine the current network attribute information according to the current network condition;

[0122] Obtain the preset size of the compressed data buffer;

[0123] Determine the compression threshold of the communication data according to the current network attribute information and the size of the compressed data buffer.

[0124] Optionally, the current network attribute information includes: current network bandwidth, current network delay.

[0125] Optionally, the encoder and decoder of the data decompression model adopt a symmetric structure.

[0126] Optionally, the encoder and decoder of the data decompression model are provided with a convolutional layer and an attention layer based on an attention mechanism.

[0127] Optionally, before determining the compression threshold of the communication data according to the current network situation, it further includes:

[0128] Obtain the data header identifier of the communication data;

[0129] When the data header identifier indicates that the communication data is video stream data, execute determining the compression threshold of the communication data according to the current network situation.

[0130] Optionally, before determining the compression threshold of the communication data according to the current network situation, it further includes:

[0131] Standardize the communication data into two-dimensional data according to the time dimension.

[0132] In an embodiment of the present invention, the first communication end obtains the communication data to be transmitted, determines the compression threshold of the communication data according to the current network situation, uses the encoder of the data decompression model to compress the communication data, and adjusts the compression rate of the communication data according to the compression threshold during the compression process, and sends the compressed communication data to the second communication end. The second communication end uses the decoder of the data decompression model to decompress the compressed communication data, realizing the compression and decompression of the communication data through the encoder and decoder of the data decompression model, improving the compression rate, and effectively utilizing network resources by combining the current network situation to control the compression rate of the communication data.

[0133] Some embodiments of the present invention further provide a device for processing communication data, which is applied to the second communication end. The second communication end is provided with a decoder of a data decompression model. The device is used for:

[0134] Obtain the compressed communication data sent by the first communication end; wherein, the compressed communication data is obtained by the first communication end using the encoder of the data decompression model to compress the communication data, and adjusts the compression rate of the communication data according to the compression threshold during the compression process; the compression threshold is determined according to the current network situation, and the compression threshold is negatively correlated with the compression rate;

[0135] Use the decoder of the data decompression model to decompress the compressed communication data.

[0136] Optionally, determining the compression threshold of the communication data according to the current network situation includes:

[0137] Determine the current network attribute information according to the current network situation;

[0138] Obtain the preset size of the compressed data buffer;

[0139] Determine the compression threshold of the communication data according to the current network attribute information and the size of the compressed data buffer.

[0140] Optionally, the current network attribute information includes: current network bandwidth, current network delay.

[0141] Optionally, the encoder and decoder of the data decompression model adopt a symmetric structure.

[0142] Optionally, the encoder and decoder of the data decompression model are provided with a convolutional layer and an attention layer based on an attention mechanism.

[0143] Optionally, before determining the compression threshold of the communication data according to the current network situation, it further includes:

[0144] Obtain the data header identifier of the communication data;

[0145] When the data header identifier indicates that the communication data is video stream data, execute determining the compression threshold of the communication data according to the current network situation.

[0146] Optionally, before determining the compression threshold of the communication data according to the current network situation, it further includes:

[0147] Standardize the communication data into two-dimensional data according to the time dimension.

[0148] In the embodiment of the present invention, the first communication end obtains the communication data to be transmitted, determines the compression threshold of the communication data according to the current network situation, uses the encoder of the data decompression model to compress the communication data, and adjusts the compression rate of the communication data according to the compression threshold during the compression process, and sends the compressed communication data to the second communication end. The second communication end uses the decoder of the data decompression model to decompress the compressed communication data, realizing the compression and decompression of communication data through the encoder and decoder of the data decompression model, improving the compression rate, and effectively utilizing network resources by combining the current network situation to control the compression rate of communication data.

[0149] Some embodiments of the present invention further provide an electronic device, including a processor, a memory, and a computer program stored on the memory and capable of running on the processor. When the computer program is executed by the processor, the above method is implemented.

[0150] Some embodiments of the present invention further provide a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by the processor, the above method is implemented.

[0151] Some embodiments of the present invention further provide a computer program product, including a computer program. When the computer program is executed by the processor, the above method is implemented.

[0152] For the device embodiments, since they are basically similar to the method embodiments, the description is relatively simple. For the relevant parts, please refer to the partial description of the method embodiments.

[0153] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Moreover, the collection, use, and processing of the relevant data need to comply with the relevant laws, regulations, and standards of the relevant countries and regions, and corresponding operation entrances are provided for the user to choose to authorize or refuse.

[0154] Each embodiment in this specification is described in a progressive manner. The key point of each embodiment is to illustrate the differences from other embodiments. For the same or similar parts among the embodiments, please refer to each other.

[0155] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a device, or a computer program product. Therefore, the embodiments of the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the embodiments of the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) containing computer-usable program codes.

[0156] Embodiments of the present invention are described with reference to the flowcharts and / or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, and the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing terminal device generate a device for implementing the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 or in multiple blocks.

[0157] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing terminal device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device that implements the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 or in multiple blocks.

[0158] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device, such that a series of operation steps are executed on the computer or other programmable terminal device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable terminal device provide steps for implementing the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 or in multiple blocks.

[0159] Although the preferred embodiments of the present invention have been described, those skilled in the art can make additional changes and modifications to these embodiments once they learn the basic creative concepts. Therefore, the appended claims are intended to be construed to include the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.

[0160] Finally, it should also be noted that in this text, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, article or terminal device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or terminal device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or terminal device comprising the above elements.

[0161] The above provides a detailed introduction to a method and apparatus for processing communication data. Specific examples are used in this text to illustrate the principles and implementation manners of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for processing communication data, characterized in that, Applied to the first communication end, where an encoder of a data decompression model is provided at the first communication end, the method includes: Obtain communication data to be transmitted; Determine the compression threshold of the communication data according to the current network condition; Use the encoder of the data decompression model to compress the communication data, and during the compression process, adjust the compression rate of the communication data according to the compression threshold; wherein, the compression threshold is negatively correlated with the compression rate; Send the compressed communication data to the second communication end; wherein, a decoder of the data decompression model is provided at the second communication end, and the second communication end is configured to use the decoder of the data decompression model to decompress the compressed communication data.

2. The method according to claim 1, wherein The determining the compression threshold of the communication data according to the current network condition includes: Determine the current network attribute information according to the current network condition; Obtain the preset size of the compressed data buffer; Determine the compression threshold of the communication data according to the current network attribute information and the size of the compressed data buffer.

3. The method according to claim 2, characterized in that, The current network attribute information includes: the current network bandwidth, the current network delay.

4. The method according to any one of claims 1 to 3, characterized in that The encoder and decoder of the data decompression model adopt a symmetric structure.

5. The method according to claim 4, wherein The encoder and decoder of the data decompression model are provided with a convolutional layer and an attention layer based on an attention mechanism.

6. The method according to any one of claims 1 to 3, characterized in that, Before the determining the compression threshold of the communication data according to the current network condition, it further includes: Obtain the data header identifier of the communication data; When the data header identifier indicates that the communication data is video stream data, execute the determining the compression threshold of the communication data according to the current network condition.

7. The method according to any one of claims 1 to 3, characterized in that Before the determining the compression threshold of the communication data according to the current network condition, it further includes: Normalize the communication data into two-dimensional data according to the time dimension.

8. A method for processing communication data, characterized in that Applied to the second communication end, where a decoder of a data decompression model is provided at the second communication end, the method includes: Obtain the compressed communication data sent by the first communication end; wherein, the compressed communication data is obtained by the first communication end using the encoder of the data decompression model to compress the communication data, and during the compression process, the compression rate of the communication data is adjusted according to the compression threshold; the compression threshold is determined according to the current network condition, and the compression threshold is negatively correlated with the compression rate; Use the decoder of the data decompression model to decompress the compressed communication data.

9. A device for processing communication data, characterized in that, Applied to the first communication end, where an encoder of a data decompression model is provided at the first communication end, the apparatus is configured to: Obtain communication data to be transmitted; Determine the compression threshold of the communication data according to the current network condition; Use the encoder of the data decompression model to compress the communication data, and during the compression process, adjust the compression rate of the communication data according to the compression threshold; wherein, the compression threshold is negatively correlated with the compression rate; Send the compressed communication data to the second communication end; wherein, a decoder of the data decompression model is provided at the second communication end, and the second communication end is configured to use the decoder of the data decompression model to decompress the compressed communication data.

10. A device for processing communication data, characterized in that, Applied to the second communication end, a decoder of the data decompression model is provided at the second communication end, and the apparatus is configured to: Obtain the compressed communication data sent by the first communication end; wherein, the compressed communication data is obtained by the first communication end using the encoder of the data decompression model to compress the communication data, and during the compression process, the compression rate of the communication data is adjusted according to a compression threshold; the compression threshold is determined according to the current network condition, and the compression threshold is negatively correlated with the compression rate; use the decoder of the data decompression model to decompress the compressed communication data.

Images