Network congestion control method and apparatus, network device, and readable storage medium

Abstract

This application relates to a network congestion control method, apparatus, network device, and readable storage medium. The method trains a network congestion control model using training parameters and outputs a congestion judgment threshold. The training parameters include multiple historical transmission durations and congestion indices corresponding to each historical transmission duration. The congestion judgment threshold is the congestion index corresponding to a preset target transmission duration. Then, based on the relationship between the current congestion index and the congestion judgment threshold, server-side configuration parameters are adjusted. The adjusted server-side configuration parameters are then used for network transmission to achieve network congestion control and improve network transmission stability.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to a network congestion control method, apparatus, network device, and readable storage medium. Background Technology

[0002] During network transmission, the actual network environment often differs significantly from theoretical scenarios. Due to the complexity of the network environment, network congestion may occur, leading to transmission delays or even interruptions. This can significantly impact the stability of services carried by the network, especially those with long processing times, where retransmission costs are substantial. For example, PXE (Preboot Execution Environment) network booting using the TFTP protocol may experience slow startup or even failure due to unstable network transmission. Therefore, to improve network transmission stability and ensure the stable operation of services carried by the network, appropriate network congestion control is necessary. Summary of the Invention

[0003] Therefore, it is necessary to provide a network congestion control method, apparatus, network device, and readable storage medium that can improve network transmission stability in response to the above-mentioned technical problems.

[0004] Firstly, a network congestion control method is provided, applied to network devices, the method comprising:

[0005] The network congestion control model is trained using training parameters, and the congestion judgment threshold is output. The training parameters include multiple historical transmission durations and the congestion index corresponding to each historical transmission duration. The congestion judgment threshold is the congestion index corresponding to a preset target transmission duration.

[0006] Based on the relationship between the current congestion index and the congestion threshold, adjust the server configuration parameters and use the adjusted server configuration parameters for network transmission.

[0007] In some embodiments, the method further includes:

[0008] After using the adjusted server configuration parameters for network transmission, obtain the corresponding transmission duration.

[0009] The training parameters are updated using the transmission duration and the current congestion index. The steps of training the network congestion control model with the training parameters and outputting the congestion judgment threshold are then performed to update the congestion judgment threshold. The updated congestion judgment threshold is used in the server configuration parameter adjustment process during the next network transmission.

[0010] In some embodiments, the network congestion control model is trained using training parameters, and a congestion determination threshold is output, including:

[0011] The network congestion control model is trained using training parameters to determine the relational formula of the network congestion control model;

[0012] Based on the relational formula of the network congestion control model and the preset target transmission duration, the corresponding congestion index is calculated, and the corresponding congestion index is determined as the congestion judgment threshold.

[0013] In some embodiments, the network congestion control model is a polynomial: y = a0 + a1x + a2x^2; where x is the transmission duration, y is the congestion index corresponding to the transmission duration, and a0, a1, and a2 are model coefficients.

[0014] In some embodiments, server configuration parameters include timeout duration;

[0015] Based on the relationship between the current congestion index and the congestion threshold, adjust the server configuration parameters, including:

[0016] If the current congestion index is greater than the congestion threshold, increase the timeout waiting time.

[0017] In some embodiments, the server configuration parameters also include the data transfer block size;

[0018] Adjusting server configuration parameters based on the relationship between the current congestion index and the congestion threshold also includes:

[0019] If the current congestion index is less than the congestion threshold, increase the data transmission block size.

[0020] In some embodiments, the method further includes:

[0021] In the event of a network transmission failure, obtain the network transmission information from the transmission protocol extension field and resume the transmission from the breakpoint based on the network transmission information; the network transmission information includes packet information, transmission offset, and file checksum.

[0022] In some embodiments, obtaining network transmission information from the transport protocol extension field and resuming interrupted transmissions based on the network transmission information includes:

[0023] Retrieve network transmission information from the transmission protocol extension field of the previous network transmission, perform file consistency verification based on the file checksum, and resume transmission from the breakpoint based on packet information and transmission offset if the verification passes; the packet information includes packet sequence number and packet size information.

[0024] Secondly, a network congestion control device is provided, the device comprising:

[0025] The model training module is used to train the network congestion control model using training parameters and output a congestion judgment threshold. The training parameters include multiple historical transmission durations and the congestion index corresponding to each historical transmission duration. The congestion judgment threshold is the congestion index corresponding to a preset target transmission duration.

[0026] The configuration adjustment module is used to adjust the server configuration parameters according to the relationship between the current congestion index and the congestion judgment threshold, and to use the adjusted server configuration parameters for network transmission.

[0027] In some embodiments, the device further includes:

[0028] The parameter acquisition module is used to obtain the transmission duration corresponding to the network transmission after the adjusted server configuration parameters are used for network transmission.

[0029] The training update module is used to update the training parameters using the transmission duration and the current congestion index, and to perform the steps of training the network congestion control model using the training parameters and outputting the congestion judgment threshold to update the congestion judgment threshold; the updated congestion judgment threshold is used in the server configuration parameter adjustment process during the next network transmission.

[0030] In some embodiments, the model training module is specifically used for:

[0031] The network congestion control model is trained using training parameters to determine the relational formula of the network congestion control model;

[0032] Based on the relational formula of the network congestion control model and the preset target transmission duration, the corresponding congestion index is calculated, and the corresponding congestion index is determined as the congestion judgment threshold.

[0033] In some embodiments, the network congestion control model is a polynomial: y = a0 + a1x + a2x^2; where x is the transmission duration, y is the congestion index corresponding to the transmission duration, and a0, a1, and a2 are model coefficients.

[0034] In some embodiments, server configuration parameters include timeout duration;

[0035] The configuration adjustment module is specifically used for:

[0036] If the current congestion index is greater than the congestion threshold, increase the timeout waiting time.

[0037] In some embodiments, the server configuration parameters also include the data transfer block size;

[0038] The configuration adjustment module is also used for:

[0039] If the current congestion index is less than the congestion threshold, increase the data transmission block size.

[0040] In some embodiments, the device further includes:

[0041] The breakpoint resume module is used to obtain network transmission information from the transmission protocol extension field in the event of network transmission failure, and to resume the transmission based on the network transmission information; the network transmission information includes packet information, transmission offset and file checksum.

[0042] In some embodiments, the breakpoint resume module is specifically used for:

[0043] Retrieve network transmission information from the transmission protocol extension field of the previous network transmission, perform file consistency verification based on the file checksum, and resume transmission from the breakpoint based on packet information and transmission offset if the verification passes; the packet information includes packet sequence number and packet size information.

[0044] Thirdly, this application also provides a network device, including a memory, a transceiver, and a processor. The memory is used to store computer programs; the transceiver is used to send and receive data under the control of the processor; and the processor is used to read the computer program from the memory and perform the following operations:

[0045] The network congestion control model is trained using training parameters, and the congestion judgment threshold is output. The training parameters include multiple historical transmission durations and the congestion index corresponding to each historical transmission duration. The congestion judgment threshold is the congestion index corresponding to a preset target transmission duration.

[0046] Based on the relationship between the current congestion index and the congestion threshold, adjust the server configuration parameters and use the adjusted server configuration parameters for network transmission.

[0047] In some embodiments, the operation further includes:

[0048] After using the adjusted server configuration parameters for network transmission, obtain the corresponding transmission duration.

[0049] The training parameters are updated using the transmission duration and the current congestion index. The steps of training the network congestion control model with the training parameters and outputting the congestion judgment threshold are then performed to update the congestion judgment threshold. The updated congestion judgment threshold is used in the server configuration parameter adjustment process during the next network transmission.

[0050] In some embodiments, the network congestion control model is trained using training parameters, and a congestion determination threshold is output, including:

[0051] The network congestion control model is trained using training parameters to determine the relational formula of the network congestion control model;

[0052] Based on the relational formula of the network congestion control model and the preset target transmission duration, the corresponding congestion index is calculated, and the corresponding congestion index is determined as the congestion judgment threshold.

[0053] In some embodiments, the network congestion control model is a polynomial: y = a0 + a1x + a2x^2; where x is the transmission duration, y is the congestion index corresponding to the transmission duration, and a0, a1, and a2 are model coefficients.

[0054] In some embodiments, server configuration parameters include timeout duration;

[0055] Based on the relationship between the current congestion index and the congestion threshold, adjust the server configuration parameters, including:

[0056] If the current congestion index is greater than the congestion threshold, increase the timeout waiting time.

[0057] In some embodiments, the server configuration parameters also include the data transfer block size;

[0058] Adjusting server configuration parameters based on the relationship between the current congestion index and the congestion threshold also includes:

[0059] If the current congestion index is less than the congestion threshold, increase the data transmission block size.

[0060] In some embodiments, the operation further includes:

[0061] In the event of a network transmission failure, obtain the network transmission information from the transmission protocol extension field and resume the transmission from the breakpoint based on the network transmission information; the network transmission information includes packet information, transmission offset, and file checksum.

[0062] In some embodiments, obtaining network transmission information from the transport protocol extension field and resuming interrupted transmissions based on the network transmission information includes:

[0063] Retrieve network transmission information from the transmission protocol extension field of the previous network transmission, perform file consistency verification based on the file checksum, and resume transmission from the breakpoint based on packet information and transmission offset if the verification passes; the packet information includes packet sequence number and packet size information.

[0064] Fourthly, this application also provides a processor-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the network congestion control methods described in the first aspect.

[0065] Fifthly, this application also provides a computer program product, including a computer program for causing a processor to execute any of the network congestion control methods described in the first aspect above.

[0066] The aforementioned network congestion control method, apparatus, network device, and readable storage medium train a network congestion control model using training parameters, outputting a congestion judgment threshold. These training parameters include multiple historical transmission durations and corresponding congestion indices. The congestion judgment threshold is the congestion index corresponding to a preset target transmission duration. Then, based on the relationship between the current congestion index and the congestion judgment threshold, server-side configuration parameters are adjusted, and network transmission is performed using the adjusted server-side configuration parameters. In other words, the network congestion control method provided in this application can perform self-learning training based on historical transmission durations and corresponding congestion indices to derive the influence of the congestion index on transmission duration, and output a reasonable judgment threshold for network congestion determination. This congestion judgment threshold serves as the basis for determining whether the network is congested. Combined with the congestion index of the current network transmission environment, the network transmission server configuration parameters are adjusted to achieve network congestion control and improve network transmission stability. Attached Figure Description

[0067] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0068] Figure 1 A flowchart illustrating a network congestion control method provided in an embodiment of this application;

[0069] Figure 2 A flowchart illustrating yet another network congestion control method provided in this application embodiment;

[0070] Figure 3 A flowchart illustrating another network congestion control method provided in an embodiment of this application;

[0071] Figure 4 This is a schematic diagram of an overall PXE startup optimization scheme provided in an embodiment of this application;

[0072] Figure 5 A schematic diagram of a congestion control training process provided in an embodiment of this application;

[0073] Figure 6 This is a schematic diagram of a service configuration adjustment process provided in an embodiment of this application;

[0074] Figure 7 A schematic diagram of a network device structure provided in an embodiment of this application;

[0075] Figure 8 A structural block diagram of a network congestion control device provided in an embodiment of this application; Detailed Implementation

[0076] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0077] During network transmission, the actual network environment often differs from the theoretical scenario in various ways. Due to the complexity of the network environment, network congestion may occur, leading to network transmission delays or even interruptions. This will significantly impact the smooth operation of services carried by the network transmission, especially for services with long processing times, where retransmission due to interruption is costly. For example, PXE network booting using the TFTP protocol may experience slow startup or even failure due to unstable network transmission.

[0078] With the rapid development of network technology, booting and managing computer systems via the network has become a viable solution, and PXE boot technology has gradually developed in this context. PXE boot allows a computer to obtain the resources needed for booting, including the operating system bootloader, installation files, and other necessary configuration information, from a server over the network during startup. This boot method makes the deployment of computer systems independent of physical storage media, improving deployment efficiency and flexibility.

[0079] Due to considerations of cost reduction and flexible configuration, the combination of a main control storage board and multiple baseband boards without storage media is increasingly meeting the needs of 5G base station systems. The baseband board's boot file can be loaded from the main control board to the baseband board via the network and successfully boot. This relies on PXE network booting, which downloads the operating system file stored on the main control board to the baseband board for execution without relying on the baseband board's data storage medium. The advantages of network booting are obvious: when configuring multiple baseband boards, it's unnecessary to back up the system file for each board; only a unified system file stored on the main control board needs to be obtained via the network. However, PXE network booting is a double-edged sword, bringing many drawbacks in practical applications. Because the operating system file is not permanently stored on the current board's storage medium, but relies entirely on retrieving the system file via the network for booting, the success rate of the baseband board booting directly affects the normal operation of the entire base station system. However, network instability can cause delays or transmission interruptions, leading to slow or even failed baseband board booting.

[0080] Specifically, the PXE boot protocol consists of a client and a server. The PXE client resides in the ROM (Read-Only Memory) of the network card. During boot, the client BIOS (Basic Input / Output System) loads the PXE client into memory for execution. The DHCP client broadcasts an IP address request message to the PXE server. The DHCP server checks the client's legitimacy; if successful, it returns the client's IP address and the location of the boot file. The PXE client then requests files from the server's TFTP service. The PXE client sequentially obtains the executable file, configuration file, kernel file, root file system, etc., via TFTP, ultimately booting the Linux kernel. In the base station system, the main control board acts as the server, and the baseband board as the client. The main control board deploys DHCP and TFTP services, while the baseband board's network card integrates the PXE client. The configuration information for both the PXE server and client in this process is typically pre-defined. If a transmission interruption occurs due to network issues, the boot process will fail. The client needs to re-initiate the entire process. Under ideal network conditions, a complete PXE interaction for a 180MB root file system takes approximately 15 seconds. Therefore, network instability significantly impacts the stable operation of network-based services.

[0081] Therefore, improving the stability of network transmission has become increasingly challenging.

[0082] The network congestion control method, apparatus, network device, and readable storage medium provided in this application train a network congestion control model using training parameters, outputting a congestion judgment threshold. The training parameters include multiple historical transmission durations and corresponding congestion indices for each historical transmission duration. The congestion judgment threshold is the congestion index corresponding to a preset target transmission duration. Then, based on the relationship between the current congestion index and the congestion judgment threshold, server-side configuration parameters are adjusted, and network transmission is performed using the adjusted server-side configuration parameters. In other words, the network congestion control method provided in this application can perform self-learning training based on historical transmission durations and corresponding congestion indices to derive the influence of the congestion index on transmission duration, and output a reasonable judgment threshold for network congestion determination. This congestion judgment threshold serves as the basis for determining whether the network is congested. Combined with the congestion index of the current network transmission environment, the network transmission server configuration parameters are adjusted to achieve network congestion control and improve network transmission stability. When applied to PXE network startup, it can improve startup stability and success rate.

[0083] In one embodiment of this application, such as Figure 1 As shown, a network congestion control method is provided. Taking the application of this method to a network device as an example, the network device carries a network transmission server, such as a TFTP server. This network device can be a server, a base station, or the main control board within a base station. The network congestion control method may include the following steps:

[0084] Step 101: Train the network congestion control model using training parameters and output the congestion judgment threshold.

[0085] The training parameters include multiple historical transmission durations and the congestion index corresponding to each historical transmission duration; the congestion judgment threshold is the congestion index corresponding to the preset target transmission duration.

[0086] For example, a data pool can be established to store various relevant information from historical network transmissions. The data pool can be understood as a configuration parameter structure. Besides variables such as the index used to maintain records, key members of this structure may include transmission duration, the congestion index corresponding to the transmission duration, server configuration parameters, and congestion thresholds. It may also include raw network environment information data (also called congestion factors) used to analyze and calculate the congestion index. It should be noted that besides the data pool method, other data formats can also be used to record and store relevant information; this application does not impose any restrictions, but uses a data pool as an example for illustration. Optionally, for each network transmission, various relevant information, such as transmission duration and the corresponding congestion index, can be saved to the data pool. This updates the training parameter set in the data pool, forming multiple configuration entries, each of which stores a corresponding set of relevant information.

[0087] The transmission duration refers to the time taken for successful network transmission using the corresponding server configuration parameters. In PXE network startup, it can refer to the TFTP interaction time. The congestion index corresponding to the transmission duration is the network congestion index when using the corresponding server configuration parameters for network transmission. The congestion index indicates the degree of network congestion and can be calculated by analyzing network environment information data.

[0088] Multiple historical transmission durations and their corresponding congestion indices from the data pool are input into the network congestion control model for self-learning training. This process reveals the influence of the congestion index on transmission duration. Combined with a preset target transmission duration, a network congestion threshold is output, which is the congestion index corresponding to the preset target transmission duration. This congestion threshold can be used as the basis for determining whether the network is congested.

[0089] In some embodiments, training parameters are used to train the network congestion control model and output a congestion determination threshold, specifically including the following steps:

[0090] The network congestion control model is trained using training parameters to determine the relational formula of the network congestion control model;

[0091] Based on the relational formula of the network congestion control model and the preset target transmission duration, the corresponding congestion index is calculated, and the corresponding congestion index is determined as the congestion judgment threshold.

[0092] In this embodiment, multiple historical transmission durations and their corresponding congestion indices from the data pool are input into the network congestion control model for self-learning training. This process reveals the influence of the congestion index on transmission duration, thus determining the model coefficients and formulas of the network congestion control model. A preset target transmission duration is then substituted into these formulas to calculate the corresponding congestion index, which serves as the network congestion threshold. The target transmission duration is the tolerance value for transmission duration in a critical state. As long as the transmission duration is controlled within this range, the network transmission is considered stable and congestion-free. Therefore, the congestion index corresponding to this transmission duration can be used as the threshold for determining whether congestion has occurred. The size of this target transmission duration is related to the file size requirements of the specific service. Different file sizes require different target transmission durations, which can be preset based on practical experience. For example, in PXE startup of a base station system, if the average recovery time for congestion startup is approximately 20 seconds, 18 seconds can be selected as the target transmission duration.

[0093] In some embodiments, the network congestion control model can be a polynomial, such as: y = a0 + a1x + a2x^2; where x is the transmission duration, y is the congestion index corresponding to the transmission duration, and a0, a1, and a2 are model coefficients.

[0094] For example, a polynomial fitting algorithm (based on the least squares method) can be implemented using a programming language. The least squares method is used to find the coefficients a0, a1, and a2 of the polynomial, minimizing the sum of squared errors between the polynomial and the sampling points (transmission duration and congestion index) in the actual data pool. Once the model coefficients are determined, the relationship of the network congestion control model can be derived.

[0095] It should be noted that the network congestion control model can also be other relational expressions, as long as they can be used to characterize the relationship between transmission time and congestion index. This application does not impose any specific restrictions.

[0096] Optionally, in the early stage of basic data research, data samples can be obtained based on packet injection experiments. Under the default server configuration, the congestion index of successful transmission and the actual transmission time are exponentially related. By training the exponential model, an initial congestion judgment threshold can be obtained.

[0097] Step 102: Based on the relationship between the current congestion index and the congestion determination threshold, adjust the server configuration parameters and use the adjusted server configuration parameters for network transmission.

[0098] When server-side configuration parameters need adjustment, such as before network transmission, these parameters can be adjusted based on the current congestion index and the congestion threshold output by the congestion control model. In this embodiment, for example, the current network environment status (e.g., busy or idle) can be determined by the relationship between the current congestion index and the congestion threshold. Based on the determination result, the corresponding server-side configuration parameters are adjusted. After the parameter adjustments take effect, network transmission is performed using the adjusted server-side configuration parameters to achieve network congestion control.

[0099] In this embodiment, the server configuration parameters of the network transmission server may include multiple parameters, such as network timeout waiting time and data transmission block size. Based on the determination result, one or more of these parameters can be adjusted. Network transmission is performed using the adjusted server configuration parameters, which include both the adjusted parameters and the unadjusted parameters.

[0100] For example, server configuration parameters can be adjusted according to a preset granularity. It should be noted that the granularity is generally set quite fine; the purpose of a small granularity is to avoid drastic fluctuations that could affect system stability, aiming to train an optimal threshold and prevent oscillating adjustments. Changing server configuration parameters assumes that the original configuration is also likely to succeed; adjustments can further improve transmission stability.

[0101] Optionally, if transmission fails using the adjusted configuration parameters, a post-processing procedure can be improved for self-recovery: after detecting a failure, current network environment data is collected again, the congestion index is calculated, data pool entries are compared, and the closest applicable server configuration parameters are matched. Optionally, if self-recovery fails, the interrupted transmission function can be used for transmission after restarting; the specific process is described in the following embodiments.

[0102] The network congestion control method provided in this application trains a network congestion control model using training parameters from a data pool, outputting a congestion judgment threshold. The training parameters include multiple transmission durations and corresponding congestion indices. Then, based on the current congestion index and the congestion judgment threshold, the server-side configuration parameters are adjusted, and network transmission is performed using these adjusted parameters to achieve network congestion control. In other words, the network congestion control method provided in this application can perform self-learning training based on multiple historical transmission durations and their corresponding congestion indices to derive the influence of the congestion index on transmission duration, and output a reasonable network congestion judgment threshold. This congestion judgment threshold serves as the basis for determining whether the network is congested. Combined with the current network transmission environment's congestion index, the network transmission server configuration parameters are adjusted to achieve network congestion control and improve network transmission stability.

[0103] Furthermore, using the relationship between transmission duration and congestion index as the training object, and then deriving the congestion index as the congestion threshold, the training model is binary, making modeling simple and easy to implement. If congestion control parameters were directly trained, with multiple parameters, the modeling difficulty would inevitably increase. Moreover, the congestion threshold is based on the results of large-scale successful record analysis in the past, which can effectively reflect the environmental state of network transmission, thereby controlling network congestion and improving network transmission stability.

[0104] In some embodiments, the method further includes:

[0105] After performing network transmission using the adjusted server configuration parameters, obtain the transmission duration corresponding to the network transmission.

[0106] The training parameters are updated using the transmission duration and the current congestion index. The steps of training the network congestion control model with the training parameters and outputting the congestion judgment threshold are then performed to update the congestion judgment threshold. The updated congestion judgment threshold is used in the server configuration parameter adjustment process during the next network transmission.

[0107] For example, after adjusting the server-side configuration parameters, the adjusted server-side configuration parameters and the corresponding current congestion index can be placed into a new identical entry in the data pool. When using the adjusted server-side configuration parameters for network transmission, after obtaining the corresponding transmission duration, the current server-side configuration parameters can be read. This configuration is used as an input parameter to match the records in the data pool, and the congestion index corresponding to the entry is matched. The transmission duration is then placed as a new historical transmission duration into the corresponding server-side configuration parameter and congestion index entry data chain to update the historical network transmission data in the data pool.

[0108] In this embodiment, after successful network transmission using the adjusted server-side configuration parameters, the corresponding network transmission duration can be obtained. This transmission duration, along with the current congestion index, is used to update the training parameters. The updated training parameters are then used to train the network congestion control model, outputting an updated congestion threshold. During the next network transmission, the updated congestion threshold is compared with the current congestion index to determine its magnitude, adjusting the server-side configuration parameters accordingly. The adjusted server-side configuration parameters are then used for the current network transmission. After each successful network transmission, the corresponding transmission duration and congestion index are obtained to update the training parameters and train the congestion control model, forming a closed-loop, continuously feedback training process that makes the trained congestion threshold more reasonable and effective.

[0109] In this embodiment, the process of training the network congestion control model using the updated training parameters can be found in the relevant description of step 101 in the aforementioned embodiment, and will not be repeated here.

[0110] The network congestion control method provided in this embodiment collects real-time network environment data to obtain the current congestion index. Based on the current congestion index and a congestion threshold, it adjusts the server-side configuration parameters. After network transmission using the adjusted server-side configuration parameters, it obtains the corresponding transmission duration. This transmission duration, along with the current congestion index, is used to update the training parameters. The updated training parameters are then used to retrain the network congestion control model, outputting an updated congestion threshold. This updated congestion threshold is then used to adjust the server-side configuration parameters for the next network transmission. This provides practical and effective training parameters for the congestion control model, performs closed-loop verification of the congestion control's rationality, and continuously feeds back to tame the congestion threshold, making it more effective. This better controls the rationality of server-side configuration parameter adjustments and continuously optimizes the effectiveness of congestion control.

[0111] Optionally, before each network transmission, the server configuration parameters can be adjusted based on the current congestion index and the congestion judgment threshold. Then, the network transmission is performed using the adjusted server configuration parameters to obtain the transmission duration corresponding to the network transmission. The congestion control model is trained using the transmission duration and the corresponding current congestion index in combination with historical training parameters to continuously correct the congestion control threshold and make it more reasonable.

[0112] The network congestion control method provided in this embodiment performs self-learning training on the network congestion control model using the latest training parameters after each network transmission. This allows for timely and effective training and updating of the congestion judgment threshold, making it more effective and thus better controlling the rationality of server-side configuration parameter adjustments. Furthermore, before each network transmission, timely adjustments to the server-side configuration parameters based on the current congestion index better adapt to the current network transmission environment and improve transmission stability. Adjusting server-side configuration parameters before network transmission and promptly feeding back the latest training parameters after transmission constitutes a closed-loop adjustment mode of continuous learning and training. This closed-loop verification of the congestion control's rationality and continuous feedback to refine the congestion judgment threshold makes it more effective, thereby better controlling the rationality of server-side configuration parameter adjustments and continuously optimizing the effectiveness of congestion control.

[0113] Please refer to Figure 2 In an optional embodiment of this application, a network congestion control method is provided, which may include the following steps:

[0114] Step 201: Train the network congestion control model using training parameters and output the congestion judgment threshold; the training parameters include multiple historical transmission durations and the congestion index corresponding to each historical transmission duration; the congestion judgment threshold is the congestion index corresponding to the preset target transmission duration.

[0115] Step 202: Based on the relationship between the current congestion index and the congestion judgment threshold, adjust the server configuration parameters and use the adjusted server configuration parameters for network transmission.

[0116] Step 203: Obtain the transmission duration corresponding to the network transmission.

[0117] Step 204: Update the training parameters using the transmission duration and the current congestion index, and execute the step of training the network congestion control model using the training parameters and outputting the congestion judgment threshold to update the congestion judgment threshold; the updated congestion judgment threshold is used in the server configuration parameter adjustment process during the next network transmission.

[0118] In this application embodiment, the specific implementation details of each of the above steps can be found in the relevant descriptions of the corresponding embodiments above, and will not be repeated here.

[0119] The network congestion control method provided in this embodiment is a closed-loop adjustment mode that continuously learns and trains. It verifies the rationality of congestion control in a closed loop, continuously feeds back to tame the congestion judgment threshold, making it more effective, thereby better controlling the rationality of server configuration parameter adjustments and continuously optimizing the effectiveness of congestion control.

[0120] In some embodiments, server configuration parameters may include timeout duration. Adjusting server configuration parameters based on the current congestion index and congestion determination threshold may include:

[0121] If the current congestion index is greater than the congestion threshold, increase the timeout waiting time.

[0122] Among them, the timeout waiting time, for example, is usually set between 5 and 30 seconds. If the timeout waiting time is set too short, it will cause the system to exit prematurely when the network is slightly unstable, which will affect the success rate. If it is set too long, the waiting time will be too long in the presence of network abnormalities, which will reduce the overall system response speed.

[0123] If the current congestion index is greater than the congestion threshold, it indicates that the network is relatively busy and there may be network congestion. The timeout waiting time on the server side should not be too short. The timeout waiting time can be adjusted according to the preset granularity to adjust the network transmission service to a conservative solution, which aims to improve the tolerance and stability of transmission failure and improve the success rate of single transmission.

[0124] The network congestion control method provided in this application improves the stability of network transmission by increasing the timeout waiting time when the network is busy.

[0125] In some embodiments, server configuration parameters may also include data transfer block size;

[0126] Adjusting server configuration parameters based on the relationship between the current congestion index and the congestion threshold can also include:

[0127] If the current congestion index is less than the congestion threshold, increase the data transmission block size.

[0128] Data transfer block size, which is the size of each data block transmitted, determines the transmission speed and throughput.

[0129] If the current congestion index is less than the congestion threshold, it means that the network is relatively idle. At this time, the data transmission block size can be increased according to the preset granularity, with the aim of improving transmission efficiency during idle periods.

[0130] The network congestion control method provided in this application can improve the stability of network transmission by increasing the timeout waiting time when the network is busy, and can also improve the transmission efficiency of network transmission by increasing the data transmission block size when the network is idle. This allows the network transmission to be flexibly adjusted according to the current actual network environment, thereby improving the transmission performance of the network.

[0131] Please refer to Figure 3 In an optional embodiment of this application, a network congestion control method is provided, which may include the following steps:

[0132] Step 301: Train the network congestion control model using training parameters and output the congestion judgment threshold.

[0133] The training parameters include multiple historical transmission durations and the congestion index corresponding to each historical transmission duration; the congestion judgment threshold is the congestion index corresponding to the preset target transmission duration.

[0134] Step 302: If the current congestion index is greater than the congestion determination threshold, increase the timeout waiting time; if the current congestion index is less than the congestion determination threshold, increase the data transmission block size.

[0135] Step 303: Use the adjusted server configuration parameters for network transmission.

[0136] The specific implementation process of each step in this embodiment can be found in the aforementioned related embodiments, and will not be repeated here.

[0137] The network congestion control method provided in this application can effectively improve the stability and efficiency of network transmission and enhance network transmission performance by increasing the timeout waiting time when the network is busy and increasing the data transmission block size when the network is idle.

[0138] In any implementation, the current congestion index can be determined in the following way:

[0139] Obtain current environmental network information data, and determine the current congestion index based on the obtained data. The environmental network information data includes at least one of the following:

[0140] Network bandwidth data: indicates the reception and transmission statistics of the corresponding network interface.

[0141] Network error messages; indicating error messages related to network interfaces, including network errors, packet loss, and retransmission information.

[0142] Size of files to be transferred: Indicates the size of the relevant files to be transferred in the transfer directory (e.g., the tftpboot directory).

[0143] Processor load data: Indicates processor load. For example, when applied to the main control board of a base station system, it indicates the CPU load of the main control board.

[0144] In some alternative implementations, in order to more accurately and effectively reflect the current network environment, current network information data can be acquired periodically, and the acquired current network information data can be classified, analyzed and integrated to determine the current congestion index.

[0145] In this embodiment, the current network environment information data, also known as congestion factor, can be periodically queried and parsed. This data may include network bandwidth data, network error information, the size of the file to be transmitted, and processor load data. The degree of network congestion can be analyzed, and corresponding configuration adjustments can be made. This network environment information reflects the overall external network environment between the sending and receiving ends. In addition to the network environment, factors such as the status of network device processors are also considered, allowing for a more accurate and effective reflection of the network transmission environment status.

[0146] In some implementations, the congestion index can be determined by weighting various congestion factors. As an example, the weighted calculation process might involve: collecting link data from a certain number of monitoring points over a specific time period (e.g., one monitoring point per second within a 10-minute timeframe), and reading all network environment information data from each monitoring point. The packet loss rate, link utilization, network latency, and corresponding CPU core utilization are then calculated. These factors are combined with the file size to be transmitted, and all monitoring points are assigned weights based on their proximity to the current time (closer points have higher weights). A weighted average is then calculated for each factor to obtain the current congestion index value.

[0147] It should be noted that, in addition to the above weighted calculation method to determine the congestion index, other data processing methods can also be used to determine the congestion index according to actual needs. This application does not impose specific restrictions on these methods.

[0148] The congestion index determination method provided in this embodiment calculates the congestion index by acquiring current network information data, so that the congestion index can effectively characterize the current network environment status.

[0149] In some embodiments, the network congestion control method may further include the following steps:

[0150] In the event of a network transmission failure, the network transmission information in the transmission protocol extension field is obtained, and the interrupted transmission is resumed based on the network transmission information. The transmission information may include packet information, transmission offset, and file checksum.

[0151] In some scenarios, the increased server-side timeout cannot compensate for the recovery of network conditions, ultimately leading to timeout and transmission interruption. This results in transmission service failures, such as network startup failures. This embodiment addresses this by adding network transmission information supporting resumable interruptions to the TFTP protocol extension fields of both the network transmission server and the client-side network device. This information could include packet information, transmission offset, and file checksum. This feature requires the client-side network device to support memory retention. When network congestion cannot be resolved by adjusting server-side configuration parameters, the network device can resume interrupted transmissions by obtaining the network transmission information from the protocol extension fields. The resumable interruption feature effectively compensates for the performance degradation caused by long system restart times and solves the problem of long retransmission times for incomplete transmissions.

[0152] In some embodiments, obtaining network transmission information from the transport protocol extension field and resuming interrupted transmissions based on the network transmission information may specifically include the following steps:

[0153] Retrieve network transmission information from the transmission protocol extension field of the previous network transmission, perform file consistency verification based on the file checksum, and resume transmission from the breakpoint based on packet information and transmission offset if the verification passes; the packet information includes packet sequence number and packet size information.

[0154] In this embodiment, the network device can obtain packet information, transmission offset, and file checksum from the recorded network transmission information. First, it verifies file consistency based on the file checksum. If the verification passes, it indicates that the file being transmitted this time is identical to the file transmitted last time. Then, it uses the packet information and detailed offset to transmit the remaining files. The specific execution process for resuming interrupted transmissions can be found in existing related technologies.

[0155] Taking the PXE network startup in a base station system as an example, the breakpoint resume feature is explained in detail:

[0156] In some situations, even increasing the TFTP server timeout cannot cover the recovery of network conditions, ultimately leading to timeout interruption of transmission and baseband board startup failure. The closed-loop design of the base station system includes a hardware watchdog function to restart the board and initiate a new network startup. The base station system relies on the hardware watchdog process to perform a hot reset operation on the failed baseband board. Some issues in this situation can be further optimized. For example, network instability after transmitting a portion of data may force an interruption, ultimately leading to a timeout. Here, it's necessary to combine this with the baseband board's data retention function in its reserved memory. That is, even if the board is only hot-reset without power loss, the data in the reserved memory outside of system management will not be refreshed. We utilize this feature to set the memory location corresponding to the files involved in each network startup and the TFTP client's stored data in the reserved memory area. Simultaneously, both the server and client TFTP protocol extension fields are updated with packet information, transmission offset, and file checksum that support resuming interrupted transfers. This way, after a transmission timeout requiring a board restart, the server board uses the packet information and detailed offset extracted from the previous transmission information in the PXE boot process log module to perform a file consistency check for resuming interrupted transfers. If the check passes, it means the file being transmitted this time is identical to the previous one. Then, it reads the transmission offset from the extension bits of the previously interrupted TFTP interaction data in the PXE boot process log module to determine the location of the previous interruption. The baseband board restarts, and the TFTP client program first checks the resuming validity flag in the retained memory. If the flag is present, it means the content transmitted in the previous data area is usable. The server and client then begin retransmitting the system file from the next byte after the last interruption point, continuing until the end of the file.

[0157] The breakpoint resume function provided in this application can effectively compensate for the extra time introduced by board restart when the transmission scenario is constantly deteriorating and cannot be solved by simply adjusting the configuration parameters of the network transmission server. This ensures the overall success rate of the system without sacrificing too much efficiency.

[0158] Please see Figures 4 to 6 To facilitate readers' understanding of the technical solutions provided in the embodiments of this application, the following will use PXE network boot as an example to provide some exemplary embodiments to illustrate the technical solutions provided in the embodiments of this application.

[0159] See Figure 4 This is a schematic diagram of an overall PXE startup optimization scheme provided in an embodiment of this application. From a software perspective, this technical solution can be mainly divided into three modules: a PXE startup recording module, a congestion control training module, and a network configuration adjustment module. Figure 4 The PXE startup record module not only provides complete startup location information but also serves as a feedback source for the data pool. It uses the service configuration of the current startup as a retrieval basis, compares it with the corresponding congestion index, and combines this with the startup time (corresponding to the transmission duration mentioned above) as input feedback to adjust the rationality of the training model in the congestion control training module. The congestion control module, based on the accumulated data pool from the startup records, continuously learns and adjusts its trained classification threshold (corresponding to the congestion judgment threshold mentioned above). This classification threshold also serves as the basis for the network configuration adjustment module to make configuration adjustments. The entire process of the network configuration adjustment module can be started asynchronously before startup, triggered manually. After the trigger switch is issued, it begins periodically collecting current network environment data. Through fixed-point collection and time-weighted calculation, it obtains the current congestion factor. Based on the classification threshold trained by the congestion control module, it makes corresponding configuration adjustments, which are applied in the immediate restart. Simultaneously, the adjusted configuration parameters and environmental parameters are stored as input in the data pool. From a macro perspective, these three modules constitute a closed-loop feedback adjustment system. The diagram uses three colors to distinguish the corresponding behaviors of the three modules.

[0160] The PXE startup process recording module is responsible for recording the current PXE configuration information, including the installation file path, file checksum, service status, network packet transmission and reception status, and filtering and storing network packets related to PXE interaction from all network interface packet transmission and reception information. This includes DHCP (Dynamic Host Configuration Protocol) packets, ARP (Address Resolution Protocol) packets, and TFTP (Trivial File Transfer Protocol) packets. The entire TFTP interaction process time in this stage serves as an important input for the congestion control training module. The file checksum and detailed TFTP packet information during the interaction process are crucial for extending the breakpoint resume function. The remaining recorded information contains all the basic data needed to troubleshoot PXE startup failures. This information includes the PXE startup environment and interaction data, allowing for quick problem location and analysis without needing to reproduce the issue, thus saving situations where there is a lack of on-site information.

[0161] See Figure 5 The diagram illustrates a congestion control training process provided in this embodiment. Each successful PXE process triggers entry into training mode. First, the time consumed by the last successful TFTP interaction is extracted from the PXE startup process recording module, and the current PXE server configuration parameters are read. This configuration is used as input parameters to match entries in the historical data pool, identifying the corresponding congestion index (corresponding to the current congestion index mentioned above) and the original historical data information used to divide the congestion index (i.e., bandwidth data, network error information, control board load, and congestion factors such as the size of system files to be transmitted). This data is then combined with historical training parameters for self-learning training. The congestion index and TFTP interaction time are passed as input to the self-learning module. Using the congestion factors of the corresponding entries in the historical data pool as data sources, the self-learning module derives the influence of congestion factors on interaction time, thereby continuously correcting the classification threshold for network congestion determination to make it more reasonable. This classification threshold is then passed to the network configuration adjustment module as the basis for determining whether the network is congested.

[0162] See Figure 6 The diagram illustrates a service configuration adjustment process provided in this application embodiment. When the network service parameter adjustment mode is triggered, the main program of the network status collection module starts running, periodically queries and parses the current network environment information data, analyzes the network congestion level, makes corresponding configuration adjustments, and stores the adjusted configuration list into a data pool for use by the training program. Network status information collection consists of the following parts:

[0163] 1. Bandwidth data: Periodically read the receive and transmit statistics of the corresponding network interface.

[0164] 2. Error messages: Periodically record error messages related to the network interface, including network errors, packet loss, and retransmission information.

[0165] 3. File size: Calculate the size of the relevant system files to be transferred in the tftpboot directory.

[0166] 4. Load data: Periodically read the CPU load of the main control board.

[0167] Network configuration adjustments involve classifying, analyzing, and integrating collected network status information. First, various error messages are filtered, focusing on frequent network errors, packet loss, and retransmission records. If no network error reports are found, the number of bytes received and sent per unit time is calculated. Combined with the size of the file to be transmitted, these factors are weighted to calculate the current network congestion index (see the related embodiments above for the specific calculation process). When the congestion index exceeds the congestion control classification threshold of the congestion control training module, the network is considered busy. When the network and load are congested, the server's timeout should not be too short. The transmission timeout duration is adjusted according to a predetermined granularity to improve stability. Conversely, when the congestion index from the network and load analysis falls within the idle range defined by the congestion control classification threshold, the data transmission block size is increased to maximize transmission efficiency when the network appears idle. After adjusting the server configuration parameters, new configuration entries are generated and saved to the data pool.

[0168] The PXE boot optimization scheme provided in this embodiment can improve network boot performance in all scenarios:

[0169] The specific scenarios are categorized as follows:

[0170] When the transmission scenario is relatively smooth, adjust the corresponding network startup service parameters to improve transmission efficiency and start up quickly; when the transmission scenario is relatively congested, adjust the network startup service to a conservative solution to increase the tolerance for transmission failures and increase the success rate of a single startup.

[0171] When the transmission scenario continues to deteriorate and cannot be solved by simply adjusting the network startup service parameters, the main control board enables the breakpoint resume function. After the baseband board system times out, it will trigger the hardware watchdog to reset the board. The breakpoint resume function can effectively compensate for the extra time introduced by the board restart, ensuring the overall success rate of the system without sacrificing too much efficiency.

[0172] The comprehensive solution also includes features to obtain complete log information related to network startup anomalies, similar to a network startup snapshot. This facilitates quick identification of the root cause of problems in scenarios where self-recovery is not possible, without relying on subsequent problem reproduction or scenario simulation for further analysis. This feature effectively covers network startup anomaly scenarios where self-rescue is not possible.

[0173] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps.

[0174] To implement the above embodiments, this application also provides a network device. This network device carries a network transmission server, such as a TFTP server. The network device can be, for example, a server, a base station, or the main control board within a base station.

[0175] Figure 7 This is a schematic diagram of the structure of a network device provided in an embodiment of this application.

[0176] like Figure 7 As shown, the network device may include a transceiver 700, a processor 710, and a memory 720.

[0177] in:

[0178] Transceiver 700 is used to receive and send data under the control of processor 710.

[0179] Among them, Figure 7In this context, the bus architecture can include any number of interconnected buses and bridges, specifically linking various circuits together, represented by one or more processors (processor 710) and memory (memory 720). The bus architecture can also link various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. The bus interface provides an interface. The transceiver 700 can be multiple elements, including transmitters and receivers, providing units for communicating with various other devices over transmission media, including wireless channels, wired channels, optical fibers, etc. The processor 710 is responsible for managing the bus architecture and general processing, and the memory 720 can store data used by the processor 710 during operation.

[0180] The processor 710 can be a central processing unit (CPU), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or a complex programmable logic device (CPLD). The processor can also adopt a multi-core architecture.

[0181] The processor 710 calls a computer program stored in memory and performs the following operations:

[0182] The network congestion control model is trained using training parameters, and the congestion judgment threshold is output. The training parameters include multiple historical transmission durations and the congestion index corresponding to each historical transmission duration. The congestion judgment threshold is the congestion index corresponding to a preset target transmission duration.

[0183] Based on the relationship between the current congestion index and the congestion threshold, adjust the server configuration parameters and use the adjusted server configuration parameters for network transmission.

[0184] In some embodiments, the operation further includes:

[0185] After using the adjusted server configuration parameters for network transmission, obtain the corresponding transmission duration.

[0186] The training parameters are updated using the transmission duration and the current congestion index. The steps of training the network congestion control model with the training parameters and outputting the congestion judgment threshold are then performed to update the congestion judgment threshold. The updated congestion judgment threshold is used in the server configuration parameter adjustment process during the next network transmission.

[0187] In some embodiments, the network congestion control model is trained using training parameters, and a congestion determination threshold is output, including:

[0188] The network congestion control model is trained using training parameters to determine the relational formula of the network congestion control model;

[0189] Based on the relational formula of the network congestion control model and the preset target transmission duration, the corresponding congestion index is calculated, and the corresponding congestion index is determined as the congestion judgment threshold.

[0190] In some embodiments, the network congestion control model is a polynomial: y = a0 + a1x + a2x^2; where x is the transmission duration, y is the congestion index corresponding to the transmission duration, and a0, a1, and a2 are model coefficients.

[0191] In some embodiments, server configuration parameters include timeout duration;

[0192] Based on the relationship between the current congestion index and the congestion threshold, adjust the server configuration parameters, including:

[0193] If the current congestion index is greater than the congestion threshold, increase the timeout waiting time.

[0194] In some embodiments, the server configuration parameters also include the data transfer block size;

[0195] Adjusting server configuration parameters based on the relationship between the current congestion index and the congestion threshold also includes:

[0196] If the current congestion index is less than the congestion threshold, increase the data transmission block size.

[0197] In some embodiments, the operation further includes:

[0198] In the event of a network transmission failure, obtain the network transmission information from the transmission protocol extension field and resume the transmission from the breakpoint based on the network transmission information; the network transmission information includes packet information, transmission offset, and file checksum.

[0199] In some embodiments, obtaining network transmission information from the transport protocol extension field and resuming interrupted transmissions based on the network transmission information includes:

[0200] Retrieve network transmission information from the transmission protocol extension field of the previous network transmission, perform file consistency verification based on the file checksum, and resume transmission from the breakpoint based on packet information and transmission offset if the verification passes; the packet information includes packet sequence number and packet size information.

[0201] Those skilled in the art will understand that Figure 7The structure shown is merely a block diagram of a portion of the structure related to the solution of this application and does not constitute a limitation on the network device to which the solution of this application is applied. Specific network devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0202] It should be noted that the network device provided in this application embodiment can implement all the method steps implemented in the above network congestion control method embodiment and can achieve the same technical effect. Here, the parts that are the same as those in the method embodiment and the beneficial effects will not be described in detail.

[0203] Based on the same inventive concept, this application also provides a network congestion control device for implementing the network congestion control method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in the network congestion control device embodiments provided below can be found in the limitations of the network congestion control method described above, and will not be repeated here.

[0204] In one exemplary embodiment, such as Figure 8 As shown, a network congestion control device 800 is provided, including: a model training module 801 and a configuration adjustment module 802.

[0205] The model training module 801 is used to train the network congestion control model using training parameters and output a congestion judgment threshold. The training parameters include multiple historical transmission durations and the congestion index corresponding to each historical transmission duration. The congestion judgment threshold is the congestion index corresponding to a preset target transmission duration.

[0206] The configuration adjustment module 802 is used to adjust the server configuration parameters according to the relationship between the current congestion index and the congestion judgment threshold, and to use the adjusted server configuration parameters for network transmission.

[0207] In some embodiments, the device further includes a parameter acquisition module 803 and a training update module 804.

[0208] The parameter acquisition module 803 is used to obtain the transmission duration corresponding to the network transmission after using the adjusted server configuration parameters for network transmission.

[0209] The training update module 804 is used to update the training parameters using the transmission duration and the current congestion index, and to perform the steps of training the network congestion control model using the training parameters and outputting the congestion judgment threshold to update the congestion judgment threshold; the updated congestion judgment threshold is used in the server configuration parameter adjustment process during the next network transmission.

[0210] In some embodiments, the model training module 801 is specifically used for:

[0211] The network congestion control model is trained using training parameters to determine the relational formula of the network congestion control model;

[0212] Based on the relational formula of the network congestion control model and the preset target transmission duration, the corresponding congestion index is calculated, and the corresponding congestion index is determined as the congestion judgment threshold.

[0213] In some embodiments, the network congestion control model is a polynomial: y = a0 + a1x + a2x^2; where x is the transmission duration, y is the congestion index corresponding to the transmission duration, and a0, a1, and a2 are model coefficients.

[0214] In some embodiments, server configuration parameters include timeout duration;

[0215] Configuration adjustment module 802 is specifically used for:

[0216] If the current congestion index is greater than the congestion threshold, increase the timeout waiting time.

[0217] In some embodiments, the server configuration parameters also include the data transfer block size;

[0218] Configuration adjustment module 802 is also used for:

[0219] If the current congestion index is less than the congestion threshold, increase the data transmission block size.

[0220] In some embodiments, the device further includes:

[0221] The breakpoint resume module 804 is used to obtain network transmission information from the transmission protocol extension field in the event of network transmission failure, and to resume the transmission based on the network transmission information; the network transmission information includes packet information, transmission offset and file checksum.

[0222] In some embodiments, the breakpoint resume module 804 is specifically used for:

[0223] Retrieve network transmission information from the transmission protocol extension field of the previous network transmission, perform file consistency verification based on the file checksum, and resume transmission from the breakpoint based on packet information and transmission offset if the verification passes; the packet information includes packet sequence number and packet size information.

[0224] It should be noted that the division of units in the embodiments of this application is illustrative and only represents one logical functional division. In actual implementation, other division methods may be used. Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units described above can be implemented in hardware or as software functional units.

[0225] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a processor-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application.

[0226] It should be noted that the apparatus provided in this embodiment of the invention can implement all the method steps implemented in the above method embodiment and can achieve the same technical effect. Therefore, the parts and beneficial effects that are the same as those in the method embodiment will not be described in detail here.

[0227] This application also provides a processor-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps provided in any of the above method embodiments.

[0228] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps provided in any of the above method embodiments.

[0229] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0230] The technical solutions provided in this application can be applied to a variety of systems. For example, applicable systems may include Long Term Evolution (LTE) systems, LTE Frequency Division Duplex (FDD) systems, LTE Time Division Duplex (TDD) systems, Long Term Evolution Advanced (LTE-A) systems, Universal Mobile Telecommunications System (UMTS), Worldwide Interoperability for Microwave Access (WiMAX) systems, 5G New Radio (NR) systems, and their evolved communication systems. These systems may include terminal equipment and network equipment. The systems may also include a core network component, such as Evolved Packet System (EPS) or 5G systems (5GS).

[0231] The network device involved in this application embodiment can be a base station, which may include multiple cells providing services to terminals. Depending on the specific application, the base station may also be called an access point, or a device in the access network that communicates with wireless terminal devices through one or more sectors on the air interface, or other names. The network device can be used to exchange received air frames with Internet Protocol (IP) packets, acting as a router between the wireless terminal device and the rest of the access network, where the rest of the access network may include an Internet Protocol (IP) communication network. The network device can also coordinate the attribute management of the air interface. For example, the network device involved in this application embodiment can be an evolved Node B (eNB or e-NodeB) in a long term evolution (LTE) system, a 5G base station (gNB) in a next generation system, or a Home evolved Node B (HeNB), relay node, femto, pico, network testing equipment, etc., and is not limited in this application embodiment. In some network architectures, network devices may include centralized unit (CU) nodes and distributed unit (DU) nodes, which may also be geographically separated.

[0232] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0233] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A network congestion control method, characterized in that, Applied to network devices, the method includes: The network congestion control model is trained using training parameters, and a congestion judgment threshold is output. The training parameters include multiple historical transmission durations and congestion indices corresponding to each historical transmission duration. The congestion judgment threshold is the congestion index corresponding to a preset target transmission duration. Based on the relationship between the current congestion index and the congestion determination threshold, the server configuration parameters are adjusted, and the adjusted server configuration parameters are used for network transmission.

2. The method according to claim 1, characterized in that, The method further includes: After performing network transmission using the adjusted server configuration parameters, the transmission duration corresponding to the network transmission is obtained. The training parameters are updated using the transmission duration and the current congestion index, and the steps of training the network congestion control model using the training parameters and outputting the congestion determination threshold are performed to update the congestion determination threshold; the updated congestion determination threshold is used in the server configuration parameter adjustment process during the next network transmission.

3. The method according to claim 1 or 2, characterized in that, The step of training the network congestion control model using training parameters and outputting a congestion determination threshold includes: The network congestion control model is trained using the training parameters to determine the relational expression of the network congestion control model; Based on the relationship of the network congestion control model and the preset target transmission duration, the corresponding congestion index is calculated, and the corresponding congestion index is determined as the congestion judgment threshold.

4. The method according to claim 3, characterized in that, The network congestion control model is a polynomial: y = a0 + a1x + a2x^2; where x is the transmission duration, y is the congestion index corresponding to the transmission duration, and a0, a1, and a2 are model coefficients.

5. The method according to claim 1, characterized in that, The server-side configuration parameters include the timeout waiting duration; The step of adjusting server configuration parameters based on the relationship between the current congestion index and the congestion determination threshold includes: If the current congestion index is greater than the congestion determination threshold, the timeout waiting time is increased.

6. The method according to claim 5, characterized in that, The server configuration parameters may also include the data transfer block size; The step of adjusting the server configuration parameters based on the relationship between the current congestion index and the congestion determination threshold also includes: If the current congestion index is less than the congestion determination threshold, the data transmission block size is increased.

7. The method according to claim 1, characterized in that, The method further includes: In the event of a network transmission failure, the network transmission information in the transmission protocol extension field is obtained, and the interrupted transmission is resumed based on the network transmission information; the network transmission information includes packet information, transmission offset, and file checksum.

8. The method according to claim 7, characterized in that, The step of obtaining network transmission information from the transmission protocol extension field and performing breakpoint resumption based on the network transmission information includes: Obtain the network transmission information from the transmission protocol extension field in the previous network transmission, perform file consistency verification based on the file checksum, and if the verification passes, resume the transmission from the breakpoint based on the packet information and the transmission offset; wherein, the packet information includes packet sequence number and packet size information.

9. The method according to claim 1, characterized in that, The method further includes: Obtain current environmental network information data, and determine the current congestion index based on the obtained current environmental network information data; the environmental network information data includes at least one of the following: Network bandwidth data; Network error message; Size of the file to be transferred; Processor load data.

10. A network device, comprising a memory, a transceiver, and a processor, wherein the memory is used to store a computer program; the transceiver is used to send and receive data under the control of the processor; and the processor is used to read the computer program in the memory and perform the following operations: The network congestion control model is trained using training parameters, and a congestion judgment threshold is output. The training parameters include multiple historical transmission durations and congestion indices corresponding to each historical transmission duration. The congestion judgment threshold is the congestion index corresponding to a preset target transmission duration. Based on the relationship between the current congestion index and the congestion determination threshold, the server configuration parameters are adjusted, and the adjusted server configuration parameters are used for network transmission.

11. The network device according to claim 10, characterized in that, The operation also includes: After performing network transmission using the adjusted server configuration parameters, the transmission duration corresponding to the network transmission is obtained. The training parameters are updated using the transmission duration and the current congestion index, and the steps of training the network congestion control model using the training parameters and outputting the congestion determination threshold are performed to update the congestion determination threshold; the updated congestion determination threshold is used in the server configuration parameter adjustment process during the next network transmission.

12. The network device according to claim 10 or 11, characterized in that, The step of training the network congestion control model using training parameters and outputting a congestion determination threshold includes: The network congestion control model is trained using the training parameters to determine the relational expression of the network congestion control model; Based on the relationship of the network congestion control model and the preset target transmission duration, the corresponding congestion index is calculated, and the corresponding congestion index is determined as the congestion judgment threshold.

13. The network device according to claim 12, characterized in that, The network congestion control model is a polynomial: y = a0 + a1x + a2x^2; where x is the transmission duration, y is the congestion index corresponding to the transmission duration, and a0, a1, and a2 are model coefficients.

14. The network device according to claim 10, characterized in that, The server-side configuration parameters include the timeout waiting duration; The step of adjusting server configuration parameters based on the relationship between the current congestion index and the congestion determination threshold includes: If the current congestion index is greater than the congestion determination threshold, the timeout waiting time is increased.

15. The network device according to claim 14, characterized in that, The server-side configuration parameters also include the data transfer block size; The step of adjusting the server configuration parameters based on the relationship between the current congestion index and the congestion determination threshold also includes: If the current congestion index is less than the congestion determination threshold, the data transmission block size is increased.

16. The network device according to claim 10, characterized in that, The operation also includes: In the event of a network transmission failure, the network transmission information in the transmission protocol extension field is obtained, and the interrupted transmission is resumed based on the network transmission information; the network transmission information includes packet information, transmission offset, and file checksum.

17. The network device according to claim 16, characterized in that, The step of obtaining network transmission information from the transmission protocol extension field and performing breakpoint resumption based on the network transmission information includes: Obtain the network transmission information from the transmission protocol extension field in the previous network transmission, perform file consistency verification based on the file checksum, and if the verification passes, resume the transmission from the breakpoint based on the packet information and the transmission offset; wherein, the packet information includes packet sequence number and packet size information.

18. The network device according to claim 10, characterized in that, The operation also includes: Obtain current environmental network information data, and determine the current congestion index based on the obtained current environmental network information data; the environmental network information data includes at least one of the following: Network bandwidth data; Network error message; Size of the file to be transferred; Processor load data.

19. A network congestion control device, characterized in that, The device includes: The model training module is used to train the network congestion control model using training parameters and output a congestion judgment threshold. The training parameters include multiple historical transmission durations and congestion indices corresponding to each historical transmission duration. The congestion judgment threshold is the congestion index corresponding to a preset target transmission duration. The configuration adjustment module is used to adjust the server configuration parameters according to the relationship between the current congestion index and the congestion determination threshold, and to use the adjusted server configuration parameters for network transmission.

20. A processor-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 9.

Images