A Scheduling and Flow Control Method Based on FEC Coding Direction
By introducing scheduling and flow control methods in FEC encoding, the RAM behind the encoder is optimized, and the problems of large resource consumption and delay are solved, resource saving and delay reduction are achieved, and it is suitable for all FEC code models and improves data transmission Reliability and efficiency.
Patent Information
- Application Number
- CN202411331136.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-24
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-09-24
AI Technical Summary
FEC encoding has problems with large resource consumption and delay in data transmission. Especially in delay-sensitive application scenarios, the RAM introduced by multiple FEC modules increases area and delay, and the delay changes lead to a longer convergence time of the scheme.
Using the scheduling and flow control method based on the FEC encoding direction, the scheduler adjusts the encoding parameters according to the data state, the transmitter compresses the original data and inputs it to the encoder, and the encoder performs FEC encoding, optimizes the RAM behind the encoder, and saves Resources, reduce latency, and introduce threshold configurations for all FEC patterns.
It realizes resource savings, reduces delays, is suitable for all FEC code types, improves the reliability and efficiency of data transmission, reduces area increase, and simplifies the implementation process.
Smart Images

Figure CN118869153B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of data processing, and in particular to a scheduling and flow control method based on FEC coding direction. Background Art
[0002] FEC coding, or forward error correction coding, is an error correction coding technology. It is used to detect and correct errors during data transmission. By redundantly encoding the data at the sending end, the receiving end can still restore the original data when receiving partially damaged data. The main principle of this coding method is to process the original data at the sending end according to a predetermined algorithm, generate redundant data, and send the original data and redundant data together to the receiving end. After receiving the data, the receiving end uses these redundant data for error detection and correction. Compared with other error control coding, FEC coding has higher error correction capability and better performance. In communication systems, FEC coding can improve the reliability of data transmission and ensure the smooth transmission of information. Common FEC coding technologies include Hamming code, RS code, LDPC code, etc. They all use different algorithms to generate redundant data and use these redundant data for error detection and correction.
[0003] FEC implementation is often not placed in the same location as the scheduler. There are usually many delays between the two. Because the read and write delays increase the total path delay, if you consider application scenarios such as ECC, at least 3 delays will be introduced, which has a great impact on delay-sensitive applications. RAM will also increase the area, especially if multiple FEC modules are instantiated, then multiple RAMs will be added accordingly, and the area increase introduced is also considerable. Moreover, the depth of RAM is also related to the delay between the transmitter and the encoder. If the delay changes, the RAM must also be adjusted accordingly, which will lead to a longer convergence time for the solution. In practical applications, if the system needs to perform complex data processing, including data transmission and error correction, then FEC coding needs to introduce RAM to optimize the efficiency and reliability of data processing. However, increasing RAM will first increase the total path delay due to read and write delays. If you consider application scenarios such as ECC, at least 3 delays will be introduced, which has a great impact on delay-sensitive applications. RAM will also increase the area, especially if multiple FEC modules are instantiated, then multiple RAMs will be added accordingly, and the area increase introduced is also considerable. Moreover, the depth of RAM is also related to the delay between the sender and the encoder. If the delay changes, the RAM must also be adjusted accordingly, which will increase the convergence time of the solution. Summary of the invention
[0004] The purpose of the present invention is to provide a scheduling and flow control method based on the FEC coding direction, which can optimize the RAM behind the encoder, reduce the delay, introduce threshold configuration, and can be applicable to all FEC code types, solving the technical problem that RAM will cause an increase in area and a longer convergence time.
[0005] A scheduling and flow control method based on FEC coding direction, comprising:
[0006] The scheduler generates scheduling enable to the transmitter through a fixed pattern scheduling method and adjusts the encoding parameters according to the data status;
[0007] After receiving the scheduling enable, the transmitter compresses the original data to obtain compressed data;
[0008] The compressed data is input to the encoder after a delay, and the encoder performs FEC encoding on the compressed data and outputs the encoded data.
[0009] Preferably, the scheduler generates a scheduling enable to the transmitter by a fixed pattern scheduling method, including:
[0010] The scheduler includes a first counter, a second counter and a constant register;
[0011] The first counter cyclically counts according to the first counting upper limit to obtain a first counting value;
[0012] The first counter sends the first count value to the second counter and the constant register;
[0013] The constant register uses the first count value as an index, selects a corresponding bit value and sends it to the second counter;
[0014] The second counter cyclically counts according to the second counting upper limit, and generates a scheduling enable according to the corresponding bit value;
[0015] The second counter outputs the scheduling enable to the transmitter, and the monitoring tool starts timing.
[0016] Preferably, adjusting the encoding parameters according to the data state includes:
[0017] The encoding parameters include the first count upper limit, the second count upper limit and register parameters;
[0018] The data status includes the current FEC coding code type and coding operation status;
[0019] The current FEC coding code type is obtained through machine learning, and the first counting upper limit, the second counting upper limit and the register parameters are adjusted in combination with the coding running status.
[0020] Preferably, after receiving the scheduling enable, the transmitter compresses the original data to obtain compressed data, including:
[0021] Organizing and optimizing the original data to obtain optimized data;
[0022] compressing the optimized data to obtain compressed data;
[0023] Cache the redundant information removed during compression;
[0024] When the transmitter does not receive the scheduling enable, the transmitter stops sending the compressed data.
[0025] Preferably, compressing the optimized data to obtain compressed data comprises:
[0026] Dividing the optimized data into a training set and a validation set;
[0027] A data compression model is constructed, and the data compression model is expressed as:
[0028]
[0029] in, is the i-th data in the training set, L() is the loss function, β is the encoding parameter and decoding parameter, μ is the initial learning rate, In order to seek partial guidance, The restored input data is calculated by the data compression module for the i-th data in the training set;
[0030] Train the data compression model;
[0031] Use the trained compression model to compress data.
[0032] Preferably, the compressed data is input to the encoder after a delay, the encoder performs FEC encoding on the compressed data and outputs the encoded data, and the monitoring tool stops timing, including:
[0033] Matching delay parameters according to the scheduling enable;
[0034] inputting the compressed data into the encoder according to the delay parameter;
[0035] The encoder performs FEC encoding on the compressed data and outputs encoded data;
[0036] The monitoring tool stops timing and obtains the convergence time.
[0037] Preferably, after obtaining the convergence time, the method further comprises:
[0038] Feeding back the convergence time to the scheduler;
[0039] Predicting an estimated convergence time according to a code type of the FEC encoding;
[0040] Comparing the estimated convergence time with the convergence time to obtain the encoding running state;
[0041] When the convergence time is greater than the estimated convergence time, the encoding operation state is a state to be optimized;
[0042] When the convergence time is less than or equal to the estimated convergence time, the encoding operation state is a normal operation state;
[0043] When the encoding operation state is a state to be optimized, the first upper counting limit, the second upper counting limit and register parameters are readjusted.
[0044] Preferably, after outputting the coded data, the method further comprises checking the coded data:
[0045] The coded data is checked according to the original data:
[0046] Clustering algorithms are used to classify data;
[0047] Calculate the hash value of each data in different categories:
[0048]
[0049] represents the XOR vector of the hash value of the encoded data, is a random vector of the same length as the standard hash value;
[0050] comparing the hash value of the encoded data with the hash value of the standard encoded data;
[0051] If the hash value of the encoded data is not equal to the hash value of the standard encoded data, the encoding process data is lost;
[0052] When data loss is detected in the coded data, redundant information in the cache is extracted to repair the coded data.
[0053] A scheduling and flow control system based on FEC coding direction, comprising:
[0054] The scheduling module is used by the scheduler to generate scheduling enable to the transmitter through a fixed pattern scheduling method;
[0055] A compression module, used for compressing the original data to obtain compressed data after the transmitter receives the scheduling enable;
[0056] The encoding module is used to input the compressed data to the encoder after delay, and the encoder performs FEC encoding on the compressed data and outputs the encoded data.
[0057] An electronic device comprises a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the scheduling and flow control method based on FEC coding direction is implemented.
[0058] Compared with the prior art, the present invention has the following beneficial effects:
[0059] The present invention discloses a scheduling and flow control method based on FEC coding direction, the method comprising: a scheduler generates a scheduling enable to a transmitter through a fixed pattern scheduling method, and adjusts coding parameters according to data status; after receiving the scheduling enable, the transmitter compresses the original data to obtain compressed data; the compressed data is input to an encoder after a delay, and the encoder performs FEC coding on the compressed data and outputs the coded data. The present invention accurately controls the flow by designing a scheduler, only requires a small number of counters and registers, optimizes the RAM behind the encoder, saves resources, reduces delay, introduces threshold configuration, and can be applied to all FEC code types.
[0060] It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS
[0061] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.
[0062] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.
[0063] Figure 1 A flow chart of a scheduling and flow control method based on FEC coding direction provided by an embodiment of the present invention;
[0064] Figure 2 A schematic diagram of an application scenario of a traditional FEC encoding implementation method provided in an embodiment of the present invention;
[0065] Figure 3 A schematic diagram of a conventional FEC coding equal width domain segment insertion provided by an embodiment of the present invention;
[0066] Figure 4 A schematic diagram of data caching and bit width conversion implemented by a conventional FEC encoding provided in an embodiment of the present invention;
[0067] Figure 5 A schematic diagram of the scheduler position of a scheduling and flow control method based on FEC coding direction provided by an embodiment of the present invention;
[0068] Figure 6 A schematic diagram of a scheduler structure of a scheduling and flow control method based on FEC coding direction provided by an embodiment of the present invention;
[0069] Figure 7 A constant register diagram of a scheduling and flow control method based on FEC coding direction provided by an embodiment of the present invention
[0070] Figure 8 A schematic diagram of scheduler information flow of a scheduling and flow control method based on FEC coding direction provided by an embodiment of the present invention;
[0071] Fig. 9 A schematic diagram of a scheduling pattern finally output by a scheduling and flow control method based on FEC coding direction provided in an embodiment of the present invention;
[0072] Fig.10 The figure is a schematic diagram of the hardware structure of an electronic device of the present invention. DETAILED DESCRIPTION
[0073] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0074] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.
[0075] In addition, the descriptions of "first", "second", etc. in the present invention are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.
[0076] FEC coding, or forward error correction coding, is an error correction coding technology. It is used to detect and correct errors during data transmission. By redundantly encoding the data at the sending end, the receiving end can still restore the original data when receiving partially damaged data. The main principle of this coding method is to process the original data at the sending end according to a predetermined algorithm, generate redundant data, and send the original data and redundant data together to the receiving end. After receiving the data, the receiving end uses these redundant data for error detection and correction. Compared with other error control coding, FEC coding has higher error correction capability and better performance. In communication systems, FEC coding can improve the reliability of data transmission and ensure the smooth transmission of information. Common FEC coding technologies include Hamming code, RS code, LDPC code, etc. They all use different algorithms to generate redundant data and use these redundant data for error detection and correction.
[0077] FEC implementation is often not placed in the same location as the scheduler. There are usually many beats of delay between the two. The read and write delays increase the total path delay. If ECC and other application scenarios are considered, at least 3 beats of delay will be introduced, which has a great impact on delay-sensitive applications. RAM will also increase the area, especially if multiple FEC modules are instantiated, which will increase multiple RAM blocks accordingly, and the area increase introduced is also considerable. Moreover, the depth of RAM is also related to the delay between the transmitter and the encoder. If the delay changes, the RAM must also be adjusted accordingly, which will lead to a longer convergence time of the solution. In actual applications, if the system needs to perform complex data processing, including data transmission and error correction, then FEC coding needs to introduce RAM to optimize the efficiency and reliability of data processing.
[0078] In the process of data transmission, in order to ensure the integrity and reliability of the data, the original data is usually encoded. Among them, forward error correction coding (FEC) is a commonly used encoding method. It adds redundant information to the data so that even if some data errors occur during the transmission process, they can be corrected through the redundant information. However, traditional FEC encoding methods often have problems such as high resource consumption and high latency.
[0079] In view of these problems, the present invention proposes a scheduling and flow control method based on FEC coding direction. The core of the method is to accurately control the scheduling and coding process of data by designing a scheduler. Specifically, the scheduler will send a scheduling enable to the transmitter according to the state of the data and the transmission requirements. After receiving the scheduling enable, the transmitter will compress the original data to obtain compressed data. Such a design effectively reduces the amount of data and saves resources for subsequent coding and transmission.
[0080] The present invention discloses a scheduling and flow control method based on FEC coding direction, the method comprising: a scheduler generates a scheduling enable to a transmitter through a fixed pattern scheduling method, and adjusts coding parameters according to data status; after receiving the scheduling enable, the transmitter compresses the original data to obtain compressed data; the compressed data is input to an encoder after a delay, and the encoder performs FEC coding on the compressed data and outputs the coded data. The present invention accurately controls the flow by designing a scheduler, only requires a small number of counters and registers, optimizes the RAM behind the encoder, saves resources, reduces delay, introduces threshold configuration, and can be applied to all FEC code types.
[0081] Example 1
[0082] A scheduling and flow control method based on FEC coding direction, refer to Figure 1 ,include:
[0083] S100, the scheduler generates a scheduling enable to the transmitter through a fixed pattern scheduling method;
[0084] S200, after receiving the scheduling enable, the transmitter compresses the original data to obtain compressed data;
[0085] S300, the compressed data is input to the encoder after a delay, and the encoder performs FEC encoding on the compressed data and outputs the encoded data;
[0086] Through refined scheduling and flow control, the reliability and efficiency of data transmission are ensured;
[0087] The addition of the scheduler is the core part of the present invention, and the scheduler is responsible for sending the scheduling enable to the transmitter;
[0088] The scheduler can accurately adjust the encoding parameters based on the current data status, including the number and size of data packets and transmission requirements; these encoding parameters determine the compression ratio, error correction capability and transmission efficiency of the data during transmission;
[0089] After receiving the scheduling enable, the transmitter starts compressing the original data;
[0090] The compression process aims to reduce the size of data, reduce transmission costs and increase transmission speed. The characteristics of the sender data and the encoding parameters are compressed to ensure that the data size is minimized while maintaining data integrity.
[0091] The scheduler uses an adaptive algorithm to dynamically adjust the scheduling strategy according to real-time conditions and data traffic. The transmitter uses an efficient compression algorithm to reduce the amount of data while preserving the integrity of the data as much as possible. The encoder supports multiple FEC coding schemes to adapt to different error rates and transmission distance requirements. The method includes a mechanism to monitor the quality of data transmission in real time and make adjustments based on the feedback results.
[0092] This embodiment provides an implementation method of traditional FEC encoding:
[0093] refer to Figure 2 ,FEC implementation is often not placed in the same place as the scheduler, and there is usually a delay of many beats between the two;
[0094] For each input of a beat of data, the encoder will generate a certain bit width of the equal width domain segment. The length of this equal width domain segment varies according to the different encoding code types;
[0095] In this embodiment, the equal-width domain segment length of RSFEC (528, 514) is 140 bits, while the equal-width domain segment length of RSFEC (544, 514) is 300 bits;
[0096] In actual application, every time a segment of data is added, the corresponding equal width is inserted into the end of the data. For example, the equal width insertion of RSFEC (544, 514) is Figure 3 ;
[0097] refer to Figure 4 , the 5140-bit payload is not an integer multiple of the 64-bit data bus, and the 300-bit parity is not an integer multiple of 64-bit. However, the data output bus requires that each beat is 64 bits wide. The usual practice is to add a RAM behind the encoder to achieve data caching and bit width conversion;
[0098] Compared with the traditional FEC encoding implementation method, the embodiment of the present invention accurately controls the flow by designing a scheduler, only requires a small number of counters and registers, optimizes the RAM behind the encoder, saves resources, reduces latency, introduces threshold configuration, and can be applied to all FEC code types.
[0099] Preferably, in S110, the scheduler generates a scheduling enable to the transmitter by a fixed pattern scheduling method, including:
[0100] The scheduler includes a first counter, a second counter and a constant register;
[0101] S111, the first counter cyclically counts according to the first counting upper limit to obtain a first counting value;
[0102] S112, the first counter sends the first count value to the second counter and the constant register;
[0103] S113, the constant register uses the first count value as an index, selects a corresponding bit value and sends it to the second counter;
[0104] S114, the second counter cyclically counts according to the second counting upper limit, and generates a scheduling enable according to the corresponding bit value;
[0105] S115, the second counter outputs the scheduling enable to the transmitter, and the monitoring tool starts timing;
[0106] The scheduler module is introduced to control the traffic output of the transmitter, which avoids the introduction of RAM, reduces latency and area, and accelerates solution iteration.
[0107] The scheduler is implemented with two counters and a constant register. It accurately controls the traffic output of the transmitter and reserves the bandwidth and time slot for inserting parity for the encoder in real time.
[0108] By precisely controlling the traffic output, the RAM behind the encoder is optimized, resources are saved, latency is reduced, and implementation is simplified.
[0109] Even if the delay between the transmitter and the encoder changes, it will not affect the scheduling implementation of the present invention.
[0110] refer to Figure 5 When the scheduler generates a 1-beat scheduling enable, the transmitter outputs a complete bus data. When the scheduler does not generate a scheduling enable, the transmitter stops sending data;
[0111] The present invention accurately controls the flow output of the transmitter and reserves bandwidth and time slots of equal width for insertion in real time for the encoder part, thereby avoiding the introduction of RAM, reducing delay and area, and accelerating the iteration of the solution.
[0112] In this embodiment, the code type is RSFEC (544, 514);
[0113] In fact, the present solution is not limited to this code type, and any other code type can be applied to the present device and method, and only the specific scheduler needs to be adjusted accordingly;
[0114] refer to Figure 6 ,The scheduler is implemented using 2 counters and a constant register;
[0115] In this embodiment, reference Figure 7 The first counter is the first counter, the counting cycle of this counter is 16, the counting range is 0~15, and then it returns to 0, each time it increases by 1, and the counting cycle is repeated; the second counter is the second counter, the counting cycle of this counter is 85, the counting range is 0~84, and then it returns to 0, each time it increases by 1, and the counting cycle is repeated; refer to Figure 7 , the constant register is a 16-bit constant register, which is used to indicate whether the current loop is scheduled for 81 beats continuously or 80 beats;
[0116] refer to Figure 8 , the initial value of the first counter is 0; the value is transferred to the constant register and the second counter respectively;
[0117] The constant register uses the value sent by the first counter as an index, selects the bit value at the corresponding position and sends it to the second counter;
[0118] The second counter generates a scheduling enable to the transmitter according to the value sent by the constant register to send data to the encoder. If the constant register sends a 1, it will first generate 81 consecutive scheduler enable beats, then stop for 4 beats; if the constant register sends a 0, it will first generate 80 consecutive scheduler enable beats, then stop for 5 beats;
[0119] When the second counter counts to 84, it generates a trigger signal to the first counter, and the first counter increases by 1, thereby starting the next round of scheduling;
[0120] The final output scheduling pattern reference Fig. 9 .
[0121] Preferably, S120 adjusts the encoding parameters according to the data state, including:
[0122] The encoding parameters include a first count upper limit, a second count upper limit and a register parameter;
[0123] S121, data status includes the current FEC coding code type and coding operation status;
[0124] S122, obtaining the current FEC coding pattern through machine learning, and adjusting the first counting upper limit, the second counting upper limit and the register parameters in combination with the coding operation status;
[0125] Make precise adjustments based on the current data status to achieve more efficient and stable data transmission;
[0126] By analyzing and learning a large amount of historical data, the machine learning model can identify the code characteristics under different data states and use them for subsequent coding parameter adjustments;
[0127] The coding parameter adjustment mechanism based on data status and machine learning technology can achieve accurate prediction of coding parameters, thereby performing monitoring and comparison to facilitate parameter adjustment.
[0128] Preferably, after receiving the scheduling enable, the transmitter compresses the original data to obtain compressed data, including:
[0129] S210, organizing and optimizing the original data to obtain optimized data;
[0130] S220, compressing the optimized data to obtain compressed data;
[0131] The original data may contain a lot of redundant information, resulting in underutilization of bandwidth. Compression mechanism is introduced to optimize data transmission efficiency and bandwidth utilization.
[0132] S230, caching the redundant information removed during the compression process;
[0133] When the transmitter does not receive the scheduling enable, the transmitter stops sending compressed data;
[0134] Stop sending data when scheduling is not enabled, thereby achieving flow control;
[0135] By compressing data, the bandwidth required for transmission is reduced and the utilization efficiency of bandwidth is improved. Compressing data can reduce the amount of data during transmission, thereby possibly reducing transmission delays. Through optimization and compression, the resources required to store the original data are reduced. Data transmission is dynamically controlled according to the scheduling enable signal, avoiding data transmission when it is not needed, saving network resources and improving the responsiveness of the system.
[0136] Preferably, the compressed data is input to the encoder after a delay, and the encoder performs FEC encoding on the compressed data and outputs the encoded data, including:
[0137] S310, matching delay parameters according to scheduling enable;
[0138] S320, inputting the compressed data into the encoder according to the delay parameter;
[0139] S330, the encoder performs FEC encoding on the compressed data and outputs the encoded data;
[0140] S340, the monitoring tool stops timing and obtains the convergence time;
[0141] The coding convergence time can be monitored in real time through monitoring tools, so as to flexibly adjust the coding parameters and improve the coding efficiency.
[0142] Convergence time refers to the time required for an algorithm or program to reach a stable state. The length of this time often depends on the setting of encoding parameters. If the parameters are set properly, the convergence time will be relatively short. Otherwise, it may lead to long waiting times and waste of computing resources. Therefore, real-time monitoring of encoding convergence time can help timely discover and solve parameter setting problems, thereby improving encoding efficiency.
[0143] After obtaining the monitoring data, it is necessary to analyze the data to find out the key factors affecting the coding efficiency, obtain the predicted values according to the neural network model, and make comparisons;
[0144] First, collect the relationship between historical convergence time and encoding information, and record other factors related to convergence time; extract meaningful features from the collected data, which should be able to reflect the dynamics of the training process and changes in model performance; select a deep learning framework such as TensorFlow, PyTorch, etc. to build a neural network model, predict the convergence time based on the input features, divide the data set into training set and test set, and use the training set to train your neural network model. You can use a standard training cycle, including forward propagation, calculating losses, backpropagation, and updating weights; use the trained neural network model to predict the estimated convergence time.
[0145] Preferably, compressing the optimized data to obtain compressed data includes:
[0146] Divide the optimization data into training set and validation set;
[0147] Construct a data compression model, which is expressed as:
[0148]
[0149] in, is the i-th data in the training set, L() is the loss function, β is the encoding parameter and decoding parameter, μ is the initial learning rate, In order to seek partial guidance, The restored input data is calculated by the data compression module for the i-th data in the training set;
[0150] Train the data compression model;
[0151] Use the trained compression model to compress data.
[0152] If the data that the system needs to process is very large, then if the data is not compressed, the system processing time will be long and the efficiency will be low. In addition, the large amount of data will also cause the error rate to increase. Therefore, compressing the optimized data can improve the data processing efficiency and accuracy.
[0153] Preferably, after obtaining the convergence time, the following steps are performed:
[0154] S341, feeding back the convergence time to the scheduler;
[0155] S342, predicting an estimated convergence time according to the FEC coding code type;
[0156] S343 compares the estimated convergence time with the convergence time to obtain the encoding operation status;
[0157] When the convergence time is greater than the estimated convergence time, the encoding operation state is in the state to be optimized;
[0158] When the convergence time is less than or equal to the estimated convergence time, the encoding operation state is a normal operation state;
[0159] S345: when the encoding operation state is in a state to be optimized, readjust the first upper count limit, the second upper count limit and the register parameters;
[0160] The first count upper limit and the second count upper limit can be configured by software, and the length and content of the constant register can be configured, so the scheduler can support multiple FEC code types at the same time;
[0161] The existing FEC encoding process has the problem of long convergence time, which affects the overall performance of the system. A method is needed to evaluate the performance status of the encoder in real time and make adjustments based on the evaluation results. At the same time, the system also needs to be able to adapt to different FEC code types to improve the versatility and flexibility of the scheduler.
[0162] By optimizing the convergence time and improving the efficiency of the FEC encoding process, the overall system performance is improved. The scheduler can adapt to a variety of FEC code types, which improves the versatility and flexibility of the system.
[0163] Ensure the encoder always operates in the best condition by monitoring the convergence time in real time and adjusting parameters as needed;
[0164] Through prediction and feedback mechanisms, the system can self-evaluate and self-optimize, reducing the need for human intervention;
[0165] The system includes an intelligent parameter adjustment algorithm, which automatically optimizes parameters based on historical data and real-time feedback. The system may be designed with adaptive strategies for a variety of scenarios to cope with different network conditions and data transmission requirements. Users may be able to customize FEC code types and related parameters according to specific requirements to provide a more personalized system configuration. This embodiment solves the problem of convergence time optimization in the FEC encoding process by introducing a convergence time measurement, prediction and feedback mechanism, and provides a flexible and efficient scheduling and flow control method.
[0166] Preferably, after outputting the coded data, the coded data is also checked:
[0167] Clustering algorithms are used to classify data;
[0168] Calculate the hash value of each data in different categories:
[0169]
[0170] represents the XOR vector of the hash value of the encoded data, is a random vector of the same length as the standard hash value;
[0171] comparing the hash value of the encoded data with the hash value of the standard encoded data;
[0172] The calculation of a hash value usually involves converting data of any size into an output of a fixed length through a hash function. This conversion process ensures that different input data is almost impossible to produce the same hash value, and the same input data always produces the same hash value.
[0173] If the hash value of the encoded data is not equal to the hash value of the standard encoded data, the encoding process data is lost;
[0174] When data loss is detected in the coded data, redundant information in the cache is extracted to repair the coded data.
[0175] Data hash values can be used to verify data integrity, perform fast data lookup, and construct secure data structures in combination with data encryption algorithms. Currently, the more common hash algorithms include MD5, SHA-1, BKDRHash, APHash, etc., which are implemented through addition and subtraction, bit operations, multiplication and division, table lookup, and hybrid implementation. Different implementation methods vary in running speed and hash effect.
[0176] In this embodiment, the original data may be missing or erroneous after being compressed, so it is necessary to enhance the integrity and reliability of data transmission. By introducing the use of redundant information in the checksum coding data stage, the automatic repair of possible data loss is achieved. This process ensures that even if problems occur during the data transmission process, the original data can be restored to the greatest extent.
[0177] During the data transmission process, data loss or corruption may occur, affecting the integrity of the data. A mechanism is needed to repair the data using redundant information when data loss is detected. After the encoded data is output, the encoded data is checked according to the original data to ensure the accuracy of the data. During the checking process, if data loss is detected in the encoded data, the system will identify and record this problem. Once data loss is detected, the system will extract the redundant information in the previous cache and use this information to repair the encoded data to restore the integrity of the data.
[0178] The verification and repair mechanism improves the reliability of data transmission and reduces the risk of data loss. Even if data is lost during data transmission, it can be repaired through redundant information to ensure data integrity. The system can automatically detect and repair data problems, improve the system's ability to handle abnormal situations, enhance robustness, and optimize the efficiency of system resource use by effectively utilizing redundant information in the cache.
[0179] In this embodiment, the system uses an efficient calibration algorithm to quickly and accurately detect data loss or damage. The system has the ability to dynamically manage redundant information and adjust redundancy according to the importance of the data and transmission conditions. Users may be able to configure calibration and repair parameters according to their needs to adapt to different application scenarios;
[0180] This embodiment solves the problem of data loss during data transmission by introducing the use of redundant information in the check coding data stage, and provides an effective data repair mechanism, thereby improving the integrity of data transmission and the robustness of the system.
[0181] Example 2
[0182] A scheduling and flow control system based on FEC coding direction, comprising:
[0183] The scheduling module is used by the scheduler to generate scheduling enable to the transmitter through a fixed pattern scheduling method;
[0184] A compression module, used for compressing the original data to obtain compressed data after the transmitter receives the scheduling enable;
[0185] The encoding module is used to compress data and input it to the encoder after delay. The encoder performs FEC encoding on the compressed data and outputs the encoded data. After the encoding is completed, the encoded data is sent to the check module for check. When data loss occurs, the redundant information in the cache is extracted to repair the encoded data;
[0186] The encoding module collects the encoding operation status, processes it and feeds it back to the scheduling module to determine whether the operation status needs to be optimized;
[0187] When the operating status needs to be optimized, the scheduling module readjusts the encoding parameters. The scheduling module flexibly adjusts the encoding parameters according to the transmission status of the monitored data, such as bandwidth utilization, packet loss rate, etc. This dynamic adjustment method enables the system to adapt to different network environments, thereby ensuring the stability and reliability of data transmission.
[0188] Example 3
[0189] An electronic device comprises a memory and a processor. The memory stores a computer program. When the processor executes the computer program, a scheduling and flow control method based on FEC coding direction is implemented.
[0190] refer to Fig.10 , the electronic device 2 includes a processor 21, a memory 22, an input device 23, and an output device 24. The processor 21, the memory 22, the input device 23, and the output device 24 are coupled via a connector, and the connector includes various interfaces, transmission lines, or buses, etc., which are not limited in the embodiments of the present invention. It should be understood that in various embodiments of the present invention, coupling refers to mutual connection in a specific manner, including direct connection or indirect connection through other devices, for example, through various interfaces, transmission lines, buses, etc.
[0191] The processor 21 may be one or more graphics processing units (GPUs). When the processor 21 is a GPU, the GPU may be a single-core GPU or a multi-core GPU. Optionally, the processor 21 may be a processor group consisting of multiple GPUs, and the multiple processors are coupled to each other via one or more buses. Optionally, the processor may also be other types of processors, etc., which are not limited in the embodiments of the present invention.
[0192] The memory 22 can be used to store computer program instructions and various computer program codes including program codes for executing the scheme of the present invention. Optionally, the memory includes but is not limited to random access memory (RAM), read-only memory (ROM), erasable programmable read only memory (EPROM), or portable read only memory (CD-ROM), which is used for related instructions and data.
[0193] The input device 23 is used to input data and / or signals, and the output device 24 is used to output data and / or signals. The output device 24 and the input device 23 may be independent devices or an integrated device.
[0194] The above is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features applied herein.
Claims
1. A scheduling and flow control method based on FEC coding direction, characterized in that: include: The scheduler generates a scheduling enable to the transmitter through a fixed pattern scheduling method; After receiving the scheduling enable, the transmitter compresses the original data to obtain compressed data; The compressed data is input to the encoder after a delay, and the encoder performs FEC encoding on the compressed data and outputs the encoded data; The scheduler generates a scheduling enable to the transmitter through a fixed pattern scheduling method, including: The scheduler includes a first counter, a second counter and a constant register; The first counter cyclically counts according to the first counting upper limit to obtain a first counting value; The first counter sends the first count value to the second counter and the constant register; The constant register uses the first count value as an index, selects a corresponding bit value and sends it to the second counter; The second counter cyclically counts according to the second counting upper limit, and generates a scheduling enable according to the corresponding bit value; The second counter outputs the scheduling enable to the transmitter, and the monitoring tool starts timing.
2. A scheduling and flow control method based on FEC coding direction according to claim 1, characterized in that: It also includes adjusting encoding parameters according to data status: The encoding parameters include the first count upper limit, the second count upper limit and register parameters; The data status includes the current FEC coding code type and coding operation status; The current FEC coding code type is obtained through machine learning, and the first counting upper limit, the second counting upper limit and the register parameters are adjusted in combination with the coding running status.
3. The method for scheduling and flow control based on FEC coding direction according to claim 1, characterized in that: After receiving the scheduling enable, the transmitter compresses the original data to obtain compressed data, including: Organizing and optimizing the original data to obtain optimized data; compressing the optimized data to obtain compressed data; Cache the redundant information removed during compression; When the transmitter does not receive the scheduling enable, the transmitter stops sending the compressed data.
4. The method for scheduling and flow control based on FEC coding direction according to claim 3, characterized in that: The compressing the optimized data to obtain compressed data comprises: Dividing the optimized data into a training set and a validation set; A data compression model is constructed, and the data compression model is expressed as: ; in, is the i-th data in the training set, L() is the loss function, β is the encoding parameter and decoding parameter, μ is the initial learning rate, In order to seek partial guidance, The restored input data is calculated by the data compression module for the i-th data in the training set; Train the data compression model; Use the trained compression model to compress data.
5. The method for scheduling and flow control based on FEC coding direction according to claim 4, characterized in that: The compressed data is input to the encoder after a delay, the encoder performs FEC encoding on the compressed data and outputs the encoded data, and the monitoring tool stops timing, including: Matching delay parameters according to the scheduling enable; inputting the compressed data into the encoder according to the delay parameter; The encoder performs FEC encoding on the compressed data and outputs encoded data; The monitoring tool stops timing and obtains the convergence time.
6. The method for scheduling and flow control based on FEC coding direction according to claim 5, characterized in that: After obtaining the convergence time, the method further comprises: Feeding back the convergence time to the scheduler; Predicting an estimated convergence time according to a code type of the FEC encoding; Comparing the estimated convergence time with the convergence time to obtain the encoding running state; When the convergence time is greater than the estimated convergence time, the encoding operation state is a state to be optimized; When the convergence time is less than or equal to the estimated convergence time, the encoding operation state is a normal operation state; When the encoding operation state is a state to be optimized, the first upper counting limit, the second upper counting limit and register parameters are readjusted.
7. The method for scheduling and flow control based on FEC coding direction according to claim 5, characterized in that: After outputting the coded data, the method further includes checking the coded data: The coded data is checked according to the original data: Clustering algorithms are used to classify data; Calculate the hash value of each data in different categories: ; represents the XOR vector of the hash value of the encoded data, is a random vector of the same length as the standard hash value; comparing the hash value of the encoded data with the hash value of the standard encoded data; If the hash value of the encoded data is not equal to the hash value of the standard encoded data, the encoding process data is lost; When data loss is detected in the coded data, redundant information in the cache is extracted to repair the coded data.
