Data transmission method and system of industrial edge gateway

Abstract

The invention relates to the technical field of data processing, in particular to a data transmission method and system for industrial edge gateways, and the method comprises the steps: obtaining industrial production equipment data, constructing a data queue, and obtaining the busy degree of each edge gateway based on the operation data of the data queue; according to the busy degree of each edge gateway, obtaining transmission results after the different data queues are transmitted to the cloud server, and calculating feedback rewards transmitted by the edge gateways by selecting the different data queues; and on the basis of the feedback rewards, using a UCB algorithm to calculate the confidence coefficient of each queue for transmitting data to the cloud server, and adjusting the data queue through which the edge gateway passes when transmitting the data to the cloud server according to the confidence coefficient. According to the method, the data transmission strategy is dynamically adjusted by monitoring the busy degree of the edge gateway and the transmission result of the data queue in real time, and rapid changes of task frequency, data volume and network conditions are better adapted, so that the data transmission efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and more specifically, to a data transmission method and system for an industrial edge gateway. Background Art

[0002] As short-range wireless communications are increasingly used in industrial edge computing scenarios, the deployment of industrial edge networks is becoming increasingly dense. In the case of ultra-dense deployment of edge networks, data transmission can easily cause serious signal collision problems during the process of transmitting data to the server. Therefore, a data transmission method that can adapt to multi-source heterogeneous devices and dense deployment in high-concurrency scenarios is needed.

[0003] The existing Chinese patent application document with publication number CN115242774A discloses a data transmission method between an edge gateway and a cloud server, including: the edge gateway sends a heartbeat packet to connect to the cloud server; when the cloud server has data to send to the edge gateway, the data of the cloud server is downloaded; when the edge gateway has information to upload to the cloud server, it is notified when the edge gateway sends a heartbeat packet, if the cloud server has no data to be sent, the edge gateway uploads the information to the cloud server; if the cloud server has data to be sent, the edge gateway first suspends uploading information, downloads the data of the cloud server, compares the downloaded data with the information to be uploaded, and if it is rollback information, the edge gateway stops uploading information; otherwise, the edge gateway uploads information; when there is no data to upload and download, it is notified when the edge gateway sends a heartbeat packet, and the communication connection is interrupted. The number of connections between the edge gateway and the cloud server during the data transmission process is reduced, the network traffic is reduced, and the edge gateway does not need too much network guarantee time.

[0004] This patent application document solves the problem that the network environment at the deployment site and location of the edge gateway is often unstable, resulting in the edge gateway's inability to maintain the network online at all times, so that the data transmission from the edge gateway to the cloud server cannot be fully guaranteed; currently, it is necessary to transmit data to a data queue that adapts to different conditions according to the edge gateway. Traditional data allocation strategies are usually based on simple preset rules. These rules may not be able to adapt to rapid changes in task frequency, data volume, network conditions, etc., resulting in the allocation strategy being not flexible enough in complex environments, thereby affecting data processing speed, data packet loss or transmission errors, and affecting data integrity and accuracy. Summary of the invention

[0005] In order to solve the problem that data transmission requirements are often dynamically affected by multiple factors, resulting in data processing speed, data packet loss or transmission errors in complex environments, and affecting data integrity and accuracy, the present invention provides solutions in the following aspects.

[0006] In a first aspect, a data transmission method of an industrial edge gateway includes: obtaining industrial production equipment data of the edge gateway, and constructing a data queue, and obtaining the busyness of each edge gateway based on the operation data of the data queue; according to the busyness of each edge gateway, obtaining the transmission results of different data queues after transmission to a cloud server, and calculating the feedback reward of the edge gateway selecting different data queues for transmission; based on the feedback reward, using the UCB algorithm to calculate the confidence of each data queue transmitting data to the cloud server, and adjusting the data queue passed by the edge gateway when transmitting data to the cloud server according to the confidence; wherein the confidence satisfies the following relationship: , where Indicates the current moment The edge gateway passes The confidence level of each data queue transmitting data to the cloud server, Indicates the number before the current time The edge gateway passes The average feedback incentive for each data queue to transmit data to the cloud server, Indicates The current busyness of the edge gateway, Indicates the current moment The edge gateway passes The number of times a data queue transmits data to the cloud server. Indicates the current moment The total number of times all data queues of the edge gateway transmit data to the cloud server, Expressed as a natural constant The logarithmic function of the base.

[0007] The effect is: by real-time monitoring of the busyness of the edge gateway and the transmission results of the data queue, dynamically adjusting the data transmission strategy, ensuring that data can be transmitted to the cloud server quickly and reliably, reducing packet loss and transmission errors, using the UCB algorithm to calculate the confidence of each data queue, and intelligently selecting the most appropriate data queue for data transmission, thereby improving the utilization of network resources, reducing network congestion, and balancing the load.

[0008] Preferably, constructing data queues, including: high-demand queues, large-capacity queues, frequent interaction queues, small data queues, and abnormal data queues;

[0009] Among them, the high-requirement queue is for processing data of large-scale, intensive tasks; the large-capacity queue is for storing data of large-scale, infrequent interactions; the frequent interaction queue is for processing data of large-scale, intensive tasks; the small data queue is for temporarily storing small-scale, infrequently interacted data; and the abnormal data queue is for giving priority to processing and responding to abnormal data.

[0010] Preferably, the busyness level includes:

[0011] Taking any edge gateway as the target gateway and any moment of the target gateway as the marking moment, the number of data tasks and the amount of data collected within the preset time before the marking moment are normalized and multiplied as the data load intensity;

[0012] Calculate the ratio of the idle time within the preset time before the marking moment to the preset time and perform exponential decay to obtain the idle time exponential decay;

[0013] The product of the data load intensity and the exponential decay of the idle time is taken as the busyness of the target gateway at the marking moment.

[0014] The effect is that by calculating the ratio of idle time to preset time and applying exponential decay, the idle state of the edge gateway before the marking moment can be dynamically reflected, so that the calculation of busyness can sensitively respond to changes in idle time.

[0015] Preferably, the busyness also includes:

[0016] Take any edge gateway as the target gateway, any moment of the target gateway as the marking moment, and take the product of the data packet loss rate and the memory usage rate in the period before the marking moment as the busyness.

[0017] The effect is: the data packet loss rate reflects the network congestion or instability, while the memory usage rate reflects the intensity of computing resources. By accurately evaluating the busyness of the edge gateway, network resources can be allocated more effectively, improving the resource utilization and throughput of the overall network and reducing resource waste.

[0018] Preferably, the feedback reward includes:

[0019] Taking any edge gateway as the target gateway, the ratio between the total number of data packets received by the server and the number of data packets correctly received is calculated to obtain the data transmission success rate of the target gateway through different data queues;

[0020] According to the sending time and receiving time of each data packet recorded between the target gateway and the server, the transmission delay of the target gateway through different data queues is obtained;

[0021] The ratio of the data transmission success rate to the transmission delay is used as the transmission efficiency, the complement of the busyness is used as the idleness, the idleness is corrected using a logarithmic function and a correction parameter, and the product of the transmission efficiency and the corrected idleness is used as the feedback incentive of the target gateway.

[0022] The effect is that by calculating the data transmission success rate and transmission delay, the transmission performance of the edge gateway through different data queues can be quantified, and then the feedback incentives of the target gateway can be analyzed, which helps to balance resource utilization, optimize data transmission strategies, and improve overall network performance.

[0023] Preferably, the feedback reward further includes:

[0024] Taking any edge gateway as the target gateway, taking the complement of the target gateway's busyness as the idleness, and using the logarithmic function and correction parameter to correct the idleness;

[0025] Obtain the network communication performance parameters of the target gateway, and calculate the percentage of the actual bandwidth used to the total bandwidth to obtain the bandwidth utilization rate;

[0026] The ratio of the corrected idle level to the bandwidth utilization is used as the feedback incentive of the target gateway.

[0027] The effect is that by calculating the bandwidth utilization rate, the ratio of the actual bandwidth used to the total bandwidth can be quantified, which helps to identify and solve the problem of bandwidth waste or shortage and optimize the efficiency of bandwidth use. It can adaptively adjust the data transmission strategy according to the current network resources and improve the system's responsiveness and flexibility.

[0028] Preferably, the step of adjusting the data queue through which the edge gateway transmits data to the cloud server includes:

[0029] Each edge gateway initializes each data queue. The initialization is that at the initial moment, the feedback incentive for each edge gateway to transmit data through each data queue is 0, then a data queue is randomly selected for data transmission, and a new feedback incentive is calculated according to the transmission result. Repeat the iteration to calculate the average feedback incentive for the edge gateway to transmit data to the cloud server through each data queue, and calculate the confidence of each edge gateway to transmit data to the cloud server through each data queue, and select the data queue with the largest confidence to transmit data to the cloud server.

[0030] In a second aspect, a data transmission system for an industrial edge gateway includes: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned data transmission method for the industrial edge gateway is implemented.

[0031] The present invention has the following effects:

[0032] 1. The present invention dynamically adjusts the data transmission strategy by real-time monitoring of the busyness of the edge gateway and the transmission results of the data queue, so that the adaptive data allocation mechanism can better adapt to the rapid changes in task frequency, data volume and network conditions, thereby improving data transmission efficiency, reducing data packet loss or transmission errors, and ensuring data integrity and accuracy.

[0033] 2. The present invention uses the UCB algorithm to calculate the confidence of each data queue and adjusts the data queue transmission according to the confidence, thereby improving the flexibility of the system based on the confidence, enabling it to quickly respond to environmental changes and business needs. At the same time, new data queues and edge gateways can be easily expanded without affecting the stability and efficiency of the existing system.

[0034] 3. The present invention constructs different types of data queues and intelligently selects data queues according to the busyness of the edge gateway and feedback rewards, so as to more reasonably allocate network resources, improve bandwidth utilization, reduce network congestion, and thus optimize the overall network performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0036] Figure 1 It is a method flow chart of steps S1 to S3 in a data transmission method of an industrial edge gateway in an embodiment of the present invention.

[0037] Figure 2 It is a structural block diagram of a data transmission system of an industrial edge gateway according to an embodiment of the present invention. DETAILED DESCRIPTION

[0038] 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 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 those skilled in the art without creative work are within the scope of protection of the present invention.

[0039] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0040] Specific implementation scenario: In modern industrial automation and intelligent manufacturing environments, edge gateways play a vital role. They are responsible for collecting data from various sensors and devices, and need to divide the data into data queues and input them into the cloud server through the edge gateway.

[0041] Reference Figure 1 , a data transmission method of an industrial edge gateway includes steps S1 to S3, which are specifically as follows:

[0042] S1: Obtain the industrial production equipment data of the edge gateway and build a data queue. Based on the operation data of the data queue, obtain the busyness of each edge gateway.

[0043] Building data queues includes but is not limited to: high-demand queues, large-capacity queues, frequent interaction queues, small data queues, and abnormal data queues; implementers can build more queues according to scenario needs.

[0044] Among them, the high-requirement queue is for processing large-scale, task-intensive data; the large-capacity queue is for storing large-scale, infrequently interacted data; the frequent interaction queue is for processing large-scale, task-intensive data; the small data queue is for temporarily storing small-scale, infrequently interacted data; and the abnormal data queue is for prioritizing the processing and response of abnormal data.

[0045] Further analysis shows that since different edge gateways are connected to different devices, the pressure on the server when the collected data is transmitted to the server is also different. In order to balance the use of resources, it is necessary to perform load analysis on the edge gateways connected to the server. Therefore, the busyness of each edge gateway is calculated based on the operating data of each edge gateway. The specific steps are as follows:

[0046] Taking any edge gateway as the target gateway and any moment of the target gateway as the marking moment, the number of data tasks and the amount of data collected within the preset time before the marking moment are normalized and multiplied as the data load intensity;

[0047] Calculate the ratio of the idle time within the preset time before the marking moment to the preset time and perform exponential decay to obtain the idle time exponential decay;

[0048] The product of the data load intensity and the exponential decay of the idle time is taken as the busyness of the target gateway at the marking moment.

[0049] Specifically, the busyness satisfies the following relationship:

[0050] ;

[0051] In the formula, Indicates How busy is an edge gateway at any moment? Indicates The idle time of an edge gateway within a preset time before any moment, Indicates the length of the preset time. Indicates The number of data tasks received by an edge gateway before the preset time at any moment, Indicates The amount of data received by an edge gateway at any time before the preset time, Expressed as a natural constant is the exponential function of the base, represents the linear normalization function.

[0052] In this embodiment, the preset time is 100 milliseconds, and the implementer can select the length of the time period according to the actual situation;

[0053] That is to say, It indicates the proportion of idle time of the edge gateway within the preset time before any moment. The larger the value, the lower the task load and the less busy the edge gateway. Conversely, the smaller the value, the higher the task load and the busier the edge gateway.

[0054] It reflects the density of requests processed by the edge gateway. The larger the value, the denser the requests processed by the edge gateway, the busier the edge gateway is, and the busier the edge gateway is. The smaller the value, the sparser the requests processed by the edge gateway, the more likely the edge gateway is to be idle, and the less busy the edge gateway is.

[0055] It reflects the load degree of each task and the consumption of communication resources. The larger the value, the greater the load of the edge gateway and the greater the consumption of communication resources, the busier the edge gateway is, and the busier the edge gateway is. The smaller the value, the smaller the load of the edge gateway and the smaller the consumption of communication resources, and the less busy the edge gateway is.

[0056] In addition, another embodiment further includes:

[0057] Take any edge gateway as the target gateway, any moment of the target gateway as the marking moment, and take the product of the data packet loss rate and the memory usage rate in the period before the marking moment as the busyness.

[0058] Specifically, the busyness satisfies the following relationship:

[0059] ;

[0060] In the formula, Indicates How busy is an edge gateway at any moment? Indicates The data packet loss rate of an edge gateway within a period of time before any moment, Indicates The memory usage of an edge gateway at any time in a period of time.

[0061] That is to say, the packet loss rate reflects network congestion or instability to a certain extent, which may be caused by excessive network load. The memory utilization rate means that the edge gateway is processing a large amount of data. The above calculation reflects the pressure on the edge gateway in processing data and maintaining network connections, which helps to ensure that the edge gateway can process data efficiently and maintain a stable network connection, thereby improving the performance and reliability of the entire system.

[0062] Further analysis shows that due to the different busyness and task loads of different edge gateways, when edge gateways select different data queues to transmit data to the cloud server, the transmission effect will also be different. In order to optimize transmission efficiency and resource allocation, it is necessary to evaluate the effect of different edge gateways in different states transmitting data to the cloud server through different data queues. Therefore, it is necessary to calculate the feedback reward for the gateway to select different data queues to transmit data based on the busyness of each edge gateway and the transmission results when the gateway selects different data queues to transmit data to the server. The specific steps are as follows.

[0063] S2: According to the busyness of each edge gateway, the transmission results of different data queues after being transmitted to the cloud server are obtained, and the feedback reward of the edge gateway for selecting different data queues for transmission is calculated.

[0064] The transmission results include but are not limited to: data transmission success rate, transmission delay, data packet loss rate, transmission rate, etc.

[0065] Taking any edge gateway as the target gateway, the ratio between the total number of data packets received by the server and the number of data packets correctly received is calculated to obtain the data transmission success rate of the target gateway through different data queues;

[0066] According to the sending time and receiving time of each data packet recorded between the target gateway and the server, the transmission delay of the target gateway through different data queues is obtained;

[0067] The ratio of data transmission success rate to transmission delay is taken as transmission efficiency, the complement of busyness is taken as idleness, the idleness is corrected using logarithmic function and correction parameter, and the product of transmission efficiency and corrected idleness is taken as feedback incentive of target gateway.

[0068] Specifically, the feedback reward satisfies the following relationship:

[0069] ;

[0070] In the formula, Indicates that at any time The edge gateway passes Feedback incentive for data queues to transmit data to the cloud server, Indicates that at any time The edge gateway passes The data transmission success rate of each data queue transmitting data to the cloud server, Indicates that at any time The edge gateway passes The transmission delay of data queues transmitting data to the cloud server, Indicates How busy is an edge gateway at any moment? Indicates the correction parameter for adjusting the busyness of the edge gateway. Expressed as a natural constant The logarithmic function of the base.

[0071] In this embodiment, the correction parameters , implementers can make choices based on actual conditions.

[0072] That is, the data transmission success rate of the edge gateway transmitting data to the cloud server through different data queues The larger the value, the higher the reliability of data transmission from the edge gateway to the cloud server through the data queue. In order to make the algorithm tend to select the data queue that can ensure the reliability of data transmission and improve the overall system performance, the first The edge gateway passes The greater the feedback incentive for the edge gateway to transmit data to the cloud server through different data queues, the higher the data transmission success rate of the edge gateway to transmit data to the cloud server through different data queues. The smaller the value, the higher the data loss rate, the more data retransmission is needed, and the greater the system burden and network congestion risk. In order to guide the algorithm to explore better options, the first The edge gateway passes The smaller the feedback incentive for each data queue to transmit data to the cloud server.

[0073] Transmission delay of edge gateway transmitting data to cloud server through different data queues The bigger, the The edge gateway passes The slower the data queue is to transmit data to the cloud server, the greater the impact on data tasks with high real-time requirements. In order to avoid further increasing the load, the algorithm should be guided to select a more efficient data queue. The edge gateway passes The smaller the feedback incentive for the edge gateway to transmit data to the cloud server through different data queues, the smaller the transmission delay of the edge gateway to transmit data to the cloud server through different data queues. The smaller, the The edge gateway passes The more timely the data queue is, the more it can meet the real-time requirements of data transmission. In order to optimize the completion effect of real-time tasks, the first The edge gateway passes The greater the feedback incentive for each data queue to transmit data to the cloud server.

[0074] The busyness of the edge gateway reflects the current load pressure and resource availability of the gateway, which will directly affect the quality and efficiency of the data received by the cloud server. When the gateway is busy, the transmission data may have problems such as increased delay, increased packet loss rate and decreased data integrity. At this time, the success rate and transmission delay of the edge gateway to transmit data to the server through different data queues may be at a similar level. If the feedback incentive is calculated only based on the success rate and transmission delay of the transmitted data, it may not be able to guide the algorithm to select a more efficient data queue. Therefore, it is necessary to calculate the feedback incentive in combination with the busyness of the edge gateway. The greater the edge business, the tighter the gateway resources are, and the performance of data transmission through the data queue may be poor. The data queue needs to be selected carefully. In this case, the feedback incentive for the edge gateway to transmit data to the cloud server through each data queue should be reduced. The smaller the edge gateway is, the more sufficient the gateway resources are and the quality of the transmitted data is more guaranteed. There is no need to excessively restrict the selection of data queues. The feedback incentive for the edge gateway to transmit data to the cloud server through each data queue should be increased. Therefore, through right Make corrections.

[0075] In addition, another embodiment further includes:

[0076] Taking any edge gateway as the target gateway, taking the complement of the target gateway's busyness as the idleness, and using the logarithmic function and correction parameter to correct the idleness;

[0077] Obtain the network communication performance parameters of the target gateway, and calculate the percentage of the actual bandwidth used to the total bandwidth to obtain the bandwidth utilization rate;

[0078] The ratio of the corrected idle level to the bandwidth utilization is used as the feedback incentive of the target gateway.

[0079] Specifically, the feedback reward satisfies the following relationship:

[0080] ;

[0081] In the formula, Indicates that at any time The edge gateway passes Feedback incentive for data queues to transmit data to the cloud server, Indicates How busy is an edge gateway at any moment? Indicates the correction parameter for adjusting the busyness of the edge gateway. Indicates that at any time The edge gateway passes The bandwidth utilization rate when each data queue transmits data to the cloud server. Expressed as a natural constant The logarithmic function of the base.

[0082] In this embodiment, the correction parameters , implementers can make choices based on actual conditions.

[0083] That is, the above calculation method reflects the network efficiency and resource utilization of the edge gateway when transmitting data. A high feedback incentive value indicates that the current data queue is a good choice for transmitting data, while a low feedback incentive value may indicate that other data queues need to be considered to optimize transmission performance.

[0084] Further analysis shows that different data queues have different characteristics. When the edge gateway is in different states, the transmission effect of transmitting data to the cloud server through different data queues is also different. In order for each edge gateway to select the most suitable data queue to achieve better data transmission effect, the present invention selects the data queue through which the edge gateway transmits data to the server according to the feedback reward for the gateway selecting different data queues to transmit data. The specific analysis is as follows:

[0085] S3: Based on the feedback reward, the UCB algorithm is used to calculate the confidence of each data queue in transmitting data to the cloud server, and the data queue through which the edge gateway transmits data to the cloud server is adjusted according to the confidence.

[0086] It should be noted that the UCB algorithm is a well-known technology to those skilled in the art and will not be described in detail.

[0087] Specifically, the confidence satisfies the following relationship:

[0088] ;

[0089] In the formula, Indicates the current moment The edge gateway passes The confidence level of each data queue transmitting data to the cloud server, Indicates the number before the current time The edge gateway passes The average feedback incentive for each data queue to transmit data to the cloud server, Indicates The current busyness of the edge gateway, Indicates the current moment The edge gateway passes The number of times a data queue transmits data to the cloud server. Indicates the current moment The total number of times all data queues of the edge gateway transmit data to the cloud server, Expressed as a natural constant The logarithmic function of the base.

[0090] That is to say, Indicates The edge gateway passes The average feedback incentive for data queues to transmit data to the cloud server. The larger the value, the higher the feedback incentive before the current moment. The edge gateway passes The better the data transmission effect of each data queue to the cloud server, the The edge gateway passes The greater the confidence of the data queue to transmit data to the cloud server, the smaller the value is, the more confidence there is in the data queue before the current moment. The edge gateway passes The worse the data transmission effect of the first data queue to the cloud server is, the less likely it is to continue to transmit data to the first data queue. The edge gateway passes The data queue transmits data to the cloud server, then The edge gateway passes The smaller the confidence of a data queue in transmitting data to the cloud server.

[0091] Indicates The edge gateway passes The uncertainty estimate of the value of the data queue is The larger the The edge gateway passes The more data queues are explored, the more the The smaller the uncertainty of the first data queue, the smaller the The edge gateway passes The confidence of each data queue increases, enabling the edge gateway to explore data queues with greater uncertainty to select data queues that are more suitable for data transmission; The smaller the description The edge gateway passes The less exploration of the data queue, the better. The greater the uncertainty of the data queue, the larger the The edge gateway passes The confidence of the data queue increases, allowing the edge gateway to explore the data queue to determine the Whether the data queue is suitable for data transmission.

[0092] On the other hand, Under the condition of unchanged The larger the The edge gateway passes The more data queues are explored outside the first data queue, the The edge gateway passes The less exploration of the data queue, the better. The greater the uncertainty of the data queue, the larger the The edge gateway passes The confidence of the data queue increases, allowing the edge gateway to explore the data queue to determine the Whether a data queue is a data queue suitable for data transmission; The smaller the The edge gateway passes The less exploration of data queues outside the first data queue, the The edge gateway passes The more data queues are explored, the more the The smaller the uncertainty of the data queue, the smaller the The edge gateway passes The confidence of each data queue increases, enabling the edge gateway to explore data queues with greater uncertainty to select data queues that are more suitable for data transmission.

[0093] An adjustment value representing the estimate of the uncertainty in the estimate of the value of the data queue, The larger the value, the greater the data transmission pressure at the current moment. Since a high-load environment may not truly reflect the performance of the data queue, the cost of exploring new data queues is higher. Therefore, it is necessary to reduce the exploration of data queues and select more known data queues with better performance. The smaller it is, the smaller the data transmission pressure is at the current moment, the transmission performance of different data queues is more stable, and the actual performance of the data queue is closer to the true potential. In this case, the cost of exploring new data queues is lower, and it is more suitable to try underutilized data queues. Therefore, it is necessary to increase the exploration of data queues and increase attempts at under-evaluated data queues to select data queues that are more suitable for data transmission.

[0094] Get the data queue that is passed when adjusting the edge gateway to transmit data to the cloud server, including the following steps:

[0095] Each edge gateway initializes each data queue, and the feedback incentive for each edge gateway to transmit data through each data queue is initialized to 0 at the initial moment. Then, a data queue is randomly selected for data transmission, and a new feedback incentive is calculated based on the transmission result. Repeat the iteration to calculate the average feedback incentive for the edge gateway to transmit data to the cloud server through each data queue, and calculate the confidence of each edge gateway transmitting data to the cloud server through each data queue, and select the data queue with the largest confidence to transmit data to the cloud server.

[0096] It should be noted that when the system starts or a new data queue is introduced, the feedback incentive of each edge gateway through each data queue is initialized to 0, which means that in the absence of historical data, all data queues are considered to have the same transmission performance. After each data transmission, the system calculates a new feedback incentive based on the transmission results (such as transmission success rate, delay, etc.). This helps the system evaluate the performance of each data queue. Over time, the system continuously repeats the process of data transmission, result collection, and feedback incentive update.

[0097] The present invention also provides a data transmission system for an industrial edge gateway. Figure 2 As shown, the system includes a processor and a memory, the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a data transmission method of an industrial edge gateway according to the first aspect of the present invention is implemented.

[0098] The system also includes other components familiar to those skilled in the art, such as a communication bus and a communication interface. The configuration and functions of these components are known in the art and will not be described in detail here.

[0099] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, apparatus or device. For example, a computer-readable storage medium may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device. Any application or module described in the present invention may be implemented using computer-readable / executable instructions that may be stored or otherwise maintained by such a computer-readable medium.

[0100] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0101] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A data transmission method for an industrial edge gateway, characterized in that: include: Obtain industrial production equipment data from edge gateways, build data queues, and obtain the busyness of each edge gateway based on the operating data of the data queues; According to the busyness of each edge gateway, the transmission results of different data queues after being transmitted to the cloud server are obtained, and the feedback reward of the edge gateway for selecting different data queues for transmission is calculated; Based on the feedback reward, the UCB algorithm is used to calculate the confidence of each data queue in transmitting data to the cloud server, and the data queue through which the edge gateway transmits data to the cloud server is adjusted according to the confidence; The confidence level satisfies the following relationship: , where Indicates the current moment The edge gateway passes The confidence level of each data queue transmitting data to the cloud server, Indicates the number before the current time The edge gateway passes The average feedback incentive for each data queue to transmit data to the cloud server, Indicates The current busyness of the edge gateway, Indicates the current moment The edge gateway passes The number of times a data queue transmits data to the cloud server. Indicates the current moment The total number of times all data queues of the edge gateway transmit data to the cloud server, Expressed as a natural constant The logarithmic function of the base.

2. The data transmission method of an industrial edge gateway according to claim 1, characterized in that: Build data queues, including high-demand queues, large-capacity queues, frequent interaction queues, small data queues, and abnormal data queues; Among them, the high-requirement queue is for processing data of large-scale, intensive tasks; the large-capacity queue is for storing data of large-scale, infrequent interactions; the frequent interaction queue is for processing data of large-scale, intensive tasks; the small data queue is for temporarily storing small-scale, infrequently interacted data; and the abnormal data queue is for giving priority to processing and responding to abnormal data.

3. The data transmission method of an industrial edge gateway according to claim 1, characterized in that: The busyness level includes: Taking any edge gateway as the target gateway and any moment of the target gateway as the marking moment, the product of the number of data tasks and the amount of data collected within the preset time before the marking moment is normalized as the data load intensity; Calculate the ratio of the idle time within the preset time before the marking moment to the preset time and perform exponential decay to obtain the idle time exponential decay; The product of the data load intensity and the exponential decay of the idle time is taken as the busyness of the target gateway at the marking moment.

4. The data transmission method of an industrial edge gateway according to claim 1, characterized in that: The busyness also includes: Take any edge gateway as the target gateway, any moment of the target gateway as the marking moment, and take the product of the data packet loss rate and the memory usage rate in the period before the marking moment as the busyness.

5. The data transmission method of an industrial edge gateway according to claim 1, characterized in that: The feedback reward includes: Taking any edge gateway as the target gateway, the ratio between the total number of data packets received by the server and the number of data packets correctly received is calculated to obtain the data transmission success rate of the target gateway through different data queues; According to the sending time and receiving time of each data packet recorded between the target gateway and the server, the transmission delay of the target gateway through different data queues is obtained; The ratio of the data transmission success rate to the transmission delay is used as the transmission efficiency, the complement of the busyness is used as the idleness, the idleness is corrected using a logarithmic function and a correction parameter, and the product of the transmission efficiency and the corrected idleness is used as the feedback incentive of the target gateway.

6. The data transmission method of an industrial edge gateway according to claim 1, characterized in that: The feedback reward also includes: Taking any edge gateway as the target gateway, taking the complement of the target gateway's busyness as the idleness, and using the logarithmic function and correction parameter to correct the idleness; Obtain the network communication performance parameters of the target gateway, and calculate the percentage of the actual bandwidth used to the total bandwidth to obtain the bandwidth utilization rate; The ratio of the corrected idle level to the bandwidth utilization is used as the feedback incentive of the target gateway.

7. The data transmission method of an industrial edge gateway according to claim 1, characterized in that: The step of adjusting the data queue through which the edge gateway transmits data to the cloud server includes: Each edge gateway initializes each data queue. The initialization is that at the initial moment, the feedback incentive for each edge gateway to transmit data through each data queue is 0, then a data queue is randomly selected for data transmission, and a new feedback incentive is calculated according to the transmission result. Repeat the iteration to calculate the average feedback incentive for the edge gateway to transmit data to the cloud server through each data queue, and calculate the confidence of each edge gateway to transmit data to the cloud server through each data queue, and select the data queue with the largest confidence to transmit data to the cloud server.

8. A data transmission system for an industrial edge gateway, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the data transmission method of the industrial edge gateway according to any one of claims 1 to 7 is implemented.

Images