Intelligent short message scheduling system and method based on 5G network slices

Abstract

The invention belongs to the technical field of intelligent short message service scheduling, and particularly relates to an intelligent short message scheduling system and method based on 5G network slices. According to the method, the short message service processing efficiency in the 5G network slice environment is effectively improved through accurate prediction and intelligent scheduling of the short message service traffic, the change trend of the short message service traffic can be accurately captured through smooth processing and exponentially weighted moving average processing of the traffic change rate time sequence, and the short message service processing efficiency is improved. Reliable data support is provided for subsequent scheduling decisions, meanwhile, through confidence evaluation and resource utilization evaluation, the predicted demand of short message service flow and the resource utilization condition can be comprehensively considered, the scheduling reasonability and effectiveness are ensured, in addition, the intelligent scheduling mechanism can automatically adjust resource allocation according to the scheduling priority, and the scheduling efficiency is improved. The sending requirement of the high-priority short message service is met, and the processing speed of the short message service and the user experience are further improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent scheduling of SMS services, and particularly relates to an intelligent SMS scheduling system and method based on 5G network slicing. Background Art

[0002] With the rapid development of 5G communication technology, SMS services, as an important part of mobile communication networks, face higher requirements for transmission efficiency and service quality. Traditional SMS scheduling methods often rely on static resource allocation and are difficult to adapt to dynamic network environments and user needs. Therefore, how to achieve intelligent scheduling of SMS services has become an urgent problem in the current field of SMS service scheduling. Currently, with the rise of 5G network slicing technology, this technology can divide the physical network into multiple virtual network slices to provide customized network resources and services for different types of services, which provides a new solution for the intelligent scheduling of SMS services.

[0003] In the prior art, although some SMS scheduling schemes based on 5G network slicing technology have been proposed, there are still some problems. For example, some schemes simply allocate resources according to the size of SMS traffic volume, ignoring the differences in network resource requirements for different SMS service types, resulting in low resource utilization. Other schemes, although considering the differences in SMS service types, lack flexibility in the resource scheduling process and are difficult to cope with sudden SMS traffic peaks. In addition, existing schemes often rely on manual experience or fixed algorithm models in SMS traffic prediction and scheduling decision-making, lacking intelligent adaptive capabilities and being unable to reflect changes in network environments and user needs in real time and accurately. Therefore, the present invention proposes an intelligent SMS scheduling method based on 5G network slicing to solve the above problems. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent SMS scheduling system and method based on 5G network slicing, which can dynamically adjust the resource allocation of SMS service network slices according to the real-time network environment and user needs, and improve resource utilization and the transmission efficiency of SMS services.

[0005] The technical solutions adopted by the present invention are specifically as follows: An intelligent SMS scheduling method based on 5G network slicing, comprising: Creating multiple types of SMS service network slices, where the SMS service network slices include low-latency slices, guaranteed slices, and elastic resource slices; Real-time monitoring of the load information of each type of SMS service network slice, and predicting the expected demand for SMS traffic in the future period based on the load information; Evaluate the confidence level of the predicted demand for SMS service traffic of each type of SMS service network slice, and determine the first scheduling condition parameter of each SMS service network slice in the future time period; Collect the resource utilization rate of each type of SMS service network slice, and determine the second scheduling condition parameter of each SMS service network slice in the future time period according to the resource utilization rate and the preset resource threshold; Perform fusion calculation on the first scheduling condition parameter and the second scheduling condition parameter to determine the scheduling priority of each SMS service network slice, and then perform intelligent scheduling on each type of SMS service network slice according to the scheduling priority.

[0006] In a preferred solution, a preemption level matrix is set between the low-latency slice, the guarantee slice, and the elastic resource slice. The level of the low-latency slice is higher than that of the guarantee slice, and the level of the guarantee slice is higher than that of the elastic resource slice; When the low-latency slice encounters high load or emergency demand, it can preferentially occupy the resources of the guarantee slice and the elastic resource slice. When the guarantee slice faces emergency service demands, it can preempt the resources of the elastic resource slice.

[0007] In a preferred solution, the step of predicting the predicted demand for SMS service traffic in the future time period based on load information includes: Collect the historical load data of each type of SMS service network slice and summarize it into a slice time series database including time stamps, service types, and peak traffic; Establish a sliding window and perform time series analysis on the historical load data within the sliding window to identify the periodic change trend and non-periodic trend of the historical load data; Under the periodic change trend, directly predict the predicted demand for SMS service traffic in the future time period according to the periodic change trend; Under the non-periodic trend, perform non-linear fitting on the historical load data to obtain the adjusted value of SMS service traffic in the future time period, and superimpose the SMS service traffic baseline value and the adjusted value of SMS service traffic to obtain the predicted demand for SMS service traffic in the future time period.

[0008] In a preferred solution, the step of performing non-linear fitting on the historical load data to obtain the adjusted value of SMS service traffic in the future time period includes: Set multiple tracking time periods within the sliding window and calculate the change rate of SMS service traffic between adjacent tracking time periods; Establish a time series of traffic change rates according to the change rate of SMS service traffic and perform smoothing processing on the time series of traffic change rates to remove abnormal fluctuation points; Perform exponential weighted moving average processing on the smoothed traffic change rate time series to obtain the predicted adjustment trend of the short message service traffic in the future period, and then calculate the adjustment value of the short message service traffic in the future period according to the adjustment trend of the short message service traffic.

[0009] In a preferred solution, the step of evaluating the confidence level of the predicted demand for the short message service traffic of each type of short message service network slice and determining the first scheduling condition parameter of each short message service network slice in the future period includes: Obtain the historical prediction deviation rate of the predicted demand for the short message service traffic and record it as the first evaluation characteristic parameter; Collect the fluctuation coefficient of the short message service traffic within the current sliding window and record it as the second evaluation characteristic parameter; Obtain the confidence level evaluation function, and input the first evaluation characteristic parameter and the second evaluation characteristic parameter into the confidence level evaluation function together, and output the confidence level evaluation value of each short message service network slice in the future period and record it as the first scheduling condition parameter; Among them, after the first scheduling condition parameter is output, immediately perform a rationality verification process to determine whether the first scheduling condition parameter is within the preset confidence level threshold range. If it exceeds the range, it indicates that the rationality verification fails, and the confidence level evaluation of the predicted demand for the short message service traffic needs to be re-performed. Otherwise, it indicates that the rationality verification is successful, and the corresponding first scheduling condition parameter is retained.

[0010] In a preferred solution, the step of determining the second scheduling condition parameter of each short message service network slice in the future period according to the resource utilization rate and the preset resource threshold includes: Obtain the resource utilization rate indicators of each type of short message network slice within the current sliding window, including CPU occupancy rate, bandwidth occupancy rate, and storage resource occupancy rate; Calculate the difference between the resource utilization rate indicator and the corresponding type of slice resource preset, and generate a resource saturation coefficient matrix; Perform normalization processing on the resource saturation coefficient matrix, and use the elements in the normalized resource saturation coefficient matrix as the second scheduling condition parameter; Among them, the lower the resource saturation coefficient, the higher the resource redundancy of the corresponding short message service network slice.

[0011] In a preferred solution, the step of performing fusion calculation on the first scheduling condition parameter and the second scheduling condition parameter to determine the scheduling priority of each short message service network slice includes: Obtain the first scheduling condition parameter and the second scheduling condition parameter, and assign initial weight coefficients to the first scheduling condition parameter and the second scheduling condition parameter; Dynamically correct the initial weight coefficient of the first scheduling condition parameter according to the historical prediction deviation rate, and output it as the first corrected weight coefficient; Dynamically correct the initial weight coefficient of the second scheduling condition parameter according to the resource saturation coefficient, and output it as the second corrected weight coefficient; Perform weighted summation on the first corrected weight coefficient and the second corrected weight coefficient, and the corresponding first scheduling condition parameter and the second scheduling condition parameter to obtain the comprehensive scheduling evaluation value of each SMS service network slice; Sort each SMS service network slice according to the size of the comprehensive scheduling evaluation value. Among them, the higher the comprehensive scheduling evaluation value of the SMS service network slice, the higher the corresponding scheduling priority.

[0012] In a preferred solution, the step of intelligently scheduling each type of SMS service network slice according to the scheduling priority includes: When the scheduling priority of the SMS service network slice changes, automatically trigger the intelligent scheduling mechanism; Check the usage of the current network slice resources, and determine whether there are idle resources or releasable resources that meet the scheduling requirements; If there are idle resources or releasable resources, allocate the required resources to each SMS service network slice in turn according to the order of the scheduling priority; If the current network slice resources cannot meet the scheduling requirements of all SMS service network slices, automatically adjust the resources occupied by the low-priority slices according to the preemption level matrix to meet the sending requirements of the high-priority SMS services.

[0013] The present invention also provides a SMS intelligent scheduling system based on 5G network slices, which uses the above-mentioned SMS intelligent scheduling method based on 5G network slices, including: A slice construction module for creating various types of SMS service network slices, where the SMS service network slices include low-latency slices, guarantee slices, and elastic resource slices; A prediction module for real-time monitoring of the load information of each type of SMS service network slice, and predicting the expected demand for SMS service traffic in the future period based on the load information; A confidence evaluation module for evaluating the confidence of the expected demand for SMS service traffic of each type of SMS service network slice, and determining the first scheduling condition parameter of each SMS service network slice in the future period; A resource utilization evaluation module for collecting the resource utilization rate of each type of SMS service network slice, and determining the second scheduling condition parameter of each SMS service network slice in the future period according to the resource utilization rate and a preset resource threshold; An intelligent scheduling module is used to perform fusion calculation on the first scheduling condition parameter and the second scheduling condition parameter, determine the scheduling priority of each SMS service network slice, and then perform intelligent scheduling on each type of SMS service network slice according to the scheduling priority.

[0014] And an electronic device, the electronic device includes: At least one processor; And a memory communicatively connected to the at least one processor; Wherein, the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the above-mentioned SMS intelligent scheduling method based on 5G network slices.

[0015] The technical effects achieved by the present invention are: Through the accurate prediction and intelligent scheduling of SMS service traffic, the present invention effectively improves the processing efficiency of SMS services in the 5G network slice environment. Through the smoothing processing and exponentially weighted moving average processing of the traffic change rate time series, the changing trend of SMS service traffic can be accurately captured, providing reliable data support for subsequent scheduling decisions. At the same time, through confidence evaluation and resource utilization evaluation, the expected demand of SMS service traffic and the resource utilization situation can be comprehensively considered to ensure the rationality and effectiveness of scheduling. In addition, the intelligent scheduling mechanism of the present invention can automatically adjust resource allocation according to the scheduling priority to meet the sending requirements of high-priority SMS services, further improving the processing speed of SMS services and the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is a schematic flowchart of the method of the present invention; Figure 2 is a schematic diagram of the system module of the present invention; Figure 3 is a schematic diagram of the structure of the electronic device of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0017] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and understandable, the following detailed description of the specific embodiments of the present invention will be made in conjunction with the accompanying drawings of the specification.

[0018] Many specific details are set forth in the following description in order to provide a thorough understanding of the present invention. However, the present invention may be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0019] Secondly, the so-called "one embodiment" or "embodiment" herein refers to specific features, structures, or characteristics that may be included in at least one implementation manner of the present invention. The "in a preferred embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it an individual or selectively mutually exclusive embodiment with other embodiments.

[0020] Please refer to Figure 1 As shown, the present invention provides a method for intelligent scheduling of short messages based on 5G network slicing, including: S1. Create multiple types of short message service network slices, where the short message service network slices include low-latency slices, guarantee slices, and elastic resource slices; In the step S1, in the 5G network where a short message sending task needs to be executed, according to different types of short message service requirements, corresponding short message service network slices are constructed. The short message service network slices cover multiple types such as low-latency slices, guarantee slices, and elastic resource slices. Multiple of each type of short message service network slice are also set to achieve precise resource allocation and scheduling for different priorities and different types of short message services. The low-latency slice mainly serves short message services with extremely high latency requirements, such as financial transaction short message notifications, emergency security alerts, etc. The guarantee slice mainly serves short message services that need to ensure delivery, such as verification code short messages, important notice short messages, etc. The elastic resource slice can dynamically adjust resources according to the traffic fluctuations of short message services to cope with sudden high-traffic short message sending demands. Among them, a preemption level matrix is set between the low-latency slice, the guarantee slice, and the elastic resource slice. The level of the low-latency slice is higher than that of the guarantee slice, and the level of the guarantee slice is higher than that of the elastic resource slice. When the low-latency slice encounters high load or emergency demands, it can preferentially occupy the resources of the guarantee slice and the elastic resource slice. When the guarantee slice faces emergency service demands, it can preempt the resources of the elastic resource slice to ensure the timely sending of critical short message services.

[0021] S2. Real-time monitor the load information of each type of short message service network slice, and predict the expected demand for short message service traffic in the future period based on the load information; In the step S2, after the short message service network slices are constructed, the load information of each type of short message service network slice is monitored in real time. By collecting and analyzing the load data of the short message service network slices, the expected demand for short message service traffic in a future period is predicted, so as to provide data support for subsequent scheduling. The future period generally ranges from the next 15 minutes to 1 hour, and the specific time length can be adjusted according to actual needs and network conditions to ensure the accuracy and practicality of the prediction results. Among them, the step of predicting the expected demand for short message service traffic in the future period based on the load information includes: Collect the historical load data of various types of SMS service network slices and summarize them into a sliced time series database containing timestamps, service types, and peak traffic; Establish a sliding window and perform time series analysis on the historical load data within the sliding window to identify the periodic change trends and non-periodic trends of the historical load data; Under the periodic change trend, directly predict the expected demand for SMS service traffic in the future period based on the periodic change trend; Under the non-periodic trend, perform non-linear fitting on the historical load data to obtain the adjustment value of SMS service traffic in the future period, and superimpose the baseline value of SMS service traffic and the adjustment value of SMS service traffic to obtain the expected demand for SMS service traffic in the future period; Specifically, when predicting the expected demand for SMS service traffic in the future period based on load information, it is first necessary to comprehensively collect the historical load data of various types of SMS service network slices, and summarize the collected historical load data into a sliced time series database containing key information such as timestamps, service types, and peak traffic to ensure the integrity and accuracy of the data. Then, the size and step length of the sliding window will be preset, and in-depth time series analysis will be performed on the collected historical load data within the sliding window to identify the periodic change trends and non-periodic change trends existing in the historical load data. In the case of identifying the periodic change trend, directly predict the expected demand for SMS service traffic in the future period based on this periodic change trend. In the case of identifying the non-periodic trend, more complex non-linear fitting processing needs to be performed on the historical load data. Based on this process, obtain the adjustment value of SMS service traffic in the future period, and superimpose the adjustment value of SMS service traffic and the baseline value of SMS service traffic ( , where represents the expected demand for SMS service traffic, represents the predicted value of the baseline traffic (obtained by prediction based on the periodic trend), and finally obtain the expected demand for SMS service traffic in the future period to ensure the comprehensiveness and accuracy of the prediction result.

[0022] It should be noted that the steps of performing non-linear fitting on the historical load data to obtain the adjustment value of SMS service traffic in the future period include: Set multiple tracking periods within the sliding window and calculate the change rate of SMS service traffic between adjacent tracking periods; Establish a time series of traffic change rates based on the change rate of SMS service traffic and perform smoothing processing on the time series of traffic change rates to remove abnormal fluctuation points; Perform exponential weighted moving average processing on the smoothed traffic change rate time series to obtain the predicted adjustment trend of the SMS service traffic in the future period. Then, based on the adjustment trend of the SMS service traffic, calculate the adjustment value of the SMS service traffic in the future period. In this embodiment, first, set multiple consecutive tracking periods within the sliding window, monitor and record the SMS service traffic within the tracking periods, and calculate the change rate of the SMS service traffic between adjacent tracking periods based on this to achieve the purpose of capturing the dynamic change characteristics of the SMS service traffic. Then, construct a traffic change rate time series based on the obtained change rate of the SMS service traffic. To ensure the stability and reliability of the data, smooth processing is performed on this traffic change rate time series to effectively remove the abnormal fluctuation points therein, thereby obtaining a smoother time series that reflects the true change trend. Then, perform exponential weighted moving average processing on the smoothed traffic change rate time series, assigning higher weights to recent data to make the prediction result closer to the current trend, thereby obtaining the predicted adjustment trend of the SMS service traffic in the future period. Finally, based on the predicted adjustment trend of the SMS service traffic, combined with specific service requirements and historical data, calculate the adjustment value of the SMS service traffic in the future period ( , where represents the future period of the adjustment value of the SMS service traffic, represents the attenuation factor, with a value of 0 < < 1, represents the traffic change rate), and then, based on the baseline value of the SMS service traffic, the expected demand for the SMS service traffic in the future period can be obtained.

[0023] S3. Evaluate the confidence level of the expected demand for the SMS service traffic of each type of SMS service network slice, and determine the first scheduling condition parameter of each SMS service network slice in the future period; In step S3, after obtaining the expected demand for the SMS service traffic in the future period, to ensure the rationality and accuracy of the scheduling decision, the confidence level of the expected demand for the SMS service traffic of each type of SMS service network slice is evaluated, and the first scheduling condition parameter of each SMS service network slice in the future period is determined through a scientific method. Among them, the steps of evaluating the confidence level of the expected demand for the SMS service traffic of each type of SMS service network slice and determining the first scheduling condition parameter of each SMS service network slice in the future period include: Obtain the historical prediction deviation rate of the expected demand for the SMS service traffic and record it as the first evaluation characteristic parameter; Collect the fluctuation coefficient of the SMS service traffic within the current sliding window and record it as the second evaluation characteristic parameter; Obtain a confidence evaluation function, and input the first evaluation characteristic parameter and the second evaluation characteristic parameter into the confidence evaluation function together, output the confidence evaluation value of each SMS service network slice in the future time period, and record it as the first scheduling condition parameter; After the first dispatch condition parameter is output, a rationality check is immediately performed to determine whether the first dispatch condition parameter is within a preset confidence threshold. If it exceeds the range, it indicates that the rationality check fails and the confidence evaluation of the estimated demand for SMS service traffic needs to be re-performed. Otherwise, it indicates that the rationality check succeeds and the corresponding first scheduling condition parameter is retained; Specifically, when performing the confidence assessment of the expected demand for SMS service traffic, first obtain the historical forecast deviation rate data of the expected demand for SMS service traffic and record it as the first evaluation characteristic parameter for subsequent analysis. Then, within the currently set sliding window range, collect the fluctuation coefficient of SMS service traffic (fluctuation coefficient = (maximum value of SMS service traffic in the current period - minimum value of SMS service traffic in the current period) / average value of SMS service traffic in the current period)), and accurately record this fluctuation coefficient data as the second evaluation characteristic parameter to provide a pre-data reference for the confidence assessment. Then obtain the confidence assessment function applicable to this scenario, and input the previously recorded first evaluation characteristic parameter and second evaluation characteristic parameter into the confidence assessment function. The confidence assessment function calculates and outputs the confidence assessment value of each SMS service network slice in the future period, and records this confidence assessment value as the first scheduling condition parameter, wherein the expression of the confidence assessment function is: ; In the formula, Indicates The confidence evaluation value of the SMS-like service network slice (the first scheduling condition parameter), Indicates The historical prediction deviation rate of SMS-like service network slices (the first evaluation characteristic parameter), Indicates The fluctuation coefficient of the SMS-like service network slice (the second evaluation characteristic parameter); It is particularly important to note that after the first scheduling condition parameter is output, a rationality check will be immediately performed to determine whether the first scheduling condition parameter is within the preset confidence threshold range. If the first scheduling condition parameter exceeds the preset confidence threshold range, it indicates that the rationality check has failed. At this time, it is necessary to re-evaluate the confidence of the estimated demand for SMS service traffic to ensure the accuracy of the evaluation result. Conversely, if the first scheduling condition parameter is within the preset confidence threshold range, it indicates that the rationality check is successful, and the corresponding first scheduling condition parameter can be retained as a prerequisite for subsequent scheduling decisions.

[0024] S4. Collect the resource utilization rates of network slices for various types of SMS services, and determine the second scheduling condition parameters for each SMS service network slice in the future period according to the resource utilization rates and preset resource thresholds. In step S4, after the first adjustment condition parameters are output, it is also necessary to collect the resource utilization rate data of various SMS service network slices, and comprehensively determine the second scheduling condition parameters for each SMS service network slice in the future period in combination with the preset resource thresholds to ensure the reasonable allocation and efficient utilization of resources. The step of determining the second scheduling condition parameters for each SMS service network slice in the future period according to the resource utilization rates and preset resource thresholds includes: Obtain the resource utilization rate indicators of various types of SMS network slices within the current sliding window, including CPU occupancy rate, bandwidth occupancy rate, and storage resource occupancy rate. Calculate the difference between the resource utilization rate indicators and the preset resource thresholds for the corresponding type of slices to generate a resource saturation coefficient matrix. Normalize the resource saturation coefficient matrix, and use the elements in the normalized resource saturation coefficient matrix as the second scheduling condition parameters. Among them, the lower the resource saturation coefficient, the higher the resource redundancy of the corresponding SMS service network slice. Specifically, first obtain the resource utilization rate indicators of various types of SMS network slices within the current sliding window. The resource utilization rate indicators cover multiple aspects such as CPU occupancy rate, bandwidth occupancy rate, and storage resource occupancy rate to ensure a comprehensive assessment of resource usage. Then calculate the difference between the obtained resource utilization rate indicators and the preset resource thresholds for the corresponding type of slices, and generate a resource saturation coefficient matrix based on this process. The resource saturation coefficient matrix can intuitively reflect the tightness of resources for each slice. Then normalize the generated resource saturation coefficient matrix. The normalization method can be min-max normalization or Z-score normalization, etc., to ensure that the element values are in the same magnitude. Finally, use the elements in the normalized resource saturation coefficient matrix as the second scheduling condition parameters for subsequent scheduling decisions. It should be noted that the lower the resource saturation coefficient, the higher the resource redundancy of the corresponding SMS service network slice, and vice versa, indicating that resources are tight and need to be scheduled and optimized first.

[0025] S5. Perform a fusion calculation on the first scheduling condition parameters and the second scheduling condition parameters to determine the scheduling priorities of each SMS service network slice, and then perform intelligent scheduling on various types of SMS service network slices according to the scheduling priorities. In step S5, after both the first scheduling condition parameter and the second scheduling condition parameter are output, the first scheduling condition parameter and the second scheduling condition parameter will be fused and calculated. After comprehensive evaluation, the scheduling priorities of each SMS service network slice are determined, and then intelligent scheduling is performed on each type of SMS service network slice according to these scheduling priorities, so as to realize the optimized management and efficient operation of the SMS service. Among them, the steps of fusing and calculating the first scheduling condition parameter and the second scheduling condition parameter to determine the scheduling priorities of each SMS service network slice include: Obtain the first scheduling condition parameter and the second scheduling condition parameter, and assign initial weight coefficients to the first scheduling condition parameter and the second scheduling condition parameter; Dynamically correct the initial weight coefficient of the first scheduling condition parameter according to the historical prediction deviation rate, and output it as the first corrected weight coefficient; Dynamically correct the initial weight coefficient of the second scheduling condition parameter according to the resource saturation coefficient, and output it as the second corrected weight coefficient; Perform weighted summation on the first corrected weight coefficient and the second corrected weight coefficient, and the corresponding first scheduling condition parameter and the second scheduling condition parameter to obtain the comprehensive scheduling evaluation value of each SMS service network slice; Sort each SMS service network slice according to the size of the comprehensive scheduling evaluation value. Among them, the higher the comprehensive scheduling evaluation value of the SMS service network slice, the higher the corresponding scheduling priority; Specifically, when fusing and calculating the first scheduling condition parameter and the second scheduling condition parameter, first, obtain the first scheduling condition parameter and the second scheduling condition parameter, and assign initial weight coefficients to the first scheduling condition parameter and the second scheduling condition parameter respectively. At the same time, the initial weight coefficient of the first scheduling condition parameter will be dynamically corrected according to the historical prediction deviation rate. Specifically, by analyzing the prediction deviation situation in historical data, the weight of the first scheduling condition parameter is adjusted to make it more in line with the actual situation, and the corrected weight coefficient is output as the first corrected weight coefficient ( , where represents the first corrected weight coefficient, represents the initial weight coefficient of the first scheduling condition parameter, represents the historical prediction deviation attenuation factor). The higher the historical prediction deviation rate, the more significant the corresponding weight attenuation, and the lower the corresponding first corrected weight coefficient, and vice versa. Then, dynamically correct the initial weight coefficient of the second scheduling condition parameter according to the resource saturation coefficient. The resource saturation coefficient reflects the utilization of current network resources. By adjusting the resource saturation coefficient, the weight of the second scheduling condition parameter can be made more reasonable, and the corrected weight coefficient in this process is output as the second corrected weight coefficient ( , where represents the second correction weight coefficient, represents the initial weight coefficient of the second scheduling condition parameter, represents the normalized resource saturation coefficient of the category of short message service network slices for the category of resource providers. Then, the first correction weight coefficient and the second correction weight coefficient are respectively weighted and summed with the corresponding first scheduling condition parameter and the second scheduling condition parameter, so as to obtain the comprehensive scheduling evaluation value of each short message service network slice. Finally, according to the magnitude of the comprehensive scheduling evaluation value, each short message service network slice is sorted. Specifically, the higher the comprehensive scheduling evaluation value of a short message service network slice, the higher its corresponding scheduling priority. Through this sorting method, it can be ensured that those short message services with higher comprehensive evaluation values are processed preferentially, thereby optimizing the overall resource scheduling efficiency.

[0026] Secondly, the steps of intelligently scheduling each type of short message service network slice according to the scheduling priority include: When the scheduling priority of the short message service network slice changes, the intelligent scheduling mechanism is automatically triggered; Check the usage of the current network slice resources to determine whether there are idle resources or releasable resources that meet the scheduling requirements; If there are idle resources or releasable resources, the required resources are allocated to each short message service network slice in turn according to the order of the scheduling priority; If the current network slice resources cannot meet the scheduling requirements of all short message service network slices, the resources occupied by the low-priority slices are automatically adjusted according to the pre-emptive level matrix to meet the sending requirements of the high-priority short message services; In this embodiment, when any change occurs in the scheduling priority of the short message service network slice, the intelligent scheduling mechanism will be automatically triggered to ensure that the demand for priority adjustment can be responded to in a timely manner. First, the actual usage of the current network slice resources will be comprehensively checked to evaluate and determine whether there are idle resources that meet the scheduling requirements, or whether there are releasable resources for reallocation. If it is found through the check that there are indeed idle resources or releasable resources, the required resources will be reasonably allocated to each short message service network slice in turn according to the order of the output scheduling priority, ensuring that high-priority services can obtain resource support first. However, in the case where the current network slice resources are insufficient to meet the scheduling requirements of all short message service network slices, the resources occupied by those low-priority slices will be automatically adjusted according to the pre-set pre-emptive level matrix, and in this way, the sending requirements of high-priority short message services will be preferentially guaranteed to ensure the smooth progress of key services.

[0027] Please refer toFigure 2 , a short message intelligent scheduling system based on 5G network slicing, using the above-mentioned short message intelligent scheduling method based on 5G network slicing, includes: A slice construction module, used to create multiple types of short message service network slices, where the short message service network slices include low-latency slices, guarantee slices, and elastic resource slices; A prediction module, used to monitor the load information of each type of short message service network slice in real time, and predict the expected demand for short message service traffic in the future period based on the load information; A confidence evaluation module, used to evaluate the confidence of the expected demand for short message service traffic of each type of short message service network slice, and determine the first scheduling condition parameter of each short message service network slice in the future period; A resource utilization evaluation module, used to collect the resource utilization rate of each type of short message service network slice, and determine the second scheduling condition parameter of each short message service network slice in the future period according to the resource utilization rate and a preset resource threshold; An intelligent scheduling module, used to perform fusion calculation on the first scheduling condition parameter and the second scheduling condition parameter, determine the scheduling priority of each short message service network slice, and then perform intelligent scheduling on each type of short message service network slice according to the scheduling priority.

[0028] In the above, the main function of the slice construction module is to create and construct multiple different types of short message service network slices. Specifically, the short message service network slices cover multiple types such as low-latency slices, guarantee slices, and elastic resource slices to meet the needs of different short message service scenarios. The prediction module is responsible for monitoring and collecting the load information of each type of short message service network slice in real time. Based on the real-time obtained load information, the prediction module can predict the expected demand for short message service traffic in the future period through advanced algorithms and models, so as to provide data support for subsequent scheduling decisions. The role of the confidence evaluation module is to evaluate the confidence of the expected demand for short message service traffic of each type of short message service network slice, and determine the first scheduling condition parameter of each short message service network slice in the future period through comprehensive analysis of various factors, ensuring the accuracy and reliability of scheduling decisions. The main task of the resource utilization evaluation module is to collect and count the resource utilization rate data of each type of short message service network slice. On this basis, combined with the preset resource threshold, determine the second scheduling condition parameter of each short message service network slice in the future period to better optimize resource allocation and utilization. The intelligent scheduling module is responsible for performing fusion calculation on the first scheduling condition parameter and the second scheduling condition parameter, determining the scheduling priority of each short message service network slice through comprehensive analysis of these two parameters, and performing intelligent scheduling on each type of short message service network slice according to the scheduling priority, so as to achieve efficient and accurate short message service management.

[0029] Please refer to Figure 3 , an electronic device, which includes: At least one processor; And a memory communicatively connected to the at least one processor; Wherein, the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the above-mentioned intelligent SMS scheduling method based on 5G network slices.

[0030] The processor of the above-mentioned electronic device can be a central processing unit (CPU), a graphics processing unit (GPU), or a digital signal processor (DSP), etc., and the memory can include a read-only memory (ROM), a random access memory (RAM), a flash memory (Flash), or a hard disk, etc. In the electronic device, the processor reads the computer program stored in the memory and executes the above-mentioned intelligent SMS scheduling method based on 5G network slices to achieve intelligent scheduling and management of the SMS service network slice. In addition, the electronic device can also include an arithmetic unit, an input device, an output device, and a network interface, etc. The arithmetic unit can provide computing support for various arithmetic and logical operations. The input device, such as a keyboard, a touch screen, etc., is used to receive user operation instructions and data input. The output device, such as a display screen, a printer, etc., is used to display the processing results and output information. The network interface is used to realize the connection and communication between the electronic device and other devices or networks, facilitating data transmission and sharing.

[0031] It should be noted that in this article, the term "including", "comprising", or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, device, article, or method including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or further includes elements inherent to such a process, device, article, or method. Without further limitations, an element defined by the statement "including one..." does not exclude the existence of another identical element in the process, device, article, or method including that element.

[0032] The above are only the preferred embodiments of the present invention. It should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention. The structures, devices, and operation methods not specifically described and explained in the present invention are implemented according to the conventional means in the art without special instructions and limitations.

Claims

1. A smart SMS scheduling method based on 5G network slicing, characterized in that: Including: Create multiple types of SMS service network slices, where the SMS service network slices include low-latency slices, guaranteed slices, and elastic resource slices; Real-time monitor the load information of each type of SMS service network slice, and predict the expected demand for SMS service traffic in the future period based on the load information; Conduct a confidence assessment on the expected demand for SMS service traffic of each type of SMS service network slice, and determine the first scheduling condition parameters of each SMS service network slice in the future period; Collect the resource utilization rate of each type of SMS service network slice, and determine the second scheduling condition parameters of each SMS service network slice in the future period according to the resource utilization rate and a preset resource threshold; Perform a fusion calculation on the first scheduling condition parameters and the second scheduling condition parameters to determine the scheduling priority of each SMS service network slice, and then perform intelligent scheduling on each type of SMS service network slice according to the scheduling priority.

2. The SMS intelligent scheduling method based on 5G network slicing according to claim 1, characterized in that: A preemption level matrix is set between the low-latency slice, the guaranteed slice, and the elastic resource slice, and the level of the low-latency slice is higher than that of the guaranteed slice, and the level of the guaranteed slice is higher than that of the elastic resource slice; When the low-latency slice encounters high load or emergency demand, it can preferentially occupy the resources of the guaranteed slice and the elastic resource slice. When the guaranteed slice faces emergency service demand, it can preempt the resources of the elastic resource slice.

3. The SMS intelligent scheduling method based on 5G network slicing according to claim 1, characterized in that: The step of predicting the expected demand for SMS service traffic in the future period based on the load information includes: Collect the historical load data of each type of SMS service network slice and summarize it into a slice time series database containing timestamps, service types, and peak traffic; Establish a sliding window and perform time series analysis on the historical load data within the sliding window to identify the periodic change trend and non-periodic trend of the historical load data; Under the periodic change trend, directly predict the expected demand for SMS service traffic in the future period according to the periodic change trend; Under the non-periodic trend, perform non-linear fitting on the historical load data to obtain the SMS service traffic adjustment value in the future period, and superimpose the SMS service traffic baseline value and the SMS service traffic adjustment value to obtain the expected demand for SMS service traffic in the future period.

4. The SMS intelligent scheduling method based on 5G network slicing according to claim 3, characterized in that: The step of performing non-linear fitting on the historical load data to obtain the SMS service traffic adjustment value in the future period includes: Set multiple tracking periods within the sliding window and calculate the SMS service traffic change rate between adjacent tracking periods; Establish a traffic change rate time series according to the SMS service traffic change rate and perform smoothing processing on the traffic change rate time series to remove abnormal fluctuation points; Perform exponentially weighted moving average processing on the smoothed traffic change rate time series to obtain the predicted SMS service traffic adjustment trend in the future period, and then calculate the SMS service traffic adjustment value in the future period according to the SMS service traffic adjustment trend.

5. The SMS intelligent scheduling method based on 5G network slicing according to claim 1, wherein: The step of conducting a confidence assessment on the expected demand for SMS service traffic of each type of SMS service network slice and determining the first scheduling condition parameters of each SMS service network slice in the future period includes: Obtain the historical prediction deviation rate of the predicted demand for SMS service traffic and record it as the first evaluation characteristic parameter; Collect the fluctuation coefficient of SMS service traffic within the current sliding window and record it as the second evaluation characteristic parameter; Obtain the confidence evaluation function, and input the first evaluation characteristic parameter and the second evaluation characteristic parameter into the confidence evaluation function together, output the confidence evaluation value of each SMS service network slice in the future period, and record it as the first scheduling condition parameter; Among them, after the first scheduling condition parameter is output, immediately perform a rationality verification process to determine whether the first scheduling condition parameter is within the preset confidence threshold range; If it exceeds the range, it indicates that the rationality verification fails, and the confidence evaluation of the predicted demand for SMS service traffic needs to be re-performed. Otherwise, it indicates that the rationality verification is successful, and the corresponding first scheduling condition parameter is retained.

6. The SMS intelligent scheduling method based on 5G network slicing according to claim 1 is characterized in that: The step of determining the second scheduling condition parameter of each SMS service network slice in the future period according to the resource utilization rate and the preset resource threshold includes: Obtain the resource utilization rate indicators of each type of SMS network slice within the current sliding window, including CPU occupancy rate, bandwidth occupancy rate, and storage resource occupancy rate; Calculate the difference between the resource utilization rate indicator and the corresponding type of slice resource, and generate a resource saturation coefficient matrix; Perform normalization processing on the resource saturation coefficient matrix, and use the elements in the normalized resource saturation coefficient matrix as the second scheduling condition parameter; Among them, the lower the resource saturation coefficient, the higher the resource redundancy of the corresponding SMS service network slice.

7. The intelligent SMS scheduling method based on 5G network slicing according to claim 1, characterized in that: The step of performing fusion calculation on the first scheduling condition parameter and the second scheduling condition parameter to determine the scheduling priority of each SMS service network slice includes: Obtain the first scheduling condition parameter and the second scheduling condition parameter, and assign initial weight coefficients to the first scheduling condition parameter and the second scheduling condition parameter; Dynamically correct the initial weight coefficient of the first scheduling condition parameter according to the historical prediction deviation rate, and output it as the first corrected weight coefficient; Dynamically correct the initial weight coefficient of the second scheduling condition parameter according to the resource saturation coefficient, and output it as the second corrected weight coefficient; Perform weighted summation on the first corrected weight coefficient and the second corrected weight coefficient with the corresponding first scheduling condition parameter and second scheduling condition parameter to obtain the comprehensive scheduling evaluation value of each SMS service network slice; Sort each SMS service network slice according to the size of the comprehensive scheduling evaluation value. Among them, the higher the comprehensive scheduling evaluation value of the SMS service network slice, the higher the corresponding scheduling priority.

8. The SMS intelligent scheduling method based on 5G network slicing according to claim 1, characterized in that: The step of performing intelligent scheduling on each type of SMS service network slice according to the scheduling priority includes: When the scheduling priority of the SMS service network slice changes, automatically trigger the intelligent scheduling mechanism; Check the usage of the current network slice resources to determine whether there are idle resources or releasable resources that meet the scheduling requirements; If there are idle resources or releasable resources, allocate the required resources to each SMS service network slice in turn according to the high and low order of the scheduling priority; If the current network slice resources cannot meet the scheduling requirements of all SMS service network slices, the resources occupied by low-priority slices are automatically adjusted according to the preemption level matrix to meet the sending requirements of high-priority SMS services.

9. A short message intelligent scheduling system based on 5G network slicing, characterized in that: The SMS intelligent scheduling method based on 5G network slices according to any one of claims 1 to 8, comprising: A slice construction module for creating various types of SMS service network slices, wherein the SMS service network slices include low-latency slices, guaranteed slices, and elastic resource slices; A prediction module for real-time monitoring of the load information of each type of SMS service network slice and predicting the expected demand for SMS service traffic in a future period based on the load information; A confidence evaluation module for performing a confidence evaluation on the expected demand for SMS service traffic of each type of SMS service network slice to determine the first scheduling condition parameter of each SMS service network slice in a future period; A resource utilization evaluation module for collecting the resource utilization rate of each type of SMS service network slice and determining the second scheduling condition parameter of each SMS service network slice in a future period according to the resource utilization rate and a preset resource threshold; An intelligent scheduling module for performing a fusion calculation on the first scheduling condition parameter and the second scheduling condition parameter to determine the scheduling priority of each SMS service network slice, and then performing intelligent scheduling on each type of SMS service network slice according to the scheduling priority.

10. An electronic device, characterized in that: The electronic device includes: At least one processor; And a memory communicatively connected to the at least one processor; Wherein, the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the SMS intelligent scheduling method based on 5G network slices according to any one of claims 1 to 8.

Images