Data transmission resource prediction method based on historical data

Abstract

The present invention relates to the field of resource prediction, and in particular to a data transmission resource prediction method based on historical data. The present invention acquires historical data transmission data when a satellite runs to different transit time windows to analyze data transmission behavior characteristics, calculates fuzzy characterization parameters of data transmission behavior, sets labels for sites, determines the dimensions of historical data transmission data required to be called based on the fuzzy characterization parameters, calls historical data transmission data of corresponding dimensions to use a pre-trained prediction model to predict the data transmission resource demand in each historical period, and verifies the validity of the prediction model for the site based on the data transmission resource reception amount in the corresponding historical period to determine whether to output the prediction result of the prediction model, determine the data transmission resource demand, or, based on the fluctuation range of the historical data transmission resource reception amount, determine the data transmission resource demand of the site. The present invention determines the resource prediction method by analyzing the historical data of the site, thereby improving the accuracy of the prediction, efficiency and utilization.

Description

Technical Field

[0001] The present invention relates to the field of resource prediction, and in particular to a digital resource prediction method based on historical data. Background Art

[0002] With the rapid development of science and technology, people's demand for communication technology is getting higher and higher. However, due to the limitations of natural conditions, the scope and effect of communication through ground base stations are limited to a certain extent. Satellite communication systems can communicate and transmit around the world and have strong adaptability, but the resource allocation problem is one of the difficulties that need to be solved in satellite communication systems. Resource allocation in satellite communication systems refers to the reasonable allocation of precious satellite communication spectrum resources to different sites or users to maximize economic and technical benefits. For satellite communication systems, excellent resource allocation methods can significantly improve the efficiency and utilization of communication systems.

[0003] China Patent Publication No.: CN118644269A, discloses a method for predicting business demand of satellite communication terminals based on artificial intelligence algorithm, including the following steps: S1: using the business demand data of each beam coverage area in history as training samples, training a prediction model to predict business demand for a period of time in the future based on the historical business demand of the beam; S2: inputting the historical business demand of the beam coverage area to be predicted into the prediction model, and the prediction model outputs the final future business demand; wherein, in step S1, the training prediction model is a Seq2Seq model using an LSTM network as an encoder and decoder. The method of the invention can face the dynamic changes in business demand, make efficient and accurate predictions, and serve as input for adjusting the on-orbit resources of communication satellites, thereby improving the resource utilization rate of communication satellites and user satisfaction.

[0004] Chinese Patent Publication No.: CN116611587A, discloses a solar energy resource prediction method based on polar-orbit-geostationary satellite fusion technology, including: obtaining polar-orbit satellite solar energy data and geostationary satellite solar energy data at the time of geostationary satellite collection; determining the geostationary satellite solar energy data at the time of polar-orbit satellite collection and the geostationary satellite solar energy data at a first prediction time according to the geostationary satellite solar energy data at the time of geostationary satellite collection; determining the spatial conversion coefficient by linear regression using the polar-orbit satellite solar energy data and the geostationary satellite solar energy data at the time of polar-orbit satellite collection; establishing a fusion model based on the spatial conversion coefficient according to the polar-orbit satellite solar energy data, the geostationary satellite solar energy data at the time of polar-orbit satellite collection and the geostationary satellite solar energy data at the first prediction time; and outputting the polar-orbit satellite solar energy data at the first prediction time within a preset spatial interval according to the fusion model.

[0005] However, there are still the following problems in the prior art:

[0006] In actual situations, when predicting the demand for data transmission resources at a site, most people rely on manual experience to apply for resources according to fixed ideas or make overall predictions based on existing models. However, after the successful resource pre-allocation, deviations may occur later, resulting in waste of resources, loss of computing power, and reduced efficiency and utilization of the communication system. Summary of the invention

[0007] To this end, the present invention provides a data transmission resource prediction method based on historical data to solve the problem that in actual situations, when predicting the data transmission resource demand of a site, most of the time, resources are applied for according to fixed ideas based on manual experience or an overall prediction is made based on existing models. However, after the successful application of resource pre-allocation, deviations may occur in the later stage, resulting in waste of resources, loss of computing power, and reduced efficiency and utilization of the communication system.

[0008] To achieve the above object, the present invention provides a data transmission resource prediction method based on historical data, which comprises:

[0009] Obtain the historical data transmission between each station and the satellite when the satellite moves to different transit time windows;

[0010] Analyzing data transmission behavior characteristics based on the historical data transmission data, calculating data transmission behavior fuzzy characterization parameters for the site based on the data transmission behavior characteristics, and setting a label for the site, wherein the data transmission behavior characteristics include task identity characteristics of data transmission resources and a receiving amount fluctuation value of the data transmission resources;

[0011] Based on the tags of each station, the historical data transmission data of the station is analyzed to determine the data transmission resource requirements of each station when the satellite operates to each transit time window, including:

[0012] Determine the dimension of the historical data transmission data to be called based on the fuzzy characterization parameter, call the historical data transmission data of the corresponding dimension to use the pre-trained prediction model to predict the data transmission resource demand in each historical period, and verify the validity of the prediction model for the site based on the data transmission resource reception amount in the corresponding historical period to determine whether to output the prediction result of the prediction model and determine the data transmission resource demand;

[0013] Or, based on the fluctuation range of the amount of data transmission resource received in a historical period, determine the data transmission resource demand of the site;

[0014] The satellite is controlled to allocate corresponding data transmission resources based on the data transmission resource demand.

[0015] Furthermore, the process of analyzing the data transmission behavior characteristics based on the historical data transmission data includes:

[0016] Determine the application content and the acceptance content of digital transmission resources;

[0017] Determining the relevance between the application content and the received content as a task identity feature of the data transmission resource;

[0018] Determine the variance of the received amount of data transmission resources in a number of historical periods, and determine the variance as a received amount fluctuation value;

[0019] The application content is text information sent by the site, the received content includes data transmission resources in several formats transmitted by the satellite to the site, and the correlation is determined based on the text information and the text description of the received content.

[0020] Furthermore, the process of calculating the fuzzy characterization parameters of the data transmission behavior of the site includes:

[0021] Determining a ratio of the task identity feature to the benchmark identity feature as an identity impact factor;

[0022] Determine the ratio of the reference reception volume fluctuation value to the reception volume fluctuation value as the reception volume fluctuation value influencing factor;

[0023] The weighted sum of the identity influencing factor and the receiving amount fluctuation value influencing factor is determined as the fuzzy characterization parameter of the data transmission behavior.

[0024] Further, a label is set for the site, wherein:

[0025] If the fuzzy characterization parameter of the digital transmission behavior is greater than or equal to the reference fuzzy characterization parameter threshold, setting the label to a clear label;

[0026] If the fuzzy characterization parameter of the digital transmission behavior is less than the reference fuzzy characterization parameter threshold, the label is set as a fuzzy label.

[0027] Furthermore, the data transmission resource demand of each site when the satellite moves to each transit time window is determined, wherein:

[0028] If the label is a fuzzy label, the dimension of the historical data transmission data to be called is determined based on the fuzzy characterization parameter, the historical data transmission data of the corresponding dimension is called to use a pre-trained prediction model to predict the data transmission resource demand in each historical period, and the validity of the prediction model for the site is verified based on the data transmission resource reception amount in the corresponding historical period to determine whether to output the prediction result of the prediction model and determine the data transmission resource demand;

[0029] If the tag is a clear tag, the data transmission resource demand of the site is determined based on the fluctuation range of the data transmission resource reception amount in the historical period.

[0030] Furthermore, the process of determining the dimension of the historical data transmission data to be called based on the fuzzy characterization parameter includes:

[0031] Determine several dimensions of historical data transmission data;

[0032] Determining the interval in which the fuzzy characterization parameter is located;

[0033] Determine that the dimension corresponding to the interval is the dimension of the historical data transmission data to be called;

[0034] Among them, the correspondence between each interval and each dimension is pre-set, and the dimensions include the satellite's orbital offset, the satellite's depression angle, the satellite's transmission power, the site's elevation angle, the site's signal shielding rate, and the relative distance between the site and the satellite.

[0035] Furthermore, the process of verifying the validity of the prediction model for the site based on the amount of data transmission resource received in the corresponding historical period includes:

[0036] Determine the amount of data transmission resources received by the site during a number of historical periods;

[0037] Determine the predicted demand for data transmission resources corresponding to the site in several historical periods;

[0038] Determine the difference between the received amount of data transmission resources and the predicted required amount of resources as a validity parameter;

[0039] If the validity parameter is less than a predetermined parameter threshold, it is determined that the prediction model is valid for the site.

[0040] Furthermore, the determining whether to output the prediction result of the prediction model comprises:

[0041] If the prediction model is valid, the prediction result of the prediction model is output;

[0042] If the prediction model is invalid, the prediction result of the prediction model is not output and the site is marked.

[0043] Furthermore, the process of determining the data transmission resource requirement of the site includes:

[0044] Determine the fluctuation range of the data transmission resource reception amount corresponding to each historical period of the site;

[0045] An average value of the fluctuation range is determined as the data transmission resource requirement.

[0046] Furthermore, the process of controlling the satellite to allocate corresponding data transmission resources based on the data transmission resource demand includes:

[0047] Determine the data transmission resource requirements of each station within each transit time window;

[0048] The data transmission resource requirements of each site are sent to the satellite to control the satellite to allocate corresponding data transmission resources.

[0049] Compared with the prior art, the present invention obtains historical data transmission data when the satellite runs to different transit time windows to analyze the data transmission behavior characteristics, calculate the fuzzy characterization parameters of the data transmission behavior, set labels for the sites, determine the dimensions of the historical data transmission data required to be called based on the fuzzy characterization parameters, call the historical data transmission data of the corresponding dimensions to use a pre-trained prediction model to predict the data transmission resource demand in each historical period, and verify the validity of the prediction model for the site based on the data transmission resource reception in the corresponding historical period to determine whether to output the prediction result of the prediction model, determine the data transmission resource demand, or, based on the fluctuation range of the historical data transmission resource reception, determine the data transmission resource demand of the site. The present invention improves the accuracy of the prediction by analyzing the historical data of the site and determining the resource prediction method, and improves the efficiency and utilization of data transmission resource acquisition of the site and the satellite.

[0050] In particular, the present invention calculates fuzzy characterization parameters of the data transmission behavior for the site based on the data transmission behavior characteristics, sets labels for the sites, and determines whether the data transmission resource requirements of each site are clear. In actual situations, when predicting data transmission resources, most of the predictions are made through manual experience, or directly through models. However, the first method is not accurate enough in complex situations, which may cause deviations in resource predictions, resulting in insufficient or wasteful resources. Although the second method has higher accuracy, performing model predictions for each site and each cycle will cause a load on the computer, resulting in a waste of computing power. Based on this, the present invention considers calculating fuzzy characterization parameters of the data transmission behavior of the site, which are used to set labels for the site, characterize the fuzziness of the data transmission behavior of the site, and adopt different prediction methods for sites with different labels, thereby improving computing power, and improving the efficiency and utilization of data transmission resource acquisition for sites and satellites.

[0051] In particular, the present invention determines the dimensions of the historical data transmission data required to be called based on the fuzzy characterization parameters. In actual situations, when using the model to predict the data transmission resources, due to the large number of transit time windows and the large number of sites, the data required to be predicted is massive, and the fuzziness of the data transmission behavior of each site is inconsistent. The superposition of multiple factors leads to serious computing power consumption. Therefore, the present invention considers using a model prediction method for sites with fuzzy labels, and considers determining the dimensions of the historical data transmission data required to be called based on the fuzzy characterization parameters. When making a prediction, the demand for data transmission resources is predicted according to the determined dimensions, thereby ensuring the accuracy of the prediction and improving the efficiency and utilization of data transmission resource acquisition of sites and satellites.

[0052] In particular, the present invention verifies the validity of the prediction model. Verification of the validity of the model is a necessary step in model prediction. In actual use, if the model is used directly without being verified, inaccurate prediction of data transmission resources may occur. Based on this, the present invention verifies the validity of the prediction model to ensure that the model is available when predicting data transmission resources, thereby improving the accuracy of the prediction and improving the efficiency and utilization of data transmission resource acquisition of sites and satellites. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 A schematic diagram of the method steps of a data transmission resource prediction method based on historical data according to an embodiment of the invention;

[0054] Figure 2 A logic block diagram of setting a label for the site according to an embodiment of the invention;

[0055] Figure 3 It is a logic block diagram of determining the data transmission resource demand of each site when the satellite moves to each transit time window according to an embodiment of the invention;

[0056] Figure 4 A logic block diagram of an embodiment of the invention for determining whether to output the prediction result of the prediction model. DETAILED DESCRIPTION

[0057] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0058] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0059] To facilitate reading of this plan, some of the terms involved in this plan are explained below:

[0060] Data transmission resources: resources involved in the data transmission process.

[0061] Transit time window: The time range in which a satellite can observe a target under the constraints of the field of view. It is the basis for the satellite to perform observation tasks on ground targets. Any observation task must be performed within the visible window. The satellite transit time window has a wide range of applications in meteorological observation, environmental monitoring, disaster warning, urban planning and other fields. For example, agricultural monitoring requires understanding the satellite transit time of specific farmland to arrange the best shooting window; when a disaster occurs, it is necessary to quickly locate the affected area to plan rescue routes and resources.

[0062] Station: Ground equipment installed on the earth for space communications. It is used to communicate with satellites and spacecraft. It is an important component of a satellite or aerospace system. Its main functions include monitoring satellite operation, receiving remote sensing and telemetry data, and processing, storing and distributing information.

[0063] See also Figures 1 to 4 As shown, Figure 1 This is a schematic diagram of the method steps of a data transmission resource prediction method based on historical data according to an embodiment of the invention. Figure 2 A logic block diagram of setting a label for the site according to an embodiment of the invention, Figure 3 This is a logic block diagram of the data transmission resource requirements of each site when the fixed satellite runs to each transit time window in an embodiment of the invention. Figure 4 The present invention provides a method for predicting data transmission resources based on historical data, which includes:

[0064] Obtain the historical data transmission data of each station and the satellite when the satellite moves to different transit time windows. It can be understood that the historical data transmission data can be recorded once every historical period. The historical period can be set to 24 hours. In implementation, the historical data transmission data is a data set of several historical periods.

[0065] Analyze the data transmission behavior characteristics based on the historical data transmission data, calculate the data transmission behavior fuzzy characterization parameters for the site based on the data transmission behavior characteristics, and set a label for the site, wherein the data transmission behavior characteristics include the task identity characteristics of the data transmission resources and the actual reception amount fluctuation value of the reception amount fluctuation value data transmission resources;

[0066] Based on the tags of each station, the historical data transmission data of the station is analyzed to determine the data transmission resource requirements of each station when the satellite operates to each transit time window, including:

[0067] Determine the dimension of the historical data transmission data to be called based on the fuzzy characterization parameter, call the historical data transmission data of the corresponding dimension to use the pre-trained prediction model to predict the data transmission resource demand in each historical period, and verify the validity of the prediction model for the site based on the data transmission resource reception amount in the corresponding historical period to determine whether to output the prediction result of the prediction model and determine the data transmission resource demand;

[0068] Or, based on the fluctuation range of the amount of data transmission resource received in a historical period, determine the data transmission resource demand of the site;

[0069] The satellite is controlled to allocate corresponding data transmission resources based on the data transmission resource demand.

[0070] Specifically, there is no limitation on the specific type of digital data, for example, it can be text data, image data, video data and control command data. Those skilled in the art need to determine it based on the specific needs of the site, which will not be elaborated here.

[0071] Specifically, in actual use, relay satellites are set up between the stations and satellites to establish communication links between the stations and satellites and provide data relay and measurement and control services. Relay satellites usually operate in geosynchronous orbits and can provide all-weather, real-time communication services for satellites and spacecraft in medium and low orbits, greatly improving the utilization efficiency and emergency response capabilities of various types of satellites.

[0072] Specifically, in implementation, the tag is not static. In implementation, an update cycle is set to calculate the fuzzy characterization parameters of the site's data transmission behavior to update the tag. It can be understood that the duration of the update cycle must not be less than the duration of the satellite's operating cycle. The satellite's operating cycle is 24 hours. The specific update cycle duration can be determined by technical personnel in this field based on actual conditions, which will not be repeated here.

[0073] Specifically, when updating the label, when recalculating the fuzzy characterization parameters of the data transmission behavior, the recorded historical data of the satellite needs to be updated synchronously to ensure the accuracy of calculating the fuzzy characterization parameters and setting the label, which will not be repeated here.

[0074] Specifically, the measurement units of the amount of data transmission resources received and the amount of data transmission resources required are the units used in computer storage and data transmission, such as GB, which will not be repeated here.

[0075] Specifically, the process of analyzing the data transmission behavior characteristics based on the historical data transmission data includes:

[0076] Determine the application content and the acceptance content of digital transmission resources;

[0077] Determining the relevance between the application content and the received content as a task identity feature of the data transmission resource;

[0078] Determine the variance of the received amount of data transmission resources in a number of historical periods, and determine the variance as a received amount fluctuation value;

[0079] The application content is text information sent by the site, the received content includes data transmission resources in several formats transmitted by the satellite to the site, and the correlation is determined based on the text information and the text record of the received content.

[0080] It can be understood that the application content is text information proposed by the site to describe the needs, and the satellite transmits corresponding data transmission resources to the site based on the corresponding needs, which will not be repeated here.

[0081] It is understandable that the received content of each site needs to be recorded in real time, including the transmission format, the transmission quantity, the data nature, and the transmitted data itself, etc. Those skilled in the art can expand it by themselves and will not be elaborated here.

[0082] Taking part of the text information of the application content as an example: "The site number is 001, and a preset number of pictures within a certain period of time in a certain area are required."

[0083] Then, the text record of the received content includes that the transmission format is pictures, the transmission quantity is the preset quantity, the data nature is the image corresponding area and the time period, and the pictures themselves transmitted.

[0084] It is understandable that the method of selecting keywords can be used to determine the relevance. The specific method of selecting keywords is not limited, as long as the keywords can be determined. For example, natural language processing (NLP) tools and keyword extraction algorithms can be used to identify keywords. In implementation, natural language processing (NLP) tools are used to select keywords, and the specific steps are as follows;

[0085] Step S21: Split the continuous text into independent words, delete common words such as "de" (of), "shi" (is), etc., and perform part-of-speech tagging on the independent words.

[0086] Step S22: Determine the occurrence probability of the independent words in the text and the occurrence probability of the independent words in the existing corpus database, and calculate the TF-IDF value.

[0087] Step S22: Determine the keywords according to the TF-IDF value.

[0088] Specifically, the calculation method of TF-IDF is prior art. The larger the TF-IDF value, the more critical the word is in the current document. Those skilled in the art can set a threshold to screen keywords, and the threshold setting can be determined by those skilled in the art according to needs and will not be elaborated here.

[0089] Specifically, the calculation method of the relevance is not limited. For example, cosine similarity and Minkowski distance can be used for calculation, as long as the relevance of the keywords can be obtained. Preferably, in implementation, the method of cosine similarity is used for calculation, and the specific process is as follows;

[0090] Step S201: Represent the two keywords to be compared as vectors and determine the dot product of the vectors.

[0091] Step S202: Determine the modulus of each vector.

[0092] Step S203: Determine that the dot product of the vectors divided by the product of the moduli of each vector is the cosine similarity.

[0093] Specifically, the closer the cosine similarity value is to 1, the more similar the two vectors are; the closer the value is to -1, the more opposite the two vectors are; and the closer the value is to 0, the less related the two vectors are.

[0094] Specifically, the process of calculating the fuzzy characterization parameters of the data transmission behavior of the site includes:

[0095] Determining a ratio of the task identity feature to the benchmark identity feature as an identity impact factor;

[0096] Determine the ratio of the reference reception volume fluctuation value to the reception volume fluctuation value as the reception volume fluctuation value influencing factor;

[0097] The weighted sum of the identity influencing factor and the receiving amount fluctuation value influencing factor is determined as the fuzzy characterization parameter of the data transmission behavior.

[0098] Specifically, in order to ensure the accuracy of the results of the fuzzy characterization parameters, it is necessary to exclude obviously abnormal data when calculating the received volume fluctuation value. For example, when there is a sudden large gap between the received volume of data transmission resources and the required volume of data transmission resources, such data is obviously abnormal and cannot be used as the data required for calculating the received volume fluctuation value.

[0099] The excessive difference between the amount of data transmission resources received and the amount of data transmission resources required is manifested in that the difference between the amount of data transmission resources received and the amount of data transmission resources required is greater than or equal to 0.8 times the required amount.

[0100] Specifically, the benchmark identity feature is calculated in advance, and a number of historical task identity features are obtained in advance, and 1.25 times the average value of the historical task identity features is determined as the benchmark identity feature.

[0101] Specifically, the benchmark reception volume fluctuation value is calculated in advance, and several historical task reception volume fluctuation values ​​are obtained in advance, and 0.85 times the average value of the historical task reception volume fluctuation values ​​is determined as the benchmark reception volume fluctuation value.

[0102] Specifically, the sum of the weight coefficients of the task identity influencing factor of the data transmission resource and the receiving volume fluctuation value influencing factor is 1, the weight coefficient of the task identity influencing factor of the data transmission resource is 0.62, and the weight coefficient of the receiving volume fluctuation value influencing factor is 0.38.

[0103] Specifically, the present invention calculates fuzzy characterization parameters of the data transmission behavior of the site based on the data transmission behavior characteristics, sets labels for the sites, and determines whether the data transmission resource requirements of each site are clear. In actual situations, when predicting data transmission resources, most of the predictions are made through manual experience, or directly through models. However, the first method is not accurate enough in complex situations, which may cause deviations in resource predictions, resulting in insufficient or wasteful resources. Although the second method has higher accuracy, model predictions for each site and each cycle will cause a load on the computer, resulting in a waste of computing power. Based on this, the present invention considers calculating fuzzy characterization parameters of the data transmission behavior of the site, which are used to set labels for the site, characterize the fuzziness of the data transmission behavior of the site, and adopt different prediction methods for sites with different labels, thereby reducing computing power and improving the acquisition efficiency and utilization rate of data transmission resources for sites and satellites.

[0104] Specifically, a label is set for the site, wherein:

[0105] If the fuzzy characterization parameter of the digital transmission behavior is greater than or equal to the reference fuzzy characterization parameter threshold, setting the label to a clear label;

[0106] If the fuzzy characterization parameter of the digital transmission behavior is less than the reference fuzzy characterization parameter threshold, the label is set as a fuzzy label.

[0107] Specifically, the threshold of the benchmark fuzzy representation parameter is selected in the interval [0.51, 0.73].

[0108] Specifically, determine the data transmission resource requirements of each station when the satellite operates to each transit time window, where:

[0109] If the label is a fuzzy label, the dimension of the historical data transmission data to be called is determined based on the fuzzy characterization parameter, the historical data transmission data of the corresponding dimension is called to use a pre-trained prediction model to predict the data transmission resource demand in each historical period, and the validity of the prediction model for the site is verified based on the data transmission resource reception amount in the corresponding historical period to determine whether to output the prediction result of the prediction model and determine the data transmission resource demand;

[0110] If the tag is a clear tag, the data transmission resource demand of the site is determined based on the fluctuation range of the data transmission resource reception amount in the historical period.

[0111] Specifically, the process of determining the dimension of the historical data transmission data to be called based on the fuzzy characterization parameters includes:

[0112] Determine several dimensions of historical data transmission data;

[0113] Determining the interval in which the fuzzy characterization parameter is located;

[0114] Determine that the dimension corresponding to the interval is the dimension of the historical data transmission data to be called;

[0115] Among them, the correspondence between each interval and each dimension is pre-set, and the dimensions include the satellite's orbital offset, the satellite's depression angle, the satellite's transmission power, the site's elevation angle, the site's signal shielding rate, and the relative distance between the site and the satellite.

[0116] It is understandable that different dimensions based on historical data transmission data can be used to predict the demand for data transmission resources under the corresponding transit time window. The more dimensions there are, the more computing power is required for the prediction, but the accuracy is relatively high.

[0117] In implementation, optionally, three intervals may be set:

[0118] For the first interval, the upper limit of the interval is the benchmark fuzzy representation parameter threshold, and the lower limit of the interval is 0.85 times the benchmark fuzzy representation parameter threshold;

[0119] For the second interval, the upper limit of the interval is 0.85 times the baseline fuzzy representation parameter threshold, and the lower limit of the interval is 0.65 times the baseline fuzzy representation parameter threshold;

[0120] For the third interval, the upper limit of the interval is 0.65 times the baseline fuzzy representation parameter threshold, and the lower limit of the interval is 0;

[0121] Wherein, the first interval, the second interval and the third interval are continuous intervals;

[0122] In the implementation, in the first interval, 4 dimensions are selected;

[0123] In the second interval, 5 dimensions are selected;

[0124] In the third interval, 6 dimensions are selected.

[0125] It is understandable that the prediction model can be pre-trained using a neural network architecture, and a model that is adapted to historical data of multiple dimensions can be trained without limiting the training process, for example,

[0126] Collect the satellite orbit offsets corresponding to several stations in different transit time windows, the satellite depression angle, the satellite transmission power, the station elevation angle, the station signal shielding rate, the relative distance between the station and the satellite, and the corresponding data transmission resource requirements as sample data, and divide them into training sets, validation sets, and test sets;

[0127] Those skilled in the art select a suitable neural network architecture to construct a model, wherein the input features of the input layer include the orbital offset of the satellite, the depression angle of the satellite, the transmission power of the satellite, the elevation angle of the site, the signal shielding rate of the site, and the relative distance between the site and the satellite. The hidden layer can use a convolutional neural network (CNN) to extract features, and the output layer outputs the data transmission resource demand;

[0128] Train the constructed model, including setting the loss function and optimizer, performing forward propagation, calculating the loss, performing back propagation, etc.;

[0129] After each training cycle, technicians in this field use the validation set to evaluate the performance of the model, adjust the hyperparameters, repeat the training, and save the model after the model converges.

[0130] Of course, those skilled in the art may also adopt other model training methods as long as they can achieve the corresponding functions, which will not be elaborated here.

[0131] Specifically, the present invention determines the dimensions of the historical data transmission data required to be called based on the fuzzy characterization parameters. In actual situations, when using the model to predict the data transmission resources, due to the large number of transit time windows and the large number of sites, the data required to be predicted is massive, and the fuzziness of the data transmission behavior of each site is inconsistent. The superposition of multiple factors leads to serious computing power consumption. Therefore, the present invention considers using a model prediction method for sites with fuzzy labels, and considers determining the dimensions of the historical data transmission data required to be called based on the fuzzy characterization parameters. When making a prediction, the demand for data transmission resources is predicted according to the determined dimensions, thereby ensuring the accuracy of the prediction, reducing computing power consumption, and improving the acquisition efficiency and utilization rate of data transmission resources for sites and satellites.

[0132] Specifically, the process of verifying the validity of the prediction model for the site based on the amount of data transmission resource received in the corresponding historical period includes:

[0133] Determine the amount of data transmission resources received by the site during a number of historical periods;

[0134] Determine the predicted demand for data transmission resources corresponding to the site in several historical periods;

[0135] Determine the difference between the received amount of data transmission resources and the predicted required amount of resources as a validity parameter;

[0136] If the validity parameter is less than a predetermined parameter threshold, it is determined that the prediction model is valid for the site.

[0137] Specifically, the purpose of setting a predetermined parameter threshold is to limit the difference between the amount of data transmission resources received and the predicted resource demand. In actual situations, the greater the difference, the lower the accuracy of the representation model. In implementation, the predetermined parameter threshold should be less than 0.3 times the amount of data transmission resources received, which will not be repeated here.

[0138] Specifically, the present invention verifies the validity of the prediction model. Verification of the validity of the model is a necessary step in model prediction. In actual use, if the model is used directly without verifying its validity, inaccurate prediction of data transmission resources may occur. Based on this, the present invention verifies the validity of the prediction model to ensure that the model is available when predicting data transmission resources, thereby improving the accuracy of the prediction and improving the efficiency and utilization of data transmission resource acquisition of sites and satellites.

[0139] If the prediction model is valid, the prediction result of the prediction model is output;

[0140] If the prediction model is invalid, the prediction result of the prediction model is not output and the site is marked.

[0141] It is understandable that if the prediction model is invalid for some sites, technicians in this field can adjust the model. Before the model adjustment is completed, in order to determine and apply for the resources of the site in advance, the site can be marked and a manual prediction method can be used.

[0142] Specifically, the process of determining the data transmission resource requirements of a site includes:

[0143] Determine the fluctuation range of the data transmission resource reception amount corresponding to each historical period of the site;

[0144] An average value of the fluctuation range is determined as the data transmission resource requirement.

[0145] It can be understood that the average value of the fluctuation range is calculated based on the data transmission resource reception amount in several historical periods corresponding to the fluctuation range, that is, the average value of the data transmission resource reception amount in several historical periods.

[0146] Specifically, the process of controlling the satellite to allocate corresponding data transmission resources based on the data transmission resource demand includes:

[0147] Determine the data transmission resource requirements of each station under each transit time window;

[0148] The data transmission resource requirements of each site are sent to the satellite to control the satellite to allocate corresponding data transmission resources.

[0149] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. A data transmission resource prediction method based on historical data, characterized in that: include: Obtain the historical data transmission between each station and the satellite when the satellite moves to different transit time windows; Analyze the data transmission behavior characteristics based on the historical data transmission data, calculate the data transmission behavior fuzzy characterization parameters for the site based on the data transmission behavior characteristics, and set a label for the site, wherein: If the fuzzy characterization parameter of the digital transmission behavior is greater than or equal to the reference fuzzy characterization parameter threshold, setting the label to a clear label; If the fuzzy characterization parameter of the digital transmission behavior is less than the reference fuzzy characterization parameter threshold, setting the label as a fuzzy label; The data transmission behavior characteristics include task identity characteristics of data transmission resources and a receiving amount fluctuation value of data transmission resources; Based on the tags of each station, the historical data transmission data of the station is analyzed to determine the data transmission resource requirements of each station when the satellite operates to each transit time window, including: If the label is a fuzzy label, the dimension of the historical data transmission data to be called is determined based on the fuzzy characterization parameter, the historical data transmission data of the corresponding dimension is called to use a pre-trained prediction model to predict the data transmission resource demand in each historical period, and the validity of the prediction model for the site is verified based on the data transmission resource reception amount in the corresponding historical period to determine whether to output the prediction result of the prediction model and determine the data transmission resource demand; If the tag is a clear tag, determining the data transmission resource demand of the site based on the fluctuation range of the data transmission resource reception amount in the historical period; Controlling the satellite to allocate corresponding data transmission resources based on the data transmission resource demand; The process of analyzing the data transmission behavior characteristics based on the historical data transmission data includes: Determine the application content and the acceptance content of digital transmission resources; Determining the relevance between the application content and the received content as a task identity feature of the data transmission resource; Determine the variance of the received amount of the data transmission resource in a number of historical periods, and determine the variance as the received amount fluctuation value; The application content is text information sent by the site, the received content includes data transmission resources in several formats transmitted by the satellite to the site, and the correlation is determined based on the text information and the text description of the received content; The process of calculating the fuzzy characterization parameters of the data transmission behavior of the site includes: Determining a ratio of the task identity feature to the benchmark identity feature as an identity impact factor; Determine the ratio of the reference reception volume fluctuation value to the reception volume fluctuation value as the reception volume fluctuation value influencing factor; The weighted sum of the identity influencing factor and the receiving amount fluctuation value influencing factor is determined as the fuzzy representation parameter of the data transmission behavior.

2. The data transmission resource prediction method based on historical data according to claim 1, characterized in that: The process of determining the dimension of the historical data transmission data to be called based on the fuzzy characterization parameter includes: Determine several dimensions of historical data transmission data; Determining the interval in which the fuzzy characterization parameter is located; Determine that the dimension corresponding to the interval is the dimension of the historical data transmission data to be called; Among them, the correspondence between each interval and each dimension is pre-set, and the dimensions include the satellite's orbital offset, the satellite's depression angle, the satellite's transmission power, the site's elevation angle, the site's signal shielding rate, and the relative distance between the site and the satellite.

3. The data transmission resource prediction method based on historical data according to claim 1, characterized in that: The process of verifying the validity of the prediction model for the site based on the amount of data transmission resource received in the corresponding historical period includes: Determine the amount of data transmission resources received by the site during a number of historical periods; Determine the predicted demand for data transmission resources corresponding to the site in several historical periods; Determine the difference between the received amount of data transmission resources and the predicted required amount of resources as a validity parameter; If the validity parameter is less than a predetermined parameter threshold, it is determined that the prediction model is valid for the site.

4. The data transmission resource prediction method based on historical data according to claim 3 is characterized in that: The step of determining whether to output the prediction result of the prediction model comprises: If the prediction model is valid, the prediction result of the prediction model is output; If the prediction model is invalid, the prediction result of the prediction model is not output and the site is marked.

5. The data transmission resource prediction method based on historical data according to claim 1, characterized in that: The process of determining the data transmission resource requirement of the site includes: Determine the fluctuation range of the data transmission resource reception amount corresponding to each historical period of the site; An average value of the fluctuation range is determined as the data transmission resource requirement.

6. The data transmission resource prediction method based on historical data according to claim 1, characterized in that: The process of controlling the satellite to allocate corresponding data transmission resources based on the data transmission resource demand includes: Determine the data transmission resource requirements of each station under each transit time window; The data transmission resource requirements of each site are sent to the satellite to control the satellite to allocate corresponding data transmission resources.

Images