Buoy terminal communication scheduling method and device, equipment and storage medium

Abstract

The invention provides a buoy terminal communication scheduling method and device, equipment and a storage medium. The method comprises the following steps: acquiring network environment data; obtaining network quality information based on the network environment data and a network quality prediction model; scheduling communication of the buoy terminal according to the network quality information; wherein the network quality prediction model is obtained by training based on a historical network environment data set. According to the method, the change of the network quality along with time can be better predicted, the service data transmission requirement which can be met by the current network is autonomously judged according to the network quality information prediction result, and the service data transmission success rate is improved. And meanwhile, according to the predicted available network quality information, the communication state is selectively quitted, so that the occupation of a satellite communication channel is reduced, and the deployment density of the buoy terminals is maximized.

Description

Technical Field

[0001] The present invention relates to the field of wireless communications, and in particular to a buoy terminal communication scheduling method, device, equipment and storage medium. Background Art

[0002] Buoy terminals are an important means of marine ecological monitoring. Their communication systems in the open ocean mostly use satellite networks such as maritime, Iridium, and Beidou short messages to transmit the business data collected by the terminals. Business data are mostly low-rate data such as sensors and meteorology, and the transmission method is mostly periodic data transmission. In recent years, with the increasing demand for image, video and other types of business data in marine observation, the buoy communication system has an increasingly strong demand for satellite networks.

[0003] Due to its advantages such as low latency and large bandwidth, low-orbit satellite networks are gradually being promoted and applied in the field of ocean buoy communications. However, the topology of low-orbit satellite networks changes rapidly, and the buoy terminals are affected by multiple factors such as satellite-to-ground links, inter-satellite routing, beam switching, as well as the movement of the buoy terminals themselves and the environment. The network quality will change dynamically over time. In addition, after the buoy terminal is connected to the network, it will continue to remain in the network state. When there is no business data to be transmitted back, it will still occupy the satellite channel, resulting in a limited number of terminals that can be served by a single satellite, and other buoy terminals in the same sea area frequently disconnect. Summary of the invention

[0004] The embodiments of the present application provide a buoy terminal communication scheduling method, device, equipment and storage medium, which are mainly used to solve the problem of business data transmission failure caused by changes in network quality of the buoy.

[0005] Other features and advantages of the present application will become apparent from the following detailed description, or may be learned in part by the practice of the present application.

[0006] According to the first aspect of an embodiment of the present application, a buoy terminal communication scheduling method is provided, including: acquiring network environment data; obtaining network quality information based on the network environment data and a network quality prediction model; scheduling buoy terminal communication according to the network quality information; wherein the network quality prediction model is trained based on a historical network environment data set.

[0007] This method can be used to predict the quality of future satellite networks, assist in determining the time of future available communication windows, and enable on-demand calls to satellite networks, thereby ensuring the success rate of business data transmission.

[0008] In one embodiment of the present application, the network environment data includes meteorological data and constellation ephemeris data. The elevation angle and azimuth angle in the constellation ephemeris data and meteorological data directly affect the quality of the satellite-to-ground communication link. The accuracy of future network quality prediction can be improved through meteorological data and constellation ephemeris data.

[0009] In one embodiment of the present application, before obtaining network quality information based on the network environment data and the network quality prediction model, the method further includes: receiving a second model parameter and updating the first model parameter. The first model parameter is the original model parameter, and the second model parameter is the updated model parameter. By updating the model parameter, the prediction accuracy of the network quality prediction model can be guaranteed.

[0010] In one embodiment of the present application, training the network quality prediction model based on the historical network environment data set includes: training the initial model based on the historical network environment data and the corresponding labels. After the initial model is trained with the historical network environment data, the desired network quality prediction model can be obtained, and the network quality prediction can be completed by combining the current network environment data.

[0011] In one embodiment of the present application, the training of the initial model includes: obtaining historical network environment data and corresponding labels to form a data set; filtering the data set based on the correlation between the historical network environment data and the labels; and training the initial model based on the filtered data set. By filtering the data set, weakly related or irrelevant data can be removed, effectively improving the prediction accuracy of the trained model.

[0012] In one embodiment of the present application, the training of the initial model includes: determining the prediction accuracy of the trained model using mean absolute error and mean square error. The mean absolute error and mean square error can be used to accurately measure the prediction effect of the trained network quality prediction model.

[0013] In one embodiment of the present application, the scheduling of buoy terminal communication according to the network quality information includes: determining a network quality threshold based on the data to be transmitted; in response to the network quality information satisfying the network quality threshold, transmitting the data to be transmitted within a time window of the network quality information. By determining whether the predicted network quality information satisfies the network quality threshold of the data to be transmitted, and then determining the time window for transmission, the transmission success rate of the service data can be guaranteed.

[0014] In one embodiment of the present application, transmitting the data to be transmitted within the time window of the network quality information specifically includes: setting a timed task in the time window, and timing triggering entering the communication state to transmit the service data. The timed triggering of entering the communication state can selectively exit the communication state, thereby reducing the occupation of the satellite communication channel.

[0015] In one embodiment of the present application, the second model parameter is generated by the cloud platform after completing the iterative training of the network quality prediction model online. By training the network quality prediction model online through the cloud platform, the update of the model parameters can effectively reduce the computing resources required for terminal local model training and optimization.

[0016] In one embodiment of the present application, the scheduling of the buoy terminal communication according to the network quality information further includes: in response to the network quality information not meeting the network quality threshold, uploading the buoy terminal status information based on a preset time interval. When the network quality data does not meet the network quality threshold for transmitting service data, the service data is not transmitted, but the status information of the buoy terminal itself, such as power, temperature and other information, is uploaded at a preset time interval to facilitate the user to determine whether the buoy terminal is offline.

[0017] According to the second aspect of an embodiment of the present application, a buoy terminal communication scheduling device is provided, including: a data acquisition module for acquiring network environment data; a network quality prediction module for obtaining network quality information based on the network environment data and a network quality prediction model, wherein the network quality prediction model is trained based on a historical network environment data set; and a communication scheduling module for scheduling buoy terminal communications according to the network quality information.

[0018] In one embodiment of the present application, the buoy terminal communication scheduling device further includes a model parameter updating module, which is used to receive the second model parameter sent by the cloud platform before the network quality prediction module predicts, and update the first model parameter. The first model parameter is the original model parameter, and the second model parameter is the updated model parameter.

[0019] In one embodiment of the present application, the buoy terminal communication scheduling device also includes a model training module for training the initial model based on historical network environment data and corresponding labels.

[0020] In one embodiment of the present application, the model training module includes: a data set acquisition module for acquiring historical network environment data and corresponding labels, and forming a data set; a data filtering module for filtering the data set based on the correlation between the historical network environment data and the labels; and a training module for training the initial model based on the filtered data set.

[0021] In one embodiment of the present application, the model training module also includes a detection feedback module for determining the prediction accuracy of the trained model using the mean absolute error and the mean square error.

[0022] In one embodiment of the present application, the communication scheduling module includes: a network quality threshold determination module that determines the network quality threshold of the data to be transmitted in combination with a pre-set strategy; a scheduling module that is used to combine the predicted network quality information and the determined network quality threshold to transmit the data to be transmitted within the time window of the network quality information.

[0023] In one embodiment of the present application, the scheduling module also includes a business data transmission module, which is used to set a scheduled task when the time window is met, and regularly trigger the entry into the communication state to transmit business data.

[0024] In one embodiment of the present application, the scheduling module further includes a status information transmission module, which uploads the buoy terminal status information based on a preset time interval in response to the network quality information not satisfying the network quality threshold.

[0025] According to a third aspect of an embodiment of the present application, an electronic device is provided, including a memory and a processor, wherein the memory stores a computer program corresponding to the buoy terminal communication scheduling method described in the first aspect and capable of being loaded and executed by the processor.

[0026] According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, on which computer program instructions are stored. When the program instructions are executed by a processor, they are used to implement the process corresponding to the buoy terminal communication scheduling method described in the first aspect.

[0027] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] The drawings herein are incorporated into the specification and constitute a part of the specification, illustrate embodiments consistent with the present application, and together with the specification are used to explain the principles of the present application. Obviously, the drawings described below are only some embodiments of the present application, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0029] Figure 1 This is a schematic diagram of a buoy terminal communication scheduling method according to an embodiment of the present application.

[0030] Figure 2 This is a flow chart of the buoy terminal communication scheduling method according to an embodiment of the present application.

[0031] Figure 3 This is a flowchart of the long short-term memory network model training of an embodiment of the present application.

[0032] Figure 4 This is a schematic diagram of the prediction of the long short-term memory network model of the embodiment of the present application, wherein: Figure 4 (a) is a schematic diagram of the predicted network rate. Figure 4 (b) is a schematic diagram of the predicted packet loss rate. Figure 4 (c) is a schematic diagram of the predicted jitter curve.

[0033] Figure 5 This is a schematic diagram of a buoy terminal communication scheduling device according to an embodiment of the present application.

[0034] Figure 6 A schematic diagram of an electronic device according to an embodiment of the present application.

[0035] Figure 7 A schematic diagram of the structure of a computer system suitable for implementing the embodiments of the present application.

[0036] Figure markings: 400-communication scheduling device, 401-data acquisition module, 402-network quality prediction module, 403-communication scheduling module, 500-electronic device, 501-memory, 502-processor; 600-computer system, 601-CPU, 602-ROM, 603-RAM, 604-bus, 605-I / O interface, 606-input part, 607-output part, 608-storage part, 609-communication part, 610-drive, 611-mobile storage medium. DETAILED DESCRIPTION

[0037] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments.

[0038] Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative work are within the scope of protection of this application. In the absence of conflict, the embodiments in this application and the features in the embodiments can be combined with each other arbitrarily. In addition, although the logical order is shown in the flow chart, in some cases, the steps shown or described can be performed in an order different from that here.

[0039] The term "comprising" and any variations thereof in the specification and claims of this application and the above drawings are intended to cover non-exclusive protection. For example, a process, method, system, product or device comprising a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products or devices.

[0040] The term "multiple" in this application may mean at least two, for example, two, three or more, and this embodiment of the application does not limit this. In the technical solution of this application, the collection, dissemination and use of data are in compliance with the requirements of relevant national laws and regulations.

[0041] In view of the fact that the network quality of buoy terminals changes dynamically over time and that they will continue to remain in the network after joining the network, they will still occupy satellite channels when there is no business data to be sent back, causing other buoy terminals in the sea area to frequently disconnect. It is necessary to make short-term or long-term predictions on the network quality, independently decide on the available communication time window, and reduce the occupation of satellite channel resources. On the one hand, this can better carry the business, and on the other hand, it can improve the utilization rate of communication links, realize independent decision-making during working hours, and call network services on demand.

[0042] At present, the network quality prediction on the market is mainly based on ground networks. The predicted network quality of the target geographical area in the future prediction period is predicted based on the load of the configured base stations in the target geographical area, the estimated number of people, and the network demand. There is no dedicated method to evaluate the quality of satellite networks. Usually, the quality of satellite networks is manually judged based on the relevant experience of technical experts to make a rough judgment on the quality of satellite networks transmitting civil aviation passenger aircraft. There is a great reliance on the technical capabilities and industry experience of experts. The judgment results of different experts in different environments may have large deviations, are highly subjective, and have low efficiency.

[0043] Based on this, an embodiment of the present application provides a buoy terminal communication scheduling method, in which a network quality prediction model predicts future network quality information and calculates the available communication time window to complete the transmission based on the data to be transmitted, thereby ensuring that the QoS requirements of business data transmission are met while reducing the occupancy of satellite communication channels.

[0044] The preferred embodiments of the present application are described below in conjunction with the drawings in the specification. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present application, and are not used to limit the present application. In addition, the embodiments and features in the embodiments of the present application may be combined with each other if there is no conflict.

[0045] Figure 1 A flowchart of a buoy terminal communication scheduling method provided in an embodiment of the present application is provided. The method can be applied to an electronic device with data processing or computing capabilities. The method can include the following steps:

[0046] Step S100: Acquire network environment data.

[0047] In the embodiment of the present application, the network environment data mainly includes meteorological data and constellation ephemeris data, which can be used for subsequent network quality prediction. Specifically, the meteorological data includes wind speed, wind direction, rainfall, cloud thickness, temperature, etc., and the constellation ephemeris data includes elevation angle and azimuth angle, etc. In the embodiment of the present application, it is preferred to use meteorological data and constellation ephemeris data to predict the network quality. This is because the elevation angle and azimuth angle in the constellation ephemeris data and the meteorological data directly affect the quality of the satellite-to-ground communication link. The higher the elevation angle and the better the weather conditions, the higher the communication link quality, and the better the corresponding network quality. In other embodiments, the transceiver performance parameters of the satellite payload antenna also affect the quality of the communication link, and can also be used as one of the prediction parameters. At this time, this parameter should also be added in the subsequent model training process.

[0048] In one embodiment of the present application, before performing network quality prediction, a second model parameter can also be received to update the first model parameter, where the first model parameter is the original model parameter and the second model parameter is the updated model parameter. In order to reduce the performance requirements for the buoy terminal and reduce the local computing resource consumption of the buoy, it is preferred to complete the model training in the cloud platform, that is, the second model parameter is generated by the cloud platform after completing the iterative training of the network quality prediction model online. When the buoy terminal is connected to the cloud platform, the second model parameter is issued, and the buoy terminal can be updated and used. For details, please refer to Figure 2 After the buoy terminal is powered on, it is determined whether it can access the low-orbit satellite network. If it can, the buoy terminal can receive the second model parameters from the cloud platform and update the local first model parameters to ensure the prediction performance of the model. If it cannot access the low-orbit satellite network, the historical network quality prediction model can be directly used for prediction.

[0049] Step S200: obtaining network quality information based on the network environment data and the network quality prediction model.

[0050] The network quality prediction algorithm based on neural network is highly targeted, and can train good model effects for users from different backgrounds and fields, and is easy to tune for specific data sets. The quality of the buoy terminal communication network changes periodically, and the neural network has good prediction results for this. As a type of recurrent neural network (RNN), the Long Short-Term Memory (LSTM) network can learn and memorize long-term dependencies, effectively capture the associated information in time series data, and is often used for time series data prediction. Based on this, the embodiment of the present application adopts the LSTM model to realize the prediction of future satellite network quality information, which can better predict the changes in network quality over time.

[0051] Figure 3 The LSTM model pre-training process of the embodiment of the present application is shown, which mainly uses the historical data of the buoy terminal, performs correlation judgment and normalization, and then divides the historical data set into 80% training set and 20% test set, and then inputs it into the constructed LSTM model for training, and finally uses the trained model for network quality prediction. The specific steps are as follows:

[0052] (1) Obtain historical network environment data and corresponding labels to form a data set.

[0053] The embodiment of the present application obtains historical data from the buoy terminal, including historical network environment data and corresponding labels, to form a data set. Among them, the historical network environment data includes: historical meteorological data (wind speed, wind direction, rainfall, cloud thickness, temperature, etc.), historical constellation ephemeris data (elevation angle and azimuth angle, etc.); the label is the historical network quality information mainly including historical network quality indicators, namely bandwidth, latency and packet loss rate. After obtaining the data set, it is divided into 80% training set and 20% test set.

[0054] (2) Filter the dataset based on the correlation between historical network environment data and labels.

[0055] In order to further improve the prediction accuracy, the embodiment of the present application analyzes the relationship between various historical meteorological data (wind speed, wind direction, rainfall, cloud thickness, temperature), historical constellation ephemeris collection and tags (bandwidth, delay, packet loss rate) through the Pearson correlation coefficient (PCC). The Pearson correlation coefficient is calculated as follows:

[0056]

[0057] in, represents the mean value of variable x, Represents the average value of variable y. Correlation coefficient r xy The larger the value, the higher the correlation between the two. xy Variables with a correlation coefficient < 0.2 can remove extremely weakly correlated or irrelevant data.

[0058] Furthermore, since the dimensions and numerical ranges of historical network quality indicators are different from those of historical constellation ephemeris data and meteorological data, directly inputting the original data into the model for prediction will lead to problems such as slow calculation and model gradient explosion. The embodiment of the present application normalizes the filtered data set and limits its numerical range to between 0 and 1, eliminating the impact of data dimensions on model prediction. For the test set and training set, the maximum-minimum normalization process is used to obtain the normalized training set X and test set T. The calculation formula is as follows:

[0059]

[0060] Among them, x norm is the normalized value of the variable, x is the original value of the variable, and x max is the maximum value of the variable, x min is the minimum value of the variable.

[0061] (3) Train the initial model based on the filtered dataset.

[0062] In this embodiment, the long-term dependency between satellite network quality and time is learned through the LSTM model. Therefore, an initial LSTM model needs to be constructed first. Specifically,

[0063] In the LSTM model, the operations of "forget", "input", "update" and "output" are introduced. The following is a detailed description of each operation:

[0064] The "forget" operation of the LSTM model is used to selectively forget previous information in order to better process new input data. The specific operation is as follows:

[0065] f t =σ(w f ·[h t-1 ,x t ]+b f )

[0066] Among them, f t is the output of the forget gate, w f is the weight matrix of the forget gate, h t-1 is the hidden state at the previous moment, x t is the input at the current moment, b f is the bias term, and σ is the sigmoid function.

[0067] The "input" operation of the LSTM model is to determine which inputs at the current moment can be saved as unit states. The specific operations are as follows:

[0068] i t =σ(w i ·[h t-1 ,x t ]+b i )

[0069]

[0070] Among them, i t is the output of the input gate, w i is the weight matrix of the input gate, b i is the bias term, is the candidate cell state, indicating the information updated to the cell state, w h is the weight matrix used to generate candidate cell states, b c is the bias term, and tanh is the hyperbolic tangent function.

[0071] The "update" operation of the LSTM model is to adjust and update model parameters such as weights and biases in real time based on the prediction error. The specific operations are as follows:

[0072]

[0073] The "output" operation of the LSTM model is used to determine which parts are used as output states and hidden states. The specific operations are as follows:

[0074] O t =σ(w 0 [h t-1 ,x t ]+b 0 )

[0075] h t =O t *tanh(C t )

[0076] Among them, O t is the output of the output gate, w 0 is the weight matrix of the output gate, b o is the bias term, h t is the hidden state at the current moment.

[0077] The Adam algorithm (Adaptive Moment Estimation) is an optimization algorithm for training deep learning models. It combines the adaptive learning rate characteristics of the RMSProp algorithm and the accelerated convergence characteristics of the Momentum algorithm. It is efficient, intuitive, and requires relatively few hyperparameter adjustments. Therefore, after completing the construction of the LSTM model, the embodiment of the present application uses the training set to train the LSTM model through the Adam algorithm. The update rules of the Adam algorithm are as follows:

[0078] Compute the first moment estimate of the gradient (momentum):

[0079] m t =β 1 m t-1 +(1-β 1 ) t

[0080] Compute the second moment estimate of the gradient:

[0081]

[0082] For m t and v t Perform bias correction (due to m t and v t are initialized to 0, so they tend to underestimate the true value in early iterations):

[0083]

[0084] Update parameters:

[0085]

[0086] Among them, t is the current number of generations, g t is the current gradient, m t and v t are estimates of the first and second order moments, β 1 and β 2 are two hyperparameters (set to 0.9 and 0.999), α is the learning rate (set to 0.01), and ∈ is a very small positive number (set to 1e-8).

[0087] (4) Use mean absolute error and mean square error to determine the prediction accuracy of the trained model.

[0088] When using the data set to train the LSTM model, the hyperparameters are adjusted through cross-validation. Finally, the performance of the model in satellite network quality information prediction is measured by two evaluation index functions: Mean Absolute Error (MAE) and Mean Square Error (MSE). The calculation formulas for the two indicators are:

[0089]

[0090]

[0091] Among them, i is the prediction point sequence number; N is the number of all samples predicted by the model; y i and Represent the predicted value and true value of the prediction model respectively. The closer MAE and MSE are to 0, the higher the prediction accuracy of the model.

[0092] The following is an example of training the LSTM model with a sampling period of 1h and a total of 200 sets of historical data sets, of which 80% are used as training sample sets and the rest as test sample sets. The input features are 6 features including rainfall, cloud thickness, wind speed, azimuth, and elevation, and the output labels are 3 labels including network rate, packet loss rate, and jitter. The LSTM model is implemented using the PyTorch framework. The LSTM neural network hyperparameter values ​​are set as follows: the number of hidden layers, the number of neurons, the learning rate, the probability of discarding neurons, the time step, the training time window, and the batch size are 2, 32, 0.001, 0.2, 15, 24, and 16 respectively. Figure 4 (a) Figure 4 (b) and Figure 4 (c) shows the schematic diagram of the predicted network rate, packet loss rate and jitter curve. The blue line in the figure represents the prediction result of the training set, the green line represents the prediction result of the test set, and the red line represents the actual value. The RMSE obtained through training is 0.01 and the R2 is 0.84, indicating that the model performs well on the training set.

[0093] Based on the trained LSTM model, in actual application, the collected network environment data can be directly input into the LSTM model to complete the prediction of future satellite network quality information. The obtained network quality information includes the predicted bandwidth B 预测 , prediction delay L 预测 And the predicted packet loss rate P 预测 .

[0094] Step S300: Schedule buoy terminal communications according to the network quality information.

[0095] Since different business data often have different requirements for network quality, after completing the prediction of future satellite network quality information, it is also necessary to determine the network quality threshold of the business data to be transmitted, and then complete the communication scheduling.

[0096] In the embodiments of the present application, the network quality threshold is mainly determined by the type of business data (such as sensor data, large file transmission data, or real-time audio and video data), which represents the minimum network quality requirement for transmitting this type of business data. Generally, the network quality threshold required for each type of business data (such as sensor data, large file transmission data, or real-time audio and video data) can be determined according to a pre-set strategy, including the minimum bandwidth B min , minimum delay L min , minimum packet loss rate P min .

[0097] When performing communication scheduling, in response to the network quality information meeting the network quality threshold, that is, B 预测 >B min And L 预测>L min And P 预测 >P min When the time window is met, a scheduled task can be set to trigger the time to enter the communication state to transmit business data. By managing the time when the buoy terminal enters the communication state through a scheduled task, it can ensure that the QoS requirements of business data transmission are met while reducing the occupancy of the satellite communication channel and improving the success rate of business data transmission.

[0098] Generally, when predicting the quality information of the satellite network in the future, the quality information of the satellite network is predicted for a period of time in the future, such as the quality information of the satellite network for the next day. Similarly, the prediction time period can be refined, for example, the satellite network quality information for the next hour can be predicted, which depends on the parameters input during the LSTM model training. If the training parameters are in hours, the prediction can be made on an hourly basis. Since it only takes a short period of time for the buoy terminal to transmit service data, as long as there is a time window in the predicted network quality information whose quality is greater than the determined network quality threshold requirement, it means that service data can be transmitted in this time window.

[0099] If the predicted network quality information does not meet the determined network quality threshold, the buoy terminal enters the communication state based on the preset time interval and uploads the buoy terminal status information. The communication state does not transmit business data. In practical applications, if the buoy terminal has not been connected to the network for a long time, the user at the rear cannot determine whether the terminal is down due to abnormal reasons. Therefore, in the embodiment of the present application, by setting the buoy terminal to access the network regularly and reporting its own status with extremely low network quality requirements, such as power, temperature and other basic information, it is convenient for users to determine whether the buoy terminal is faulty at the rear. In this way, even if the terminal fails to access the network, it is still likely to see the terminal's access request in the satellite access network and the core network. The user can judge that the buoy terminal is working normally based on this phenomenon.

[0100] In one embodiment, when the set trigger time arrives, an attempt is also made to obtain the second model parameters from the cloud platform, and the next future satellite network quality prediction is performed.

[0101] It should be noted that, in this embodiment, when the buoy terminal does not enter the communication state, its communication module is turned off to enter the low power consumption mode, and the other functional modules (such as timing, future network quality information prediction, etc.) work normally.

[0102] In the communication scheduling process, based on the determined available time window, the buoy terminal only enters the communication state in the scheduled time window and selectively exits the communication state, reducing the occupancy of the satellite communication channel and maximizing the deployment density of the buoy terminal. Through this scheduling method, it can ensure that the QoS requirements of business data transmission are met while reducing the occupancy of the satellite communication channel.

[0103] In the buoy terminal, the performance data of the satellite network is continuously monitored, and the real-time data of the network quality is recorded to identify abnormal situations. These data not only help to adjust the communication strategy in real time, but are also important inputs for LSTM model training, which can continuously optimize the prediction model. Specifically, during the time period when the buoy terminal returns business data, the network quality of this time period is continuously monitored. If the network quality meets the theoretical requirements for business data transmission, but business data cannot be returned, the business data threshold for network quality can be updated and adjusted later, that is, the minimum network quality requirement can be adjusted.

[0104] Figure 5 A buoy terminal communication scheduling device 400 of another embodiment of the present application is shown, which mainly includes: a data acquisition module 401, used to acquire network environment data; a network quality prediction module 402, used to obtain network quality information based on the network environment data and a network quality prediction model, wherein the network quality prediction model is obtained by training based on a historical network environment data set; a communication scheduling module 403, used to schedule buoy terminal communications according to the network quality information.

[0105] Specifically, the data acquisition module is the data entry of the entire device and is mainly responsible for acquiring various necessary network environment data, mainly including meteorological data and constellation ephemeris data. This data provides a basis for subsequent network quality information prediction.

[0106] The buoy terminal communication scheduling device also includes a model parameter updating module, which is used to receive the second model parameters sent by the cloud platform before the network quality prediction module predicts, and update the first model parameters.

[0107] In one embodiment of the present application, the buoy terminal communication scheduling device also includes a model training module, which is used to train the initial model based on historical network environment data and corresponding labels. Specifically, the model training module includes a data set acquisition module, a data filtering module, a training module and a prediction accuracy detection module, wherein the data set acquisition module is used to acquire historical network environment data and corresponding labels, and form a data set; the data filtering module is used to filter the data set based on the correlation between historical network environment data and labels; the training module is used to train the initial model based on the filtered data set; the detection feedback module is used to determine the prediction accuracy of the trained model using the mean absolute error and mean square error. In the network quality prediction module, based on the data acquired by the data acquisition module, combined with the network quality prediction model, the future network quality information is predicted, the network status is understood in advance, and forward-looking guidance is provided for business scheduling.

[0108] Since different types of business data have different requirements for network quality, when performing communication scheduling, it is necessary to first judge the business data to be transmitted, and then determine the minimum network quality threshold corresponding to this type of business data. The communication scheduling module of this embodiment includes a network quality threshold determination module and a scheduling module, wherein the network quality threshold determination module determines the network quality threshold of the data to be transmitted in combination with a pre-set strategy; the scheduling module combines the predicted network quality information and determines the network quality threshold, and transmits the data to be transmitted within the time window of the network quality information that meets the threshold.

[0109] Furthermore, after determining that the network quality information meets the network quality threshold, the scheduling module can set a timer task in the time window that meets the requirements, and trigger the timing to enter the communication state to transmit business data. By managing the time to enter the communication state through the timer task, it can be ensured that while meeting the QoS requirements of business data transmission, the occupancy of the satellite communication channel is reduced, thereby improving the success rate of business data transmission. After determining that the network quality information does not meet the network quality threshold, the scheduling module enters the communication state to upload the buoy terminal status information based on the preset time interval, and the communication state does not transmit business data. Uploading the buoy terminal status information makes it easier for users to determine whether the buoy terminal is faulty / offline at the rear.

[0110] In one embodiment of the present application, the buoy terminal communication scheduling device also includes a network quality detection module, which can continuously monitor the performance data of the satellite network and record the real-time data of the network quality to identify abnormal situations to adjust the communication strategy. At the same time, this part of the data can be uploaded to the cloud platform for iterative optimization of the network quality prediction model.

[0111] Reference Figure 6 As shown, according to an embodiment of the present application, an electronic device 500 includes: a memory 501 and a processor 502, wherein the memory 501 stores a computer program corresponding to the buoy terminal communication scheduling method described in the first aspect and can be loaded and executed by the processor 502.

[0112] Figure 7 A schematic diagram of the structure of a computer system suitable for implementing an electronic device of an embodiment of the present application is shown.

[0113] It should be noted that Figure 7 The computer system 600 of the receiving system shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0114] like Figure 7As shown, the computer system 600 includes a central processing unit (CPU) 601, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 602 or the program loaded from the storage part 608 to the random access memory (RAM) 603, such as executing the method described in the above embodiment. In the RAM 603, various programs and data required for system operation are also stored. The CPU 601, ROM 602 and RAM 603 are connected to each other through a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0115] The following components are connected to the I / O interface 605: an input section 606 including a keyboard, a mouse, etc.; an output section 607 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 608 including a hard disk, etc.; and a communication section 609 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to the I / O interface 605 as needed. A removable storage medium 611, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 610 as needed so that a computer program read therefrom is installed into the storage section 608 as needed.

[0116] In particular, according to an embodiment of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 609, and / or installed from the mobile storage medium 611. When the computer program is executed by the central processing unit (CPU) 601, various functions defined in the system of the present application are executed.

[0117] It should be noted that the computer-readable medium shown in the embodiment of the present application may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, - but not limited to - an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by an instruction execution system, device or device or used in combination with it. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, wherein a computer-readable program code is carried. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, which may send, propagate, or transmit programs for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0118] The flowchart and block diagram in the accompanying drawings illustrate the possible architecture, functions and operations of the system, method and computer program product according to various embodiments of the present application. Wherein, each box in the flowchart or block diagram can represent a module, a program segment, or a part of the code, and the above-mentioned module, program segment, or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0119] The units involved in the embodiments described in this application may be implemented by software or hardware, and the units described may also be set in a processor. The names of these units do not, in some cases, constitute limitations on the units themselves.

[0120] As another aspect, the present application also provides a computer program product or a computer program, which includes a computer instruction stored in a computer-readable storage medium. A processor of a computer device reads the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the computer device executes the buoy terminal communication scheduling method described in the above embodiment.

[0121] As another aspect, the present application also provides a computer-readable medium, which may be included in the electronic device described in the above embodiment; or may exist independently without being assembled into the electronic device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by an electronic device, the electronic device implements the buoy terminal communication scheduling method described in the above embodiment.

[0122] It should be noted that, although several modules or units of the equipment for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiments of the present application, the features and functions of two or more modules or units described above can be embodied in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into being embodied by multiple modules or units.

[0123] Through the description of the above implementation methods, it is easy for those skilled in the art to understand that the example implementation methods described here can be implemented by software or by combining software with necessary hardware. Therefore, the technical solution according to the implementation methods of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which can be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the implementation methods of the present application.

[0124] For those skilled in the art, the specific meanings of the above terms in the present invention can be understood in specific situations; the drawings in the embodiments are used to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings herein can be arranged and designed in various different configurations.

[0125] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A buoy terminal communication scheduling method, characterized in that: include: Get network environment data; Obtaining network quality information based on the network environment data and the network quality prediction model; Scheduling buoy terminal communications according to the network quality information; in The network quality prediction model is trained based on a historical network environment data set.

2. The buoy terminal communication scheduling method according to claim 1, characterized in that: The network environment data includes meteorological data and constellation ephemeris data.

3. The buoy terminal communication scheduling method according to claim 1 or 2, characterized in that: Before obtaining the network quality information based on the network environment data and the network quality prediction model, the method further includes: The second model parameter is received, and the first model parameter is updated.

4. The buoy terminal communication scheduling method according to claim 1, characterized in that: The network quality prediction model obtained by training based on the historical network environment data set includes: The initial model is trained based on historical network environment data and corresponding labels.

5. The buoy terminal communication scheduling method according to claim 4, characterized in that: The training of the initial model includes: Obtain historical network environment data and corresponding labels to form a data set; Filtering datasets based on historical network environment data and label relevance; The initial model is trained based on the filtered dataset.

6. The buoy terminal communication scheduling method according to claim 1, characterized in that: The training of the initial model also includes: using the mean absolute error and the mean square error to determine the prediction accuracy of the trained model.

7. The buoy terminal communication scheduling method according to claim 1, characterized in that: The scheduling of buoy terminal communication according to the network quality information includes: Determining a network quality threshold based on the data to be transmitted; In response to the network quality information satisfying the network quality threshold, the to-be-transmitted data is transmitted within a time window of the network quality information.

8. The terminal communication scheduling method according to claim 7, characterized in that: Transmitting the data to be transmitted within the time window of the network quality information specifically includes: A scheduled task is set in the time window to trigger the communication state to transmit business data.

9. The buoy terminal communication scheduling method according to claim 7 or 8, characterized in that: The scheduling of buoy terminal communication according to the network quality information further includes: In response to the network quality information not satisfying the network quality threshold, uploading the buoy terminal status information based on a preset time interval.

10. The buoy terminal communication scheduling method according to claim 3, characterized in that: The second model parameters are generated after the cloud platform completes the iterative training of the network quality prediction model online.

11. A buoy terminal communication dispatching device, characterized in that: include: A data acquisition module, used to acquire network environment data; A network quality prediction module, used to obtain network quality information based on the network environment data and a network quality prediction model, wherein the network quality prediction model is obtained by training based on a historical network environment data set; The communication scheduling module is used to schedule the buoy terminal communication according to the network quality information.

12. An electronic device, characterized in that: It comprises a memory and a processor, wherein the memory stores a computer program corresponding to the buoy terminal communication scheduling method as claimed in any one of claims 1 to 10, which can be loaded by the processor and executed.

13. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the program instructions are executed by the processor, they are used to implement the process corresponding to the buoy terminal communication scheduling method described in any one of claims 1 to 10.

Images