Satellite communication channel optimization method and device, terminal equipment and storage medium

Abstract

The invention discloses an optimization method and device of a satellite communication channel, terminal equipment and a storage medium, and belongs to the technical field of satellite communication, and the method comprises the steps: through a channel state prediction model, according to a first channel state parameter of the satellite communication channel at a current moment and second channel state parameters of a plurality of historical moments, carrying out the optimization of the satellite communication channel; and identifying the change trend of the satellite communication channel, and predicting the predicted channel state parameter of the satellite communication channel at the next moment. According to the first channel state parameter and the predicted channel state parameter, when it is judged that a fluctuation risk exists in a satellite communication channel, a target modulation coding scheme is generated; and finally, switching the modulation coding mechanism of the satellite communication system according to the target modulation coding scheme, thereby realizing real-time optimization of the satellite communication channel, and solving the problems that the modulation coding mechanism of the satellite communication system is lagged in adjustment and the satellite communication quality is difficult to guarantee.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of satellite communication, and in particular to a satellite communication channel optimization method and device, a terminal device and a storage medium. BACKGROUND

[0002] Current satellite communication systems usually rely on adaptive modulation and coding (AMC) technology to optimize spectrum utilization and interference resistance. AMC technology has been widely used in satellite communication by adjusting the modulation mode and coding scheme according to the signal-to-noise ratio (SNR) and bit error rate (BER) of the channel. In the traditional AMC method, the satellite terminal reports the signal-to-noise ratio to the ground station through feedback signals, and the ground station selects the most appropriate modulation and coding scheme according to these feedback signals, thereby optimizing the channel environment and improving communication quality.

[0003] However, due to the large time delay of the satellite communication link, the feedback channel environment is different from the actual channel environment, so this feedback delay defect causes the modulation and coding mechanism of the satellite communication system to have adjustment lag, making it difficult to guarantee the quality of satellite communication. SUMMARY

[0004] The present application provides a satellite communication channel optimization method, device, terminal device and storage medium, which can solve the problem of adjustment lag in the modulation and coding mechanism of the satellite communication system in the prior art, which makes it difficult to guarantee the quality of satellite communication.

[0005] An embodiment of the present application provides a satellite communication channel optimization method, comprising:

[0006] obtaining a first channel state parameter of a satellite communication channel at a current time and a second channel state parameter at a plurality of historical times;

[0007] inputting the first channel state parameter and the second channel state parameter into a preset channel state prediction model, so that the channel state prediction model identifies the trend of the satellite communication channel according to the first channel state parameter and the second channel state parameter, and predicts a predicted channel state parameter of the satellite communication channel at a next time;

[0008] when it is determined that the satellite communication channel has a fluctuation risk according to the first channel state parameter and the predicted channel state parameter, generating a target modulation and coding scheme according to the predicted channel state parameter;

[0009] switching the modulation and coding mechanism of the satellite communication system according to the target modulation and coding scheme;

[0010] The construction of the channel state prediction model comprises:

[0011] acquire a plurality of historical channel state parameters of the satellite communication channel;

[0012] acquire a plurality of continuous historical channel state parameters in a preset time length, construct a channel state time sequence, and take a next historical channel state parameter of a last historical channel state parameter in the channel state time sequence as a predicted value of the channel state time sequence;

[0013] construct an initial prediction model based on a long short-term memory network;

[0014] train the initial prediction model by using the channel state time sequence and the corresponding predicted value, evaluate the initial prediction model by using a preset loss function in each round of training, generate a corresponding loss function value, optimize model parameters of the initial prediction model by using a back propagation algorithm when the loss function value does not converge, and take the last optimized initial prediction model as the channel state prediction model when the loss function value converges.

[0015] Further, the first channel state parameter includes a first signal-to-noise ratio at a current time, and the predicted channel state parameter includes a second signal-to-noise ratio at a next time.

[0016] determining that the satellite communication channel has a fluctuation risk according to the first channel state parameter and the predicted channel state parameter includes:

[0017] calculating a signal-to-noise ratio fluctuation value according to the first signal-to-noise ratio and the second signal-to-noise ratio.

[0018] determining that the satellite communication channel has a fluctuation risk when the signal-to-noise ratio fluctuation value is greater than a preset threshold.

[0019] Further, generating a target modulation and coding scheme according to the predicted channel state parameter includes:

[0020] inputting the predicted channel state parameter into a preset channel optimization model, so that the channel optimization model outputs a target modulation and coding scheme according to the predicted channel state parameter, with the goal of optimizing the channel state of the satellite communication channel.

[0021] Further, the construction of the channel optimization model includes:

[0022] constructing a satellite channel simulation model, an agent and a reward function; wherein the reward function is used to guide the agent to generate a modulation and coding scheme for optimizing the channel state.

[0023] randomly extract a plurality of historical channel state parameters as initial state space values;

[0024] The satellite channel simulation model is taken as a simulation environment for interaction of an agent, and the initial state space is sequentially input into the satellite channel simulation model, so that the agent interacts with the satellite channel simulation model to randomly generate a modulation and coding scheme as an initial action space value;

[0025] The satellite channel simulation model is used to generate a response state space value according to the initial state space value and the initial action space value, and the reward function is used to calculate a reward function value of the initial action space value;

[0026] The initial state space value, the initial action space value, the reward function value and the response state space value are taken as training samples;

[0027] The agent is iteratively trained by using a plurality of training samples, in each round of training process, the agent is evaluated by using a Q learning formula to generate a Q value of the agent, and when the Q value converges, the agent is taken as the channel optimization model.

[0028] An embodiment of the present application further provides a satellite communication channel optimization device, which comprises:

[0029] A channel state acquisition module is configured to acquire a first channel state parameter of a satellite communication channel at a current time and a second channel state parameter at a plurality of historical times;

[0030] A channel state prediction module is configured to input the first channel state parameter and the second channel state parameter into a preset channel state prediction model, so that the channel state prediction model identifies a change trend of the satellite communication channel according to the first channel state parameter and the second channel state parameter, and predicts a predicted channel state parameter of the satellite communication channel at a next time;

[0031] A channel state optimization module is configured to generate a target modulation and coding scheme according to the predicted channel state parameter when it is determined that the satellite communication channel has a fluctuation risk according to the first channel state parameter and the predicted channel state parameter;

[0032] A channel control module is configured to switch a modulation and coding mechanism of a satellite communication system according to the target modulation and coding scheme;

[0033] The construction of the channel state prediction model comprises:

[0034] A plurality of historical channel state parameters of the satellite communication channel are acquired;

[0035] acquire a plurality of historical channel state parameters in a preset time length, construct a channel state time sequence, and take a next historical channel state parameter of a last historical channel state parameter in the channel state time sequence as a predicted value of the channel state time sequence;

[0036] construct an initial prediction model based on a long short-term memory network;

[0037] train the initial prediction model by using the channel state time sequence and the corresponding predicted value, evaluate the initial prediction model by using a preset loss function in each training round, generate a corresponding loss function value, optimize model parameters of the initial prediction model by using a back propagation algorithm when the loss function value does not converge, and take the last optimized initial prediction model as the channel state prediction model when the loss function value converges.

[0038] Further, the first channel state parameter includes a first signal-to-noise ratio at a current time point, and the predicted channel state parameter includes a second signal-to-noise ratio at a next time point.

[0039] The channel state optimization module determines that the satellite communication channel has a fluctuation risk according to the first channel state parameter and the predicted channel state parameter, including:

[0040] According to the first signal-to-noise ratio and the second signal-to-noise ratio, a signal-to-noise ratio fluctuation value is calculated.

[0041] When the signal-to-noise ratio fluctuation value is greater than a preset threshold, it is determined that the satellite communication channel has a fluctuation risk.

[0042] Further, the channel state optimization module generates a target modulation and coding scheme according to the predicted channel state parameter, including:

[0043] The predicted channel state parameter is input into a preset channel optimization model, so that the channel optimization model outputs a target modulation and coding scheme according to the predicted channel state parameter, with the goal of optimizing the channel state of the satellite communication channel.

[0044] Further, the construction of the channel optimization model includes:

[0045] A satellite channel simulation model, an agent and a reward function are constructed, wherein the reward function is used to guide the agent to generate a modulation and coding scheme for optimizing the channel state.

[0046] A plurality of historical channel state parameters are randomly extracted as initial state space values.

[0047] The satellite channel simulation model is taken as a simulation environment for interaction of the agent, and the initial state space is input into the satellite channel simulation model in sequence, so that the agent interacts with the satellite channel simulation model to randomly generate a modulation and coding scheme as an initial action space value;

[0048] The satellite channel simulation model is used to generate a response state space value according to the initial state space value and the initial action space value, and the reward function value of the initial action space value is calculated by using the reward function;

[0049] The initial state space value, the initial action space value, the reward function value and the response state space value are taken as training samples;

[0050] The agent is iteratively trained by using a plurality of training samples, in each training process, the agent is evaluated by using a Q learning formula to generate a Q value of the agent, and when the Q value converges, the agent is taken as the channel optimization model.

[0051] The application further provides a terminal device, comprising:

[0052] One or more processors;

[0053] A memory coupled to the processor, configured to store one or more programs;

[0054] When the one or more programs are executed by the one or more processors, the one or more processors implement the optimization method of the satellite communication channel as described in the above application embodiments.

[0055] The application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the optimization method of the satellite communication channel as described in the above application embodiments.

[0056] By implementing the application, the following beneficial effects are achieved:

[0057] The application provides a satellite communication channel optimization method and device, a terminal equipment and a storage medium. BRIEF DESCRIPTION OF DRAWINGS

[0058] In order to more clearly illustrate the technical solutions of the present application, the drawings needed in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and all other drawings obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0059] Figure 1 is a flow diagram of a satellite communication channel optimization method provided by an embodiment of the present application;

[0060] Figure 2 is a structural diagram of a satellite communication channel optimization device provided by an embodiment of the present application;

[0061] Figure 3 is a structural diagram of a terminal equipment provided by an embodiment of the present application. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical solutions and advantages of the present application clearer, the technical solutions in the present application will be described clearly and completely in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0063] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of this application; the use of the terms "including," "comprising," or "having" and variations thereof herein is intended to be broad and encompass the terms "consisting of" and "consisting essentially of" and variations thereof. Unless otherwise noted, the terms "including" and / or "comprising" when used herein shall mean "including, but not limited to."

[0064] In the description of the embodiments of the present application, the technical terms "first", "second", etc. are only used to distinguish different objects, and cannot be understood as indicating or implying relative importance or implicitly indicating the number, specific order or primary and secondary relationship of the indicated technical features. In the description of the embodiments of the present application, the meaning of "multiple" is more than two, unless otherwise explicitly and specifically limited.

[0065] Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The occurrence of the phrase in various places in the specification is not necessarily all referring to the same embodiment, nor is it necessarily referring to a separate or alternative embodiment to the other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with each other.

[0066] In the description of the embodiments of the present application, the term "and / or" is only a description of the association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B, which can represent the three cases of A alone, A and B together, and B alone. In addition, the character " / " herein generally represents an "or" relationship between the front and rear associated objects.

[0067] In the description of the embodiments of the present application, the term "multiple" refers to more than two (including two), and similarly, "multiple groups" refers to more than two groups (including two groups), and "multiple pieces" refers to more than two pieces (including two pieces).

[0068] In the description of the embodiments of the present application, unless otherwise explicitly specified and limited, the technical terms "mounting", "connecting", "connecting", "fixing" and the like should be understood in a broad sense, for example, it can be fixedly connected, or it can be detachably connected, or it can be integrated; it can be mechanical connection, or it can be electrical connection; it can be directly connected, or it can be indirectly connected through an intermediate medium; it can be the internal communication of two elements or the interaction relationship between two elements. For those skilled in the art, the specific meaning of the above terms in the embodiments of the present application can be understood according to the specific circumstances.

[0069] Reference Figure 1 To solve the problems in the prior art, an embodiment of the present application provides an optimization method of a satellite communication channel, comprising:

[0070] S1, acquiring a first channel state parameter of a satellite communication channel at a current time and a second channel state parameter at a plurality of historical times;

[0071] In a preferred embodiment of the present application, the first channel state parameter and the second channel state parameter include: signal-to-noise ratio, bit error rate, time delay, network load and interference condition at the corresponding time.

[0072] S2, inputting the first channel state parameter and the second channel state parameter into a preset channel state prediction model, so that the channel state prediction model identifies a change trend of the satellite communication channel according to the first channel state parameter and the second channel state parameter, and predicts a predicted channel state parameter of the satellite communication channel at a next time;

[0073] The construction of the channel state prediction model comprises:

[0074] S21, acquiring a plurality of historical channel state parameters of the satellite communication channel;

[0075] S22, acquiring a plurality of continuous historical channel state parameters with a preset time length, constructing a channel state time sequence, and taking a next historical channel state parameter of a last historical channel state parameter in the channel state time sequence as a predicted value of the channel state time sequence;

[0076] S23, constructing an initial prediction model based on a long short-term memory network;

[0077] S24, training the initial prediction model by using the channel state time sequence and the corresponding predicted value, and in each round of training, evaluating the initial prediction model by a preset loss function, generating a corresponding loss function value, optimizing model parameters of the initial prediction model by a back propagation algorithm when the loss function value does not converge, and taking the last optimized initial prediction model as the channel state prediction model when the loss function value converges.

[0078] In a preferred embodiment of the present application, since LSTM (Long Short-Term Memory) has the advantage of being good at processing time series data, LSTM is used to build a channel state prediction model to capture long-term and short-term trends of channel state fluctuations. First, historical data of signal-to-noise ratio, bit error rate, delay, load and interference conditions are collected, which are input into the model as a training set. Then, through data cleaning and feature engineering, such as removing noise, filling missing values, and standardizing data, the data is ensured to be suitable for model training. Next, LSTM (Long Short-Term Memory) is used to process time series data, which can capture long-term dependencies and accurately predict future changes in signals. During training, data is input into the network in batches by time steps, using mean square error (MSE) as the loss function, and the model parameters are optimized through the backpropagation algorithm. During training, the cross-validation method is used to prevent overfitting, and the hyperparameters (such as learning rate, batch size, number of training rounds, etc.) are adjusted to improve model performance. After training, the test set is used to evaluate the generalization ability of the model, and the communication strategy of the system is adjusted according to the prediction results of the model, such as modulation mode and coding scheme, to ultimately achieve adaptive adjustment and optimization of the dynamic channel environment.

[0079] Preferably, the first channel state parameter comprises a first signal-to-noise ratio at the current time, and the predicted channel state parameter comprises a second signal-to-noise ratio at the next time.

[0080] According to the first channel state parameter and the predicted channel state parameter, it is determined that the satellite communication channel has a fluctuation risk, comprising:

[0081] S21, according to the first signal-to-noise ratio and the second signal-to-noise ratio, a signal-to-noise ratio fluctuation value is calculated.

[0082] S22, when the signal-to-noise ratio fluctuation value is greater than a preset threshold, it is determined that the satellite communication channel has a fluctuation risk.

[0083] In a preferred embodiment of the present application, by subtracting the first signal-to-noise ratio from the second signal-to-noise ratio, a signal-to-noise ratio fluctuation value is obtained, and when the signal-to-noise ratio fluctuation value ΔSNR_th>3dB is detected, it is determined that the satellite communication channel has a fluctuation risk.

[0084] S3, when it is determined according to the first channel state parameter and the predicted channel state parameter that the satellite communication channel has a fluctuation risk, a target modulation and coding scheme is generated according to the predicted channel state parameter.

[0085] Preferably, the target modulation and coding scheme is generated according to the predicted channel state parameter, comprising:

[0086] S31, input the predicted channel state parameter into a preset channel optimization model, so that the channel optimization model, aiming at optimizing the channel state of the satellite communication channel, outputs a target modulation and coding scheme according to the predicted channel state parameter.

[0087] Preferably, the construction of the channel optimization model comprises:

[0088] S311, a satellite channel simulation model, an agent and a reward function are constructed, wherein the reward function is used to guide the agent to generate a modulation and coding scheme for optimizing the channel state;

[0089] S312, historical channel state parameters are randomly extracted multiple times as initial state space values;

[0090] S313, the satellite channel simulation model is used as a simulation environment for interaction of the agent, and the initial state space is sequentially input into the satellite channel simulation model, so that the agent interacts with the satellite channel simulation model to randomly generate a modulation and coding scheme as an initial action space value;

[0091] S314, the satellite channel simulation model is used to generate a response state space value according to the initial state space value and the initial action space value, and the reward function value of the initial action space value is calculated by using the reward function;

[0092] S315, the initial state space value, the initial action space value, the reward function value and the response state space value are used as training samples;

[0093] S316, the agent is iteratively trained by using a plurality of training samples, in each training process, the agent is evaluated by using a Q learning formula to generate a Q value of the agent, and when the Q value converges, the agent is used as the channel optimization model.

[0094] In a preferred embodiment of the application, the channel optimization model is constructed by a deep reinforcement learning model (such as Q-learning, deep Q network DQN, etc.), and the agent of the deep reinforcement learning model can autonomously learn and optimize the communication strategy in a given environment. The goal of reinforcement learning is to maximize the cumulative reward of the agent through interaction with the environment, that is, to maximize the overall performance of the system by selecting the best modulation and demodulation strategy and coding and decoding scheme.

[0095] The core of reinforcement learning is to learn the optimal strategy through continuous trial and error process, which can be described by the following Q learning update formula in mathematics:

[0096]

[0097] wherein S t is the t-th pre-constructed state space. A t is the action space composed of actions randomly selected by the agent according to the t-th state space St, E represents the expected value (average of long-term rewards), z is the future time step, R throughput is the data throughput, P decoding is the decoding power consumption, γ is a discount factor with a value range of [0.9, 0.99], and η is a weighting coefficient of the DRL dynamic optimization of throughput and power consumption.

[0098] Specifically, in the present embodiment, the state space is defined as:

[0099]

[0100] The action space is defined as:

[0101] A = M x C;

[0102] wherein,

[0103]

[0104] wherein SNR is the predicted signal-to-noise ratio, BER is the predicted bit error rate, delay is the predicted delay, load is the predicted load, and interference is the interference condition; M is the modulation order, including QPSK, 16QAM, and 64QAM; C is the encoding mode, including LDPC and Turbo; n represents the total length of the encoding (i.e., the number of bits per code word), and k is the number of information bits (i.e., the number of data bits transmitted). n and k determine the performance of the encoding scheme and affect the bit error rate and bandwidth utilization of the system.

[0105] Further, the reward function is a crucial part of reinforcement learning, which is used to evaluate the goodness of the actions taken by the agent. In a satellite communication system, the reward function can be designed by considering multiple indicators such as communication quality, spectrum utilization, and delay.

[0106] R t = λ1SNR - λ2*bit error rate + λ3throughput;

[0107] wherein R t is the reward function, λ1, λ2, and λ3 are weighting coefficients, SNR is the signal-to-noise ratio, bit error rate is the bit error rate, and throughput is the data throughput. The goal is to guide the modulation and coding scheme generated by the agent to reduce the bit error rate of the channel, improve the signal-to-noise ratio and throughput, and optimize the channel state.

[0108] By training the agent model, it can adaptively adjust the strategy under different channel conditions. During the training process, the agent not only needs to deal with regular signal attenuation and interference, but also needs to handle complex delay changes and spectrum congestion and other dynamic environmental changes. Therefore, large-scale training and simulation of the agent are needed to ensure that it can make quick and accurate decisions in different environments.

[0109] The training process usually uses a simulated environment to conduct large-scale experiments to optimize the decision-making ability of the agent. During the training process, the agent adjusts its strategy through multiple interactions with the environment until it reaches the optimal long-term return, that is, maximizes the stability and efficiency of the system.

[0110] Specifically, in the initial exploration stage, the agent selects a random action according to the tth pre-constructed state space St t Randomly selected actions generate action space A t . Randomly selected actions (such as randomly switching modulation methods, coding schemes) collect "state-action-reward-new state" samples (St t , At t , Rt t , St+1 t+1 ) through interaction with the environment, and accumulate initial experience.

[0111] In the policy optimization stage, based on the collected samples, the Q value is updated through the Q learning formula, for example, when the agent selects action At t (such as 16QAM+LDPC) in state St t , obtains reward Rt t and enters new state St+1 t+1 , then Q(St t , At t ) is updated, and the selection probability of "high-reward action" is strengthened. When the Q value fluctuation is less than the threshold (such as Q value change <1% in continuous 1000 iterations), it is considered that the offline training converges.

[0112] For example, when there is a fluctuation risk in the satellite communication channel, the DRL agent calculates the optimal coding scheme through the value function, calculates the optimal coding method and modulation order through the Q value, and selects them as the target modulation and coding scheme. At the same time, an improved quasi-cyclic LDPC check matrix design is adopted:

[0113]

[0114] I pi,j is a p-order unit matrix cyclically shifted p i,j times, and p i,j is a special parameter of LDPC coding. When LDPC is selected, the DRL dynamically generates p i,jCooperate with M to optimize system performance through joint reward function.

[0115] Quasi-cyclic LDPC check matrix H realizes low complexity coding and decoding through cyclic shift of unit matrix. It is used to generate adaptive error correction code at the transmitting end, and is decoded at the receiving end through improved Min-Sum decoding algorithm, especially when the signal-to-noise ratio is low, the p i,j parameter is adjusted to enhance the error correction capability, and ensure good performance under different channel conditions.

[0116] with adaptive min-sum decoding algorithm:

[0117]

[0118] λ is a trainable parameter, and β is an adaptive scaling factor, and β changes with SNR, so that the number of decoding iterations is reduced. The receiving end adopts the improved Min-Sum algorithm with adaptive β to decode. μ c→v is the message passing value from check node c to variable node v, and μ v′→c is the message passing value from variable node v' to check node c. Wherein v' is a variable node connected with the check node c, and v'≠v. N(c)\v defines the local calculation range, and finally realizes efficient and low-power decoding under complex channel. On the premise of ensuring decoding accuracy, the calculation complexity and power consumption are significantly reduced, and it is especially suitable for high-dynamic channel environment of satellite communication.

[0119] S4, according to the target modulation coding scheme, switching the modulation and coding mechanism of the satellite communication system;

[0120] In a preferred embodiment of the present application, the embodiment adopts a hierarchical response mechanism, and the physical layer of the satellite communication system completes the switching of the target modulation coding scheme within 20ms, and the MAC layer dynamically adjusts the frame structure according to the urgency.

[0121] In summary, the present application provides an optimization method for a satellite communication channel, which predicts the change trend of the satellite communication channel according to the first channel state parameter of the satellite communication channel at the current time and the second channel state parameter at a plurality of historical times through a channel state prediction model, predicts the predicted channel state parameter of the satellite communication channel at the next time, and overcomes the defect that the satellite communication link sometimes has a large time delay. Further, when it is judged that the satellite communication channel has a fluctuation risk according to the first channel state parameter and the predicted channel state parameter, a target modulation coding scheme is generated; finally, according to the target modulation coding scheme, the modulation and coding mechanism of the satellite communication system is switched, the real-time optimization of the satellite communication channel is realized, the problem of adjustment lag of the modulation and coding mechanism of the satellite communication system is solved, and the satellite communication quality is ensured.

[0122] ReferenceFigure 2 The application provides a satellite communication channel optimization device, which is an embodiment of the application and comprises:

[0123] A channel state acquisition module is configured to acquire a first channel state parameter of a satellite communication channel at a current time and a second channel state parameter at a plurality of historical times.

[0124] A channel state prediction module is configured to input the first channel state parameter and the second channel state parameter into a preset channel state prediction model, so that the channel state prediction model identifies a change trend of the satellite communication channel according to the first channel state parameter and the second channel state parameter and predicts a predicted channel state parameter of the satellite communication channel at a next time.

[0125] A channel state optimization module is configured to generate a target modulation and coding scheme according to the predicted channel state parameter when it is determined that the satellite communication channel has a fluctuation risk according to the first channel state parameter and the predicted channel state parameter.

[0126] A channel control module is configured to switch a modulation and coding mechanism of a satellite communication system according to the target modulation and coding scheme.

[0127] The construction of the channel state prediction model comprises the following steps.

[0128] A plurality of historical channel state parameters of the satellite communication channel are acquired.

[0129] A plurality of continuous historical channel state parameters are acquired at a preset time length, a channel state time sequence is constructed, and a next historical channel state parameter of a last historical channel state parameter in the channel state time sequence is taken as a predicted value of the channel state time sequence.

[0130] An initial prediction model based on a long short-term memory network is constructed.

[0131] The initial prediction model is trained by using the channel state time sequence and the corresponding predicted value, and in each round of training, the initial prediction model is evaluated by using a preset loss function to generate a corresponding loss function value. When the loss function value does not converge, the model parameters of the initial prediction model are optimized by using a back propagation algorithm, and when the loss function value converges, the last optimized initial prediction model is taken as the channel state prediction model.

[0132] Further, the first channel state parameter comprises a first signal-to-noise ratio at the current time, and the predicted channel state parameter comprises a second signal-to-noise ratio at the next time.

[0133] The channel state optimization module determines that the satellite communication channel has a fluctuation risk according to the first channel state parameter and the predicted channel state parameter, including:

[0134] According to the first signal-to-noise ratio and the second signal-to-noise ratio, a signal-to-noise ratio fluctuation value is calculated.

[0135] When the signal-to-noise ratio fluctuation value is greater than a preset threshold, it is determined that the satellite communication channel has a fluctuation risk.

[0136] Further, the channel state optimization module generates a target modulation and coding scheme according to the predicted channel state parameter, including:

[0137] The predicted channel state parameter is input into a preset channel optimization model, so that the channel optimization model outputs a target modulation and coding scheme according to the predicted channel state parameter, with the goal of optimizing the channel state of the satellite communication channel.

[0138] Further, the construction of the channel optimization model includes:

[0139] A satellite channel simulation model, an agent, and a reward function are constructed, wherein the reward function is used to guide the agent to generate a modulation and coding scheme that optimizes the channel state;

[0140] Randomly extract historical channel state parameters multiple times as initial state space values;

[0141] The satellite channel simulation model is used as a simulation environment for agent interaction, and the initial state space is sequentially input into the satellite channel simulation model, so that the agent interacts with the satellite channel simulation model to randomly generate a modulation and coding scheme as an initial action space value;

[0142] The satellite channel simulation model generates a response state space value according to the initial state space value and the initial action space value, and the reward function value of the initial action space value is calculated using the reward function;

[0143] The initial state space value, the initial action space value, the reward function value, and the response state space value are used as training samples;

[0144] A number of training samples are used to iteratively train the agent, and in each round of training process, the agent is evaluated using a Q-learning formula to generate a Q value of the agent, and when the Q value converges, the agent is used as the channel optimization model.

[0145] It can be understood that the above device embodiment is corresponding to the method embodiment of the present application, and can realize the optimization method of the satellite communication channel provided by any one of the above method embodiments of the present application.

[0146] It should be noted that the device embodiments described above are only schematic, and part or all of the modules can be selected according to actual needs to achieve the purpose of the present embodiment. In addition, in the device embodiment provided by the present application, the connection relationship between the modules indicates that there is a communication connection between them, which can be realized as one or more communication buses or signal lines. Those skilled in the art can understand and implement it without creative labor.

[0147] Referring to Figure 3 The terminal device provided by the embodiment of the present application comprises:

[0148] one or more processors;

[0149] a memory coupled to the processor for storing one or more programs;

[0150] When the one or more programs are executed by the one or more processors, the one or more processors implement the optimization method of the satellite communication channel as described above.

[0151] The processor is used to control the overall operation of the terminal device to complete all or part of the steps of the optimization method of the satellite communication channel described above. The memory is used to store various types of data to support the operation of the terminal device, which can include, for example, instructions for operating any application or method on the terminal device, and application-related data. The memory can be realized by any type of volatile or non-volatile storage device or their combination, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0152] In an exemplary embodiment, the terminal device can be implemented by one or more Application Specific Integrated Circuit (ASIC), Digital Signal Processor (DSP), Digital Signal Processing Device (DSPD), Programmable Logic Device (PLD), Field Programmable Gate Array (FPGA), controller, microcontroller, microprocessor or other electronic elements for executing the optimization method of a satellite communication channel as described in any of the above embodiments and achieving the technical effects consistent with the above method.

[0153] In another exemplary embodiment, a computer readable storage medium including a computer program is also provided, which, when executed by a processor, implements the steps of the optimization method of a satellite communication channel as described in any of the above embodiments. For example, the computer readable storage medium can be the above-mentioned memory including the computer program, which can be executed by the processor of the terminal device to complete the optimization method of a satellite communication channel as described in any of the above embodiments and achieve the technical effects consistent with the above method.

[0154] The above is the preferred embodiment of the present application, it should be noted that for those skilled in the art, without departing from the principles of the present application, can make a number of improvements and refinements, these improvements and refinements are also considered to be within the scope of the present application.

Claims

1. A method of optimizing a satellite communication channel, characterized by, The method comprises the following steps: acquiring a first channel state parameter of a satellite communication channel at a current time and a second channel state parameter at a plurality of historical times; inputting the first channel state parameter and the second channel state parameter into a preset channel state prediction model, so that the channel state prediction model identifies a change trend of the satellite communication channel according to the first channel state parameter and the second channel state parameter, and predicts a predicted channel state parameter of the satellite communication channel at a next time; when it is determined that the satellite communication channel has a fluctuation risk according to the first channel state parameter and the predicted channel state parameter, generating a target modulation and coding scheme according to the predicted channel state parameter; switching a modulation and coding mechanism of a satellite communication system according to the target modulation and coding scheme; wherein the construction of the channel state prediction model comprises: acquiring a plurality of historical channel state parameters of the satellite communication channel; acquiring a plurality of continuous historical channel state parameters with a preset time length, constructing a channel state time sequence, and taking a next historical channel state parameter of the last historical channel state parameter in the channel state time sequence as a predicted value of the channel state time sequence; constructing an initial prediction model based on a long short-term memory network; training the initial prediction model by using the channel state time sequence and the corresponding predicted value, and in each round of training, evaluating the initial prediction model by using a preset loss function to generate a corresponding loss function value, optimizing the model parameters of the initial prediction model by using a back propagation algorithm when the loss function value does not converge, and taking the last optimized initial prediction model as the channel state prediction model when the loss function value converges.

2. A method for optimizing a satellite communication channel as claimed in claim 1, wherein, The first channel state parameter comprises a first signal-to-noise ratio at the current time, and the predicted channel state parameter comprises a second signal-to-noise ratio at the next time; determining that the satellite communication channel has a fluctuation risk according to the first channel state parameter and the predicted channel state parameter comprises: calculating a signal-to-noise ratio fluctuation value according to the first signal-to-noise ratio and the second signal-to-noise ratio; when the signal-to-noise ratio fluctuation value is greater than a preset threshold, determining that the satellite communication channel has a fluctuation risk.

3. A method of optimizing a satellite communication channel as claimed in claim 2, characterized in that, The generation of the target modulation and coding scheme according to the predicted channel state parameter comprises: inputting the predicted channel state parameter into a preset channel optimization model, so that the channel optimization model outputs a target modulation and coding scheme according to the predicted channel state parameter, with the goal of optimizing the channel state of the satellite communication channel.

4. A method of optimizing a satellite communication channel as claimed in claim 3, characterized in that, The construction of the channel optimization model comprises: constructing a satellite channel simulation model, an agent and a reward function; wherein the reward function is used to guide the agent to generate a modulation and coding scheme for optimizing the channel state; randomly extracting a plurality of historical channel state parameters as initial state space values; The satellite channel simulation model is taken as a simulation environment for interaction of the agent, and the initial state space is sequentially input into the satellite channel simulation model, so that the agent interacts with the satellite channel simulation model to randomly generate a modulation and coding scheme as an initial action space value; The satellite channel simulation model is used to generate a response state space value according to the initial state space value and the initial action space value, and the reward function value of the initial action space value is calculated by using the reward function; The initial state space value, the initial action space value, the reward function value and the response state space value are taken as training samples; The agent is iteratively trained by using a plurality of training samples, in each round of training process, the agent is evaluated by using a Q learning formula to generate a Q value of the agent, and when the Q value converges, the agent is taken as the channel optimization model.

5. An apparatus for optimizing a satellite communication channel, characterized by Comprise: The channel state acquisition module is configured to acquire a first channel state parameter of the satellite communication channel at a current time and a second channel state parameter at a plurality of historical times. The channel state prediction module is configured to input the first channel state parameter and the second channel state parameter into a preset channel state prediction model, so that the channel state prediction model identifies a change trend of the satellite communication channel according to the first channel state parameter and the second channel state parameter, and predicts a predicted channel state parameter of the satellite communication channel at a next time. The channel state optimization module is configured to generate a target modulation and coding scheme according to the predicted channel state parameter when it is determined that the satellite communication channel has a fluctuation risk according to the first channel state parameter and the predicted channel state parameter. The channel control module is configured to switch a modulation and coding mechanism of the satellite communication system according to the target modulation and coding scheme. The construction of the channel state prediction model comprises: Acquire a plurality of historical channel state parameters of the satellite communication channel. Acquire a plurality of continuous historical channel state parameters with a preset time length, construct a channel state time sequence, and take a next historical channel state parameter of the last historical channel state parameter in the channel state time sequence as a prediction value of the channel state time sequence. Construct an initial prediction model based on a long short-term memory network. Train the initial prediction model by using the channel state time sequence and the corresponding prediction value, and in each round of training, evaluate the initial prediction model by using a preset loss function to generate a corresponding loss function value, optimize the model parameters of the initial prediction model by using a back propagation algorithm when the loss function value does not converge, and take the last optimized initial prediction model as the channel state prediction model when the loss function value converges.

6. An apparatus for optimizing a satellite communication channel as recited in claim 5, wherein, The first channel state parameter comprises a first signal-to-noise ratio at a current time, and the predicted channel state parameter comprises a second signal-to-noise ratio at a next time. The channel state optimization module determines that the satellite communication channel has a fluctuation risk according to the first channel state parameter and the predicted channel state parameter. According to the first signal-to-noise ratio and the second signal-to-noise ratio, a signal-to-noise ratio fluctuation value is calculated; When the signal-to-noise ratio fluctuation value is greater than a preset threshold, it is determined that the satellite communication channel has a fluctuation risk.

7. An apparatus for optimizing a satellite communication channel as recited in claim 6, wherein, The channel state optimization module generates a target modulation and coding scheme according to the predicted channel state parameter, including: The predicted channel state parameter is input into a preset channel optimization model, so that the channel optimization model outputs a target modulation and coding scheme according to the predicted channel state parameter, with the goal of optimizing the channel state of the satellite communication channel.

8. An apparatus for optimizing a satellite communication channel as recited in claim 7, wherein, The construction of the channel optimization model includes: A satellite channel simulation model, an agent, and a reward function are constructed, wherein the reward function is used to guide the agent to generate a modulation and coding scheme that optimizes the channel state. Historical channel state parameters are randomly extracted multiple times as initial state space values; The satellite channel simulation model is used as a simulation environment for agent interaction, and the initial state space is sequentially input into the satellite channel simulation model, so that the agent interacts with the satellite channel simulation model to randomly generate a modulation and coding scheme as an initial action space value; The satellite channel simulation model generates a response state space value according to the initial state space value and the initial action space value, and the reward function is used to calculate the reward function value of the initial action space value; The initial state space value, the initial action space value, the reward function value, and the response state space value are used as training samples; A number of training samples are used to iteratively train the agent, and in each round of training process, the agent is evaluated using a Q-learning formula to generate a Q value of the agent, and when the Q value converges, the agent is used as the channel optimization model.

9. A terminal device, comprising: It includes: One or more processors; Memory coupled to the processor for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the optimization method of the satellite communication channel according to any one of claims 1-4.

10. A storage medium having stored thereon a computer program, characterized in that The computer program is executed by the processor to implement the optimization method of the satellite communication channel according to any one of claims 1-4.

Images