Adaptive spectrum 5G communication module and use method thereof

Abstract

The invention relates to the technical field of 5G communication modules, in particular to a 5G communication module with a self-adaptive frequency spectrum and a use method of the 5G communication module. Comprising a spectrum sensing and identification module, a spectrum measurement and reconstruction module based on compressed sensing, a self-adaptive spectrum selection and dynamic spectrum allocation control module, a dynamic channel modeling and prediction module, a comprehensive optimization algorithm module and a feedback-assisted channel estimation control module. Wherein the spectrum sensing and recognition module adopts a deep learning algorithm to improve the precision and speed of spectrum sensing, and meets the requirements of 5G high data rate and low delay; the specific implementation mode comprises a convolutional neural network (CNN) structure, the convolutional neural network is adopted for spectrum sensing, and the spectrum state can be accurately and rapidly recognized in a complex dynamic environment; by training a large amount of spectrum data, the CNN can extract spectrum features and effectively distinguish target signals and interference signals, and the precision and speed of spectrum sensing are significantly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of 5G communication modules, and specifically to a spectrum-adaptive 5G communication module and a method for using the module. Background Art

[0002] With the rapid development of the fifth generation mobile communication technology (5G), the efficient use of spectrum resources has become an important research topic in the field of wireless communications. 5G communication systems need to meet the requirements of high speed, low latency and large capacity, so the demand for spectrum resources has increased significantly. However, traditional spectrum management methods usually rely on pre-allocated fixed frequency bands, which are difficult to cope with the increasingly complex and dynamically changing communication environment, resulting in inefficient use of spectrum resources and cannot fully meet the needs of 5G communications.

[0003] In the prior art, spectrum sensing and channel estimation are key links that affect spectrum utilization efficiency. Traditional spectrum sensing methods, such as energy detection method and cooperative sensing method, have problems such as low sensing accuracy and susceptibility to noise and interference, making it difficult to accurately identify and distinguish target signals from interference signals in complex environments. In addition, the non-flat fading channels in the 5G communication environment make it difficult for traditional channel estimation methods (such as pilot-based estimation methods) to cope with complex and changeable communication environments, resulting in large channel estimation errors, which in turn affects system performance. Therefore, in response to the above problems, a 5G communication module with adaptive spectrum and a method for using the module are proposed. Summary of the invention

[0004] The purpose of the present invention is to provide a 5G communication module with adaptive spectrum and a method of using the same to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A 5G communication module with adaptive spectrum, comprising: a spectrum sensing and identification module, a spectrum measurement and reconstruction module based on compressed sensing, an adaptive spectrum selection and dynamic spectrum allocation control module, a dynamic channel modeling and prediction module, a comprehensive optimization algorithm module and a feedback-assisted channel estimation control module;

[0007] The spectrum sensing and identification module uses a deep learning algorithm to improve the accuracy and speed of spectrum sensing to meet the requirements of 5G high data rate and low latency. The specific implementation method includes a convolutional neural network (CNN) structure for feature extraction and classification of spectrum data. By training the data set D = {(Y i , S i )}, where Y i is the spectrum data, S i is the label (target signal or interference signal), and optimizes the loss function to train the CNN model:

[0008]

[0009] where θ CNN are the parameters of the CNN model, is the loss function, f CNN is the output of CNN; set N = 1000, by using the spectrum sensing module based on deep learning, the experimental data shows that the recognition accuracy can reach more than 90% in complex environments.

[0010] Preferably, the spectrum measurement and reconstruction module based on compressed sensing uses sparse representation theory to achieve efficient sampling and reconstruction of spectrum signals. The specific implementation method includes the following algorithm:

[0011]

[0012] in is the reconstructed spectrum signal, Y is the received spectrum data, Φ is the measurement matrix, λ is the regularization parameter, and the spectrum signal X is measured using the random measurement matrix Φ, and the minimum To reconstruct the signal X, set ||X||1=0.05, the optimized algorithm significantly reduces the measurement time and resource consumption while ensuring accuracy.

[0013] Preferably, the adaptive spectrum selection and dynamic spectrum allocation control module uses a reinforcement learning algorithm to optimize the allocation and utilization efficiency of spectrum resources in real time. The specific implementation method includes the following algorithm:

[0014]

[0015] Where Q(s, a) is the value function of taking action a in state s, r is the reward, γ is the discount factor, and s′ is the next state. By establishing a Markov decision process (MDP) model to describe the relationship between state, action and reward, the optimized strategy π * Ability to adjust the allocation of spectrum resources in real time to maximize system throughput or minimize communication delay.

[0016] Preferably, the dynamic channel modeling and prediction module uses a recurrent neural network (RNN) structure to perform real-time modeling and prediction of complex and changeable 5G communication channels. The specific implementation method includes the following algorithm:

[0017]

[0018] in is the channel estimate at time t, h(t-1) is the channel state at time t-1, and y(t) is the signal received at time t. The long short-term memory network (LSTM) is trained to capture the long-term dependency of the channel state and predict the channel response at the next moment. The optimized model can effectively cope with high-speed mobile environments and multipath fading.

[0019] Preferably, the integrated optimization algorithm module jointly optimizes spectrum sensing, channel estimation and data transmission processes to maximize the performance of the communication system; by constructing a joint optimization problem Where S is the spectrum selection matrix, H(S) is the channel matrix that depends on the spectrum selection, X is the transmitted data, Y is the received data, and λ1 and λ2 are regularization parameters that improve the performance of the system in various scenarios.

[0020] Preferably, the feedback-assisted channel estimation control module adjusts the parameters of the channel estimation algorithm in real time through a two-way feedback mechanism between the base station and the user equipment to adapt to different mobile speeds and channel environment changes. The specific implementation method includes the following algorithm:

[0021]

[0022] in is the current estimated channel response, h feedback is the feedback channel information, and α is the update rate. Through the semi-supervised learning method, the feedback information from the user equipment is used to adjust the parameters of the channel model to improve the accuracy and stability of the channel estimation.

[0023] Preferably, the invention further comprises a computer program product for implementing the above method, wherein the computer program product stores a computer-readable storage medium having a computer program code capable of executing the above method.

[0024] Preferably, it also includes: a communication system, which includes multiple 5G communication modules with adaptive spectrum, used to build an efficient and stable 5G communication network and provide high-performance communication services.

[0025] Preferably, the spectrum sensing module based on deep learning includes the introduction of convolutional neural network (CNN) and recurrent neural network (RNN) to process complex and changeable spectrum signals and dynamic interference environments; the spectrum reconstruction module based on compressed sensing achieves efficient reconstruction and utilization of spectrum signals by optimizing sparse representation theory.

[0026] A method for using a 5G communication module with adaptive spectrum includes the following steps:

[0027] Step 1: receiving multiple signal samples;

[0028] Step 2: preprocessing the signal samples to extract spectrum features;

[0029] Step 3: Use spectrum characteristics to perform spectrum sensing and determine available frequency bands;

[0030] Step 4: Dynamically adjust spectrum allocation strategy to optimize communication performance;

[0031] Among them, in the preprocessing step, signal denoising is first performed, and then the fast Fourier transform (FFT) is used to extract spectrum features; in spectrum sensing, the convolutional neural network (CNN) is first used to classify the spectrum features, and then the spectrum occupancy is determined based on the classification results; when dynamically adjusting the spectrum allocation strategy, the Q learning algorithm is first used to optimize the spectrum allocation strategy, and then the strategy is updated based on historical spectrum usage data; optimizing communication performance includes: minimizing communication delay and maximizing data transmission rate.

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] 1. In the present invention, convolutional neural network (CNN) is used for spectrum sensing, which can accurately and quickly identify spectrum status in complex dynamic environments. By training a large amount of spectrum data, CNN can extract spectrum features and effectively distinguish target signals from interference signals, significantly improving the accuracy and speed of spectrum sensing.

[0034] 2. In the present invention, the Q learning algorithm is introduced to optimize the spectrum allocation strategy, making the use of spectrum resources more efficient. By establishing a Markov decision process (MDP) model, the system can dynamically adjust the spectrum allocation strategy according to the real-time spectrum sensing results to maximize the system throughput or minimize the communication delay, thereby improving the utilization of spectrum resources.

[0035] 3. In the present invention, a method combining compressed sensing and deep learning is used for channel estimation, which can accurately estimate the channel state in a complex and changeable communication environment. Through the joint optimization of sparse representation and deep neural network (DNN), the system can reconstruct the channel response at a low sampling rate, reduce the channel estimation error, and improve the performance of the communication system.

[0036] 4. The present invention can adapt to complex and changeable communication environments, such as high-speed mobile scenarios and multipath fading environments. By dynamically adjusting the parameters of the spectrum sensing and channel estimation algorithms, the robustness and flexibility of the system are enhanced, and the communication quality is guaranteed.

[0037] 5. In the present invention, by means of a comprehensive optimization algorithm, spectrum sensing, channel estimation and data transmission processes are jointly optimized. The system can improve communication performance in a variety of scenarios, reduce communication delays, increase data transmission rates, and meet the high performance requirements of 5G communications. DETAILED DESCRIPTION

[0038] Embodiment 1: The present invention provides a technical solution:

[0039] A 5G communication module with adaptive spectrum, comprising: a spectrum sensing and identification module, a spectrum measurement and reconstruction module based on compressed sensing, an adaptive spectrum selection and dynamic spectrum allocation control module, a dynamic channel modeling and prediction module, a comprehensive optimization algorithm module and a feedback-assisted channel estimation control module;

[0040] The spectrum sensing and identification module uses a deep learning algorithm to improve the accuracy and speed of spectrum sensing to meet the requirements of 5G high data rate and low latency. The specific implementation method includes a convolutional neural network (CNN) structure for feature extraction and classification of spectrum data. By training the data set D = {(Y i , S i )}, where Y i is the spectrum data, S i is the label (target signal or interference signal), and optimizes the loss function to train the CNN model:

[0041]

[0042] where θ CNN are the parameters of the CNN model, is the loss function, f CNN is the output of CNN; set N = 1000, by using the spectrum sensing module based on deep learning, the experimental data shows that the recognition accuracy can reach more than 90% in complex environments.

[0043] The spectrum sensing and recognition module in the present invention greatly improves the accuracy and speed of spectrum sensing by adopting deep learning algorithms, especially convolutional neural networks (CNN). Experimental data show that the recognition accuracy can reach more than 90% in complex environments. By optimizing the loss function to train the CNN model, the system can quickly and accurately identify the spectrum state and effectively distinguish the target signal from the interference signal, thereby significantly improving the spectrum utilization efficiency of the 5G communication system and meeting the requirements of high data rate and low latency.

[0044] The spectrum measurement and reconstruction module based on compressed sensing uses sparse representation theory to achieve efficient sampling and reconstruction of spectrum signals. The specific implementation method includes the following algorithms:

[0045]

[0046] in is the reconstructed spectrum signal, Y is the received spectrum data, Φ is the measurement matrix, λ is the regularization parameter, and the spectrum signal X is measured using the random measurement matrix Φ, and the minimum To reconstruct the signal X, set ||X||1=0.05, the optimized algorithm significantly reduces the measurement time and resource consumption while ensuring accuracy.

[0047] The spectrum measurement and reconstruction module based on compressed sensing in the present invention uses sparse representation theory to achieve efficient sampling and reconstruction of spectrum signals. By using a random measurement matrix to measure the spectrum signal and reconstructing the signal through an optimization algorithm, the measurement time and resource consumption are significantly reduced. While ensuring the accuracy of signal reconstruction, the computational complexity and power consumption of the system are reduced, and the performance and efficiency of the 5G communication system in complex dynamic environments are improved.

[0048] The adaptive spectrum selection and dynamic spectrum allocation control module uses a reinforcement learning algorithm to optimize the allocation and utilization efficiency of spectrum resources in real time. The specific implementation method includes the following algorithms:

[0049]

[0050] Where Q(s, a) is the value function of taking action a in state s, r is the reward, γ is the discount factor, and s′ is the next state. By establishing a Markov decision process (MDP) model to describe the relationship between state, action and reward, the optimized strategy π * Ability to adjust the allocation of spectrum resources in real time to maximize system throughput or minimize communication delay.

[0051] The adaptive spectrum selection and dynamic spectrum allocation control module in the present invention optimizes the allocation of spectrum resources in real time by introducing a reinforcement learning algorithm. The system can dynamically adjust the spectrum allocation strategy based on real-time spectrum sensing results and historical usage data to maximize system throughput or minimize communication delay. By establishing a Markov decision process (MDP) model, the system can flexibly adjust the allocation of spectrum resources under different communication requirements, thereby improving spectrum utilization and the overall performance of the communication system.

[0052] The dynamic channel modeling and prediction module uses a recurrent neural network (RNN) structure to perform real-time modeling and prediction of complex and changeable 5G communication channels. The specific implementation method includes the following algorithms:

[0053]

[0054] in is the channel estimate at time t, h(t-1) is the channel state at time t-1, and y(t) is the signal received at time t. The long short-term memory network (LSTM) is trained to capture the long-term dependency of the channel state and predict the channel response at the next moment. The optimized model can effectively cope with high-speed mobile environments and multipath fading.

[0055] The channel estimation module in the present invention adopts a method combining compressed sensing and deep learning, which can accurately estimate the channel state in a complex and changeable communication environment. Through the joint optimization of sparse representation and deep neural network (DNN), the system can reconstruct the channel response at a low sampling rate, reduce the channel estimation error, and improve the performance of the communication system. In high-speed mobile scenarios and multipath fading environments, the robustness and flexibility of the system are significantly enhanced, ensuring the communication quality.

[0056] The integrated optimization algorithm module jointly optimizes spectrum sensing, channel estimation and data transmission processes to maximize the performance of the communication system; by constructing a joint optimization problem Where S is the spectrum selection matrix, H(S) is the channel matrix that depends on the spectrum selection, X is the transmitted data, Y is the received data, and λ1 and λ2 are regularization parameters that improve the performance of the system in various scenarios.

[0057] The joint optimization algorithm in the present invention can improve communication performance in a variety of scenarios by comprehensively optimizing spectrum sensing, channel estimation and data transmission processes. By minimizing communication delays and maximizing data transmission rates, the system can meet the high-performance requirements of 5G communications and improve user experience. The optimization algorithm achieves fast and reliable data transmission while ensuring efficient spectrum utilization.

[0058] The feedback-assisted channel estimation control module adjusts the parameters of the channel estimation algorithm in real time through a two-way feedback mechanism between the base station and the user equipment to adapt to different mobile speeds and channel environment changes. The specific implementation method includes the following algorithms:

[0059]

[0060] in is the current estimated channel response, h feedback is the feedback channel information, and α is the update rate. Through the semi-supervised learning method, the feedback information from the user equipment is used to adjust the parameters of the channel model to improve the accuracy and stability of the channel estimation.

[0061] The feedback mechanism and semi-supervised learning method in the present invention utilize feedback information from user equipment, so that the system can adjust the channel model and spectrum allocation strategy in real time to achieve efficient resource management and utilization. The system can adaptively adjust parameters in a dynamic environment, significantly improving the resource management efficiency and communication performance of the system.

[0062] The invention also includes a computer program product for implementing the above method, wherein the computer program product stores a computer readable storage medium having a computer program code capable of executing the above method.

[0063] It also includes: a communication system, which includes multiple 5G communication modules with adaptive spectrum, used to build an efficient and stable 5G communication network and provide high-performance communication services.

[0064] The spectrum sensing module based on deep learning includes the introduction of convolutional neural network (CNN) and recurrent neural network (RNN) to process complex and changeable spectrum signals and dynamic interference environments; the spectrum reconstruction module based on compressed sensing achieves efficient reconstruction and utilization of spectrum signals by optimizing sparse representation theory.

[0065] A method for using a 5G communication module with adaptive spectrum includes the following steps:

[0066] Step 1: receiving multiple signal samples;

[0067] Step 2: preprocessing the signal samples to extract spectrum features;

[0068] Step 3: Use spectrum characteristics to perform spectrum sensing and determine available frequency bands;

[0069] Step 4: Dynamically adjust spectrum allocation strategy to optimize communication performance;

[0070] Among them, in the preprocessing step, signal denoising is first performed, and then the fast Fourier transform (FFT) is used to extract spectrum features; in spectrum sensing, the convolutional neural network (CNN) is first used to classify the spectrum features, and then the spectrum occupancy is determined based on the classification results; when dynamically adjusting the spectrum allocation strategy, the Q learning algorithm is first used to optimize the spectrum allocation strategy, and then the strategy is updated based on historical spectrum usage data; optimizing communication performance includes: minimizing communication delay and maximizing data transmission rate.

[0071] This article uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only used to help understand the method and core ideas of the present invention. The above is only a preferred implementation of the present invention. It should be pointed out that due to the limitations of textual expression and the objective existence of infinite specific structures, ordinary technicians in this technical field can make several improvements, modifications or changes without departing from the principles of the present invention, and can also combine the above technical features in an appropriate manner; these improvements, modifications, changes or combinations, or the direct application of the inventive concept and technical solution to other occasions without improvement, should be regarded as the protection scope of the present invention.

Claims

1. A 5G communication module with adaptive spectrum, characterized in that: include: Spectrum sensing and identification module, spectrum measurement and reconstruction module based on compressed sensing, adaptive spectrum selection and dynamic spectrum allocation control module, dynamic channel modeling and prediction module, comprehensive optimization algorithm module and feedback-assisted channel estimation control module; The spectrum sensing and identification module uses a deep learning algorithm to improve the accuracy and speed of spectrum sensing to meet the requirements of 5G high data rate and low latency. The specific implementation method includes a convolutional neural network (CNN) structure for feature extraction and classification of spectrum data. By training the data set D = {(Y i , S i )}, where Y i is the spectrum data, S i is the label (target signal or interference signal), and optimizes the loss function to train the CNN model: where θ CNN are the parameters of the CNN model, is the loss function, f CNN is the output of CNN.

2. The 5G communication module with adaptive spectrum according to claim 1, characterized in that: The spectrum measurement and reconstruction module based on compressed sensing uses sparse representation theory to achieve efficient sampling and reconstruction of spectrum signals. The specific implementation method includes the following algorithms: in is the reconstructed spectrum signal, Y is the received spectrum data, Φ is the measurement matrix, λ is the regularization parameter, and the spectrum signal X is measured using the random measurement matrix Φ, and the minimum To reconstruct the signal X, set ||X||1=0.05, the optimized algorithm significantly reduces the measurement time and resource consumption while ensuring accuracy.

3. The 5G communication module with adaptive spectrum according to claim 1, characterized in that: The adaptive spectrum selection and dynamic spectrum allocation control module uses a reinforcement learning algorithm to optimize the allocation and utilization efficiency of spectrum resources in real time. The specific implementation method includes the following algorithms: Where Q(s, a) is the value function of taking action a in state s, r is the reward, γ is the discount factor, and s′ is the next state; By building a Markov decision process (MDP) model to describe the relationship between state, action and reward, the optimized strategy π * Ability to adjust the allocation of spectrum resources in real time to maximize system throughput or minimize communication delay.

4. The 5G communication module with adaptive spectrum according to claim 1, characterized in that: The dynamic channel modeling and prediction module uses a recurrent neural network (RNN) structure to perform real-time modeling and prediction of complex and changeable 5G communication channels. The specific implementation method includes the following algorithms: in is the channel estimate at time t, h(t-1) is the channel state at time t-1, and y(t) is the signal received at time t. The long short-term memory network (LSTM) is trained to capture the long-term dependency of the channel state and predict the channel response at the next moment. The optimized model can effectively cope with high-speed mobile environments and multipath fading.

5. The 5G communication module with adaptive spectrum according to claim 1, characterized in that: The integrated optimization algorithm module jointly optimizes spectrum sensing, channel estimation and data transmission processes to maximize the performance of the communication system; by constructing a joint optimization problem Where S is the spectrum selection matrix, H(S) is the channel matrix that depends on the spectrum selection, X is the transmitted data, Y is the received data, and λ1 and λ2 are regularization parameters that improve the performance of the system in various scenarios.

6. The 5G communication module with adaptive spectrum according to claim 1, characterized in that: The feedback-assisted channel estimation control module adjusts the parameters of the channel estimation algorithm in real time through a two-way feedback mechanism between the base station and the user equipment to adapt to different mobile speeds and channel environment changes. The specific implementation method includes the following algorithms: in is the current estimated channel response, h feedback is the feedback channel information, and α is the update rate. Through the semi-supervised learning method, the feedback information from the user equipment is used to adjust the parameters of the channel model to improve the accuracy and stability of the channel estimation.

7. The 5G communication module with adaptive spectrum according to claim 1, characterized in that: The invention also includes a computer program product for implementing the above method, wherein the computer program product stores a computer readable storage medium having a computer program code capable of executing the above method.

8. The 5G communication module with adaptive spectrum according to claim 1, characterized in that: Also includes: A communication system includes multiple 5G communication modules with adaptive spectrum, which are used to build an efficient and stable 5G communication network and provide high-performance communication services.

9. The 5G communication module with adaptive spectrum according to claim 1, characterized in that: The spectrum sensing module based on deep learning includes the introduction of convolutional neural network (CNN) and recurrent neural network (RNN) to process complex and changeable spectrum signals and dynamic interference environments; the spectrum reconstruction module based on compressed sensing achieves efficient reconstruction and utilization of spectrum signals by optimizing sparse representation theory.

10. A method for using a 5G communication module with adaptive spectrum, characterized in that: The following steps are involved: Step 1: receiving multiple signal samples; Step 2: preprocessing the signal samples to extract spectrum features; Step 3: Use spectrum characteristics to perform spectrum sensing and determine available frequency bands; Step 4: Dynamically adjust spectrum allocation strategy to optimize communication performance; Among them, in the preprocessing step, signal denoising is first performed, and then the fast Fourier transform (FFT) is used to extract spectrum features; in spectrum sensing, the convolutional neural network (CNN) is first used to classify the spectrum features, and then the spectrum occupancy is determined based on the classification results; when dynamically adjusting the spectrum allocation strategy, the Q learning algorithm is first used to optimize the spectrum allocation strategy, and then the strategy is updated based on historical spectrum usage data; optimizing communication performance includes: minimizing communication delay and maximizing data transmission rate.

Images