Wireless signal interference suppression method and device based on deep learning
By building a deep learning model, combining convolutional neural networks and long-term memory networks, the problem of poor wireless signal separation effect in complex interference environments is solved, and stronger interference suppression and signal separation effect is achieved, improving the performance and reliability of wireless communication systems.
Patent Information
- Application Number
- CN202510538998.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-27
- Publication Date
- 2025-08-01
AI Technical Summary
Traditional wireless signal interference suppression methods are not effective in complex interference environments, especially with limited adaptability to non-stationary and unknown interferences.
Using a deep learning-based method, we build a hybrid architecture combining convolutional neural networks with long and short-term memory networks, combining data acquisition, preprocessing, model training and interference suppression steps, we automatically learn the characteristics and interference modes of wireless signals to achieve the separation of target signals and interfering signals.
It improves the adaptability and signal separation capabilities of wireless communication systems in complex interference environments, and enhances the accuracy and effectiveness of signals.
Smart Images
Figure CN120415591A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of wireless communication technologies, and in particular to a method and apparatus for suppressing wireless signal interference based on deep learning. Background Art
[0002] In modern wireless communication systems, wireless signals face various interferences. These interferences may come from other wireless devices operating on the same frequency, multipath effects, man-made interference sources, etc. The existence of interference seriously affects the quality of wireless communication. For example, it reduces the signal-to-noise ratio of the signal, increases the bit error rate, and thus affects the accuracy and effectiveness of data transmission.
[0003] Traditional interference suppression methods include filtering techniques, equalization techniques, etc. However, these traditional methods often have limitations in complex interference environments. For example, filtering techniques have poor suppression effects on non-stationary interferences, and equalization techniques require accurate channel estimation and have limited adaptability to unknown interferences. Therefore, it is necessary to design a method and apparatus for suppressing wireless signal interference based on deep learning. Summary of the Invention
[0004] The purpose of the present invention is to solve the drawbacks existing in the prior art, and to propose a method and apparatus for suppressing wireless signal interference based on deep learning.
[0005] To achieve the above purpose, the present invention adopts the following technical solutions: A method for suppressing wireless signal interference based on deep learning, including the following steps:
[0006] S1: Data acquisition: Use multiple receiving antennas to receive wireless signals containing interference, sample the received signals to convert them into digital signals, and collect signal samples of a certain time length to form an original signal dataset;
[0007] S2: Data preprocessing: Perform normalization and framing processing on the original signal dataset;
[0008] S3: Deep learning model construction: Construct a deep learning model for learning the characteristics and interference patterns of wireless signals. The input of the model is the preprocessed signal data, and the output is the estimated value of the target signal or the suppression result of the interference signal;
[0009] S4: Model training: Divide the preprocessed signal dataset into a training set, a validation set, and a test set, define a loss function, and use an optimization algorithm to train the deep learning model;
[0010] S5: Interference suppression: Preprocess the newly received wireless signals containing interference and input them into the trained deep learning model, and perform denormalization processing on the signals output by the model to obtain the final wireless signals after interference suppression.
[0011] As a further description of the above technical solution:
[0012] The signals collected in step S1 include target signals and interference signals. The target signals include one or more of Wi-Fi, Bluetooth, and LTE, and the interference signals include one or more of noise, narrowband interference, and pulse interference.
[0013] As a further description of the above technical solution:
[0014] The normalization process in step S2 normalizes the amplitude of the signal to the interval [-1, 1], and the frame segmentation process divides the continuous signal into segments according to a certain frame length.
[0015] As a further description of the above technical solution:
[0016] In step S3, a hybrid architecture combining a convolutional neural network (CNN) and a long short-term memory network (LSTM) is adopted. The CNN part is used to extract local features of the signal, and the LSTM part is used to process the temporal characteristics of the signal. A fully connected layer is set between the CNN and the LSTM.
[0017] As a further description of the above technical solution:
[0018] In step S4, the training set, validation set, and test set are divided according to the ratios of 70%, 15%, and 15% respectively. The mean squared error function is used as the loss function, and the stochastic gradient descent is used as the optimization algorithm.
[0019] A wireless signal interference suppression device based on deep learning, comprising:
[0020] A data acquisition module: used to acquire signal data in a wireless communication environment, including target signals and interference signals;
[0021] A data preprocessing module: preprocesses the acquired signal data, including operations such as signal normalization, filtering, and sampling rate adjustment;
[0022] A deep learning model processing module: constructs a deep learning model for learning the characteristics and interference patterns of wireless signals;
[0023] A model training module: collects labeled data and trains the model using a suitable loss function and optimization algorithm;
[0024] An interference suppression module: inputs the wireless signal to be processed into the trained model and outputs an estimated value of the target signal after interference suppression.
[0025] As a further description of the above technical solution:
[0026] The deep learning model processing module includes a storage module for storing deep learning model parameters and a calculation module for performing model calculations.
[0027] The present invention has the following beneficial effects:
[0028] Compared with the prior art, the present invention automatically learns the characteristics and interference patterns of wireless signals through a deep learning model, can adapt to complex interference environments and variable signal characteristics, and has stronger adaptability than traditional methods. Moreover, the deep learning model has powerful feature extraction and pattern recognition capabilities, can more accurately identify and separate target signals and interference signals, thereby achieving more effective interference suppression and improving the performance and reliability of wireless communication systems. Description of the Drawings
[0029] Figure 1 is a flowchart of the method of the present invention;
[0030] Figure 2 is a structural block diagram of the present invention. Detailed Embodiments
[0031] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.
[0032] Refer to Figure 1-2 , the wireless signal interference suppression method based on deep learning provided by the present invention includes the following steps:
[0033] S1: Data acquisition: Use multiple receiving antennas to receive wireless signals containing interference, sample the received signals to convert them into digital signals, and collect signal samples of a certain time length to form an original signal dataset; the collected signals include target signals and interference signals, and the target signals include one or more of Wi-Fi, Bluetooth, and LTE, and the interference signals include one or more of noise, narrowband interference, and pulse interference.
[0034] In the actual environment of a cellular network, signal data containing co-channel interference is collected through a base station and a mobile terminal. The collected data includes uplink signals and downlink signals; in the actual environment of a wireless local area network, signal data containing multipath interference is collected through a wireless access point and a terminal device. The collected data includes signal samples of different channels and different positions.
[0035] S2: Data preprocessing: Normalize and frame the original signal dataset. Normalization is to normalize the amplitude of the signal to the range [-1, 1], and framing is to segment the continuous signal according to a certain frame length.
[0036] Specifically, in the data preprocessing unit, the normalization module performs normalization according to the following formula:
[0037] Let the original signal sample be x i , i = 1, 2, ···, n, where n is the number of signal samples. The normalized signal y i is calculated according to the formula , where min(x) and max(x) are the minimum and maximum values in the original signal samples respectively.
[0038] The framing module frames the normalized signal according to the frame length L. For example, L = 1024 sampling points is one frame.
[0039] S3: Deep learning model construction: Construct a deep learning model for learning the characteristics and interference patterns of wireless signals. The input of the model is the preprocessed signal data, and the output is the estimated value of the target signal or the suppression result of the interference signal. The deep learning model construction adopts a hybrid architecture combining a convolutional neural network (CNN) and a long short-term memory network (LSTM). The CNN part is used to extract the local features of the signal, and the LSTM part is used to process the temporal characteristics of the signal. A fully connected layer is set between the CNN and the LSTM.
[0040] Specifically, the CNN part is mainly used to extract the local features of the signal. Its convolutional layers use convolutional kernels of different sizes, such as 3×3, 5×5, etc., to capture signal features at different scales. For example, smaller convolutional kernels can capture the detailed features in the signal, while larger convolutional kernels can capture more macroscopic features.
[0041] The LSTM part is used to process the temporal characteristics of the signal. Since wireless signals are sequential signals that change over time, the LSTM can effectively capture the dependencies of the signal on the time axis.
[0042] A fully connected layer is set between the CNN and the LSTM, which is used to map the local features extracted by the CNN to the feature space suitable for LSTM processing.
[0043] S4: Model training: Divide the preprocessed signal dataset into a training set, a validation set, and a test set, define a loss function, and use an optimization algorithm to train the deep learning model. The training set, validation set, and test set are divided according to the ratios of 70%, 15%, and 15% respectively. The loss function uses the mean squared error function, and the optimization algorithm uses stochastic gradient descent.
[0044] Divide the preprocessed signal dataset into a training set, a validation set, and a test set. Usually, according to a certain ratio, such as 70% for the training set, 15% for the validation set, and 15% for the test set.
[0045] Define a loss function, such as the mean squared error (MSE) function, to measure the error between the signal after interference suppression output by the model and the original interference-free signal.
[0046] Use an optimization algorithm, such as stochastic gradient descent, to train the deep learning model. During the training process, adjust the parameters of the model, such as the learning rate, convolution kernel parameters, etc., according to the performance of the validation set to prevent overfitting.
[0047] During the model training process, continuously evaluate the performance of the model on the validation set. When the mean squared error on the validation set does not decrease for 5 consecutive training epochs, stop the training.
[0048] S5: Interference suppression: Preprocess the newly received wireless signal containing interference and input it into the trained deep learning model. Perform denormalization on the signal output by the model to obtain the final wireless signal after interference suppression.
[0049] After the deep learning model is trained, process the newly received wireless signal containing interference according to the data preprocessing steps. Input the processed signal into the trained deep learning model, and the model outputs an estimate of the signal after interference suppression. Perform denormalization on the output signal to convert it to the actual signal amplitude range to obtain the final wireless signal after interference suppression.
[0050] A wireless signal interference suppression device based on deep learning, comprising:
[0051] Data acquisition module: Used to acquire signal data in the wireless communication environment, including target signals and interference signals;
[0052] Data preprocessing module: Preprocess the acquired signal data, including operations such as signal normalization, filtering, and sampling rate adjustment;
[0053] Deep learning model processing module: Construct a deep learning model to learn the characteristics and interference patterns of wireless signals. The deep learning model processing module includes a storage module for storing deep learning model parameters and a calculation module for performing model calculations;
[0054] Model training module: Collect labeled data and train the model using a suitable loss function and optimization algorithm;
[0055] Interference suppression module: Input the wireless signal to be processed into the trained model and output an estimated value of the target signal after interference suppression.
[0056] Finally, it should be noted that the above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements on some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.
Claims
1. A method for suppressing wireless signal interference based on deep learning, characterized in that: It includes the following steps: S1: Data acquisition: Use multiple receiving antennas to receive wireless signals containing interference, sample the received signals to convert them into digital signals, and collect signal samples of a certain time length to form an original signal dataset; S2: Data preprocessing: Perform normalization and framing on the original signal dataset; S3: Deep learning model construction: Construct a deep learning model for learning the characteristics of wireless signals and interference patterns. The input of the model is the preprocessed signal data, and the output is the estimated value of the target signal or the suppression result of the interference signal; S4: Model training: Divide the preprocessed signal dataset into a training set, a validation set, and a test set, define a loss function, and use an optimization algorithm to train the deep learning model; S5: Interference suppression: Preprocess the newly received wireless signals containing interference and input them into the trained deep learning model, and perform denormalization on the signals output by the model to obtain the final wireless signals after interference suppression.
2. The method for suppressing wireless signal interference based on deep learning according to claim 1, characterized in that: The signals collected in step S1 include target signals and interference signals. The target signals include one or more of Wi-Fi, Bluetooth, and LTE, and the interference signals include one or more of noise, narrowband interference, and impulse interference.
3. The method for suppressing wireless signal interference based on deep learning according to claim 1, wherein: In step S2, the normalization process normalizes the amplitude of the signal to the interval [-1, 1], and the framing process divides the continuous signal according to a certain frame length.
4. The method for suppressing wireless signal interference based on deep learning according to claim 1, characterized in that: In step S3, a hybrid architecture combining a convolutional neural network (CNN) and a long short-term memory network (LSTM) is adopted. The CNN part is used to extract the local features of the signal, and the LSTM part is used to process the temporal characteristics of the signal. A fully connected layer is set between the CNN and the LSTM.
5. The method for suppressing wireless signal interference based on deep learning according to claim 1, wherein: In step S4, the training set, the validation set, and the test set are divided at a ratio of 70%, 15%, and 15% respectively. The mean squared error function is used as the loss function, and the stochastic gradient descent is used as the optimization algorithm.
6. A wireless signal interference suppression device based on deep learning, characterized in that, It includes: Data acquisition module: Used to collect signal data in the wireless communication environment, including target signals and interference signals; Data preprocessing module: Preprocess the collected signal data, including operations such as signal normalization, filtering, and sampling rate adjustment; Deep learning model processing module: Construct a deep learning model for learning the characteristics of wireless signals and interference patterns; Model training module: Collect labeled data and use a suitable loss function and optimization algorithm to train the model; Interference suppression module: Input the wireless signals to be processed into the trained model and output the estimated value of the target signal after suppressing interference; 7. The wireless signal interference suppression device based on deep learning according to claim 1, characterized in that: The deep learning model processing module includes a storage module for storing the parameters of the deep learning model and a calculation module for performing model calculations.