Resource data sensing method and device and communication equipment
By identifying interference types through multi-dimensional environmental parameters and interference classification models, dynamically selecting anti-interference algorithm combinations, and optimizing the resource data perception process, the problems of data distortion and performance drop in resource data perception in existing technologies are solved, and high-confidence resource data perception is achieved.
Patent Information
- Application Number
- CN202511440983.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-10
- Publication Date
- 2026-01-06
AI Technical Summary
Existing resource data sensing methods cannot dynamically adjust strategies for different types of interference, resulting in high distortion rates of the sensed data. In particular, traditional anti-interference algorithms experience a sharp drop in performance in high-speed movement and low signal-to-noise ratio scenarios in complex communication environments.
An environmental feature vector is generated using multi-dimensional environmental parameters. Interference types are identified through an interference classification model. An anti-interference strategy library is pre-built. The best combination of anti-interference algorithms is dynamically selected. The resource data perception process is optimized through confidence assessment and strategy iteration to ensure data confidence and anti-interference effectiveness.
It enables accurate perception of resource data in complex environments, reduces data distortion rate, improves data confidence, adapts to high-speed movement and low signal-to-noise ratio scenarios, and ensures high reliability of resource scheduling.
Smart Images
Figure CN121284637A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of communication technology, and specifically relates to a resource data sensing method, apparatus, and communication equipment. Background Technology
[0002] With the rapid development of 5G / 6G communication networks, vehicle-to-everything (V2X) networks, and the industrial internet, accurate perception of resource data (such as spectrum, computing power, network status, and spatial location) has become a core foundation for supporting intelligent network scheduling and improving service quality. In real-world communication environments, various types of interference exist (such as narrowband interference, broadband interference, sudden electromagnetic interference, and Doppler frequency shift interference). Existing resource data perception methods often employ single anti-interference algorithms (such as fixed filtering or energy detection), failing to dynamically adjust strategies for different interference types, resulting in high data distortion rates. For example, in the industrial internet, sudden electromagnetic interference generated by motors, electric arcs, and other equipment can mask the true resource status; traditional energy detection methods have false alarm rates exceeding 30% and data distortion rates exceeding 40%. In high-speed mobile scenarios, Doppler frequency shift causes signal frequency deviations, which existing fixed compensation strategies struggle to eliminate, resulting in a center frequency deviation of over 10Hz in spectrum resource perception, severely impacting resource utilization. Therefore, there is an urgent need for a resource data perception method, device, and communication equipment with strong anti-interference capabilities. Summary of the Invention
[0003] In view of the above-mentioned shortcomings in the prior art, the present invention provides a resource data sensing method, apparatus and communication device to solve the problems in the background art.
[0004] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: This invention discloses a resource data sensing method, comprising the following steps: S1. The sensing node acquires multi-dimensional environmental parameters of the communication environment, generates an environmental feature vector based on the multi-dimensional environmental parameters, and identifies the type of interference in the resource data of the communication environment based on the trained interference classification model and the environmental feature vector. S2. Pre-build an anti-interference strategy library. The anti-interference strategy library stores at least one set of anti-interference algorithm combinations associated with each type of interference. The anti-interference algorithm combinations corresponding to each type of interference are sorted from high to low according to a preset interference suppression effect index. Based on the interference type identified in step S1, the anti-interference algorithm combination that ranks first in the anti-interference strategy library is selected. S3. Based on the dynamic parameters in the environmental feature vector, the sensing node uses the first-ranked anti-interference algorithm combination to collect target resource data, and preprocesses the collected target resource data to obtain noise-reduced resource data; S4. Evaluate the confidence level of the denoised resource data. If the confidence level reaches a preset threshold, output the resource data. If the confidence level does not reach the preset threshold, trigger strategy iteration, select other anti-interference algorithm combinations corresponding to the current interference type from the anti-interference strategy library in order of sorting, re-collect data, and update the interference classification model until resource data with a confidence level of 1 is output.
[0005] Further, in step S1, the sensing node is at least one of the following: an in-vehicle terminal in the Internet of Vehicles, an edge computing gateway in the Industrial Internet, or a distributed sensing unit of a 6G base station; the hardware configuration of the sensing node meets the requirements of a sampling rate ≥ 5kHz, a processor computing power ≥ 1TOPS, and supports at least three anti-interference algorithms running in parallel. The environmental feature vector includes real-time signal-to-noise ratio, interference signal strength, node moving speed, channel fading coefficient, and service delay sensitivity; The types of interference include narrowband interference, broadband interference, burst electromagnetic interference, and Doppler frequency offset interference.
[0006] Further, in step S1, the interference classification model is a convolutional neural network model. The training process of the convolutional neural network model includes collecting sample data containing environmental feature vectors and known interference types, dividing the training set and validation set in a 7:3 ratio, using the environmental feature vector as input and the interference type label as output, and training the convolutional neural network using the cross-entropy loss function until the accuracy of the convolutional neural network model on the validation set is ≥90%.
[0007] Further, in step S2, the anti-interference strategy library and the anti-interference algorithm combinations associated with each interference type include: The combination of wavelet packet denoising algorithm and notch filtering algorithm is used to target the narrowband interference. The wavelet packet denoising algorithm has a decomposition layer of 3 to 5 layers, and the center frequency error of the notch filtering algorithm does not exceed ±5kHz. The combination of adaptive FIR filtering algorithm and spectral subtraction is used to target the narrowband interference. The adaptive FIR filtering algorithm has a filter order of 32 and a cutoff frequency accuracy of no more than ±2kHz. The spectral subtraction has a noise estimation window size of 50ms and an attenuation coefficient of 0.5 to 0.8. The combination of adaptive beamforming algorithm and Kalman filtering algorithm is used to target the broadband interference, wherein the interference suppression ratio of the adaptive beamforming algorithm is ≥20dB. The combination of orthogonal frequency division multiplexing (OFDM) algorithm and frequency domain equalization algorithm is used to target the broadband interference, wherein the number of subcarriers in the OFDM algorithm is ≥1024 and the subcarrier spacing is 15kHz, and the equalization order of the frequency domain equalization algorithm is 16 and the convergence speed is ≤10ms. The combination of chaotic spread spectrum algorithm and energy accumulation detection algorithm is used to deal with the sudden electromagnetic interference, wherein the spread spectrum gain of the chaotic spread spectrum algorithm is ≥10dB, and the detection threshold of the energy accumulation detection algorithm is dynamically adjusted from -90dBm to -70dBm. A combination of frequency hopping communication algorithm and pulse compression algorithm is used to address the sudden electromagnetic interference, wherein the frequency hopping communication algorithm has a frequency hopping rate ≥ 500 hops / second and a frequency hopping bandwidth of 10MHz; the pulse compression algorithm has a compression ratio ≥ 10:1. The combination of frequency offset pre-compensation algorithm and cyclic prefix extension algorithm is used to address the Doppler frequency offset interference, wherein the compensation accuracy of the frequency offset pre-compensation algorithm does not exceed ±10Hz, and the extension length of the cyclic prefix extension algorithm is 1.5 to 2 times that of the original prefix. The combination of coherent demodulation algorithm and Doppler frequency offset tracking algorithm is used to target the Doppler frequency offset interference, wherein the carrier synchronization error of the coherent demodulation algorithm is ≤5Hz, the tracking range of the Doppler frequency offset tracking algorithm is ±50Hz, and the update frequency is 100Hz. The interference suppression effect index is a quantitative standard determined based on historical data, including the signal-to-noise ratio improvement, the amplitude of interference signal attenuation, and the degree of improvement in bit error rate.
[0008] Further, in step S3, the dynamic parameters include node moving speed and signal-to-noise ratio; the adjustment rule for the sampling frequency in the sensing parameters is as follows: when the node moving speed does not exceed 60km / h, the sampling frequency is 1kHz; when the node moving speed exceeds 60km / h but does not exceed 120km / h, the sampling frequency is increased stepwise to 2kHz to 5kHz; when the signal-to-noise ratio is less than 10dB, the oversampling mode is activated, and the sampling frequency in the oversampling mode is twice the default value.
[0009] Furthermore, in step S3, the target resource data includes at least one of the following: Spectrum resource data: spectrum occupancy rate, idle spectrum bandwidth, channel center frequency offset; Network node data: cell load, user plane latency, uplink and downlink throughput; Computing resource data: CPU utilization of computing nodes, memory usage, and task scheduling latency; Spatial resource data: node geographic location coordinates, coverage overlap, obstacle occlusion coefficient, and distance between multiple nodes.
[0010] Further, in step S3, the preprocessing includes outlier removal, smoothing filtering, and normalization. The outlier removal involves removing data that deviates from the mean by more than three times the standard deviation. The smoothing filtering uses a sliding window of 5 to 10 sampling points. The normalization process maps the data to the interval between 0 and 1.
[0011] Further, the confidence level of the denoised resource data in step S4 includes: detecting data continuity by calculating the deviation rate between the current resource data and the historical baseline data; when the deviation rate is ≤15% and there is no data loss or abnormal jump within 5 consecutive sampling periods, the confidence level of the resource data is determined to be ≥85%, and the resource data is output; otherwise, if the confidence level is <85%, the strategy iteration is triggered. The strategy iteration includes: if the confidence level is <85%, triggering the strategy iteration, selecting other anti-interference algorithm combinations corresponding to the current interference type from the anti-interference strategy library in sorted order to re-collect data; recording the environmental feature vector and strategy effect parameters during the iteration process, wherein the strategy effect parameters are the actual results of executing the current anti-interference algorithm combination, including the improvement value of resource data signal-to-noise ratio, data acquisition latency, and bit error rate; after every 10 triggering of the strategy iteration, updating the weight parameters of the interference classification model using a federated learning framework, only uploading the model gradient parameters during the update.
[0012] This invention discloses a resource data sensing device, applied to a sensing node, comprising: The environmental feature extraction unit is used to collect multi-dimensional environmental parameters and generate environmental feature vectors. Interference identification unit: Built-in trained convolutional neural network model, used to identify the current interference type; Strategy selection unit: stores an anti-interference strategy library, in which the anti-interference algorithm combinations associated with each interference type are sorted according to a preset interference suppression effect index, and are used to select the corresponding sorted anti-interference algorithm combination according to the interference type. Data acquisition and preprocessing unit: used to adjust sensing parameters based on dynamic parameters, acquire target resource data using the selected anti-interference algorithm combination, and perform preprocessing; Confidence assessment and optimization unit: used to assess the confidence of data and trigger strategy iteration, outputting optimized resource data; The resource data sensing device performs the method as described in any one of claims 1 to 8.
[0013] The present invention provides a communication device, comprising a memory, a processor, and a computer program stored in the memory, wherein when the processor executes the computer program, it implements the resource data sensing method as described in any one of claims 1 to 8.
[0014] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention identifies interference types in real time through an interference classification model. Compared with traditional single anti-interference algorithms, this invention can adapt to diverse interference such as narrowband, broadband, and Doppler frequency offset. It selects the best combination of anti-interference algorithms ranked by interference suppression effect index from a pre-built anti-interference strategy library to achieve precise binding between interference type and anti-interference strategy, ensuring maximum interference suppression effect. Compared with a fixed single algorithm, this invention reduces data distortion rate in complex environments. Through the closed-loop logic of "confidence assessment - strategy iteration - model update", it quantitatively evaluates data deviation rate and continuity. When the standard is not met, it automatically switches to the suboptimal algorithm and updates the model to ensure that the confidence of the output data is ≥85%, providing highly reliable data support for resource scheduling. 2. Existing technologies rely on fixed sensing parameters, which cause a sharp drop in performance in dynamic scenarios such as high-speed movement and low signal-to-noise ratio. This invention dynamically adjusts sensing parameters based on node movement speed and signal-to-noise ratio, enabling sensing behavior to adapt to environmental changes in real time. While ensuring data integrity, it reduces invalid sampling and can be stably applied to various complex scenarios such as vehicle terminals, industrial gateways, and 6G base stations. Attached Figure Description
[0015] Figure 1 This is a flowchart illustrating a resource data sensing method according to an embodiment of the present invention. Figure 2 This is a diagram illustrating the composition of a resource data sensing device according to an embodiment of the present invention. Figure 3 This is a structural diagram of a communication device according to an embodiment of the present invention. Detailed Implementation
[0016] To enable those skilled in the art to better understand the present invention, the technical solution of the present invention will be further described below in conjunction with the accompanying drawings and embodiments.
[0017] The accompanying drawings are for illustrative purposes only and are schematic diagrams, not actual images. They should not be construed as limiting the scope of this patent. To better illustrate the embodiments of the present invention, some parts in the drawings may be omitted, enlarged, or reduced, and do not represent the actual dimensions of the product. It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.
[0018] In the accompanying drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components. In the description of the present invention, it should be understood that if terms such as "upper," "lower," "left," "right," "inner," and "outer" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the terms used to describe positional relationships in the accompanying drawings are only for illustrative purposes and should not be construed as limiting the present patent. For those skilled in the art, the specific meaning of the above terms can be understood according to the specific circumstances.
[0019] In the description of this invention, unless otherwise explicitly specified and limited, the term "connection" or similar designation indicating a connection between components should be interpreted broadly. For example, it can refer to a fixed connection, a detachable connection, or an integral part; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can refer to the internal communication between two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.
[0020] Example 1: like Figure 1 As shown, the present invention provides a resource data sensing method, characterized by comprising the following steps: S1. The sensing node acquires multi-dimensional environmental parameters of the communication environment, generates an environmental feature vector based on the multi-dimensional environmental parameters, and identifies the interference type of the resource data of the communication environment based on the trained interference classification model and the environmental feature vector, making interference identification more accurate and improving the interference type identification accuracy by ≥15%. The sensing node is at least one of the following: an in-vehicle terminal in the Internet of Vehicles, an edge computing gateway in the Industrial Internet, or a distributed sensing unit of a 6G base station; the hardware configuration of the sensing node meets the requirements of a sampling rate ≥ 5kHz, stable operation at a working temperature of -40℃ to 85℃, processor computing power ≥ 1TOPS, and supports the parallel operation of at least 3 anti-interference algorithms, which are implemented through multi-threaded scheduling, with thread switching time ≤ 1ms; Among them, the environmental feature vector includes real-time signal-to-noise ratio (calculated by the ratio of signal power to noise power, with a sampling interval of 1ms), interference signal strength, node moving speed, channel fading coefficient, and service delay sensitivity; The types of interference include narrowband interference, broadband interference, burst electromagnetic interference, and Doppler frequency offset interference; S2. Pre-build an anti-interference strategy library. The anti-interference strategy library stores at least one set of anti-interference algorithm combinations associated with each type of interference. The anti-interference algorithm combinations corresponding to each type of interference are sorted from high to low according to a preset interference suppression effect index. Based on the interference type identified in step S1, the anti-interference algorithm combination that ranks first in the anti-interference strategy library is selected. The anti-interference strategy library is stored in an SQLite database and supports more than 100 query responses per second. S3. Based on the dynamic parameters in the environmental feature vector, the sensing node uses the first-ranked anti-interference algorithm combination to collect target resource data, and preprocesses the collected target resource data to obtain noise-reduced resource data; S4. Evaluate the confidence level of the denoised resource data. If the confidence level reaches a preset threshold, output the resource data. If the confidence level does not reach the preset threshold, trigger strategy iteration, select other anti-interference algorithm combinations corresponding to the current interference type from the anti-interference strategy library in order of sorting, re-collect data, and update the interference classification model until resource data with a confidence level of 1 is output.
[0021] Through a closed-loop process of "interference identification - strategy matching - dynamic adjustment - iterative optimization", the system achieves accurate perception of resource data in complex environments, improving data confidence and anti-interference response speed compared to existing fixed strategy methods.
[0022] In step S1, the interference classification model is a convolutional neural network (CNN) model. The training process of the CNN model includes collecting more than 10,000 sets of sample data containing environmental feature vectors and known interference types, with 70% used as the training set and 30% as the validation set. The CNN is trained using the environmental feature vector as input and the interference type label as output, employing the cross-entropy loss function until the accuracy of the CNN model on the validation set is ≥90%. By clearly defining the model structure and training standards, the accuracy and lightweight nature of the interference classification model are ensured. Compared with traditional machine learning models (such as SVM), the interference identification speed is improved by ≥2 times and the accuracy is improved by ≥10%.
[0023] In step S2, the anti-interference strategy library and the anti-interference algorithm combinations associated with each interference type include: The combination of wavelet packet denoising algorithm and notch filtering algorithm is used to target the narrowband interference. The wavelet packet denoising algorithm has a decomposition layer of 3 to 5 layers, and the center frequency error of the notch filtering algorithm does not exceed ±5kHz. The combination of adaptive FIR filtering algorithm and spectral subtraction is used to target the narrowband interference. The adaptive FIR filtering algorithm has a filter order of 32 and a cutoff frequency accuracy of no more than ±2kHz. The spectral subtraction has a noise estimation window size of 50ms and an attenuation coefficient of 0.5 to 0.8. The combination of adaptive beamforming algorithm and Kalman filtering algorithm is used to target the broadband interference, wherein the interference suppression ratio of the adaptive beamforming algorithm is ≥20dB. The combination of orthogonal frequency division multiplexing (OFDM) algorithm and frequency domain equalization algorithm is used to target the broadband interference, wherein the number of subcarriers in the OFDM algorithm is ≥1024 and the subcarrier spacing is 15kHz, and the equalization order of the frequency domain equalization algorithm is 16 and the convergence speed is ≤10ms. The combination of chaotic spread spectrum algorithm and energy accumulation detection algorithm is used to deal with the sudden electromagnetic interference, wherein the spread spectrum gain of the chaotic spread spectrum algorithm is ≥10dB, and the detection threshold of the energy accumulation detection algorithm is dynamically adjusted from -90dBm to -70dBm. A combination of frequency hopping communication algorithm and pulse compression algorithm is used to address the sudden electromagnetic interference, wherein the frequency hopping communication algorithm has a frequency hopping rate ≥ 500 hops / second and a frequency hopping bandwidth of 10MHz; the pulse compression algorithm has a compression ratio ≥ 10:1. The combination of frequency offset pre-compensation algorithm and cyclic prefix extension algorithm is used to address the Doppler frequency offset interference, wherein the compensation accuracy of the frequency offset pre-compensation algorithm does not exceed ±10Hz, and the extension length of the cyclic prefix extension algorithm is 1.5 to 2 times that of the original prefix. The combination of coherent demodulation algorithm and Doppler frequency offset tracking algorithm is used to target the Doppler frequency offset interference, wherein the carrier synchronization error of the coherent demodulation algorithm is ≤5Hz, the tracking range of the Doppler frequency offset tracking algorithm is ±50Hz, and the update frequency is 100Hz. The interference suppression effect index is a quantitative standard determined based on historical data, including the signal-to-noise ratio improvement, the amplitude of interference signal attenuation, and the degree of improvement in bit error rate.
[0024] Compared to a single algorithm, the anti-interference strategy library, combined with anti-interference algorithms associated with various interference types, ensures the targetedness and effectiveness of the anti-interference strategy.
[0025] In step S3, the dynamic parameters include node moving speed and signal-to-noise ratio (SNR). The adjustment rule for the sampling frequency among the sensing parameters is as follows: when the node moving speed does not exceed 60 km / h, the sampling frequency is 1 kHz; when the node moving speed exceeds 60 km / h but does not exceed 120 km / h, the sampling frequency is increased stepwise to 2 kHz to 5 kHz to ensure data continuity in high-speed moving scenarios; when the SNR is less than 10 dB, an oversampling mode is activated, and the sampling frequency in the oversampling mode is twice the default value, improving the data acquisition success rate in weak signal environments. By dynamically adjusting the sampling frequency, invalid sampling is reduced while ensuring data integrity. In high-speed vehicle networking scenarios, data continuity can be improved by ≥30%, and the acquisition success rate in industrial low SNR environments can be improved by ≥50%.
[0026] In step S3, the target resource data includes at least one of the following: Spectrum resource data: spectrum occupancy rate, idle spectrum bandwidth, channel center frequency offset; Network node data: cell load, user plane latency, uplink and downlink throughput; Computing resource data: CPU utilization of computing nodes, memory usage, and task scheduling latency; Spatial resource data: node geographic location coordinates, coverage overlap, obstacle occlusion coefficient, and distance between multiple nodes.
[0027] In step S3, the preprocessing includes outlier removal, smoothing filtering, and normalization. Outlier removal involves removing data that deviates from the mean by more than three times the standard deviation. The smoothing filtering uses a sliding window of 5 to 10 sampling points. The normalization process maps the data to the interval between 0 and 1.
[0028] In step S4, evaluating the confidence level of the denoised resource data includes: calculating the deviation rate between the current resource data and the historical baseline data (deviation rate = |current value - baseline value| / baseline value × 100%, historical baseline data is updated daily at 3:00 AM, calculated based on stable data from the past 7 days) to detect data continuity; when the deviation rate is ≤15% and there is no data loss or abnormal jumps within 5 consecutive sampling periods (jump amplitude >30% is considered abnormal), the confidence level of the resource data is determined to be ≥85%, and the resource data is output; otherwise, if the confidence level is <85%, the strategy iteration is triggered. The strategy iteration includes: if the confidence level is <85%, triggering the strategy iteration, selecting other anti-interference algorithm combinations corresponding to the current interference type from the anti-interference strategy library in sorted order to re-collect data; recording the environmental feature vector and strategy effect parameters during the iteration process, wherein the strategy effect parameters are the actual results of executing the current anti-interference algorithm combination, including the improvement value of resource data signal-to-noise ratio, data acquisition latency, and bit error rate; after every 10 triggering of the strategy iteration, updating the weight parameters of the interference classification model using a federated learning framework, only uploading the model gradient parameters during the update to avoid leakage of original data.
[0029] like Figure 2 As shown, the present invention provides a resource data sensing device applied to a sensing node, characterized in that it comprises: The environmental feature extraction unit is used to collect multi-dimensional environmental parameters and generate environmental feature vectors. Interference identification unit: Built-in trained convolutional neural network model, used to identify the current interference type; Strategy selection unit: stores an anti-interference strategy library, in which the anti-interference algorithm combinations associated with each interference type are sorted according to a preset interference suppression effect index, and are used to select the corresponding sorted anti-interference algorithm combination according to the interference type. Data acquisition and preprocessing unit: used to adjust sensing parameters based on dynamic parameters, acquire target resource data using the selected anti-interference algorithm combination, and perform preprocessing; Confidence assessment and optimization unit: used to assess the confidence of data and trigger strategy iteration, outputting optimized resource data; The resource data sensing device performs the method as described in any one of claims 1 to 8.
[0030] The method and apparatus are based on the same concept of the application. Since the methods and apparatus solve problems in similar ways, the implementation of the apparatus and methods can refer to each other, and the repeated parts will not be described again.
[0031] like Figure 3 As shown, the present invention provides a communication device, characterized in that it includes a memory, a processor, and a computer program stored in the memory, wherein when the processor executes the computer program, it implements the resource data sensing method as described in any one of claims 1 to 8.
[0032] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.
[0033] The above are merely embodiments of the present invention. The circuits, electronic components, and modules involved are all prior art, fully achievable by those skilled in the art, and require no further explanation. The scope of protection in this application does not involve improvements to the software and methods. Commonly known structures and characteristics in the solutions are not described in detail here. Those skilled in the art are aware of all common technical knowledge in the field prior to the application date or priority date, are aware of all prior art in that field, and have the ability to apply conventional experimental methods prior to that date. Those skilled in the art can, under the guidance of this application, improve and implement this solution based on their own capabilities. Some typical known structures or methods should not be obstacles for those skilled in the art to implement this application. It should be noted that those skilled in the art can make several modifications and improvements without departing from the structure of the present invention. These should also be considered within the scope of protection of the present invention, and will not affect the effectiveness of the implementation of the present invention or the practicality of the patent.
Claims
1. A resource data-aware method, characterized by, The method comprises the following steps: S1. The sensing node obtains multi-dimensional environment parameters of the communication environment, generates an environment feature vector according to the multi-dimensional environment parameters, and identifies the interference type of the resource data of the communication environment according to the environment feature vector based on a trained interference classification model; S2. A pre-constructed anti-interference strategy library stores at least one set of anti-interference algorithm combinations associated with each interference type, and the anti-interference algorithm combinations corresponding to each interference type are sorted from high to low according to a preset interference suppression effect index. According to the interference type identified in step S1, the anti-interference algorithm combination at the top of the sorting is selected from the anti-interference strategy library; S3. Adjust the sensing parameter based on the dynamic parameter in the environment feature vector, and collect the target resource data using the anti-interference algorithm combination at the top of the sorting. The collected target resource data is preprocessed to obtain noise-reduced resource data; S4. Evaluate the confidence of the noise-reduced resource data. If the confidence reaches a preset threshold, output the resource data; If the confidence does not reach the preset threshold, trigger strategy iteration, sequentially select other anti-interference algorithm combinations corresponding to the current interference type from the anti-interference strategy library according to the sorting, re-collect data, and update the interference classification model until the resource data with confidence up to standard is output.
2. A resource data awareness method as claimed in claim 1, characterized by, In step S1, The sensing node is at least one of the following: a vehicle terminal in a vehicle network, an edge computing gateway in an industrial internet, and a distributed sensing unit of a 6G base station; the hardware configuration of the sensing node satisfies a sampling rate ≥ 5 kHz, a processor computing power ≥ 1 TOPS, and supports at least 3 anti-interference algorithms running in parallel; The environment feature vector includes real-time signal-to-noise ratio, interference signal strength, node moving speed, channel fading coefficient, and business time delay sensitivity; The interference type includes narrowband interference, wideband interference, burst electromagnetic interference, and Doppler frequency offset interference.
3. The method of claim 1, wherein, In step S1, the interference classification model is a convolutional neural network model, and the training process of the convolutional neural network model includes collecting sample data containing environment feature vectors and known interference types, dividing the training set and the validation set in a ratio of 7:3, taking the environment feature vector as the input and the interference type label as the output, training the convolutional neural network using the cross-entropy loss function, and stopping when the accuracy of the convolutional neural network model on the validation set is ≥ 90%.
4. The method of claim 1, wherein, In step S2, The anti-interference algorithm combinations associated with each interference type in the anti-interference strategy library include: A combination of wavelet packet denoising algorithm and notch filter algorithm for the narrowband interference, wherein the decomposition layer number of the wavelet packet denoising algorithm is 3 to 5 layers, and the center frequency error of the notch filter algorithm is not more than ± 5 kHz; A combination of adaptive FIR filter algorithm and spectral subtraction for the narrowband interference, wherein the filter order of the adaptive FIR filter algorithm is 32 orders, the cutoff frequency accuracy is not more than ± 2 kHz, the noise estimation window size of the spectral subtraction is 50 ms, and the attenuation coefficient is 0.5 to 0.8; The adaptive beamforming algorithm is combined with the Kalman filtering algorithm to suppress the wideband interference, and the interference suppression ratio of the adaptive beamforming algorithm is greater than or equal to 20 dB. The orthogonal frequency division multiplexing algorithm is combined with the frequency domain equalization algorithm to suppress the wideband interference, the number of subcarriers of the orthogonal frequency division multiplexing algorithm is greater than or equal to 1024, the subcarrier spacing is 15 kHz, the equalization order of the frequency domain equalization algorithm is 16, and the convergence speed is less than or equal to 10 ms. The chaotic spread spectrum algorithm is combined with the energy accumulation detection algorithm to suppress the burst electromagnetic interference, the spread spectrum gain of the chaotic spread spectrum algorithm is greater than or equal to 10 dB, and the detection threshold of the energy accumulation detection algorithm is dynamically adjusted in the range of -90 dBm to -70 dBm. The frequency hopping communication algorithm is combined with the pulse compression algorithm to suppress the burst electromagnetic interference, the frequency hopping rate of the frequency hopping communication algorithm is greater than or equal to 500 hops per second, and the frequency hopping bandwidth is 10 MHz; the compression ratio of the pulse compression algorithm is greater than or equal to 10:
1. The frequency offset pre-compensation algorithm is combined with the cyclic prefix extension algorithm to suppress the Doppler frequency offset interference, the compensation accuracy of the frequency offset pre-compensation algorithm is not more than ±10 Hz, and the extension length of the cyclic prefix extension algorithm is 1.5 to 2 times the original prefix. The coherent demodulation algorithm is combined with the Doppler frequency offset tracking algorithm to suppress the Doppler frequency offset interference, the carrier synchronization error of the coherent demodulation algorithm is less than or equal to 5 Hz, the tracking range of the Doppler frequency offset tracking algorithm is ±50 Hz, and the update frequency is 100 Hz. The interference suppression effect index is a quantitative standard determined based on historical data, including signal-to-noise ratio improvement value, interference signal attenuation amplitude, and error code rate improvement degree.
5. The method of claim 1, wherein, In step S3, the dynamic parameters include node moving speed and signal-to-noise ratio; and the adjustment rule of the sampling frequency in the perception parameter is that when the node moving speed is not more than 60 km / h, the sampling frequency is 1 kHz; when the node moving speed is more than 60 km / h and not more than 120 km / h, the sampling frequency is stepped up to 2 kHz to 5 kHz; and when the signal-to-noise ratio is less than 10 dB, an oversampling mode is started, and the sampling frequency in the oversampling mode is twice the default value.
6. The method of claim 1, wherein, In step S3, the target resource data includes at least one of the following: Spectrum resource data: spectrum occupancy, idle spectrum bandwidth, channel center frequency offset; Network node data: cell load, user plane latency, uplink and downlink throughput; Computing power resource data: CPU utilization of computing power node, memory occupancy, task scheduling latency; Spatial resource data: node geographic position coordinates, coverage range overlap degree, obstacle shielding coefficient, and multi-node spacing.
7. The method of claim 1, wherein, In step S3, the preprocessing includes outlier rejection, smoothing filtering, and normalization processing.
8. The method of claim 1, wherein, In step S4, evaluating the confidence of the noise-reduced resource data comprises: detecting data continuity by calculating the deviation rate of the current resource data from the historical baseline data; when the deviation rate is ≤ 15% and there is no loss or abnormal jump in data in the last 5 consecutive sampling periods, it is determined that the confidence of the resource data is ≥ 85%, the resource data is output, otherwise the confidence is < 85%, and the policy iteration is triggered; The policy iteration comprises: if the confidence is < 85%, triggering policy iteration, selecting other anti-interference algorithm combinations corresponding to the current interference type from the anti-interference strategy library in order according to the ranking to re-collect data; recording the environmental feature vector and the policy effect parameter in the iteration process, the policy effect parameter being the actual result of executing the anti-interference algorithm combination, including resource data signal-to-noise ratio improvement value, data collection delay, and error rate; every 10 times the policy iteration is triggered, the weight parameters of the interference classification model are updated using a federated learning framework, and only model gradient parameters are uploaded during the updating. 9.A resource data awareness device applied to a sensing node, characterized in that, Comprise: An environmental feature extraction unit configured to collect multi-dimensional environmental parameters to generate an environmental feature vector; An interference identification unit comprising a trained convolutional neural network model, configured to identify the current interference type; A policy selection unit configured to store an anti-interference strategy library, wherein the anti-interference algorithm combinations associated with each interference type in the anti-interference strategy library are ranked according to a preset interference suppression effect index, and the anti-interference algorithm combinations corresponding to the ranking are selected according to the interference type; A data collection preprocessing unit configured to adjust perception parameters based on dynamic parameters, collect target resource data using the selected anti-interference algorithm combination, and perform preprocessing; A confidence evaluation and optimization unit configured to evaluate data confidence and trigger policy iteration, and output optimized resource data; The resource data perception device executes the method according to any one of claims 1-8.
10. A communication device, characterized by A computer program product comprising a memory, a processor, and a computer program stored on the memory, wherein the processor executes the computer program to implement the resource data perception method according to any one of claims 1-8.