5G signal dynamic adjustment method and device based on environmental perception
Through environmental perception technology and intelligent algorithms, a model of environmental parameters and signal transmission influence is built, and 5G signals are dynamically adjusted, which solves the problems of signal adjustment lag and inefficiency in traditional methods, and realizes the stability of 5G signals in different environments and Efficient transmission.
Patent Information
- Application Number
- CN202410775876.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-17
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-06-17
AI Technical Summary
Traditional 5G signal adjustment methods are difficult to cope with complex environmental changes, resulting in signal adjustment lag and inefficiency, and it is impossible to dynamically adjust signal transmission strategies based on real-time environmental data.
Using an environmental perception-based method, a mapping relationship model of environmental parameters and signal transmission impact is constructed through multiple sensors, and a multi-stage joint convolution classifier and adaptive ephemeral algorithm is used to generate and optimize signal adjustment strategies. , realize dynamic adjustment of 5G signals.
It improves the transmission stability and adaptability of 5G signals, ensures high-quality transmission of signals under different environmental conditions, and enhances the network's response capabilities and policy accuracy.
Smart Images

Figure CN118488473B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of signal processing technology, and in particular to a method and device for dynamically adjusting 5G signals based on environmental perception. Background Art
[0002] Environmental changes within the 5G network coverage area have a significant impact on the quality of 5G signal transmission. For example, fluctuations in environmental parameters such as temperature, humidity, air pressure and light will cause changes in key performance indicators such as 5G signal strength, latency and bit error rate, thus affecting users' network experience and communication quality.
[0003] Traditional 5G signal adjustment methods usually rely on static signal configuration strategies, which are difficult to cope with complex environmental changes. In addition, these methods often ignore the collection and analysis of real-time environmental data and cannot dynamically adjust signal transmission strategies according to real-time environmental changes, resulting in lag and inefficiency in signal adjustment. Summary of the invention
[0004] The present application provides a method and device for dynamically adjusting 5G signals based on environmental perception, which is used to dynamically adjust 5G signals in an environmental perception manner to improve the transmission stability of 5G signals.
[0005] In a first aspect, the present application provides a 5G signal dynamic adjustment method based on environment perception, and the 5G signal dynamic adjustment method based on environment perception includes:
[0006] Through multiple sensors, historical data of the 5G network coverage area is collected and preprocessed to obtain historical environmental data sets and historical 5G signal data for each sensor;
[0007] Performing feature analysis and 5G signal transmission impact calculation on the historical environmental data set and the historical 5G signal data, and constructing a mapping relationship model between each environmental parameter and the 5G signal transmission impact;
[0008] Acquire multiple real-time environmental data sets and real-time 5G signal data in the 5G network coverage area, and perform feature extraction to obtain multiple real-time environmental parameter features and real-time 5G signal features;
[0009] By using the mapping relationship model, feature vector mapping is performed on the multiple real-time environmental parameter features and the real-time 5G signal features to generate multiple environment-signal relationship feature vectors;
[0010] Inputting the plurality of environment-signal relationship feature vectors into a preset multi-level joint convolution classifier to perform signal adjustment strategy analysis, and outputting a plurality of initial signal adjustment strategies;
[0011] The multiple initial signal adjustment strategies are optimized through strategy feedback using an adaptive ephemera algorithm to obtain multiple target signal adjustment strategies, and the 5G signal is dynamically adjusted in the 5G network coverage area using the multiple target signal adjustment strategies.
[0012] In a second aspect, the present application provides a 5G signal dynamic adjustment device based on environment perception, and the 5G signal dynamic adjustment device based on environment perception includes:
[0013] The acquisition module is used to collect and preprocess historical data of the 5G network coverage area through various sensors to obtain historical environmental data sets and historical 5G signal data for each sensor;
[0014] A construction module is used to perform feature analysis and 5G signal transmission impact calculation on the historical environmental data set and the historical 5G signal data, and to construct a mapping relationship model between each environmental parameter and the 5G signal transmission impact;
[0015] A feature extraction module is used to obtain multiple real-time environmental data sets and real-time 5G signal data in the 5G network coverage area, and perform feature extraction to obtain multiple real-time environmental parameter features and real-time 5G signal features;
[0016] A vector mapping module, used to perform feature vector mapping on the multiple real-time environmental parameter features and the real-time 5G signal features through the mapping relationship model to generate multiple environment-signal relationship feature vectors;
[0017] A strategy analysis module, used for inputting the plurality of environment-signal relationship feature vectors into a preset multi-level joint convolution classifier to perform signal adjustment strategy analysis and output a plurality of initial signal adjustment strategies;
[0018] A feedback optimization module is used to perform strategy feedback optimization on the multiple initial signal adjustment strategies through an adaptive ephemera algorithm to obtain multiple target signal adjustment strategies, and dynamically adjust the 5G signal in the 5G network coverage area through the multiple target signal adjustment strategies.
[0019] In the technical solution provided by this application, by deploying a variety of sensors, real-time monitoring and data collection of temperature, humidity, air pressure and light environment parameters in the 5G network coverage area are realized. The sensor transmits the collected data to the central processing unit in real time through a low-latency network transmission protocol, which effectively ensures the real-time and accuracy of the environmental data. The ability of real-time environmental perception enables the 5G network to respond to environmental changes in a timely manner, thereby improving the accuracy and dynamic adaptability of signal adjustment. By analyzing historical environmental data and historical 5G signal data, a mapping relationship model between environmental parameters and 5G signal transmission effects is constructed. The mapping relationship model can accurately describe the specific impact of each environmental parameter on signal transmission, and after obtaining real-time environmental data, the key real-time environmental parameter features and 5G signal features are extracted through feature extraction technology. By inputting the extracted environment-signal relationship feature vector into a multi-level joint convolution classifier, a signal adjustment strategy analysis is performed to generate an initial signal adjustment strategy. The adaptive mayfly algorithm is used to perform strategy feedback optimization on the initial signal adjustment strategy to obtain the optimized target signal adjustment strategy, realizing the automatic generation and optimization of the signal adjustment strategy. The accuracy and adaptability of the strategy are also improved through intelligent algorithms, thereby ensuring the stability and high-quality transmission of 5G signals under different environmental conditions. BRIEF DESCRIPTION OF THE DRAWINGS
[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.
[0021] Figure 1 This is a schematic diagram of an embodiment of a method for dynamically adjusting a 5G signal based on environment perception in an embodiment of the present application;
[0022] Figure 2 This is a schematic diagram of an embodiment of a 5G signal dynamic adjustment device based on environmental perception in an embodiment of the present application. DETAILED DESCRIPTION
[0023] The embodiments of the present application provide a method and device for dynamic adjustment of 5G signals based on environmental perception. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0024] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 , an embodiment of the 5G signal dynamic adjustment method based on environment perception in the embodiment of the present application includes:
[0025] Step S101: collect and preprocess historical data of the 5G network coverage area through multiple sensors to obtain historical environmental data sets and historical 5G signal data for each sensor;
[0026] It is understandable that the execution subject of the present application can be a 5G signal dynamic adjustment device based on environment perception, or a terminal or a server, which is not limited here. The embodiment of the present application is described by taking the server as the execution subject as an example.
[0027] Specifically, multiple sensors are configured in the 5G network coverage area, and these sensors include temperature sensors, humidity sensors, air pressure sensors, and light sensors. Through multiple sensors, historical environmental data are collected for the 5G network coverage area respectively, and the initial environmental data set of each sensor is obtained. Data denoising and data cleaning are performed on the initial environmental data set of each sensor respectively, and the historical environmental data set of each sensor is obtained, wherein the historical environmental data set includes: historical temperature data set, historical humidity data set, historical air pressure data set, and historical light data set. Through the temperature sensors configured in the 5G network coverage area, the temperature changes in the area are continuously monitored to obtain the initial temperature data set. Data denoising is performed on the initial temperature data to remove abnormal data points and interference data, and data cleaning is performed to process missing values and erroneous data to obtain a complete and accurate historical temperature data set. Through humidity sensors, air pressure sensors, and light sensors, the humidity, air pressure, and light intensity of the 5G network coverage area are monitored respectively, and the corresponding initial humidity data set, initial air pressure data set, and initial light data set are obtained. Through the same data denoising and data cleaning steps, these initial data sets are processed respectively, and finally the historical humidity data set, historical air pressure data set, and historical light data set are obtained. At the same time, signal detection is performed on the 5G network coverage area to obtain initial 5G signal data. The initial 5G signal data is converted into frequency domain signals, and the time domain signal data is converted into frequency domain signal data through Fourier transform or other frequency domain conversion methods to obtain historical 5G signal data.
[0028] Step S102: Perform feature analysis on the historical environmental data set and the historical 5G signal data and calculate the impact of 5G signal transmission, and build a mapping relationship model between each environmental parameter and the impact of 5G signal transmission;
[0029] Specifically, curve fitting is performed for each historical environmental data set. By selecting an appropriate fitting method, such as polynomial fitting or exponential fitting, discrete historical environmental data points are converted into continuous historical data curves to better reflect the trend of environmental parameters changing over time. Curve feature points are identified for the historical data curves of each environmental parameter. By adopting algorithms such as peak detection and inflection point detection, representative curve feature points are extracted from the fitted historical data curve to form a curve feature set for each environmental parameter. Signal transmission index analysis is performed on historical 5G signal data to obtain a signal transmission index set, which includes signal strength, signal-to-noise ratio, transmission rate, etc. The Pearson correlation coefficient is used to calculate the 5G signal transmission influence coefficient between the curve feature set and the signal transmission index set of each environmental parameter. The Pearson correlation coefficient is a commonly used statistical method for measuring the degree of linear correlation between two variables. By calculating the correlation coefficient between the curve feature set and the signal transmission index set of each environmental parameter, the transmission influence coefficient between each environmental parameter and the 5G signal transmission influence is obtained, and these coefficients reflect the specific degree of influence of different environmental parameters on 5G signal transmission. According to the transmission impact coefficient, the signal transmission impact weight modeling is performed for each environmental parameter and the impact on 5G signal transmission, and the mapping relationship model between each environmental parameter and the impact on 5G signal transmission is obtained. These models accurately reflect the impact of different environmental parameters on 5G signal transmission.
[0030] Step S103: Acquire multiple real-time environmental data sets and real-time 5G signal data in the 5G network coverage area, and perform feature extraction to obtain multiple real-time environmental parameter features and real-time 5G signal features;
[0031] Specifically, multiple real-time environmental data sets and real-time 5G signal data in the 5G network coverage area are obtained through multiple sensors. These data include real-time environmental parameters such as temperature, humidity, air pressure and light, as well as information such as the strength, quality and transmission rate of the 5G signal. The real-time environmental data set is linearly transformed. By applying the linear transformation method, the original environmental data is converted into multiple linear feature sets, each of which contains multiple linear features. The covariance matrix elements of these linear features are calculated. The covariance matrix is an important tool for describing the relationship between multiple variables. By calculating the covariance matrix elements of the linear features, the covariance matrix of each linear feature set is constructed. The covariance matrix can reflect the correlation and mutual influence between different linear features. According to the covariance matrix, the correlation between multiple linear features is calculated to obtain the feature correlation of each linear feature set. The feature correlation reflects the relationship and dependence between different environmental parameters. By analyzing the feature correlation, multiple eigenvalues of each linear feature set are calculated, and these eigenvalues reflect the importance and weight of different environmental parameters in the overall data structure. Based on the calculated eigenvalues, the eigenvector of each linear feature set is generated. In order to unify and standardize these feature vectors, the principal component normalization is performed to obtain the principal component feature weights. Multiple real-time environmental data sets are multiplied by the principal component feature weights to obtain multiple target products. The target product reflects the comprehensive role and influence of different environmental parameters in the feature extraction process. By adding multiple target products, multiple real-time environmental parameter features are obtained. The real-time 5G signal data is converted into a frequency domain signal, and the time domain signal data is converted into frequency domain 5G signal data through Fourier transform or other frequency domain conversion methods. Based on the frequency domain signal data, feature recognition is performed to extract key signal features.
[0032] Step S104: Perform feature vector mapping on multiple real-time environmental parameter features and real-time 5G signal features through a mapping relationship model to generate multiple environment-signal relationship feature vectors;
[0033] Specifically, feature normalization and feature sequence conversion are performed on multiple real-time environmental parameter features, and each environmental parameter feature is normalized to the same scale range, and the dimension difference between different features is eliminated to obtain a first environmental parameter feature sequence. Similarly, feature normalization and feature sequence conversion are performed on real-time 5G signal features to ensure that the 5G signal features have a consistent scale during feature extraction and analysis, and a first 5G signal feature sequence is obtained. The first environmental parameter feature sequence and the first 5G signal feature sequence are initialized and spliced to obtain an initial spliced feature sequence. By arranging each environmental parameter feature and the corresponding 5G signal feature in sequence, the initial spliced feature sequence can fully reflect the current environmental conditions and 5G signal status. The mapping relationship model is used to perform feature weighting on each real-time environmental parameter feature and real-time 5G signal feature in the initial spliced feature sequence. Different features are weighted by the preset weight coefficients in the mapping relationship model to reflect the importance and influence of different environmental parameters and signal features in signal transmission, and multiple weighted fusion feature sequences are obtained. Feature vector mapping is performed on multiple weighted fusion feature sequences. Through the mapping relationship model, the feature information in the weighted fusion feature sequence is converted into an environment-signal relationship feature vector. The feature vector can accurately reflect the specific impact of the current environmental conditions on 5G signal transmission. The feature vector mapping process includes linear or nonlinear transformation of each feature in the weighted fusion feature sequence, so that the generated environment-signal relationship feature vector can better adapt to the subsequent signal adjustment algorithm and strategy optimization.
[0034] Step S105, inputting multiple environment-signal relationship feature vectors into a preset multi-level joint convolution classifier to perform signal adjustment strategy analysis, and outputting multiple initial signal adjustment strategies;
[0035] Specifically, multiple environment-signal relationship feature vectors are input into a preset multi-level joint convolution classifier. The classifier is composed of multiple convolution modules, each of which is connected to a pooling layer, and each pooling layer is connected to a classifier. During the processing, an environment-signal relationship feature vector is received by each convolution module, and a convolution operation is performed on the feature vector. The convolution operation calculates the input feature vector point by point through multiple filters to generate multiple first feature vectors. Each convolution module uses different filters to extract local features and patterns in the feature vector to obtain a richer feature representation. Pooling operations are performed on multiple first feature vectors through the pooling layer. The pooling operation reduces the dimension of the feature vector through a downsampling method, such as maximum pooling or average pooling, retains important features and suppresses noise, and obtains multiple second feature vectors. Signal adjustment strategy decisions are made on multiple second feature vectors through the classifier. Each classifier uses a pre-trained model to classify according to the input feature vector and outputs multiple initial signal adjustment strategies. These strategies include base station transmit power adjustment strategy, spectrum allocation adjustment strategy, antenna direction adjustment strategy and transmission mode switching strategy. The base station transmit power adjustment strategy aims to dynamically adjust the base station transmit power according to the environment and signal characteristics to optimize signal coverage and transmission quality; the spectrum allocation adjustment strategy reasonably allocates spectrum resources according to the real-time environment and network load conditions to improve spectrum utilization efficiency and network performance; the antenna direction adjustment strategy adjusts the antenna direction to make the signal coverage range and transmission path more in line with the current environmental requirements, thereby improving the stability and efficiency of signal transmission; the transmission mode switching strategy switches between different transmission modes to adapt to different environmental conditions and business needs, ensuring the reliability and efficiency of signal transmission.
[0036] Step S106: Perform strategy feedback optimization on multiple initial signal adjustment strategies through an adaptive ephemera algorithm to obtain multiple target signal adjustment strategies, and dynamically adjust the 5G signal in the 5G network coverage area through the multiple target signal adjustment strategies.
[0037] Specifically, the 5G signal is adjusted in the 5G network coverage area based on multiple initial signal adjustment strategies, and the signal status data of the area is obtained. These signal status data include key indicators such as signal strength, coverage, and transmission rate, which reflect the effect of the current signal adjustment strategy and the operation status of the 5G network. Through the adaptive ephemera algorithm, the multiple initial signal adjustment strategies are initialized according to the acquired signal status data to generate an initial population. The initial population includes multiple individuals, each of which represents a group of signal adjustment strategies. Each individual in the initial population contains strategy parameters such as base station transmit power adjustment, spectrum allocation, antenna direction adjustment, and transmission mode switching. These individuals will serve as the basic unit in the optimization process of the adaptive ephemera algorithm. The fitness value of each individual is calculated separately. The fitness value is an important indicator for evaluating individual performance and reflects the effect of each signal adjustment strategy group in actual application. The fitness value is calculated based on the signal status data, including aspects such as the improvement of signal strength, the expansion of coverage, and the improvement of transmission rate. According to the fitness value, multiple individuals are divided into populations, and individuals with higher fitness values are divided into target populations. The target population is optimized through crossover and mutation operations. New individuals are generated by exchanging some of the policy parameters of two individuals, thereby introducing new features and possible better solutions. The mutation operation is to make small random adjustments to the policy parameters of individuals to increase the diversity of the population and explore new solution spaces. Through these operations, multiple new individuals, namely the second individuals, are generated. Multiple second individuals are optimized. Through iterative calculations, the individual with the highest fitness value is found, which is called the optimal individual. The optimal individual represents the signal adjustment strategy group with the best effect under the current environmental conditions. Multiple target signal adjustment strategies corresponding to the optimal individual are obtained, and the 5G signal is dynamically adjusted in the 5G network coverage area through the target signal adjustment strategy. These adjustment strategies dynamically optimize the base station transmission power, spectrum allocation, antenna direction and transmission mode, so that the 5G network can maintain an efficient and stable operating state under different environmental conditions, ensuring that users can get a good network experience.
[0038] In the embodiment of the present application, by deploying a variety of sensors, real-time monitoring and data collection of temperature, humidity, air pressure and light environment parameters in the 5G network coverage area are realized. The sensor transmits the collected data to the central processing unit in real time through a low-latency network transmission protocol, which effectively ensures the real-time and accuracy of the environmental data. The ability of real-time environmental perception enables the 5G network to respond to environmental changes in a timely manner, thereby improving the accuracy and dynamic adaptability of signal adjustment. By analyzing historical environmental data and historical 5G signal data, a mapping relationship model between environmental parameters and 5G signal transmission effects is constructed. The mapping relationship model can accurately describe the specific impact of each environmental parameter on signal transmission, and after obtaining real-time environmental data, the key real-time environmental parameter features and 5G signal features are extracted through feature extraction technology. By inputting the extracted environment-signal relationship feature vector into a multi-level joint convolution classifier, a signal adjustment strategy analysis is performed to generate an initial signal adjustment strategy. The adaptive mayfly algorithm is used to perform strategy feedback optimization on the initial signal adjustment strategy to obtain the optimized target signal adjustment strategy, realizing the automatic generation and optimization of the signal adjustment strategy. The accuracy and adaptability of the strategy are also improved through intelligent algorithms, thereby ensuring the stability and high-quality transmission of 5G signals under different environmental conditions.
[0039] In a specific embodiment, the process of executing step S101 may specifically include the following steps:
[0040] (1) Deploy a variety of sensors in the 5G network coverage area, including temperature sensors, humidity sensors, air pressure sensors, and light sensors;
[0041] (2) Using a variety of sensors, collect historical environmental data for the 5G network coverage area to obtain the initial environmental data set for each sensor;
[0042] (3) Denoising and cleaning the initial environmental data set of each sensor to obtain the historical environmental data set of each sensor, where the historical environmental data set includes: historical temperature data set, historical humidity data set, historical air pressure data set, and historical light data set;
[0043] (4) Perform signal detection on the 5G network coverage area to obtain initial 5G signal data, and perform frequency domain signal conversion on the initial 5G signal data to obtain historical 5G signal data.
[0044] Specifically, when configuring sensors, consider the distribution of sensors to ensure that the entire 5G network area can be covered. Sensors can be installed near base stations, around antennas, and at the edge of signal coverage. Each sensor needs to be connected to the central data processing system to transmit the collected data in real time. Through these sensors, historical environmental data is collected for the 5G network coverage area to obtain the initial environmental data set for each sensor. During the historical environmental data collection process, the temperature sensor records the changes in air temperature over a period of time to generate an initial temperature data set; the humidity sensor records the changes in air humidity levels to generate an initial humidity data set; the pressure sensor records the changes in atmospheric pressure to generate an initial pressure data set; and the light sensor records the changes in light intensity to generate an initial light data set. Data denoising and data cleaning are performed on the initial environmental data set of each sensor to ensure the accuracy and reliability of the data. Data denoising is to remove noise and outliers in the data, which may be caused by sensor failure, external interference or other accidental factors. Data cleaning is to deal with missing values and erroneous data in the data, and ensure the integrity and consistency of the data through methods such as interpolation, filling or deletion. After data denoising and data cleaning, the historical environmental data sets of each sensor are obtained. The historical temperature data set contains all temperature data recorded by the temperature sensor over a period of time, the historical humidity data set contains humidity data recorded by the humidity sensor, the historical air pressure data set contains atmospheric pressure data recorded by the air pressure sensor, and the historical light data set contains light intensity data recorded by the light sensor. Signal detection is performed on the 5G network coverage area to obtain initial 5G signal data. Signal detection is performed on 5G base stations and user terminal devices. The base station can record parameters such as signal strength, signal-to-noise ratio, and transmission rate, and the user terminal device can feedback the received signal quality. Frequency domain signal conversion is performed on the initial 5G signal data. The signal data in the time domain is converted into frequency domain data through Fourier transform, so as to analyze the characteristics of the signal at different frequencies and help identify the interference and attenuation characteristics in the signal. For example, by analyzing the historical temperature data set and the historical 5G signal data, it can be found that the increase in temperature will lead to increased signal attenuation, so that the base station's transmission power needs to be increased in hot weather to compensate for signal loss. By analyzing historical humidity data sets and historical 5G signal data, it can be found that high humidity environments increase the multipath effect of signals, so it is necessary to adjust the spectrum allocation strategy to reduce interference. Dynamically adjust 5G signals according to different environmental conditions to optimize network performance.
[0045] In a specific embodiment, the process of executing step S102 may specifically include the following steps:
[0046] (1) Perform curve fitting on the historical environmental data sets to obtain the historical data curve of each environmental parameter;
[0047] (2) Identify the characteristic points of the historical data curve of each environmental parameter to obtain the curve feature set of each environmental parameter;
[0048] (3) Analyze the signal transmission index of historical 5G signal data to obtain a set of signal transmission indexes;
[0049] (4) Using the Pearson correlation coefficient, the 5G signal transmission influence coefficient between the curve feature set of each environmental parameter and the signal transmission index set is calculated respectively, and the transmission influence coefficient between each environmental parameter and the 5G signal transmission influence is obtained;
[0050] (5) Based on the transmission impact coefficient, the signal transmission impact weight modeling is performed on the impact of each environmental parameter on 5G signal transmission to obtain a mapping relationship model between each environmental parameter and the impact of 5G signal transmission.
[0051] Specifically, curve fitting is performed on the historical environmental data sets to obtain the historical data curves of each environmental parameter. The discrete environmental data is converted into a continuous function expression to more accurately describe the changing trend of environmental parameters over time. For example, assuming that the historical temperature data set is represented as ,in Indicates The temperature value at a time point, Indicates time points. Using the polynomial fitting method, the temperature data is represented as a polynomial function:
[0052] ;
[0053] in, is the fitting coefficient, which is solved by the least squares method. Polynomial fitting is performed on humidity, air pressure and light data to obtain historical humidity curves, historical air pressure curves and historical light curves respectively. After obtaining the historical data curves of each environmental parameter, it is necessary to identify the curve feature points of these curves to extract representative feature points and form a curve feature set. Curve feature points include local extreme points, inflection points and trend change points, etc. These feature points can reflect the key changes of environmental parameters. For example, for the temperature curve, the local maximum and minimum points are found by the first-order derivative, and the inflection point is found by the second-order derivative. The feature points constitute the curve feature set of temperature. Similar methods can be used for humidity, air pressure and light curves to obtain the corresponding curve feature sets respectively. The signal transmission index analysis is performed on the historical 5G signal data to obtain a signal transmission index set. The signal transmission index includes signal strength, signal-to-noise ratio, transmission rate, etc. These indicators can fully reflect the transmission performance of 5G signals. For example, assume that the historical 5G signal data is represented as {Si, ti}. Where Si represents the signal strength at the i-th time point. The signal transmission index at each time point is calculated by statistical analysis and frequency domain analysis methods to form a signal transmission index set. The Pearson correlation coefficient is used to calculate the correlation between the curve feature set of each environmental parameter and the signal transmission index set. The Pearson correlation coefficient is a statistic that measures the degree of linear correlation between two variables. Its calculation formula is:
[0054] ;
[0055] in, and Respectively represent the environmental parameter characteristic value and the signal transmission index value, and are their means respectively. By calculating the Pearson correlation coefficient between different environmental parameter feature sets and signal transmission index sets, the transmission influence coefficient between each environmental parameter and the 5G signal transmission effect is obtained. These coefficients reflect the specific degree of influence of environmental parameters on 5G signal transmission. Based on the transmission influence coefficient, the signal transmission influence weight modeling is performed for each environmental parameter and the influence on 5G signal transmission, and the mapping relationship model between each environmental parameter and the influence on 5G signal transmission is obtained. Assume that the transmission influence coefficient is expressed as , respectively represent the influence coefficients of temperature, humidity, air pressure and light on signal transmission, and the following mapping relationship model can be established
[0056] ;
[0057] in, Indicates 5G signal transmission performance, Represent the characteristic values of temperature, humidity, air pressure and light respectively are the corresponding weight coefficients, which can be adjusted by the transmission influence coefficient so that the model can accurately reflect the impact of environmental parameters on signal transmission. For example, by analyzing the environmental and signal data of a certain 5G network coverage area, it is found that temperature has the most significant impact on signal transmission, while humidity and air pressure have little impact, and light has almost no effect on signal transmission. The calculated Pearson correlation coefficient is Based on these coefficients, the weight coefficients are set to Substituting these coefficients into the mapping relationship model, we can get:
[0058] ;
[0059] This model shows that temperature changes have the greatest impact on 5G signal transmission performance. Therefore, under high temperature conditions, special attention should be paid to the impact of temperature on the signal, and corresponding adjustment measures may need to be taken, such as increasing the transmission power of the base station to compensate for signal loss.
[0060] In a specific embodiment, the process of executing step S103 may specifically include the following steps:
[0061] (1) Obtain multiple real-time environmental data sets and real-time 5G signal data in the 5G network coverage area;
[0062] (2) performing linear transformation on multiple real-time environment data sets respectively to obtain multiple linear feature sets, each of which contains multiple linear features;
[0063] (3) Calculate the covariance matrix elements of multiple linear features, and construct the covariance matrix of each linear feature set based on the covariance matrix elements;
[0064] (4) Calculate the correlation between multiple linear features according to the covariance matrix to obtain the feature correlation of each linear feature set, and calculate multiple eigenvalues corresponding to the multiple linear features according to the feature correlation;
[0065] (5) Generate a feature vector for each linear feature set based on multiple eigenvalues, and perform principal component normalization on the feature vector to obtain the principal component feature weight;
[0066] (6) Multiplying the plurality of real-time environment data sets by the principal component feature weights respectively to obtain a plurality of target products, and performing sum operation on the plurality of target products to obtain a plurality of real-time environment parameter features;
[0067] (7) Perform frequency domain signal conversion on the real-time 5G signal data to obtain frequency domain 5G signal data, and perform feature recognition on the frequency domain 5G signal data to obtain real-time 5G signal features.
[0068] Specifically, by arranging temperature sensors, humidity sensors, air pressure sensors, and light sensors in the area covered by the 5G network, environmental data can be collected in real time. These data include parameters such as temperature, humidity, air pressure, and light, which can fully reflect the environmental conditions of the area. At the same time, real-time 5G signal data such as signal strength, signal-to-noise ratio, and transmission rate are obtained through base stations and user terminal devices. Linear transformation is performed on multiple real-time environmental data sets to obtain multiple linear feature sets. By performing a linear transformation on the original data, it is mapped to a new feature space, thereby simplifying the subsequent analysis process. Calculate the covariance matrix elements of multiple linear features, and construct the covariance matrix of each linear feature set based on the covariance matrix elements. The covariance matrix is a matrix used to describe the relationship between different variables in a data set, reflecting the linear correlation between variables. Assume that the linear feature set is , the elements of the covariance matrix can be expressed as:
[0069] ;
[0070] in, Features and The covariance between and For the The characteristic value of the data point, and is the mean of the feature, is the number of samples. By calculating the elements of the covariance matrix, the covariance matrix of each linear feature set is constructed. According to the covariance matrix, the correlation between multiple linear features is calculated to obtain the feature correlation of each linear feature set. The feature correlation can be expressed by the standardized form of the covariance matrix, that is, the correlation coefficient matrix. The calculation formula of the correlation coefficient is:
[0071] ;
[0072] in, Features and The correlation coefficient between and Characteristics and The standard deviation of . By calculating the feature correlation, multiple eigenvalues corresponding to multiple linear features are calculated. Based on the eigenvalues, the eigenvectors of each linear feature set are generated, and the eigenvectors are normalized to obtain the principal component feature weights. Principal component analysis is a commonly used dimensionality reduction method that projects data into the principal component space by calculating eigenvectors and eigenvalues. Assume that the eigenvalues are , the eigenvector is
[0073] ;
[0074] in, is the normalized feature vector. Through normalization, the principal component feature weights are obtained, which reflect the importance of each feature in the principal component space. Multiple real-time environmental data sets are multiplied by the principal component feature weights to obtain multiple target products, and these target products are summed to obtain multiple real-time environmental parameter features. The real-time 5G signal data is converted into frequency domain signal to obtain frequency domain 5G signal data. The signal data in the time domain is converted into frequency domain data through Fourier transform, and the characteristics of the signal at different frequencies are analyzed. Assume that the real-time 5G signal data is , the frequency domain conversion formula is:
[0075] ;
[0076] in, is the frequency domain signal, is the time domain signal value, For time point, is the frequency. Through frequency domain signal conversion, frequency domain 5G signal data is obtained, and feature recognition is performed on these data to obtain real-time 5G signal features. Feature recognition includes features such as the amplitude spectrum and phase spectrum of the signal, which can fully reflect the transmission performance of the 5G signal.
[0077] In a specific embodiment, the process of executing step S104 may specifically include the following steps:
[0078] (1) performing feature normalization and feature sequence conversion on multiple real-time environmental parameter features to obtain a first environmental parameter feature sequence, and performing feature normalization and feature sequence conversion on real-time 5G signal features to obtain a first 5G signal feature sequence;
[0079] (2) Initializing and splicing the first environmental parameter feature sequence and the first 5G signal feature sequence to obtain an initial spliced feature sequence;
[0080] (3) Each real-time environmental parameter feature and real-time 5G signal feature in the initial spliced feature sequence is weighted by the mapping relationship model to obtain multiple weighted fusion feature sequences;
[0081] (4) Perform feature vector mapping on multiple weighted fusion feature sequences to generate multiple environment-signal relationship feature vectors.
[0082] Specifically, multiple real-time environmental data sets and real-time 5G signal data in the 5G network coverage area are obtained. These environmental data sets include parameters such as temperature, humidity, air pressure and light, while 5G signal data include indicators such as signal strength, signal-to-noise ratio and transmission rate. These data are collected in real time by sensors and base station equipment deployed at different locations to form time series data, ensuring that the environmental conditions and signal transmission performance of the network coverage area can be fully reflected. Multiple real-time environmental parameter features are normalized. Data of different features are mapped to a common scale range to eliminate dimensional differences and make each feature comparable. The minimum-maximum normalization or standardization method is usually used. For example, for temperature data, it is normalized to the range of 0 to 1, and humidity, air pressure and light data are similarly processed. These features are converted into feature sequences. Continuous feature values are arranged in chronological order to form a feature sequence. The information of the time dimension is retained so that the model can take into account the time changes of the features. For example, the temperature, humidity, air pressure and light data within a certain time period are arranged in chronological order to form a time series data set containing these features. The same feature normalization and feature sequence conversion processing is performed on the real-time 5G signal features. Data such as signal strength, signal-to-noise ratio, and transmission rate also need to be normalized and arranged in chronological order to form a complete time series data set. After completing feature normalization and sequence conversion, the first environmental parameter feature sequence and the first 5G signal feature sequence are initialized and spliced. The environmental parameter features are combined with the 5G signal features to form a comprehensive feature sequence. Through the splicing operation, an initial spliced feature sequence containing all features is formed, so that the environmental parameters and signal features at each time point can correspond. The mapping relationship model is used to perform feature weighting on each real-time environmental parameter feature and real-time 5G signal feature of the initial spliced feature sequence. The mapping relationship model assigns different weights to different features based on the previously constructed mapping relationship between environmental parameters and signal transmission performance. These weights reflect the importance and influence of each feature in signal transmission. Through feature weighting, the influence of different features is combined to generate a weighted fusion feature sequence. For example, if temperature has a greater impact on signal transmission, the weight of the temperature feature will be relatively high during the weighting process. After feature weighting processing, multiple weighted fusion feature sequences are obtained. Feature vector mapping is performed on the weighted fusion feature sequence. The weighted feature data is converted into vector form so that these data can be used for subsequent signal adjustment and optimization. For example, assume that the environmental parameters in a certain period of time include temperature 30 degrees, humidity 70%, air pressure 1000hPa, and light intensity 500Lux, and the 5G signal data includes signal strength -65dBm, signal-to-noise ratio 25dB, and transmission rate 100Mbps.Normalize these data, normalize the temperature, humidity, air pressure and light data to the range of 0 to 1, and obtain the normalized environmental parameter characteristics. Perform similar normalization on the signal strength, signal-to-noise ratio and transmission rate data to obtain the normalized 5G signal characteristics. Arrange the normalized environmental parameter characteristics and 5G signal characteristics in chronological order to form a feature sequence. Initialize and splice the feature sequence to form an initial spliced feature sequence. Through the mapping relationship model, perform feature weighting on each feature in the initial spliced feature sequence to obtain a weighted fusion feature sequence. Perform feature vector mapping on the weighted fusion feature sequence to generate an environment-signal relationship feature vector.
[0083] In a specific embodiment, the process of executing step S105 may specifically include the following steps:
[0084] (1) Inputting multiple environment-signal relationship feature vectors into a preset multi-level joint convolution classifier, the multi-level joint convolution classifier includes multiple convolution modules, each convolution module is connected to a pooling layer, and each pooling layer is connected to a classifier;
[0085] (2) receiving an environment-signal relationship feature vector through each convolution module, and performing a convolution operation on the environment-signal relationship feature vector to obtain a plurality of first feature vectors;
[0086] (3) performing pooling operations on the multiple first eigenvectors through the pooling layer to obtain multiple second eigenvectors;
[0087] (4) Using a classifier, signal adjustment strategy decisions are made for each of the multiple second feature vectors, and multiple initial signal adjustment strategies are output. The multiple initial signal adjustment strategies include: a base station transmit power adjustment strategy, a spectrum allocation adjustment strategy, an antenna direction adjustment strategy, and a transmission mode switching strategy.
[0088] Specifically, multiple environment-signal relationship feature vectors are input into a preset multi-level joint convolution classifier. The multi-level joint convolution classifier consists of multiple convolution modules, each of which is connected to a pooling layer, and each of which is connected to a classifier. The convolution module extracts local features of the input feature vector, and the pooling layer is used to reduce the dimension of the feature vector, enhance the robustness and generalization ability of the feature. The classifier makes a final classification decision based on the extracted features, and outputs a signal adjustment strategy. Each convolution module receives an environment-signal relationship feature vector and performs a convolution operation on it. The convolution operation performs local calculations on the input feature vector through a set of convolution kernels to extract representative local features. These convolution kernels can be regarded as sliding windows, which slide on the feature vector and calculate the weighted sum of the local area to generate multiple first feature vectors. Assuming that the size of the convolution kernel is 3×3, the process of the convolution operation can be regarded as performing a local weighted summation on the input feature vector, and then performing a nonlinear transformation through an activation function to obtain a convolution output. The multiple first feature vectors are pooled through the pooling layer. The dimension of the feature vector is reduced by downsampling method while retaining important feature information. Commonly used pooling methods include maximum pooling and average pooling. Maximum pooling takes the maximum value in the local area as the pooling result, while average pooling takes the average value of the local area as the pooling result. Through the pooling operation, multiple second feature vectors are obtained. These feature vectors have small dimensions but retain the key features extracted by the convolution operation. The signal adjustment strategy decision is made on multiple second feature vectors through the classifier. The classifier classifies the input feature vector and outputs different signal adjustment strategies. The classifier can be a fully connected layer that maps the input feature vector to a predefined category through linear transformation and activation function. Each category corresponds to a signal adjustment strategy, including base station transmit power adjustment strategy, spectrum allocation adjustment strategy, antenna direction adjustment strategy and transmission mode switching strategy. For example, assume that the input environment-signal relationship feature vector includes data such as temperature, humidity, air pressure, light, signal strength, signal-to-noise ratio and transmission rate. These feature vectors are input to the first convolution module, and the convolution operation is performed through a 3×3 convolution kernel to extract local features. Assuming that the temperature characteristics vary greatly in different areas, the convolution operation can extract the characteristics of these changing areas, such as the impact of local temperature rise on signal strength. After the convolution operation, the convolution output is pooled through the maximum pooling layer, and the pooled result retains the key regional characteristics of temperature change. The pooled result is input into the classifier, which processes it through the fully connected layer and makes classification decisions based on the pattern of the feature vector. Assuming that the features extracted by the convolution and pooling operations show that rising temperature will cause the signal strength to decrease, the classifier can decide that the base station's transmit power needs to be increased to compensate for the signal loss. If the pooling result shows that rising humidity will cause the signal-to-noise ratio to decrease, the classifier can decide to adjust the spectrum allocation strategy to reduce interference.Through the processing of multi-stage joint convolution classifiers, multiple initial signal adjustment strategies are obtained. These strategies include base station transmit power adjustment strategy, spectrum allocation adjustment strategy, antenna direction adjustment strategy and transmission mode switching strategy. For example, the base station transmit power adjustment strategy can decide to increase or decrease the base station transmit power according to the current environmental characteristics and signal characteristics to optimize signal coverage and transmission quality. The spectrum allocation adjustment strategy can dynamically adjust the spectrum allocation according to the current interference situation and spectrum utilization to improve spectrum efficiency and reduce interference. The antenna direction adjustment strategy can dynamically adjust the antenna direction according to user distribution and signal quality to optimize signal coverage and transmission path. The transmission mode switching strategy can select the appropriate transmission mode according to the current business needs and network conditions to ensure the stability and efficiency of signal transmission.
[0089] In a specific embodiment, the process of executing step S106 may specifically include the following steps:
[0090] (1) Adjusting the 5G signal in the 5G network coverage area based on multiple initial signal adjustment strategies, and obtaining signal status data in the 5G network coverage area;
[0091] (2) Using an adaptive ephemera algorithm, a plurality of initial signal adjustment strategies are group initialized according to the signal state data to generate an initial population, where the initial population includes a plurality of first individuals, and each first individual includes a set of signal adjustment strategies;
[0092] (3) Calculate the fitness value of each first individual respectively, and divide the multiple first individuals into populations according to the fitness values to obtain multiple target populations;
[0093] (4) Perform crossover and mutation operations on multiple target populations to generate multiple second individuals, and perform optimization on the multiple second individuals to obtain the optimal individual;
[0094] (5) Obtain multiple target signal adjustment strategies corresponding to the optimal individuals, and dynamically adjust the 5G signal in the 5G network coverage area through the multiple target signal adjustment strategies.
[0095] Specifically, the signal of the 5G network coverage area is adjusted based on multiple initial signal adjustment strategies. The adjustment strategies may include base station transmission power adjustment, spectrum allocation adjustment, antenna direction adjustment, and transmission mode switching. The signal status data of the 5G network coverage area is obtained. The signal status data includes key indicators such as signal strength, signal-to-noise ratio, and transmission rate, which can fully reflect the effect of the current signal adjustment strategy. The signal status data are analyzed by the adaptive mayfly algorithm, and the multiple initial signal adjustment strategies are initialized to generate an initial population. The adaptive mayfly algorithm is an optimization algorithm that imitates the adaptation and evolution process of mayflies in nature. The initial population includes multiple individuals, each of which represents a set of signal adjustment strategies. The diversity of these individuals ensures a wide coverage of the search space, thereby increasing the possibility of finding the optimal strategy. The fitness value of each individual is an important indicator for evaluating its performance. The fitness value reflects the effect of the signal adjustment strategy under the current environmental conditions. For example, if a strategy can significantly improve the signal strength and transmission rate, its fitness value will be higher. The calculation formula of the fitness value can be expressed as:
[0096] ;
[0097] in, Indicates The fitness value of an individual, Indicates signal strength, represents the signal-to-noise ratio, Indicates the transmission rate, are the weight coefficients of these indicators respectively. According to the fitness value of each individual, multiple individuals are divided into populations to obtain multiple target populations. The purpose of population division is to divide individuals into two categories, high fitness and low fitness, according to the fitness value, so as to concentrate resources to optimize high fitness individuals. High fitness individuals will be the main objects of crossover and mutation operations. In crossover and mutation operations, genes are exchanged between high fitness individuals to generate new individuals. The crossover operation is to generate new individuals by combining part of the genes of two high fitness individuals, thereby introducing new features and possible better solutions. The mutation operation is to introduce random changes in the genes of individuals to increase the diversity of the population and explore new solution spaces. Through these operations, multiple new individuals, namely the second individuals, are generated. Multiple second individuals are optimized to find the individual with the highest fitness value, namely the optimal individual. The optimal individual represents the signal adjustment strategy that works best under the current environmental conditions. Through this optimization process, the signal adjustment strategy can be continuously improved and optimized to ensure that the signal transmission performance of the 5G network is in the best state. Obtain multiple target signal adjustment strategies corresponding to the optimal individuals, and use these strategies to dynamically adjust the signal in the 5G network coverage area. For example, if the strategy of the optimal individual includes increasing the base station transmission power, reallocating spectrum resources, and adjusting the antenna direction, these adjustments will be immediately applied to the 5G network to optimize signal coverage and transmission quality. For example, assume that after the initial adjustment of a certain 5G network coverage area, the signal strength is -70dBm, the signal-to-noise ratio is 15dB, and the transmission rate is 50Mbps. Through the initial signal adjustment strategy, after adjusting the base station's transmission power and spectrum allocation, the signal strength is increased to -65dBm, the signal-to-noise ratio is increased to 20dB, and the transmission rate is increased to 75Mbps. After obtaining these signal status data, the adaptive ephemera algorithm is used to initialize the population and generate multiple individuals, each of which represents a different set of signal adjustment strategies. The fitness value of each individual is calculated separately, and it is found that one of the individuals has the highest fitness value. The individuals are divided into populations according to the fitness value, and the individuals with high fitness are taken as the main objects of optimization. New individuals are generated through crossover and mutation operations. Assume that in a crossover operation, the genes of two high-fitness individuals are combined to generate a new individual whose adjustment strategy includes further optimizing the base station's transmission power and spectrum allocation, while adjusting the antenna direction to cover more users. These new individuals are optimized to find the individual with the highest fitness value. Assume that the signal adjustment strategy of this individual can increase the signal strength to -60dBm, the signal-to-noise ratio to 25dB, and the transmission rate to 100Mbps. The strategy of this optimal individual will be applied to the 5G network to achieve dynamic adjustment of the signal and optimize network coverage and transmission quality.
[0098] The above describes the 5G signal dynamic adjustment method based on environment perception in the embodiment of the present application. The following describes the 5G signal dynamic adjustment device based on environment perception in the embodiment of the present application. Figure 2 In the embodiment of the present application, an embodiment of a 5G signal dynamic adjustment device based on environment perception includes:
[0099] The acquisition module 201 is used to collect and pre-process historical data of the 5G network coverage area through multiple sensors to obtain historical environmental data sets and historical 5G signal data for each sensor;
[0100] A construction module 202 is used to perform feature analysis on the historical environmental data set and the historical 5G signal data and calculate the impact of 5G signal transmission, and to construct a mapping relationship model between each environmental parameter and the impact of 5G signal transmission;
[0101] The feature extraction module 203 is used to obtain multiple real-time environmental data sets and real-time 5G signal data in the 5G network coverage area, and perform feature extraction to obtain multiple real-time environmental parameter features and real-time 5G signal features;
[0102] A vector mapping module 204 is used to perform feature vector mapping on multiple real-time environmental parameter features and real-time 5G signal features through a mapping relationship model to generate multiple environment-signal relationship feature vectors;
[0103] A strategy analysis module 205, used for inputting a plurality of environment-signal relationship feature vectors into a preset multi-level joint convolution classifier to perform signal adjustment strategy analysis and output a plurality of initial signal adjustment strategies;
[0104] The feedback optimization module 206 is used to perform strategy feedback optimization on multiple initial signal adjustment strategies through an adaptive ephemera algorithm to obtain multiple target signal adjustment strategies, and dynamically adjust the 5G signal in the 5G network coverage area through the multiple target signal adjustment strategies.
[0105] Through the collaboration of the above components and the deployment of multiple sensors, real-time monitoring and data collection of temperature, humidity, air pressure and light environment parameters in the 5G network coverage area can be achieved. The sensor transmits the collected data to the central processing unit in real time through a low-latency network transmission protocol, effectively ensuring the real-time and accuracy of the environmental data. The ability of real-time environmental perception enables the 5G network to respond to environmental changes in a timely manner, thereby improving the accuracy and dynamic adaptability of signal adjustment. By analyzing historical environmental data and historical 5G signal data, a mapping relationship model between environmental parameters and 5G signal transmission effects is constructed. The mapping relationship model can accurately describe the specific impact of each environmental parameter on signal transmission, and after obtaining real-time environmental data, the key real-time environmental parameter features and 5G signal features are extracted through feature extraction technology. By inputting the extracted environment-signal relationship feature vector into the multi-level joint convolution classifier, the signal adjustment strategy analysis is performed to generate the initial signal adjustment strategy. The adaptive mayfly algorithm is used to perform strategy feedback optimization on the initial signal adjustment strategy to obtain the optimized target signal adjustment strategy, realizing the automatic generation and optimization of the signal adjustment strategy. The accuracy and adaptability of the strategy are also improved through intelligent algorithms, thereby ensuring the stability and high-quality transmission of 5G signals under different environmental conditions.
[0106] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.
[0107] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program codes.
[0108] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.
Claims
1. A 5G signal dynamic adjustment method based on environment perception, characterized in that: The 5G signal dynamic adjustment method based on environment perception includes: Through multiple sensors, historical data of the 5G network coverage area is collected and preprocessed to obtain historical environmental data sets and historical 5G signal data for each sensor; Performing feature analysis and 5G signal transmission impact calculation on the historical environmental data set and the historical 5G signal data, and constructing a mapping relationship model between each environmental parameter and the 5G signal transmission impact; Acquire multiple real-time environmental data sets and real-time 5G signal data in the 5G network coverage area, and perform feature extraction to obtain multiple real-time environmental parameter features and real-time 5G signal features; By using the mapping relationship model, feature vector mapping is performed on the multiple real-time environmental parameter features and the real-time 5G signal features to generate multiple environment-signal relationship feature vectors; Inputting the plurality of environment-signal relationship feature vectors into a preset multi-level joint convolution classifier to perform signal adjustment strategy analysis, and outputting a plurality of initial signal adjustment strategies; The multiple initial signal adjustment strategies are optimized by strategy feedback through an adaptive ephemera algorithm to obtain multiple target signal adjustment strategies, and the 5G signal is dynamically adjusted in the 5G network coverage area through the multiple target signal adjustment strategies; specifically including: adjusting the 5G signal in the 5G network coverage area based on the multiple initial signal adjustment strategies, and obtaining the signal status data of the 5G network coverage area; initializing the multiple initial signal adjustment strategies according to the signal status data through an adaptive ephemera algorithm to generate an initial population, wherein the initial population includes multiple first individuals, and each first individual includes a group of signal adjustment strategies; calculating the fitness value of each first individual respectively, and dividing the multiple first individuals into populations according to the fitness value to obtain multiple target populations; performing crossover and mutation operations on the multiple target populations to generate multiple second individuals, and optimizing the multiple second individuals to obtain the optimal individual; obtaining multiple target signal adjustment strategies corresponding to the optimal individual, and dynamically adjusting the 5G signal in the 5G network coverage area through the multiple target signal adjustment strategies.
2. The method for dynamic adjustment of 5G signals based on environment perception according to claim 1, characterized in that: The historical data of the 5G network coverage area is collected and preprocessed by multiple sensors to obtain the historical environmental data set and historical 5G signal data of each sensor, including: Configure multiple sensors in the 5G network coverage area, wherein the multiple sensors include a temperature sensor, a humidity sensor, an air pressure sensor, and a light sensor; By using the multiple sensors, historical environmental data of the 5G network coverage area is collected respectively to obtain an initial environmental data set for each sensor; Performing data denoising and data cleaning on the initial environmental data set of each sensor respectively to obtain a historical environmental data set of each sensor, wherein the historical environmental data set includes: a historical temperature data set, a historical humidity data set, a historical air pressure data set, and a historical light data set; Signal detection is performed on the 5G network coverage area to obtain initial 5G signal data, and frequency domain signal conversion is performed on the initial 5G signal data to obtain historical 5G signal data.
3. The method for dynamic adjustment of 5G signals based on environment perception according to claim 2, characterized in that: The performing feature analysis and 5G signal transmission impact calculation on the historical environmental data set and the historical 5G signal data, and constructing a mapping relationship model between each environmental parameter and the 5G signal transmission impact, includes: Performing curve fitting on the historical environmental data sets respectively to obtain historical data curves for each environmental parameter; Identify the curve feature points of the historical data curve of each environmental parameter to obtain a curve feature set of each environmental parameter; Performing signal transmission index analysis on the historical 5G signal data to obtain a signal transmission index set; The Pearson correlation coefficient is used to calculate the 5G signal transmission influence coefficient between the curve feature set of each environmental parameter and the signal transmission index set, and the transmission influence coefficient between each environmental parameter and the 5G signal transmission influence is obtained; According to the transmission impact coefficient, signal transmission impact weight modeling is performed on each environmental parameter and its impact on 5G signal transmission to obtain a mapping relationship model between each environmental parameter and its impact on 5G signal transmission.
4. The method for dynamic adjustment of 5G signals based on environment perception according to claim 1, characterized in that: The acquiring of multiple real-time environmental data sets and real-time 5G signal data in the 5G network coverage area, and performing feature extraction to obtain multiple real-time environmental parameter features and real-time 5G signal features, includes: Acquire multiple real-time environmental data sets and real-time 5G signal data in the 5G network coverage area; Performing linear transformation on the multiple real-time environment data sets respectively to obtain multiple linear feature sets, each of which contains multiple linear features; Calculating covariance matrix elements of the plurality of linear features, and constructing a covariance matrix of each linear feature set according to the covariance matrix elements; Calculating the correlation between the multiple linear features according to the covariance matrix to obtain the feature correlation of each linear feature set, and calculating multiple eigenvalues corresponding to the multiple linear features according to the feature correlation; Generating a feature vector of each linear feature set according to the multiple feature values, and performing principal component normalization on the feature vector to obtain a principal component feature weight; Multiplying the plurality of real-time environment data sets by the principal component feature weights respectively to obtain a plurality of target products, and performing sum operation on the plurality of target products to obtain a plurality of real-time environment parameter features; The real-time 5G signal data is subjected to frequency domain signal conversion to obtain frequency domain 5G signal data, and feature recognition is performed on the frequency domain 5G signal data to obtain real-time 5G signal features.
5. The method for dynamic adjustment of 5G signals based on environment perception according to claim 1, characterized in that: The method of performing feature vector mapping on the multiple real-time environmental parameter features and the real-time 5G signal features through the mapping relationship model to generate multiple environment-signal relationship feature vectors includes: Performing feature normalization and feature sequence conversion on the multiple real-time environmental parameter features to obtain a first environmental parameter feature sequence, and performing feature normalization and feature sequence conversion on the real-time 5G signal features to obtain a first 5G signal feature sequence; Initialize and splice the first environmental parameter characteristic sequence and the first 5G signal characteristic sequence to obtain an initial spliced characteristic sequence; By using the mapping relationship model, feature weighting is performed on each real-time environmental parameter feature and real-time 5G signal feature in the initial spliced feature sequence to obtain multiple weighted fusion feature sequences; Perform feature vector mapping on the multiple weighted fusion feature sequences to generate multiple environment-signal relationship feature vectors.
6. The method for dynamic adjustment of 5G signals based on environment perception according to claim 1, characterized in that: The step of inputting the plurality of environment-signal relationship feature vectors into a preset multi-stage joint convolution classifier to perform signal adjustment strategy analysis and outputting a plurality of initial signal adjustment strategies comprises: Inputting the multiple environment-signal relationship feature vectors into a preset multi-stage joint convolution classifier, wherein the multi-stage joint convolution classifier includes multiple convolution modules, each convolution module is connected to a pooling layer, and each pooling layer is connected to a classifier; Receiving an environment-signal relationship feature vector through each convolution module, and performing a convolution operation on the environment-signal relationship feature vector to obtain a plurality of first feature vectors; Performing pooling operations on the multiple first feature vectors respectively through the pooling layer to obtain multiple second feature vectors; The classifier makes signal adjustment strategy decisions for the multiple second feature vectors respectively, and outputs multiple initial signal adjustment strategies, which include: base station transmission power adjustment strategy, spectrum allocation adjustment strategy, antenna direction adjustment strategy and transmission mode switching strategy.
7. A 5G signal dynamic adjustment device based on environment perception, characterized in that: The 5G signal dynamic adjustment device based on environment perception includes: The acquisition module is used to collect and preprocess historical data of the 5G network coverage area through various sensors to obtain historical environmental data sets and historical 5G signal data for each sensor; A construction module is used to perform feature analysis and 5G signal transmission impact calculation on the historical environmental data set and the historical 5G signal data, and to construct a mapping relationship model between each environmental parameter and the 5G signal transmission impact; A feature extraction module is used to obtain multiple real-time environmental data sets and real-time 5G signal data in the 5G network coverage area, and perform feature extraction to obtain multiple real-time environmental parameter features and real-time 5G signal features; A vector mapping module, used to perform feature vector mapping on the multiple real-time environmental parameter features and the real-time 5G signal features through the mapping relationship model to generate multiple environment-signal relationship feature vectors; A strategy analysis module, used for inputting the plurality of environment-signal relationship feature vectors into a preset multi-level joint convolution classifier to perform signal adjustment strategy analysis and output a plurality of initial signal adjustment strategies; A feedback optimization module is used to perform strategy feedback optimization on the multiple initial signal adjustment strategies through an adaptive ephemera algorithm to obtain multiple target signal adjustment strategies, and dynamically adjust the 5G signal of the 5G network coverage area through the multiple target signal adjustment strategies; specifically comprising: adjusting the 5G signal of the 5G network coverage area based on the multiple initial signal adjustment strategies, and obtaining the signal status data of the 5G network coverage area; initializing the multiple initial signal adjustment strategies according to the signal status data through an adaptive ephemera algorithm to generate an initial population, wherein the initial population includes multiple first individuals, and each first individual includes a group of signal adjustment strategies; calculating the fitness value of each first individual respectively, and dividing the multiple first individuals into populations according to the fitness value to obtain multiple target populations; performing crossover and mutation operations on the multiple target populations to generate multiple second individuals, and optimizing the multiple second individuals to obtain the optimal individual; obtaining multiple target signal adjustment strategies corresponding to the optimal individual, and dynamically adjusting the 5G signal of the 5G network coverage area through the multiple target signal adjustment strategies.
8. The device for dynamically adjusting 5G signals based on environment perception according to claim 7, characterized in that: The acquisition module is specifically used for: Configure multiple sensors in the 5G network coverage area, wherein the multiple sensors include a temperature sensor, a humidity sensor, an air pressure sensor, and a light sensor; By using the multiple sensors, historical environmental data of the 5G network coverage area is collected respectively to obtain an initial environmental data set for each sensor; Performing data denoising and data cleaning on the initial environmental data set of each sensor respectively to obtain a historical environmental data set of each sensor, wherein the historical environmental data set includes: a historical temperature data set, a historical humidity data set, a historical air pressure data set, and a historical light data set; Signal detection is performed on the 5G network coverage area to obtain initial 5G signal data, and frequency domain signal conversion is performed on the initial 5G signal data to obtain historical 5G signal data.
9. The 5G signal dynamic adjustment device based on environment perception according to claim 7 is characterized in that: The building blocks are specifically used for: Performing curve fitting on the historical environmental data sets respectively to obtain historical data curves for each environmental parameter; Identify the curve feature points of the historical data curve of each environmental parameter to obtain a curve feature set of each environmental parameter; Performing signal transmission index analysis on the historical 5G signal data to obtain a signal transmission index set; The Pearson correlation coefficient is used to calculate the 5G signal transmission influence coefficient between the curve feature set of each environmental parameter and the signal transmission index set, and the transmission influence coefficient between each environmental parameter and the 5G signal transmission influence is obtained; According to the transmission impact coefficient, signal transmission impact weight modeling is performed on each environmental parameter and its impact on 5G signal transmission to obtain a mapping relationship model between each environmental parameter and its impact on 5G signal transmission.
