An intelligent transmission system for positioning information based on 5G communication network

Abstract

The present invention discloses a positioning information intelligent transmission system based on a 5G communication network, which relates to the field of 5G communication technology, and includes a spectrum sensing module, which is responsible for collecting channel occupancy, network traffic, and user behavior patterns to form a network status data set, and predicting spectrum occupancy through an AI algorithm based on the network status data set, and intelligently allocating spectrum resources; a modulation and coding optimization module, which dynamically adjusts the modulation mode and coding rate when a decrease in channel quality is detected; a multimodal fusion positioning module, which is responsible for integrating multiple signal sources such as 5G signal strength, Wi-Fi signal strength, Bluetooth signal strength, and GPS coordinates, and fusing data of different modes through a deep learning algorithm; an intelligent decision-making module, which is responsible for comprehensively analyzing the data of the spectrum sensing module, the modulation and coding optimization module, and the multimodal fusion positioning module, predicting network interference, and executing corresponding preventive measures; and a data feedback module, which is responsible for efficiently transmitting the optimized positioning information to the target device.

Description

Technical Field

[0001] The present invention relates to the field of 5G communication technology, and in particular to a positioning information intelligent transmission system based on a 5G communication network. Background Art

[0002] With the commercial deployment of the fifth-generation mobile communication system 5G, its high speed, low latency and large connection capabilities provide technical support for emerging applications such as the Internet of Things, autonomous driving and telemedicine. The spectrum resource management, channel quality assurance and positioning services of 5G networks have become research hotspots. Traditional positioning technology mainly relies on the global positioning system GPS. However, in urban canyons, indoor environments and areas with severe signal obstruction, the GPS signal is weakened, resulting in a significant decrease in positioning accuracy and reliability. For this reason, academia and industry are actively exploring multi-modal positioning methods that integrate 5G signals, Wi-Fi, Bluetooth and other wireless signals, hoping to provide high-precision and high-reliability positioning services in various complex environments. At the same time, the dynamic and uncertain nature of 5G networks requires intelligent management of spectrum resources and adaptive optimization of modulation coding to ensure efficient and stable data transmission.

[0003] Although 5G networks have made major breakthroughs in theory and technology, they still face many challenges in practical applications. First, due to the high-density deployment of 5G networks and the application of multi-antenna technology, channel quality fluctuates frequently. How to dynamically adjust the modulation and coding strategy to cope with sudden channel deterioration while ensuring service quality is a difficult point. Secondly, although multimodal positioning improves positioning accuracy, the diversity of signal sources also brings complexity to data fusion. How to effectively integrate positioning information from different modes and build a positioning model with strong robustness and good generalization is a technical bottleneck that needs to be overcome. In addition, existing systems often lack an effective feedback mechanism, making it difficult to adjust system parameters in real time according to actual transmission effects, resulting in limited performance optimization. Summary of the invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a positioning information intelligent transmission system based on the 5G communication network to solve the problem that the channel quality fluctuates frequently due to the high-density deployment of the 5G network and the application of multi-antenna technology. How to dynamically adjust the modulation and coding strategy to cope with sudden channel deterioration while ensuring the service quality is a difficulty. Furthermore, although multimodal positioning improves the positioning accuracy, the diversity of signal sources also brings about the complexity of data fusion. How to effectively integrate positioning information from different modes and construct a positioning model with strong robustness and good generalization is a technical bottleneck that needs to be overcome. In addition, existing systems often lack an effective feedback mechanism, and it is difficult to adjust system parameters in real time according to the actual transmission effect, resulting in limited performance optimization.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, an embodiment of the present invention provides a positioning information intelligent transmission system based on a 5G communication network, which includes:

[0008] The spectrum sensing module is responsible for collecting channel occupancy, network traffic, and user behavior patterns to form a network status data set. Based on the network status data set, it uses AI algorithms to predict spectrum occupancy and intelligently allocate spectrum resources.

[0009] The modulation and coding optimization module dynamically adjusts the modulation mode and coding rate according to the network status and channel quality when the channel quality is detected to be degraded;

[0010] The multimodal fusion positioning module is responsible for integrating multiple signal sources such as 5G signal strength, Wi-Fi signal strength, Bluetooth signal strength, and GPS coordinates, and fusing data of different modalities through deep learning algorithms;

[0011] Intelligent decision-making module, responsible for comprehensive analysis of data from spectrum sensing module, modulation and coding optimization module and multi-modal fusion positioning module, predicting network interference and taking corresponding preventive measures;

[0012] The data feedback module is responsible for efficiently transmitting the optimized positioning information to the target device, while collecting feedback information to optimize system performance.

[0013] As a preferred solution of the positioning information intelligent transmission system based on 5G communication network described in the present invention, the spectrum sensing module is responsible for collecting channel occupancy, network traffic, and user behavior patterns to form a network status data set, and the specific steps are as follows:

[0014] Collect channel occupancy information through spectrum analyzer;

[0015] Continuously monitor network traffic of network devices through Nagios;

[0016] Collect user behavior patterns through GPS;

[0017] The collected data is integrated into the network status data set, and the expression is:

[0018] D={CQ, FQ, BQ, OQ, PQ, AQ, UQ, LQ};

[0019] Among them, D represents the network status dataset, CQ represents the channel occupancy, FQ represents the carrier frequency, BQ represents the bandwidth, OQ represents the channel occupancy rate, PQ represents the peak traffic, AQ represents the average traffic, UQ represents the number of users using the network, and LQ represents the user location.

[0020] As a preferred solution of the intelligent transmission system of positioning information based on 5G communication network described in the present invention, wherein: the spectrum occupancy is predicted by an AI prediction model based on the network status data set, and spectrum resources are intelligently allocated, and the specific steps are as follows:

[0021] Preprocess the network status dataset;

[0022] The channel occupancy, carrier frequency, bandwidth, channel occupancy rate, peak flow, average flow, number of active users and user location are classified and combined, and the expression is:

[0023] D(t)={O(t), F(t), U(t)};

[0024] Where D(t) represents the network status data set collected at time point t, O(t) represents the channel occupancy matrix at time point t, F(t) represents the network traffic vector at time point t, and U(t) represents the user behavior pattern matrix at time point t;

[0025] Use the deep neural network LSTM model as the basis of the AI ​​prediction model;

[0026] The forget gate, input gate, new cell state, cell state update, output gate, hidden state update, and predicted output of the AI ​​prediction model are calculated as follows:

[0027]

[0028] i t =σ(W i [D(t);ht-1"+b i )

[0029] c t =tanh(W c [D(t);h t-1 ]+b c )

[0030] c t =f t ⊙c t-1 +i t ⊙c t

[0031] o t =σ(W o [D(t);h t-1 ]+b o )

[0032] h t =o t ⊙tanh(c t )

[0033] y t =W y h t +b y

[0034] Among them, W f , W i , W i , W o and W y , corresponding to the weight matrices of the forget gate, input gate, new unit state, output gate and predicted output, respectively, b f 、b i 、b c 、b o 、b y They correspond to the forget gate, input gate, new unit state, output gate and bias vector of predicted output, σ and tanh represent Sigmoid and hyperbolic tangent activation functions, respectively. t-1 represents the output of the LSTM unit at the previous time step t-1, ⊙ represents the element-wise product, y t Represents the output at the last time point t;

[0035] The formula is integrated based on the predicted output, and the expression is:

[0036]

[0037] in, represents the predicted spectrum occupancy matrix at time t+Δt, l represents the LSTM neural network layer, W and b represent the weight and bias parameters of the AI ​​prediction model;

[0038] Based on the prediction results and network status dataset D(t), and formulate a spectrum resource allocation plan.

[0039] As a preferred solution of the intelligent transmission system of positioning information based on 5G communication network described in the present invention, when the channel quality is detected to be reduced, the modulation mode and coding rate are dynamically adjusted according to the network status and channel quality. The specific steps are as follows:

[0040] Collect channel quality parameter Q(t);

[0041] Set the channel quality threshold θ Q , when Q(t)<θ Q When , it is determined that the channel quality has deteriorated;

[0042] Principal component analysis is used to extract features from the network status data set D(t) and the channel quality parameter Q(t). The two extracted features are fused to obtain the comprehensive positioning feature Φ(t), which is expressed as:

[0043]

[0044] Wherein, Φ(t) represents the comprehensive positioning feature at time point t, PCA(D(t)) represents the feature vector obtained by principal component analysis of the network status dataset D(t), PCA(Q(t)) represents the feature vector obtained by principal component analysis of the channel quality parameter Q(t), α represents the weight coefficient when adjusting the fusion of the network status dataset D(t) and the channel quality parameter Q(t), β represents the parameter for controlling the width of the Gaussian kernel function, X and Y represent the principal components of the network status dataset D(t) and the channel quality parameter Q(t), z represents the variable in the integral, and exp represents the exponential function;

[0045] The comprehensive positioning feature Φ(t) is weighted and normalized to obtain Φ′(t), which is expressed as:

[0046]

[0047] Among them, Φ′(t) represents the comprehensive positioning feature after weighted normalization, W Φ，i represents the weight of the i-th feature, NI represents the normalization constant, which ensures that each component of Φ(t) is within a reasonable range;

[0048] Based on the weighted normalized comprehensive positioning feature Φ′(t), a nonlinear function is constructed to dynamically adjust the modulation mode M(t) and the coding rate R(t). The expression is:

[0049]

[0050] Where M and R represent the set modulation mode set and coding rate set, arg max represents the parameter for finding the maximum value, Indicates that m is an element in the modulation mode set, It indicates that r is an element in the coding rate set, and Γ represents a nonlinear function.

[0051] As a preferred solution of the positioning information intelligent transmission system based on 5G communication network described in the present invention, wherein: the multiple signal sources integrating 5G signal strength, Wi-Fi signal strength, Bluetooth signal strength, and GPS coordinates are specifically steps:

[0052] Collect 5G signal strength, Wi-Fi signal strength, and Bluetooth signal strength through sensors;

[0053] Collect GPS coordinates through map navigation applications;

[0054] The collected data are integrated into the original signal data, and the collected data are standardized. The signal data after the standardization is set as X n ;

[0055] Use convolutional neural network CNN to embed features of each signal type and obtain the feature vector of each modality, which is expressed as:

[0056]

[0057] Among them, V i represents the eigenvector of the i-th mode, i is the index variable, Represents the nth sample of the original signal data of the i-th mode;

[0058] The adaptive attention mechanism is used to dynamically adjust the weights of different signal sources. The expression is:

[0059]

[0060] Among them, A i represents the attention weight of the i-th modality, exp represents the natural exponential function, V i and V j Represent the eigenvectors of the i-th mode and the j-th mode, w i and table w j Respectively represent the V i and V j The weight vector corresponding to the feature vector is Represents the weight vector w i The transpose of Represents the weight vector w j The transpose of

[0061] Adaptive attention weights are used to fuse the feature vectors of all modalities to obtain a comprehensive feature vector, which is expressed as:

[0062] Z=∑ i A i V i ;

[0063] Among them, Z represents the comprehensive feature vector.

[0064] As a preferred solution of the positioning information intelligent transmission system based on 5G communication network described in the present invention, wherein: the deep learning algorithm fuses data of different modes, and the specific steps are:

[0065] The comprehensive feature vector Z is input into the fully connected layer of the convolutional neural network CNN for final positioning, and the expression is:

[0066]

[0067] in, Represents the predicted positioning coordinates, FC represents the fully connected layer;

[0068] Based on the feature vector V i Construct an anomaly detection module to determine whether the signal source is abnormal. The expression is:

[0069]

[0070] Among them, E i represents the abnormal probability of the i-th signal source, σ represents the activation function Sigmoid, w e represents the weight vector of the signal source, b e represents the bias term of the signal source, and T represents the transposition term;

[0071] Set the threshold EY, when E i >EY, the signal source i is considered abnormal and its weight needs to be adjusted;

[0072] When a signal source is marked as abnormal, its weight is dynamically adjusted, and the expression is:

[0073] Z′=∑ i A′ i V i ;

[0074] Among them, Z′ represents the adjusted comprehensive feature vector, A′ i represents the adjusted attention weight of the i-th modality;

[0075] Use the adjusted comprehensive feature vector Z′ to re-predict the final positioning, expressed as:

[0076]

[0077] in, Represents the final predicted positioning coordinates.

[0078] As a preferred solution of the positioning information intelligent transmission system based on 5G communication network described in the present invention, the specific steps of comprehensively analyzing the data of the spectrum sensing module, the modulation and coding optimization module and the multi-modal fusion positioning module and predicting network interference are as follows:

[0079] Spectrum sensing data, modulation coding data and multimodal fusion positioning data are collected from the spectrum sensing module, modulation coding optimization module and multimodal fusion positioning module, and the collected data are aggregated. The expression is:

[0080] DJ={D SP , D MO , D ML};

[0081] Among them, DJ represents the analysis data set, D SP represents spectrum sensing data, D MO Indicates modulation coded data, D ML Represents multi-modal fusion positioning data;

[0082] Long short-term memory network LSTM is used as the basis of the interference prediction model;

[0083] An interference prediction model is constructed based on spectrum sensing data, modulation coding data and multi-modal fusion positioning data. The expression is:

[0084] I(t+Δt)=LSTM(D SP (t), D MO (t), D ML (t);θ);

[0085] Where I(t+Δt) represents the predicted network interference level at time t+Δt, LSTM represents the interference prediction model, and D SP (t) represents the data provided by the spectrum sensing module at time t, D MO (t) represents the data provided by the modulation and coding optimization module at time t, D ML (t) represents the data provided by the multimodal fusion positioning module at time t, and θ represents the parameters of the interference prediction model.

[0086] As a preferred solution of the positioning information intelligent transmission system based on 5G communication network of the present invention, the specific steps of executing corresponding preventive measures are as follows:

[0087] Set dynamic threshold IY;

[0088] The decision function is constructed based on I(t+Δt) and the dynamic threshold IY, and the expression is:

[0089]

[0090] Among them, H(I(t+Δt)) represents the decision function, and its value range is [0, 1];

[0091] When the value of H(I(t+Δt)) is equal to 0, it means that no preventive measures are taken;

[0092] When the value of H(I(t+Δt)) is equal to 1, it means taking preventive measures;

[0093] The system collects real-time data and uses the stochastic gradient descent algorithm SGD to continuously update the model parameters θ, which is continuously updated by minimizing the prediction error of the interference prediction model.

[0094] As a preferred solution of the positioning information intelligent transmission system based on 5G communication network described in the present invention, wherein: the optimized positioning information is efficiently transmitted to the target device, and the specific steps are:

[0095] Based on the channel quality parameter Q(t) and the signal-to-noise ratio SNR, the data transmission rate is calculated as follows:

[0096]

[0097] Among them, R(t) represents the data transmission rate at time t, Q(t) represents the channel quality parameter at time t, SNR(t) represents the signal-to-noise ratio at time t, reflecting the signal quality. represents the adjustment factor, which is used to adjust the influence of Q(t) on the transmission rate, e represents the base of the natural logarithm, and log2 represents the logarithmic function with base 2;

[0098] The data packet transmission delay time is monitored in real time through timestamps, and the location of the 5G base station is determined using GPS satellite signals to obtain the positioning error.

[0099] As a preferred solution of the positioning information intelligent transmission system based on 5G communication network described in the present invention, wherein: the feedback information is collected at the same time to optimize the system performance, and the specific steps are as follows:

[0100] Based on the transmission delay time and positioning error, the system performance index is calculated as follows:

[0101]

[0102] in, represents the average positioning error, represents the average transmission delay, ΔL(t) represents the positioning error at time t, ΔT(t) represents the transmission delay at time t, and T represents the total time window length;

[0103] Based on the average positioning error and average transmission delay Dynamically adjust the modulation and coding parameters, spectrum allocation, and weight coefficients in the positioning algorithm. The expression is:

[0104]

[0105] Among them, θ new represents the new system parameters, θ old represents the original system parameters, η represents the learning rate, represents the gradient of the disturbance prediction model parameter θ, and λ is the balancing factor.

[0106] The beneficial effects of the present invention are as follows: the present invention realizes the high efficiency, stability and accuracy of positioning information transmission under 5G network through intelligent spectrum sensing and allocation, dynamic modulation and coding optimization, multimodal data fusion, network interference prediction and prevention, and closed-loop data feedback mechanism, improves spectrum utilization, ensures data transmission quality, enhances the robustness of positioning services, and can still provide high-precision positioning in complex environments, especially, significantly improving user experience and overall system performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0107] 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 only 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.

[0108] Figure 1 This is a flow chart of the intelligent transmission system for positioning information based on 5G communication network in Example 1.

[0109] Figure 2 This is a flow chart of the strategy function in Example 1. DETAILED DESCRIPTION

[0110] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0111] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0112] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0113] Example 1, reference Figure 1 and Figure 2 , which is the first embodiment of the present invention, provides a positioning information intelligent transmission system based on a 5G communication network, comprising the following steps:

[0114] The S1 spectrum sensing module is responsible for collecting channel occupancy, network traffic, and user behavior patterns to form a network status data set. Based on the network status data set, it uses an AI algorithm to predict spectrum occupancy and intelligently allocate spectrum resources.

[0115] Collect channel occupancy information through spectrum analyzer;

[0116] Continuously monitor network traffic of network devices through Nagios;

[0117] Collect user behavior patterns through GPS;

[0118] The collected data is integrated into the network status data set, and the expression is:

[0119] D={CQ, FQ, BQ, OQ, PQ, AQ, UQ, LQ};

[0120] Among them, D represents the network status dataset, CQ represents the channel occupancy, FQ represents the carrier frequency, BQ represents the bandwidth, OQ represents the channel occupancy rate, PQ represents the peak flow rate, AQ represents the average flow rate, UQ represents the number of users using the network, and LQ represents the user location;

[0121] Preprocess the network status data set, including denoising, outlier removal and standardization;

[0122] The channel occupancy, carrier frequency, bandwidth, channel occupancy rate, peak flow, average flow, number of active users and user location are classified and combined, and the expression is:

[0123] D(t)={O(t), F(t), U(t)};

[0124] Where D(t) represents the network status data set collected at time point t, O(t) represents the channel occupancy matrix at time point t, F(t) represents the network traffic matrix at time point t, and U(t) represents the user behavior pattern matrix at time point t;

[0125] Channel occupancy matrix, including carrier frequency, bandwidth and channel occupancy;

[0126] Network traffic matrix, including peak and average traffic;

[0127] User behavior pattern matrix, including the number of active users and user locations;

[0128] Use the deep neural network LSTM model as the basis of the AI ​​prediction model;

[0129] The forget gate, input gate, new cell state, cell state update, output gate, hidden state update, and predicted output of the AI ​​prediction model are calculated as follows:

[0130]

[0131] i t = σ(W i [D(t); h t-1 」+ b i )

[0132] c t = tanh(W c [D(t); h t-1 + b c )

[0133] c t = f t ⊙ c t-1 + i t ⊙ c t

[0134] o t = σ(W o [D(t); h t-1 + b o )

[0135] h t = o t ⊙ tanh(c t )

[0136] y t = W y h t + b y

[0137] Among them, W f , W i , W i , W o and W y correspond to the weight matrices of the forget gate, input gate, new cell state, output gate, and predicted output respectively, b f , b i , b c , b o , b y correspond to the bias vectors of the forget gate, input gate, new cell state, output gate, and predicted output respectively, σ and tanh represent the Sigmoid and hyperbolic tangent activation functions respectively, h t-1 represents the output of the LSTM cell at the previous time step t - 1, that is, the hidden state vector, ⊙ represents element-wise multiplication, y t represents the output at the last time point t;

[0138] Integrate the formula based on the predicted output, and the expression is:

[0139]

[0140] in, represents the predicted spectrum occupancy matrix at time t+Δt, l represents the LSTM neural network layer, W and b represent the weight and bias parameters of the AI ​​prediction model;

[0141] Based on the prediction results and network status data set D(t), formulate spectrum resource allocation plan;

[0142] Based on the prediction results The loss function is used to calculate the difference between the predicted value and the actual value. The expression is:

[0143]

[0144] Where L represents the difference between the predicted value and the actual value, and N represents the number of elements in the channel occupancy matrix;

[0145] f t 、i t , o t The value range is [0, 1], indicating the activation degree of the gating signal. t 、i t , o t When the value is 1, it indicates a fully open state, otherwise, it indicates a fully closed state;

[0146] and c t The value range of is [-1, 1], which is determined by the hyperbolic tangent activation function;

[0147] h t The value range of is [-1, 1], which is determined by the output gate and the hyperbolic tangent function;

[0148] y t The value range of is [0, 1], indicating the occupancy ratio.

[0149] S2 modulation and coding optimization module, when it detects that the channel quality has deteriorated, it dynamically adjusts the modulation mode and coding rate according to the network status and channel quality;

[0150] Collect channel quality parameters Q(t), including signal-to-noise ratio SNR and bit error rate BER;

[0151] Set the channel quality threshold θ Q , when Q(t)<θ Q When , it is determined that the channel quality has deteriorated;

[0152] When the channel quality deteriorates, the signal-to-noise ratio (SNR) decreases or the bit error rate (BER) increases, the traditional fixed modulation and coding strategies are no longer applicable because they cannot guarantee the accuracy and efficiency of data transmission. At this time, the dynamic adjustment of the modulation mode and coding rate can intelligently select the most appropriate modulation mode and coding rate according to the real-time network status and channel quality to maintain or optimize the transmission performance of positioning information.

[0153] Principal component analysis is used to extract features from the network status data set D(t) and the channel quality parameter Q(t). The two extracted features are fused to obtain the comprehensive positioning feature Φ(t), which is expressed as:

[0154]

[0155] Wherein, Φ(t) represents the comprehensive positioning feature at time point t, PCA(D(t)) represents the feature vector obtained after principal component analysis of the network status dataset D(t), PCA(Q(t)) represents the feature vector obtained after principal component analysis of the channel quality parameter Q(t), α represents the weight coefficient when adjusting the fusion of the network status dataset D(t) and the channel quality parameter Q(t), β represents the parameter controlling the width of the Gaussian kernel function, which affects the calculation of the similarity of the two feature vectors, X and Y represent the principal components of the network status dataset D(t) and the channel quality parameter Q(t), z represents the variable in the integral, and exp represents the exponential function;

[0156] The comprehensive positioning feature Φ(t) is weighted and normalized to ensure that features of different scales compete fairly in decision making, and Φ′(t) is obtained, which is expressed as:

[0157]

[0158] Among them, Φ′(t) represents the comprehensive positioning feature after weighted normalization, W Φ，i represents the weight of the i-th feature, NI represents the normalization constant, which ensures that each component of Φ(t) is within a reasonable range;

[0159] Based on the weighted normalized comprehensive positioning feature Φ′(t), a nonlinear function is constructed to dynamically adjust the modulation mode M(t) and the coding rate R(t). The expression is:

[0160]

[0161] Where M and R represent the set modulation mode set and coding rate set, arg max represents the parameter for finding the maximum value, Indicates that m is an element in the modulation mode set, Indicates that r is an element in the coding rate set, Γ represents a nonlinear function, which is used to evaluate the performance index of the AI ​​prediction model;

[0162] The nonlinear function Γ is expressed as:

[0163]

[0164] Among them, αL represents the steepness parameter of the logistic function, which determines the sensitivity of Φ′(t) to performance evaluation. A larger αL value will make the function's response sharper, which means that the change of network status has a more significant impact on performance. βL represents the offset of the logistic function, which determines the center point of the curve. γL represents the intensity of the influence of modulation mode and coding rate on performance evaluation. δ represents the steepness parameter of another logistic function, which is used for the influence of modulation mode and coding rate and determines m m / rSensitivity to performance evaluation.

[0165] The S3 multi-modal fusion positioning module integrates multiple signal sources such as 5G signal strength, Wi-Fi signal strength, Bluetooth signal strength, and GPS coordinates, and fuses data of different modalities through deep learning algorithms;

[0166] Collect 5G signal strength, Wi-Fi signal strength, and Bluetooth signal strength through sensors;

[0167] Collect GPS coordinates through map navigation applications;

[0168] The collected data are integrated into the original signal data, and the collected data are standardized. The signal data after the standardization is set as X n ;

[0169] Use convolutional neural network CNN to embed features of each signal type and obtain the feature vector of each modality, which is expressed as:

[0170]

[0171] Among them, V i The eigenvector representing the i-th mode, i is an index variable with a value range of 1, 2, 3, and 4, corresponding to 5G signal strength, Wi-Fi signal strength, Bluetooth signal strength, and GPS coordinates, respectively. Represents the nth sample of the original signal data of the i-th mode;

[0172] The adaptive attention mechanism is used to dynamically adjust the weights of different signal sources. The expression is:

[0173]

[0174] Among them, Ai represents the attention weight of the i-th modality, exp represents the natural exponential function, V i and V j Represent the eigenvectors of the i-th mode and the j-th mode, w i and table w j Respectively represent the V i and V j The weight vector corresponding to the feature vector is Represents the weight vector w i The transpose of Represents the weight vector w j The transpose of

[0175] Adaptive attention weights are used to fuse the feature vectors of all modalities to obtain a comprehensive feature vector, which is expressed as:

[0176] Z=∑ i A i V i ;

[0177] Among them, Z represents the comprehensive feature vector;

[0178] The comprehensive feature vector Z is input into the fully connected layer of the convolutional neural network CNN for final positioning, and the expression is:

[0179]

[0180] in, Represents the predicted positioning coordinates, FC represents the fully connected layer;

[0181] Based on the feature vector V i Construct an anomaly detection module to determine whether the signal source is abnormal. The expression is:

[0182]

[0183] Among them, E i represents the abnormal probability of the i-th signal source, E i The value range is 0 to 1, σ represents the activation function Sigmoid, w e represents the weight vector of the signal source, b e represents the bias term of the signal source, and T represents the transposition term;

[0184] Set the threshold EY, when E i >EY, the signal source i is considered abnormal and its weight needs to be adjusted;

[0185] When a signal source is marked as abnormal, its weight is dynamically adjusted, and the expression is:

[0186] Z′=∑i A′ i V i ;

[0187] Among them, Z′ represents the adjusted comprehensive feature vector, A′ i represents the adjusted attention weight of the i-th modality;

[0188] Use the adjusted comprehensive feature vector Z′ to re-predict the final positioning, expressed as:

[0189]

[0190] in, Represents the final predicted positioning coordinates.

[0191] S4 intelligent decision-making module, responsible for comprehensive analysis of data from the spectrum sensing module, modulation and coding optimization module, and multi-modal fusion positioning module, predicting network interference and taking corresponding preventive measures;

[0192] Spectrum sensing data, modulation coding data and multimodal fusion positioning data are collected from the spectrum sensing module, modulation coding optimization module and multimodal fusion positioning module, and the collected data are aggregated. The expression is:

[0193] DJ={D SP , D MO , D ML};

[0194] Among them, DJ represents the analysis data set, D SP represents spectrum sensing data, D MO Indicates modulation coded data, D ML Represents multi-modal fusion positioning data;

[0195] Long short-term memory network LSTM is used as the basis of the interference prediction model;

[0196] An interference prediction model is constructed based on spectrum sensing data, modulation coding data and multi-modal fusion positioning data to predict the interference level at a future time point. The expression is:

[0197] I(t+Δt)=LSTM(D SP (t), D MO (t), D ML (t);θ);

[0198] Where I(t+Δt) represents the predicted network interference level at time t+Δt, LSTM represents the interference prediction model, and D SP (t) represents the data provided by the spectrum sensing module at time t, D MO (t) represents the data provided by the modulation and coding optimization module at time t, DML (t) represents the data provided by the multimodal fusion positioning module at time t, and θ represents the parameters of the interference prediction model;

[0199] The value range of I(t+Δt) is [0, 1], where a value close to 0 indicates low network interference and a value close to 1 indicates high network interference;

[0200] Set dynamic threshold IY;

[0201] A decision function is constructed based on I(t+Δt) and the dynamic threshold IY. The decision function will decide whether to take preventive measures based on the value of I(t+Δt). The expression is:

[0202]

[0203] Among them, H(I(t+Δt)) represents the decision function, and its value range is [0, 1];

[0204] When the value of H(I(t+Δt)) is equal to 0, it means that no preventive measures are taken;

[0205] When the value of H(I(t+Δt)) is equal to 1, it means taking preventive measures;

[0206] The system collects real-time data and uses the stochastic gradient descent algorithm SGD to continuously update the model parameters θ by minimizing the prediction error of the interference prediction model;

[0207] Preventive measures include:

[0208] AI algorithms are used to predict which frequency bands will be less interfered with in the future, and these frequency bands are used first for data transmission, while reducing spectrum occupancy on interfered frequency bands.

[0209] Before the time period when interference is expected to occur, a portion of spectrum resources is reserved and not allocated in case of emergency;

[0210] In areas where interference is predicted to be high, adjust the spectrum reuse strategy to reduce spectrum usage in overlapping coverage areas and avoid co-channel interference.

[0211] The S5 data feedback module is responsible for efficiently transmitting the optimized positioning information to the target device, while collecting feedback information to optimize system performance;

[0212] Based on the channel quality parameter Q(t) and the signal-to-noise ratio SNR, the data transmission rate is calculated as follows:

[0213]

[0214] Among them, R(t) represents the data transmission rate at time t, Q(t) represents the channel quality parameter at time t, SNR(t) represents the signal-to-noise ratio at time t, reflecting the signal quality. represents the adjustment factor, which is used to adjust the influence of Q(t) on the transmission rate, e represents the base of the natural logarithm, and log2 represents the logarithmic function with base 2;

[0215] The data packet transmission delay time is monitored in real time through timestamps, and the location of the 5G base station is determined using GPS satellite signals to obtain the positioning error;

[0216] Based on the transmission delay time and positioning error, the system performance index is calculated as follows:

[0217]

[0218] in, represents the average positioning error, represents the average transmission delay, ΔL(t) represents the positioning error at time t, that is, the distance difference between the actual position and the predicted position, ΔT(t) represents the transmission delay at time t, that is, the time difference from the sending to the receiving of the data packet, and T represents the total time window length;

[0219] Based on the average positioning error and average transmission delay Dynamically adjust the modulation and coding parameters, spectrum allocation, and weight coefficients in the positioning algorithm. The expression is:

[0220]

[0221] Among them, θ new represents the new system parameters, θ old represents the original system parameters, η represents the learning rate, represents the gradient of the interference prediction model parameter θ, and λ is a balancing factor used to adjust the importance of positioning accuracy and transmission delay;

[0222] The value range of R(t) is (0, ∞), which represents the data transmission rate in bits per second;

[0223] The value range of ΔL(t) and ΔT(t) is (0, ∞), which represent the positioning error and transmission delay in meters and seconds respectively;

[0224] and The value range of is also (0, ∞), representing the average positioning error and average transmission delay respectively.

[0225] This embodiment also provides a computer device, which is suitable for the case of a positioning information intelligent transmission system based on a 5G communication network, including: a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the positioning information intelligent transmission system based on a 5G communication network as proposed in the above embodiment.

[0226] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0227] This embodiment also provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the intelligent transmission system for positioning information based on a 5G communication network as proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic memory, a flash memory, a disk or an optical disk.

[0228] In summary, the present invention achieves the high efficiency, stability and accuracy of positioning information transmission under 5G network through: intelligent spectrum sensing and allocation, dynamic modulation and coding optimization, multimodal data fusion, network interference prediction and prevention, and closed-loop data feedback mechanism, improves spectrum utilization, ensures data transmission quality, and enhances the robustness of positioning services. In particular, it can still provide high-precision positioning in complex environments, which significantly improves user experience and overall system performance.

[0229] Example 2, referring to Table 1, is the second example of the present invention. In order to further verify the advancement of the present invention, experimental simulation data of a positioning information intelligent transmission system based on a 5G communication network are provided.

[0230] In order to verify the effectiveness of the intelligent transmission system of positioning information based on the 5G communication network, the following embodiments were designed. First, 5G base stations, Wi-Fi access points, Bluetooth beacons, and GPS receivers were deployed in a simulated urban environment. The experiment lasted for two weeks, and data was collected twice a day, at 9 a.m. and 5 p.m., respectively. A total of 6 sets of data were collected. During the experiment, 10 mobile terminal devices were used, distributed in different geographical locations, to simulate the behavior patterns of real users.

[0231] Data Collection:

[0232] Use a spectrum analyzer to collect channel occupancy information, including carrier frequency, bandwidth, and channel occupancy rate;

[0233] Use Nagios software to monitor network traffic of network devices and record peak and average traffic;

[0234] GPS records the number of users using the network and their locations;

[0235] Data preprocessing and AI prediction:

[0236] The collected data were denoised, outliers removed, and standardized;

[0237] We built an LSTM model and trained it to predict spectrum occupancy and intelligently allocate spectrum resources.

[0238] Adjust the modulation and coding strategy to adapt to changes in channel quality and optimize data transmission performance;

[0239] Multimodal fusion positioning:

[0240] Integrates 5G signal strength, Wi-Fi signal strength, Bluetooth signal strength and GPS coordinates;

[0241] The positioning accuracy is improved by fusing data of different modes through deep learning algorithms;

[0242] Smart Decision-Making and Preventive Measures:

[0243] Predict network interference and take preventive measures, such as adjusting spectrum usage strategies;

[0244] Collect real-time data, update model parameters, and improve forecast accuracy;

[0245] Data feedback and system optimization:

[0246] Calculate data transmission rate, monitor transmission delay and positioning error;

[0247] Dynamically adjust system parameters to optimize modulation coding, spectrum allocation and positioning algorithms;

[0248] The details are shown in Table 1:

[0249] Table 1 Experimental record table

[0250]

[0251]

[0252] By observing the above table, we can clearly see the performance trend of the system within two weeks. The average channel occupancy rate is maintained between 0.45 and 0.52, indicating that the spectrum resources are effectively managed. The average traffic fluctuates between 35 and 39 Mbps, indicating that the data transmission capacity is stable. The signal-to-noise ratio remains at 17 to 18 dB, reflecting good signal quality. The positioning error is between 2.2 and 2.6 meters on average, confirming the high positioning accuracy of the system.

[0253] Compared with traditional static spectrum allocation and fixed modulation and coding strategies, the dynamic modulation and coding and intelligent spectrum allocation of the present invention significantly reduce positioning errors and improve data transmission rates. For example, the positioning error under traditional methods is usually around 5 meters, while this system reduces it to within 2.5 meters, and the data transmission rate is also increased by about 20%. This shows that by intelligently sensing the network status and dynamically adjusting the system parameters, the positioning information transmission performance of the 5G communication network can be effectively improved.

[0254] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A positioning information intelligent transmission system based on 5G communication network, characterized in that: include, The spectrum sensing module is responsible for collecting channel occupancy, network traffic, and user behavior patterns to form a network status data set. Based on the network status data set, it uses AI algorithms to predict spectrum occupancy and intelligently allocate spectrum resources. The modulation and coding optimization module dynamically adjusts the modulation mode and coding rate according to the network status and channel quality when the channel quality is detected to be degraded; The multimodal fusion positioning module is responsible for integrating multiple signal sources such as 5G signal strength, Wi-Fi signal strength, Bluetooth signal strength, and GPS coordinates, and fusing data of different modalities through deep learning algorithms; Intelligent decision-making module, responsible for comprehensive analysis of data from spectrum sensing module, modulation and coding optimization module and multi-modal fusion positioning module, predicting network interference and taking corresponding preventive measures; The data feedback module is responsible for efficiently transmitting the optimized positioning information to the target device, while collecting feedback information to optimize system performance; The spectrum sensing module is responsible for collecting channel occupancy, network traffic, and user behavior patterns to form a network status data set. The specific steps are as follows: Collect channel occupancy information through spectrum analyzer; Continuously monitor network traffic of network devices through Nagios; Collect user behavior patterns through GPS; The collected data is integrated into the network status data set, and the expression is: D={CQ, FQ, BQ, OQ, PQ, AQ, UQ, LQ}; Among them, D represents the network status data set, CQ represents the channel occupancy, FQ represents the carrier frequency, BQ represents the bandwidth, OQ represents the channel occupancy rate, PQ represents the peak flow, AQ represents the average flow, UQ represents the number of users using the network, and LQ represents the user location; The specific steps of predicting spectrum occupancy through an AI prediction model based on a network status data set and intelligently allocating spectrum resources are as follows: Preprocess the network status dataset; The channel occupancy, carrier frequency, bandwidth, channel occupancy rate, peak flow, average flow, number of active users and user location are classified and combined, and the expression is: D(t)={O(t), F(t), U(t)}; Where D(t) represents the network status data set collected at time point t, O(t) represents the channel occupancy matrix at time point t, F(t) represents the network traffic vector at time point t, and U(t) represents the user behavior pattern matrix at time point t; Use the deep neural network LSTM model as the basis of the AI ​​prediction model; Calculates the forget gate, input gate, new cell state, cell state update, output gate, hidden state update, and predicted output of the AI ​​prediction model; The formula is integrated based on the predicted output, and the expression is: in, represents the predicted spectrum occupancy matrix at time t+Δt, l represents the LSTM neural network layer, W and b represent the weight and bias parameters of the AI ​​prediction model; Based on the prediction results and network status dataset D(t), and formulate a spectrum resource allocation plan.

2. The positioning information intelligent transmission system based on the 5G communication network according to claim 1, characterized in that: When the channel quality is detected to be degraded, the modulation mode and coding rate are dynamically adjusted according to the network status and channel quality. The specific steps are as follows: Collect channel quality parameter Q(t); Set the channel quality threshold θ Q , when Q(t)<θ Q When , it is determined that the channel quality has deteriorated; Principal component analysis is used to extract features from the network status data set D(t) and the channel quality parameter Q(t). The two extracted features are fused to obtain the comprehensive positioning feature Φ(t), which is expressed as: Wherein, Φ(t) represents the comprehensive positioning feature at time point t, PCA(D(t)) represents the feature vector obtained by principal component analysis of the network status dataset D(t), PCA(Q(t)) represents the feature vector obtained by principal component analysis of the channel quality parameter Q(t), α represents the weight coefficient when adjusting the fusion of the network status dataset D(t) and the channel quality parameter Q(t), β represents the parameter for controlling the width of the Gaussian kernel function, X and Y represent the principal components of the network status dataset D(t) and the channel quality parameter Q(t), z represents the variable in the integral, and exp represents the exponential function; The comprehensive positioning feature Φ(t) is weighted and normalized to obtain Φ'(t), which is expressed as: Among them, Φ'(t) represents the comprehensive positioning feature after weighted normalization, W Φ,i represents the weight of the i-th feature, NI represents the normalization constant, which ensures that each component of Φ(t) is within a reasonable range; Based on the weighted normalized comprehensive positioning feature Φ'(t), a nonlinear function is constructed to dynamically adjust the modulation mode M(t) and the coding rate R(t). The expression is: Where M and R represent the set modulation mode set and coding rate set, arg max represents the parameter for finding the maximum value, Indicates that m is an element in the modulation mode set, It indicates that r is an element in the coding rate set, and Γ represents a nonlinear function.

3. The positioning information intelligent transmission system based on the 5G communication network as claimed in claim 2, characterized in that: The multiple signal sources of integrating 5G signal strength, Wi-Fi signal strength, Bluetooth signal strength, and GPS coordinates are specifically steps: Collect 5G signal strength, Wi-Fi signal strength, and Bluetooth signal strength through sensors; Collect GPS coordinates through map navigation applications; The collected data are integrated into the original signal data, and the collected data are standardized. The signal data after the standardization is set as X n ; Use convolutional neural network CNN to embed features of each signal type and obtain the feature vector of each mode, which is expressed as: Among them, V i represents the eigenvector of the i-th mode, i is the index variable, Represents the nth sample of the original signal data of the i-th mode; The adaptive attention mechanism is used to dynamically adjust the weights of different signal sources. The expression is: Among them, A i represents the attention weight of the i-th modality, exp represents the natural exponential function, V i and V j Represent the eigenvectors of the i-th mode and the j-th mode, w i and table w j Respectively represent the V i and V j The weight vector corresponding to the feature vector is Represents the weight vector w i The transpose of Represents the weight vector w j The transpose of Adaptive attention weights are used to fuse the feature vectors of all modalities to obtain a comprehensive feature vector, which is expressed as: Z=Σ i And i In i ; Among them, Z represents the comprehensive feature vector.

4. The positioning information intelligent transmission system based on the 5G communication network as claimed in claim 3, characterized in that: The deep learning algorithm fuses data of different modalities. The specific steps are: The comprehensive feature vector Z is input into the fully connected layer of the convolutional neural network CNN for final positioning, and the expression is: in, Represents the predicted positioning coordinates, FC represents the fully connected layer; Based on the feature vector V i Construct an anomaly detection module to determine whether the signal source is abnormal. The expression is: Among them, E i represents the abnormal probability of the i-th signal source, σ represents the activation function Sigmoid, w e represents the weight vector of the signal source, b e represents the bias term of the signal source, and T represents the transposition term; Set the threshold EY, when E i >EY, the signal source i is considered abnormal and its weight needs to be adjusted; When a signal source is marked as abnormal, its weight is dynamically adjusted, and the expression is: Z′=∑ i A′ i In i ; Among them, Z′ represents the adjusted comprehensive feature vector, A′ i represents the adjusted attention weight of the i-th modality; Use the adjusted comprehensive feature vector Z′ to re-predict the final positioning, expressed as: in, Represents the final predicted positioning coordinates.

5. The positioning information intelligent transmission system based on the 5G communication network as claimed in claim 4, characterized in that: The module is responsible for comprehensively analyzing the data of the spectrum sensing module, the modulation and coding optimization module, and the multi-modal fusion positioning module to predict network interference. The specific steps are as follows: Spectrum sensing data, modulation coding data and multimodal fusion positioning data are collected from the spectrum sensing module, modulation coding optimization module and multimodal fusion positioning module, and the collected data are aggregated. The expression is: DJ={D SP ,D MO ,D ML }; Among them, DJ represents the analysis data set, D SP represents spectrum sensing data, D MO Indicates modulation coded data, D ML Represents multi-modal fusion positioning data; Long short-term memory network LSTM is used as the basis of the interference prediction model; An interference prediction model is constructed based on spectrum sensing data, modulation coding data and multi-modal fusion positioning data. The expression is: I(t+Δt)=LSTM(D SP (t),D MO (t),D ML (t);θ); Where I(t+Δt) represents the predicted network interference level at time t+Δt, LSTM represents the interference prediction model, and D SP (t) represents the data provided by the spectrum sensing module at time t, D MO (t) represents the data provided by the modulation and coding optimization module at time t, D ML (t) represents the data provided by the multimodal fusion positioning module at time t, and θ represents the parameters of the interference prediction model.

6. The positioning information intelligent transmission system based on the 5G communication network as claimed in claim 5, characterized in that: The specific steps of implementing corresponding preventive measures are as follows: Set dynamic threshold IY; The decision function is constructed based on I(t+Δt) and the dynamic threshold IY, and the expression is: Among them, H(I(t+Δt)) represents the decision function, and its value range is [0, 1]; When the value of H(I(t+Δt)) is equal to 0, it means that no preventive measures are taken; When the value of H(I(t+Δt)) is equal to 1, it means taking preventive measures; The system collects real-time data and uses the stochastic gradient descent algorithm SGD to continuously update the model parameters θ, which is continuously updated by minimizing the prediction error of the interference prediction model.

7. The positioning information intelligent transmission system based on the 5G communication network as claimed in claim 6, characterized in that: The optimized positioning information is efficiently transmitted to the target device, and the specific steps are as follows: Based on the channel quality parameter Q(t) and the signal-to-noise ratio SNR, the data transmission rate is calculated as follows: Where R(t) represents the data transmission rate at time t, Q(t) represents the channel quality parameter at time t, SNR(t) represents the signal-to-noise ratio at time t, represents the adjustment factor; The data packet transmission delay time is monitored in real time through timestamps, and the location of the 5G base station is determined using GPS satellite signals to obtain the positioning error.

8. The positioning information intelligent transmission system based on the 5G communication network as claimed in claim 7, characterized in that: The specific steps of collecting feedback information and optimizing system performance are as follows: Based on the transmission delay time and positioning error, the system performance index is calculated as follows: in, represents the average positioning error, represents the average transmission delay, ΔL(t) represents the positioning error at time t, ΔT(t) represents the transmission delay at time t, and T represents the total time window length; Based on the average positioning error and average transmission delay Dynamically adjust the modulation and coding parameters, spectrum allocation, and weight coefficients in the positioning algorithm. The expression is: Among them, θ new represents the new system parameters, θ old represents the original system parameters, η represents the learning rate, represents the gradient of the disturbance prediction model parameter θ, and λ is the balancing factor.

Images