Method for processing electrical signals of a building intercommunication

Abstract

The application provides a building internal communication electric signal processing method, relates to the technical field of signal processing, and specifically comprises the following steps: preparing a building internal communication electric signal data set and constructing a propagation response model; introducing a curvature adjustment coefficient, designing a non-Euclidean manifold embedding structure, training a non-Euclidean convolutional encoder and decoder, realizing signal coding and reconstruction; obtaining a disturbance response relationship matrix through modal disturbance analysis, combining a disturbance direction sensitivity matrix and a modal self-inhibition factor to calculate a fusion weight, and optimizing and fusing a multimodal signal; and the building internal communication electric signal processing model is also provided, precise expression and enhancement of the communication signal under a complex structure are realized, and the stability and anti-interference ability of the building internal communication electric signal are effectively improved.

Description

Technical Field

[0001] This invention belongs to the field of signal processing technology, and in particular relates to a method for processing electrical signals for building interior communication. Background Technology

[0002] With the rapid development of emerging technologies such as smart buildings, IoT communication, and indoor positioning, wireless communication systems inside buildings are undertaking more and more tasks of real-time data transmission and high-reliability connection. In application scenarios such as multi-source sensing fusion, intelligent security, and high-density terminal deployment, how to accurately acquire and effectively process communication electrical signals inside buildings has become a key factor in improving system performance and service quality. However, the closed nature of the building environment, multipath effects, channel obstruction, and structural complexity pose serious challenges to traditional communication signal modeling and analysis methods.

[0003] Existing methods for modeling and enhancing electrical signals in building interiors mostly rely on rule-based modeling, static path loss models, or traditional graph structure propagation mechanisms. These methods are ill-suited to complex building structures, multipath interference, and frequency drift, resulting in limited communication quality prediction and signal enhancement effects. For example, patent CN107370708A proposes a static propagation loss modeling method based on building path graphs, which predicts signal coverage by presetting wall attenuation parameters. However, this method cannot dynamically adjust to environmental changes and does not establish a coupling mechanism between path propagation and environmental parameters, making it difficult to meet the communication prediction accuracy requirements under changes in building structure or dynamic deployment of nodes. Summary of the Invention

[0004] This invention provides a method for processing electrical signals for building interior communication, aiming to construct a building communication signal processing framework with dynamic modeling and multimodal fusion capabilities. The method acquires electrical signal data from within the building; based on this data, it establishes a path propagation loss strategy and propagation power mechanism, extracting intermediate propagation state results reflecting the modulation characteristics of the building structure; it embeds a non-Euclidean convolution strategy into a non-Euclidean convolution encoder and decoder structure to perform embedding mapping and training on the electrical signals, generating reconstructed modal signals with structural distribution representation capabilities to improve signal representation accuracy in nonlinear propagation environments; in the modal perturbation modeling stage, it introduces perturbations into each modal signal and observes the response changes of other modes to construct a modal perturbation response relationship matrix and a perturbation direction sensitivity matrix, and introduces a modal self-inhibition factor to adjust the interference coupling between modes; it weights and fuses multiple modal signals according to fusion weight coefficients to obtain the fused electrical signals for the building interior communication; finally, through the building interior communication signal processing model, it completes the optimization and enhancement processing of communication signals in complex building scenarios.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for processing electrical signals for internal building communication, comprising the following specific steps:

[0006] S1. Acquire communication electrical signal data in the building's internal communication environment. The data includes transmitted signals, reflected signals, transmitted signals, and building structure-related information. Construct a dataset containing propagation paths, frequency band parameters, building material properties, and channel responses.

[0007] S2. Based on the communication electrical signal data in the building's internal communication environment, design building response parameters and combine them with frequency modulation loss terms to establish a path propagation loss strategy, perform building propagation response modeling, and obtain intermediate propagation state results describing signal modulation characteristics.

[0008] S3. Based on the intermediate propagation state results and communication electrical signal data, a curvature adjustment coefficient is introduced to perform non-Euclidean manifold embedding and construct a non-Euclidean convolution strategy; the non-Euclidean convolution encoder and decoder are trained based on the signal reconstruction error to obtain network weight parameters; the trained encoder and decoder are used to encode and reconstruct the embedded signal to obtain multiple reconstructed modal signals.

[0009] S4. Introduce a disturbance for each reconstructed modal signal and measure the degree of response change of other modal signals. Calculate the disturbance influence value between modal signals and construct the modal disturbance response relationship matrix.

[0010] S5. Based on the modal disturbance response relationship matrix, construct the disturbance direction sensitivity matrix, introduce the modal self-inhibition factor, calculate the fusion weight coefficient, and perform fusion processing on multiple reconstructed modal signals to obtain the fused building internal communication electrical signal.

[0011] S6. Construct a building internal communication electrical signal processing model, input the building internal communication electrical signal dataset, and sequentially go through steps S2 to S5, and iterate until convergence to complete the optimization processing of the building internal communication electrical signals.

[0012] Preferably, in step S1, the construction of the building's internal communication electrical signal dataset first involves preprocessing the acquired raw building internal communication electrical signal data, including signal amplitude normalization, time series alignment, noise filtering, and frequency domain conversion, extracting representative short-time features of amplitude, phase, and frequency, and constructing a standardized spatiotemporal feature sequence. After data preprocessing, further collection of building structural information related to signal propagation is performed, including building materials, wall thickness, floor layout, door and window distribution, and spatial zoning parameters. Spatial structural information is extracted using architectural drawings, BIM models, or laser scanning equipment. Simultaneously, frequency band information, including the transmission frequency, is extracted by combining communication protocol parameters and hardware settings. The system obtains the frequency, carrier frequency, and channel bandwidth, along with historical communication records and environmental change parameters, to extract channel state evolution characteristics. Based on multi-source information, and according to the three-dimensional spatial positions of the transmitting and receiving nodes, it marks the start, end, and intermediate traversing media of the propagation path, identifies walls, boundaries, and openings traversed in the path, and establishes a mapping relationship between the path topology and building structure. Finally, it encodes and integrates the signal preprocessing features, building structure information, frequency band parameters, historical communication data, and propagation path diagram structure in a multi-dimensional manner according to time, space, and propagation attributes, generating a building internal communication electrical signal dataset organized in a structured format. This provides a unified standard data input for subsequent propagation modeling and multimodal processing.

[0013] Preferably, the communication signals inside buildings have complex path structures, non-stationary channel states, and dynamic coupling characteristics that make the frequency response prone to deviation during propagation. The signals are affected by spatial obstruction, multipath reflection, and building material properties, resulting in significant time variability and structural sensitivity. Based on these characteristics, this invention proposes a response modeling mechanism that integrates environmental parameters, frequency modulation, and propagation path topology information. This mechanism is used to construct a propagation state expression model with structural awareness and dynamic adaptability, supporting communication modeling, signal enhancement, and modal fusion processing tasks in complex building interior scenarios.

[0014] Furthermore, in step S2, the specific process of building propagation response modeling includes:

[0015] S21. Couple environmental variables with the propagation path depth of signals within the building structure, and model their joint effect using a nonlinear mapping mechanism to generate building response parameters that dynamically reflect the building's internal response characteristics. The calculation formula is as follows:

[0016]

[0017] Where E is the building response parameter, θ env Here, D represents the propagation path depth of the signal within the building, which is an environmental variable.

[0018] S22. Based on the quadratic function relationship between the frequency adjustment coefficient of the building's internal communication electrical signal and the frequency deviation of the center frequency relative to the actual frequency collected by the communication path nodes, a frequency modulation loss term is defined, and the mathematical model is as follows:

[0019]

[0020] Where F is the frequency modulation loss term, β F f0 is the frequency adjustment coefficient for the internal communication electrical signals of the building, where f0 is the center frequency, and f0 is the frequency of the internal communication electrical signals. c This refers to the actual frequency collected by each path node during the communication process;

[0021] S23. Based on the frequency modulation loss term, and combined with the relationship between the signal propagation path depth inside the building and the building response parameters, construct a path propagation loss strategy.

[0022] S24. Based on the path propagation loss strategy, and considering path complexity constraints, construct a propagation power mechanism:

[0023] In the formula, P eff (t) represents the propagation power mechanism, A(t) represents the transmission power of the communication signal at time t, R represents the path complexity factor, and L(D) represents the path propagation loss strategy.

[0024] S25. By constructing a dynamic mapping mechanism between the path loss weight adjustment factor and the propagation power mechanism, the intermediate propagation state results for spatial node pairs are obtained. The mathematical model is as follows:

[0025] w ij =exp(-λ·L(D) ij ));

[0026] S ij =P eff (t)·w ij ;

[0027] In the formula, D ij w represents the propagation path depth between node i and node j. ij L(D) represents the path loss weight, λ is the path loss weight adjustment factor, and L(D) is the path loss weight. ij S represents the path propagation loss strategy between node i and node j. ij This represents the propagation state matrix value between nodes.

[0028] Furthermore, in step S2, during the building propagation response modeling process, firstly, environmental variables are coupled with the propagation path depth of the building's internal communication signals within the building structure to obtain building response parameters. These parameters dynamically reflect the degree of structural environmental influence on the building's internal communication signals at different path depths, providing an important foundation for subsequent propagation modulation modeling. Secondly, considering the energy loss caused by frequency drift in different paths of the building's internal communication signals, a frequency adjustment coefficient is introduced. The frequency modulation loss term is calculated by the deviation between the center frequency and the acquisition frequency of each path node. This loss term accurately describes the impact of frequency band offset on the propagation quality of the building's internal communication signals, and sensitivity control is achieved through the frequency adjustment coefficient. Then, based on the relationship between the propagation path depth of the building's internal communication signals and the building structure, a node topology is constructed, and a path propagation loss strategy is established to represent the connections between each path. Furthermore, by combining building response parameters, frequency modulation loss terms, and propagation path depth, and integrating the transmission power and path complexity factor of the communication signal inside the building at different times, a propagation power mechanism is established. This mechanism can dynamically adjust the propagation intensity of the communication signal inside the building on different paths within the building. Finally, based on the constructed propagation power mechanism and path loss expression weight function, a dynamic mapping mechanism between propagation path depth and path loss weight adjustment factor is used to generate the inter-node propagation state matrix value to represent the propagation state, thereby obtaining the intermediate propagation state expression result describing the propagation process of the communication signal inside the building. The overall modeling strategy, by combining propagation path depth, frequency adjustment coefficient, and building environment parameters, forms a dynamically adjustable building propagation modeling mechanism, providing high-quality propagation state prior information for subsequent non-Euclidean embedding, non-Euclidean encoding and decoding processing, and modal perturbation modeling.

[0029] Preferably, the communication electrical signals inside a building exhibit non-uniform structural connections, spatial topological distortion, and significant multipath interference during propagation. The distribution of the signal in different building areas is limited by local geometric relationships and propagation path folding effects, making it difficult to achieve a continuous and distinguishable structural representation in Euclidean space. Based on the non-Euclidean geometric characteristics of this propagation space, this invention employs a non-Euclidean convolutional coding structure to perform geometric embedding and structural modeling of the communication electrical signals inside a building, in order to support subsequent disturbance response extraction and multimodal collaborative modeling tasks.

[0030] Furthermore, in step S3, the specific steps for constructing the non-Euclidean convolution strategy and training the non-Euclidean encoder and decoder include:

[0031] S31. Construct an embedding mapping function to map the input signal to a non-Euclidean space, and establish a geometric metric model based on the structural mapping relationship between the embedded samples, thereby obtaining the geodesic relationship between samples in the embedding space. The mathematical model is as follows:

[0032]

[0033] Where, x i Let z be the i-th input signal. i Let φ(·) be the embedding vector representation of the i-th input signal, and σ be the embedding mapping function. emb W is a nonlinear embedding mapping function. emb b represents the weights of the nonlinear embedding mapping function. emb Here, n represents the bias of the nonlinear embedding mapping function, and n is the embedding dimension. Let be an n-dimensional Euclidean space, and → be a function mapping symbol. For the target non-Euclidean manifold space;

[0034] S32. By introducing a tensor quantization mechanism that reflects the logarithmic mapping relationship between embedding points, and combining it with a geometric distance metric with a local curvature adjustment coefficient, an adjacency weight function is constructed to describe the local structural relationship of the embedding space:

[0035]

[0036] Wherein K(z) i , z j ) is the adjacency weight function. Indicates from z j To z i The logarithmic mapping tensor, κ i For point z i The curvature adjustment coefficient at point τ is the bandwidth;

[0037] S33. Based on the adjacency weight function, a local perceptual graph structure is generated, and a non-Euclidean convolution strategy for embedding spatial structures is constructed. The mathematical model is as follows:

[0038]

[0039] in, Let σ be the neighborhood of node i, σ be the differentiable processing module of the nonlinear transformation, l be the l-th layer in the non-Euclidean convolution strategy, and W be the... (l) Let l be the weight matrix of the l-th layer. For the neighborhood The original feature map representation of a node in layer l. The output features are from the convolution.

[0040] S34. By introducing a non-Euclidean space convolution strategy into the encoder and decoder structures, a non-Euclidean convolutional encoder and decoder structure adapted to the embedded manifold space is constructed. The reconstruction error is used as the target loss function to train the non-Euclidean convolutional encoder and decoder structure. The mathematical model is as follows:

[0041]

[0042] Where N is the number of training samples, Enc(x) i ) represents the encoder function, Dec(·) represents the decoder function, Dec(Enc(·)) represents the non-Euclidean convolution encoder and decoder functions, and L rec The target loss function;

[0043] S35. Based on the trained non-Euclidean convolutional encoder and decoder structure, the input signal is encoded and reconstructed to obtain multiple reconstructed modal signals. The mathematical model is as follows:

[0044] z i =Enc(x i );

[0045]

[0046] In the formula, The reconstructed result of the modal signal, z m This represents the m-th reconstructed mode signal.

[0047] Furthermore, in step S3, for the embedding and reconstruction method of the non-Euclidean convolutional coding structure, firstly, based on the intermediate propagation state results and the building's internal communication electrical signal data, an embedding mapping function is constructed to map the building's internal communication electrical signal data from the original space to a non-Euclidean manifold space, used to capture the structural distribution characteristics of the building's internal communication electrical signal data in geometric space; secondly, a tensor logarithmic mapping mechanism is introduced, combined with curvature adjustment coefficients and bandwidth parameters, to construct an adjacency weight function, measuring the local geometric proximity relationship between signals, effectively addressing the connection degradation problem under building structure occlusion; subsequently, a local perception map is constructed based on the adjacency weight, and a non-Euclidean convolution is designed. The strategy involves layer-by-layer feature extraction and aggregation of electrical signals used in building interiors to enhance their expressive power in irregular spaces. Next, a convolutional strategy is embedded into an autoencoder structure to construct encoder and decoder modules. Training is then performed with reconstruction error as the target, improving the coding stability and reconstruction accuracy of the electrical signal data features. Finally, the trained coding structure is used to reconstruct the electrical signal data, obtaining a clear and separable embedded representation. This provides a stable structural prior for subsequent modal perturbation modeling and multimodal fusion, significantly improving the integrity and reliability of the representation of electrical signals used in complex building interior environments.

[0048] Preferably, the communication electrical signals inside the building exhibit asymmetric modal coupling structure, fluctuating interference response with changes in spatial structure, and significant differences in frequency range and path penetration characteristics among different modes during multimodal propagation. The interaction between modes changes under the influence of disturbance conditions, exhibiting unstable and nonlinear coupling characteristics. Based on the structural properties of this type of dynamic response relationship, this invention adopts a disturbance-driven modal response mapping mechanism to model the communication electrical signals inside the building, so as to realize the expression of the relationship between multimodal signals and support the weight calculation in the subsequent fusion strategy.

[0049] Furthermore, in step S4, the process of constructing the modal perturbation response relation matrix includes:

[0050] S41. Based on the reconstructed multiple modal signals, a multimodal signal set is constructed. A perturbation is added to each modal signal in the set to obtain a perturbated multimodal signal set. The mathematical model is as follows:

[0051] z′ m =z m +δ m ;

[0052] In the formula, z m δ represents the original signal of the m-th mode. m For the introduced perturbation, z′ m The mode signal after perturbation;

[0053] S42, using the modal input z before perturbation m and the perturbed modal input z′ m A response difference mapping mechanism for mode pairs is established to accurately quantify the perturbation influence values ​​between modes. The mathematical model is as follows:

[0054] ΔR m,n =R n (z′ m )-R n (z m );

[0055] Among them, R n (·) represents the response function of mode n, ΔR m，n This represents the perturbation effect of mode m on mode n;

[0056] S43. The disturbance influence values ​​between modes are matrix-integrated according to their index positions to construct a uniquely determined modal disturbance response relationship matrix. The mathematical model is as follows:

[0057] R = [ΔR] m,n ] M×M ;

[0058] In the formula, R represents the intermodal disturbance response matrix, and M represents the total number of modes.

[0059] Furthermore, in step S4, a response difference mapping mechanism is constructed based on the reconstructed multimodal building internal communication electrical signals. First, perturbations are injected into the reconstructed multimodal building internal communication electrical signals respectively, and their impact on the reconstruction output of other modes is observed, thereby quantifying the perturbation transmission relationship between modes and capturing the interdependence of different modes under changes in building structure. Second, the perturbation response values ​​between all modes are matrix-integrated according to the index order to construct a modal perturbation response relationship matrix, reflecting the interference connection and dynamic linkage characteristics between different modal communication electrical signals inside the building, solving the problem of the unquantifiable interaction relationship between modes during the fusion process. Finally, this response relationship matrix is ​​used as the basis for calculating modal orientation sensitivity and modal self-inhibition factor to support the dynamic allocation strategy of fusion weights, improve the accuracy and adaptability of the fusion process, enhance the controllability and robustness of multimodal building internal communication electrical signals fusion modeling in the building scene, and provide structured weight support for subsequent modal building internal communication electrical signal enhancement and perception tasks.

[0060] Preferably, the communication electrical signals inside the building exhibit characteristics such as dynamic fluctuations in modal response stability with environmental changes during multimodal fusion, inconsistent influence weights of each mode on the fusion result under structural occlusion, frequency shift, and path distribution, and time-varying and structurally dependent actual contributions of different modes to the fusion effect. Based on the characteristics of uneven modal weight response and dynamic adjustment driven by the environment, this invention adopts a weight calculation mechanism based on perturbation response relationship to perform multimodal fusion processing on the communication electrical signals inside the building, so as to achieve adaptive allocation of modal contributions and improve the stability and expressiveness of fusion modeling.

[0061] Furthermore, in step S5, the specific steps for the fusion processing of electrical signals for building-internal communication include:

[0062] S51. By normalizing the relative intensity ratio of intermodal disturbance response values ​​in a specific modal direction, a direction sensitivity matrix is ​​constructed to characterize the distribution characteristics of intermodal directional disturbances. The mathematical model is as follows:

[0063]

[0064] In the formula, S m，n ΔR represents the sensitivity coefficient of mode m to directional perturbations of mode n. m，n This represents the impact value of the disturbance.

[0065] S52. The mean-inverse metric is applied to the response intensity of each mode in the direction sensitivity matrix to construct a mode self-inhibition factor for measuring the mode's adaptability to external disturbances. The mathematical model is as follows:

[0066]

[0067] In the formula, α m The self-inhibition factor represents mode m;

[0068] S53. Normalize the self-inhibition factors of each modality and construct a fusion weight coefficient to measure the relative contribution of multimodal information. The mathematical model is as follows:

[0069]

[0070] In the formula, w m This represents the fusion weight coefficient for the m-th mode;

[0071] S54. Based on the reconstructed modal signals and their corresponding fusion weighting factors, a weighted combination is performed to obtain the fused building internal communication electrical signal. The mathematical model is as follows:

[0072]

[0073] In the formula, Z represents the fused internal communication electrical signal of the building.

[0074] Furthermore, in step S5, the multimodal building internal communication electrical signals are weighted and fused based on the disturbance response relationship between modes to obtain the fused output result. First, a disturbance direction sensitivity matrix is ​​constructed based on the disturbance response value to describe the direction and intensity of disturbance effects between modes, solving the problem of difficulty in modeling the influence path between modes under changes in building structure. Second, a modal self-suppression factor is introduced to measure the ability of each mode to suppress its own disturbance response, serving as a reference index for evaluating its stability, thereby addressing the problem of strong fluctuations and large noise interference in specific modes of building internal communication electrical signals. Then, the final fusion weight coefficient is calculated by combining the disturbance direction sensitivity matrix and the modal self-suppression factor to comprehensively reflect the contribution of each mode to the building internal communication electrical signals in the multimodal fusion process, improving the accuracy and interpretability of the fusion strategy. Finally, based on the fusion weight, the representations of multiple modal signals are weighted and fused to obtain the fused building internal communication electrical signals. This fusion method effectively improves the integration capability and stability of communication electrical signals inside buildings under multimodal conditions through a response structure-driven weight allocation mechanism, and enhances the consistency of signal expression and fusion robustness in complex building scenarios.

[0075] Preferably, in step S6, for the optimized processing of the building's internal communication electrical signals, a building internal communication electrical signal processing model is constructed, and propagation response modeling, non-Euclidean encoding reconstruction, modal perturbation analysis, and multi-modal fusion processing are executed sequentially. First, the building internal communication electrical signal dataset is input. In the propagation response modeling stage, based on the propagation path parameters and building structure information of the building internal communication electrical signals, a path propagation loss strategy and a propagation power mechanism are constructed to obtain intermediate propagation state results describing the modulation characteristics of the building structure. Subsequently, in the non-Euclidean encoding and decoding stage, an embedding mapping function and an adjacency weight function are constructed based on the intermediate propagation state results. A non-Euclidean convolution strategy is established and embedded into the non-Euclidean convolution encoder and decoder structure. The encoder and decoder are trained with signal reconstruction error as the optimization objective to obtain reconstructed multiple modal signals. Then… In the modal disturbance analysis stage, disturbances are introduced into the internal communication signals of each modality, and their impact on the changes in the responses of other modalities is observed. The disturbance influence values ​​between modalities are calculated, and a modal disturbance response relationship matrix is ​​constructed. Then, in the fusion processing stage, based on the modal disturbance response relationship matrix, the disturbance direction sensitivity and modal self-inhibition factor are calculated, and fusion weight coefficients are generated. The multiple modal internal communication signals are then weighted and fused to obtain the fused internal communication signals. By integrating the above stages end-to-end, and combining iterative training mechanisms and structured optimization strategies, the model can dynamically adjust the propagation state modeling, self-encoding reconstruction, and multimodal fusion strategies according to changes in the building environment. Finally, the enhanced and optimized internal communication signals are output, effectively improving the stability and availability of communication signals in complex building environments.

[0076] Compared with existing technologies, this invention constructs a communication signal modeling scheme that integrates building structure information, propagation path parameters, and frequency band characteristics. It proposes a path propagation loss strategy, introduces building response parameters and frequency modulation loss terms, and achieves dynamic modeling of the propagation state of communication signals within complex building structures. This significantly enhances the ability of the internal communication signal model to express the heterogeneity of the building environment and multipath effects. In the encoding and reconstruction stage of the internal communication signal, a non-Euclidean convolution strategy based on curvature adjustment coefficients is constructed. Through embedding mapping functions and a non-Euclidean convolution encoder and decoder structure, the embedding and reconstruction of the internal communication signal are completed, improving the model's performance. The system effectively solves the problem of unstable fusion caused by mutual interference of multimodal signals by introducing disturbances into each modal signal, measuring the response changes of other modes, constructing a modal disturbance response relationship matrix and a direction sensitivity matrix, and introducing a modal self-inhibition factor to adjust the response adaptability between modes. By combining the fusion weight coefficient to perform weighted fusion of multiple modal signals, the system finally achieves accurate modeling and robust fusion of multi-source building internal communication electrical signals in complex building communication environments, significantly improving the expression accuracy and fusion robustness of the building internal communication electrical signal processing system. Attached Figure Description

[0077] Figure 1 This is a flowchart of the building interior communication signal processing method provided by the present invention.

[0078] Figure 2 This is a structural diagram of the building propagation response modeling provided by the present invention.

[0079] Figure 3 This is a structural diagram of the training of non-Euclidean embedding and non-Euclidean convolutional encoder and decoder provided by the present invention.

[0080] Figure 4 This is a structural diagram of the modeling of modal disturbance response relationship provided by the present invention.

[0081] Figure 5 This is a structural diagram of the fusion processing of electrical signals for internal building communication provided by the present invention.

[0082] Figure 6 This is a comparison diagram of the optimized building internal communication electrical signal provided by the present invention and the original building internal communication electrical signal. Detailed Implementation

[0083] This invention provides a method for processing electrical signals for building interior communication, aiming to construct a building communication signal processing framework with dynamic modeling and multimodal fusion capabilities. The method acquires electrical signal data from within the building; based on this data, it establishes a path propagation loss strategy and propagation power mechanism, extracting intermediate propagation state results reflecting the modulation characteristics of the building structure; it embeds a non-Euclidean convolution strategy into a non-Euclidean convolution encoder and decoder structure to perform embedding mapping and training on the electrical signals, generating reconstructed modal signals with structural distribution representation capabilities to improve signal representation accuracy under nonlinear propagation environments; in the modal perturbation modeling stage, it introduces perturbations into each modal signal and observes the response changes of other modes to construct a modal perturbation response relationship matrix and a perturbation direction sensitivity matrix, and introduces a modal self-inhibition factor to adjust the interference coupling between modes; it weights and fuses multiple modal signals according to fusion weight coefficients to obtain the fused electrical signals for the building interior communication; finally, through the building interior communication signal processing model, it completes the optimization and enhancement processing of communication signals in complex building scenarios.

[0084] Please see Figure 1 As shown in the figure, a method for processing electrical signals for internal building communication in this application includes the following specific steps.

[0085] S1. Acquire communication electrical signal data in the building's internal communication environment. The data includes transmitted signals, reflected signals, transmitted signals, and building structure-related information. Construct a dataset containing propagation paths, frequency band parameters, building material properties, and channel responses.

[0086] Further, in step S1, for the building's internal communication electrical signal dataset, a multi-channel wireless signal acquisition device is used to perform communication electrical signal data acquisition tasks within the actual building structure. The types of communication electrical signals acquired include transmitted signals, reflected signals, and penetrating signals. Communication protocols include Wi-Fi, NB-IoT, and 5G indoor communication systems. The signals are represented in complex form and stored in binary format. The building structure covers concrete walls, brick structures, glass curtain walls, and metal partitions. Interference channels include electromagnetic equipment interference, structural obstruction interference, and background radiation interference. The acquisition areas include corridor areas, underground spaces, wall mezzanines, and enclosed rooms. The signal-to-noise ratio range is set to 10dB to 25dB. After acquisition, data preprocessing operations are performed on the original building's internal communication electrical signals, including time alignment, amplitude normalization, high-frequency noise filtering, frequency domain transformation, and window function envelope processing, to improve signal structure consistency and feature stability. The sampling rate is set to 20MHz, the quantization bit width to 16 bits, the sampling duration for each signal segment to 2 milliseconds, and the number of sampling points to 4×10. 4 The total number of communication signal samples is 2×10 5Each sample is accompanied by labeled fields, including propagation path category, spatial structure type, and interference source type, ultimately constructing a building interior communication electrical signal dataset with structural integrity, clear labels, and consistent parameters.

[0087] S2. Based on the communication electrical signal data in the building's internal communication environment, design building response parameters and combine them with frequency modulation loss terms to establish a path propagation loss strategy, perform building propagation response modeling, and obtain intermediate propagation state results describing signal modulation characteristics.

[0088] Furthermore, in step S2, the building propagation response is modeled, and the process is as follows: Figure 2 As shown, the specific steps for building propagation response modeling are as follows.

[0089] S21. Couple environmental variables with the propagation path depth of signals within the building structure, and model their joint effect using a nonlinear mapping mechanism to generate building response parameters that dynamically reflect the building's internal response characteristics. The calculation formula is as follows:

[0090]

[0091] Where E is the building response parameter, quantifying the intensity of the building structure's response to the signal propagation process, and θ env Here, D represents the propagation path depth of the signal inside the building, measured in meters (m).

[0092] In this embodiment, the environmental variable θ env The mathematical model is as follows:

[0093] θ env =θ0 + γ1·T + γ2·H;

[0094] In the formula, θ0 is the basic response constant, which is set as the initial response value of the building under standard environmental conditions, with a value range of [0.5, 1.5], and is set to 0.8 in this embodiment; γ1 is the temperature adjustment coefficient, which represents the degree of gain of the building response factor for every 1°C increase in temperature, with the unit being ( / °C), and a value range of [0.01, 0.05], and is set to 0.02 in this embodiment; γ2 is the humidity adjustment coefficient, which represents the degree of gain of the response factor for every 1% increase in humidity, with the unit being ( / %), and a value range of [0.001, 0.01], and is set to 0.005 in this embodiment; T is the real-time temperature inside the building, and in this embodiment, T = 26, with the unit being degrees Celsius (°C); h is the real-time relative humidity inside the building, and in this embodiment, H = 60, with the unit being percentage (%).

[0095] In this embodiment, the propagation path depth D inside the building is obtained using a three-dimensional spatial distance calculation formula:

[0096]

[0097] Among them, (x i y i , z i Let (x) be the spatial coordinates of the i-th communication node. j y j , z j ) represents the spatial coordinates of the j-th communication node.

[0098] S22. Based on the quadratic function relationship between the frequency adjustment coefficient of the building's internal communication electrical signal and the frequency deviation of the center frequency relative to the actual frequency collected by the communication path nodes, a frequency modulation loss term is defined, and the mathematical model is as follows:

[0099]

[0100] Where F is the frequency modulation loss term, representing the degree of influence of frequency offset on signal energy attenuation, β F The frequency adjustment coefficient for the internal communication electrical signal of the building is defined as [0.01, 0.1], and is set to 0.05 in this embodiment. f0 is the center frequency, which in this embodiment is f0 = 2.4 GHz. c The actual frequency collected by each path node during the communication process, with a value range of [2.38, 2.42], and the unit is GHz.

[0101] S23. Based on the frequency modulation loss term, and combined with the relationship between the signal propagation path depth inside the building and the building response parameters, construct a path propagation loss strategy.

[0102] The mathematical model for the path propagation loss strategy L(D) is as follows:

[0103] L(D) = E·DF·D 2 ;

[0104] In the formula, L(D) represents the path propagation loss strategy, indicating the attenuation trend of the signal in a specific building structure.

[0105] S24. Based on the path propagation loss strategy, and considering path complexity constraints, construct a propagation power mechanism:

[0106]

[0107] In the formula, P eff (t) is the propagation power mechanism, representing the effective transmission capability of the signal under path propagation conditions, where A(t) is the transmission power of the communication signal at time t, and R is the path complexity factor.

[0108] The mathematical model for the transmission power A(t) is:

[0109] A(t)=A0·(1+δ·sin(ωt));

[0110] In the formula, A0 is the basic transmit power amplitude, which is set to 1.0W in this embodiment; δ is the power disturbance coefficient, which controls the dynamic fluctuation amplitude of the power, and its value range is [0,0.5], which is set to 0.2 in this embodiment; ω is the periodic angular frequency, which controls the rate of power change, and its unit is revolutions per second (rad / s).

[0111] The mathematical model for the path complexity factor R is:

[0112] R = R0 + ρ·N wall ;

[0113] In the formula, R0 is the basic complexity factor, representing the minimum complexity under the simplest propagation path, with a value range of [1,2]. In this embodiment, it is set to 1.5. ρ is the structural impedance adjustment coefficient, controlling the influence of each structural unit traversed on the path complexity, with a value range of [0.1,1.0]. In this embodiment, it is set to 0.5. N wall N represents the number of walls or structural units traversed in the propagation path, a non-negative integer determined by the building's spatial layout. In this embodiment, N is set to... wall =4.

[0114] S25. By constructing a dynamic mapping mechanism between the path loss weight adjustment factor and the propagation power mechanism, the intermediate propagation state results for spatial node pairs are obtained. The mathematical model is as follows:

[0115] w ij =exp(-λ·L(D) ij ));

[0116] S ij =P eff (t)·w ij ;

[0117] In the formula, D ij w represents the propagation path depth between node i and node j. ij Let L(D) be the path loss weight, and λ be the path loss weight adjustment factor. In this embodiment, λ = 0.2. ij S represents the path propagation loss strategy between node i and node j. ij This represents the propagation state matrix value between nodes.

[0118] S3. Based on the intermediate propagation state results and communication electrical signal data, a curvature adjustment coefficient is introduced to perform non-Euclidean manifold embedding and construct a non-Euclidean convolution strategy; the non-Euclidean convolution encoder and decoder are trained based on the signal reconstruction error to obtain network weight parameters; the trained encoder and decoder are used to encode and reconstruct the embedded signal respectively to obtain multiple reconstructed modal signals.

[0119] Furthermore, in step S3, a non-Euclidean convolutional strategy is constructed, and non-Euclidean encoders and decoders are trained, as follows: Figure 3 As shown, the specific steps for constructing a non-Euclidean convolution strategy and training a non-Euclidean convolutional encoder and decoder are as follows.

[0120] S31. Construct an embedding mapping function to map the input signal to a non-Euclidean space, and establish a geometric metric model based on the structural mapping relationship between the embedded samples, thereby obtaining the geodesic relationship between samples in the embedding space. The mathematical model is as follows:

[0121]

[0122] Where, x i Let z be the i-th input signal, with an initial dimension of 128. i Let φ(·) be the embedding vector representation of the i-th input signal, and σ be the embedding mapping function. emb W is a nonlinear embedding mapping function used to embed signals from Euclidean space to non-Euclidean manifolds. emb b represents the weights of the nonlinear embedding mapping function. emb The bias is for the nonlinear embedding mapping function. Let n be the n-dimensional Euclidean space, i.e., the original feature space of the input signal. The features have a linear structure and a fixed dimension. → represents the function mapping symbol. Let n be the target non-Euclidean manifold space, and n be the embedding dimension. In this embodiment, n = 64. To complete the embedding mapping function φ(·) from arrive Mapping;

[0123] Nonlinear mapping function σ emb The mathematical model is as follows:

[0124]

[0125] Wherein, α is the suppression adjustment coefficient, which controls the strength of the nonlinear suppression term in the activation function, and its value ranges from [0.1, 2.0]. In this embodiment, α is set to 0.5. β is the scaling sensitivity factor, which adjusts the rate of influence of the squared input value in the exponential function, and its value ranges from [0.1, 1.0]. In this embodiment, β is set to 0.8.

[0126] S32. By introducing a tensor quantization mechanism that reflects the logarithmic mapping relationship between embedding points, and combining it with a geometric distance metric with a local curvature adjustment coefficient, an adjacency weight function is constructed to describe the local structural relationship of the embedding space:

[0127]

[0128] Wherein K(z) i , z j ) is the adjacency weight function. Indicates from z j To z i The logarithmic mapping tensor, κ i For point z i The curvature adjustment coefficient at the point is τ, which is the bandwidth and ranges from [0.5, 2.0]. In this embodiment, τ = 1.0.

[0129] Curvature adjustment coefficient κ i The mathematical model is as follows:

[0130]

[0131] In the formula, κ0 is the basic bandwidth constant, with a value range of [0.5, 1.0], and is set to 0.8 in this embodiment; γ is the perturbation amplification factor, which controls the influence of neighborhood perturbations on the bandwidth, with a value range of [0.1, 2.0], and is set to 0.5 in this embodiment. Indicates from point z i To its neighboring point z j The logarithmic mapping tensor, Let i be the neighborhood of point i. This represents the size of the neighborhood, i.e., the number of neighboring nodes.

[0132] Adjacent point set The mathematical model is as follows:

[0133]

[0134] Among them, ||z i -z j′ ||2 represents the Euclidean distance between node i and node j′ in the embedding space, where j′≠i, and argmin (k) Select the k nearest points, where k is the number of neighboring nodes. In this embodiment, k = 8.

[0135] S33. Based on the adjacency weight function, a local perceptual graph structure is generated, and a non-Euclidean convolution strategy for embedding spatial structures is constructed. The mathematical model is as follows:

[0136]

[0137] in, σ represents the neighborhood of node i; σ is the differentiable processing module for nonlinear transformation; l is the l-th layer in the non-Euclidean convolution strategy. In this embodiment, the graph structure contains 3 layers of non-Euclidean convolution units, i.e., l = 0, 1, 2, with each layer having a feature dimension of 64; W (l) This is the weight matrix for the l-th layer, used to perform weighted transformation on the features of neighboring nodes in the l-th layer. It serves as a learnable parameter in the neural network structure and is dynamically updated during training via the backpropagation algorithm. In this embodiment, the weight matrix... For the neighborhood The original feature map representation of a node at layer l. In this embodiment, the input features are at layer l = 0. Directly take the embedding vector z i As initial feature input; The mathematical model for the differentiable processing module σ of the nonlinear transformation is: (This refers to the output features of the convolution.)

[0138]

[0139] Where η is the feature enhancement adjustment coefficient, which controls the degree of nonlinear stretching of the activation function. In this embodiment, η is set to 0.5. ζ is the feature suppression adjustment factor, which adjusts the suppression effect when the input amplitude is small. In this embodiment, ζ is set to 0.1.

[0140] S34. By introducing a non-Euclidean space convolution strategy into the encoder and decoder structures, a non-Euclidean convolutional encoder and decoder structure adapted to the embedded manifold space is constructed. The reconstruction error is used as the target loss function to train the non-Euclidean convolutional encoder and decoder structure. The mathematical model is as follows:

[0141]

[0142] Where N is the number of training samples, which is set to N = 512 in this embodiment, Enc(x i ) represents the encoder function, Dec(·) represents the decoder function, Dec(Enc(·)) represents the non-Euclidean convolution encoder and decoder functions, and L rec The target loss function;

[0143] Encoder function Enc(x) i The mathematical model for ) is:

[0144] Enc(x i )=σ(W (2) ·σ(W (1) ·σ(W (0) ·x i )));

[0145] Among them, W (0) Input signal x i In this embodiment, the linear transformation weight matrix of the first-layer embedding mapping is... W (1) In this embodiment, the linear transformation weight matrix of the second-layer embedding mapping is... W (2) In this embodiment, the linear transformation weight matrix of the third-layer embedding mapping is...

[0146] The mathematical model for the decoder function Dec(·) is:

[0147]

[0148] The mathematical model for the non-Euclidean convolutional encoder and decoder function Dec(Enc(·)) is as follows:

[0149]

[0150] S35. Based on the trained non-Euclidean convolutional encoder and decoder structure, the input signal is encoded and reconstructed to obtain multiple reconstructed modal signals. The mathematical model is as follows:

[0151] z i =Enc(x i );

[0152]

[0153] In the formula, The reconstructed result of the modal signal, z m This represents the m-th reconstructed mode signal.

[0154] S4. Introduce a disturbance for each reconstructed modal signal and measure the degree of change in the response of other modal signals. Calculate the disturbance influence value between modal signals and construct the modal disturbance response relationship matrix.

[0155] Furthermore, in step S4, the modal perturbation response relationship matrix is ​​constructed, and the process is as follows: Figure 4 As shown, the specific steps for constructing the modal disturbance response relationship matrix are as follows.

[0156] S41. Based on the reconstructed multiple modal signals, a multimodal signal set is constructed. A perturbation is added to each modal signal in the set to obtain a perturbated multimodal signal set. The mathematical model is as follows:

[0157] z′ m =z m +δ m ;

[0158] In the formula, z m δ represents the original signal of the m-th mode. m To introduce the disturbance, δ is set in this embodiment. m It follows a Gaussian distribution with zero mean and a standard deviation of 0.01, i.e. The disturbance amplitude is controlled between [-0.05, 0.05], z′ m This represents the mode signal after the disturbance.

[0159] S42, using the modal input z before perturbation m and the perturbed modal input z′ m A response difference mapping mechanism for mode pairs is established to accurately quantify the perturbation influence values ​​between modes. The mathematical model is as follows:

[0160] ΔR m,n =R n (z′ m )-R n (z m );

[0161] Among them, R n (·) represents the response function of mode n, ΔR m,n This represents the perturbation effect of mode m on mode n;

[0162] Response function R n The mathematical model for (·) is:

[0163]

[0164] Among them, v n Let n be the response direction vector of the nth mode. In this embodiment, Set as a normalized vector; normalized vector v n The mathematical model is as follows:

[0165]

[0166] Where d is the modal input feature dimension, which is set to 64 in this embodiment.

[0167] S43. The disturbance influence values ​​between modes are matrix-integrated according to their index positions to construct a uniquely determined modal disturbance response relationship matrix. The mathematical model is as follows:

[0168] R = [ΔR] m,n ] M×M

[0169] In the formula, R represents the intermodal disturbance response matrix, and M represents the total number of modes. In this embodiment, M is set to 8. The disturbance response values ​​are combined according to the index position to form a disturbance response relationship matrix R with dimension M×M.

[0170] S5. Based on the modal disturbance response relationship matrix, construct the disturbance direction sensitivity matrix, introduce the modal self-inhibition factor, calculate the fusion weight coefficient, and perform fusion processing on multiple reconstructed modal signals to obtain the fused building internal communication electrical signal.

[0171] Furthermore, in step S5, the fusion processing of electrical signals for internal building communications follows the following procedure: Figure 5 As shown, the specific steps for the fusion processing of electrical signals for internal building communication are as follows.

[0172] S51. By normalizing the relative intensity ratio of intermodal disturbance response values ​​in a specific modal direction, a direction sensitivity matrix is ​​constructed to characterize the distribution characteristics of intermodal directional disturbances. The mathematical model is as follows:

[0173]

[0174] In the formula, S m，n ΔR represents the sensitivity coefficient of mode m to directional perturbations of mode n. m，n This represents the impact value of the disturbance.

[0175] S52. The mean-inverse metric is applied to the response intensity of each mode in the direction sensitivity matrix to construct a mode self-inhibition factor for measuring the mode's adaptability to external disturbances. The mathematical model is as follows:

[0176]

[0177] In the formula, α m This represents the self-inhibition factor of mode m.

[0178] S53. Normalize the self-inhibition factors of each modality and construct a fusion weight coefficient to measure the relative contribution of multimodal information. The mathematical model is as follows:

[0179]

[0180] In the formula, w m This represents the fusion weight coefficient for the m-th mode.

[0181] S54. Based on the reconstructed modal signals and their corresponding fusion weighting factors, a weighted combination is performed to obtain the fused building internal communication electrical signal. The mathematical model is as follows:

[0182]

[0183] In the formula, Z represents the fused internal communication electrical signal of the building.

[0184] S6. Construct a building internal communication electrical signal processing model, input the building internal communication electrical signal dataset, and sequentially go through steps S2 to S5, and iterate until convergence to complete the optimization processing of the building internal communication electrical signals.

[0185] In step S6, for the building's internal communication electrical signal processing model, a dataset containing building structural information, propagation path information, and multimodal communication features is first input, and the processing flows defined in steps S2 to S5 are executed sequentially. Specifically, in the propagation response modeling stage, based on the propagation path characteristics and structural parameters of the communication signal in the building environment, a propagation model reflecting the modulation capability of the building structure is constructed, the response features of the communication path are extracted, and intermediate propagation states for subsequent modeling are generated. Subsequently, in the non-Euclidean geometric encoding stage, a mapping mechanism is constructed using the structural relationships between signals, and a graph convolution strategy with local curvature adjustment capability is introduced to construct a non-Euclidean convolution encoder. The encoder and decoder are jointly trained to achieve low-dimensional embedding and signal reconstruction while preserving the signal's geometric structure features. In the modal perturbation analysis stage, small perturbations are introduced into each modal signal, and the response changes of other modes are observed to quantify the interference relationship between modes. A modal perturbation response relationship matrix is ​​constructed to characterize the mutual influence characteristics between modes. Then, in the signal fusion stage, a directional sensitivity matrix between modes is constructed by normalizing the perturbation response intensity, and the self-inhibition capability of each mode is calculated based on its average distribution. Further, fusion weight coefficients are constructed to achieve weighted fusion processing of multimodal signals, resulting in a stable and enhanced building communication electrical signal. This processing flow integrates training in an end-to-end manner, combined with a structured loss mechanism and optimization algorithm, enabling the model to dynamically adjust propagation modeling and modal fusion strategies in different building environments, thereby improving the adaptability, stability, and usability of communication signals in complex indoor scenarios.

[0186] Furthermore, in step S6, the building interior communication electrical signal processing model was developed using the Python programming language and implemented using the PyTorch framework. The input was communication electrical signal data with a signal sample dimension of 128. The model training used the RMSprop optimizer, with a learning rate of 0.0005, a batch size of 32, and 150 training epochs. The model was trained using mean squared error (MSE) as the loss function and iteratively trained through a non-Euclidean encoder structure. During the training process, the loss function converged to a stable interval

[10] . -3 10 -4 This indicates that the model has been able to recover the amplitude structure and variation trend of the communication electrical signals inside the building relatively well, and has completed the mapping and reconstruction of the communication electrical signals inside the building.

[0187] Furthermore, in step S6, the original building's internal communication electrical signals are input into the constructed communication electrical signal processing model for processing, and the processing effect is as follows: Figure 6 As shown in the figure, the horizontal axis represents time in seconds, and the vertical axis represents the signal amplitude (unitless). The figure reveals that the original communication signal exhibits significant temporal fluctuations and noise interference, affecting its stability and clarity. However, after optimization using the building interior communication signal processing model proposed in this invention, the signal fluctuation amplitude is significantly reduced, the overall curve is smoother, noise interference is effectively suppressed, and signal clarity and continuity are greatly improved. This verifies the effectiveness and robustness of the model in improving the quality of communication signals in complex building propagation environments.

[0188] The above are merely preferred embodiments of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the inventive concept of the present invention, and these modifications and improvements all fall within the protection scope of the present invention.

Claims

1. A method of processing electrical communication signals within a building, characterized by, Specifically comprising the following steps: S1, obtaining communication electric signal data in the building internal communication environment, the data including transmitted signals, reflected signals, penetrated signals and building structure related information, constructing a data set containing propagation path, frequency band parameter, building material property and channel response; S2, according to the communication electric signal data in the building internal communication environment, designing building response parameters and combining frequency modulation loss term, establishing path propagation loss strategy, performing building propagation response modeling, and obtaining intermediate propagation state results describing signal modulation characteristics; S3, according to the intermediate propagation state results and the communication electric signal data, introducing a curvature adjustment coefficient, performing non-euclidean manifold embedding, and constructing a non-euclidean convolution strategy; Training the non-euclidean convolution encoder and decoder based on signal reconstruction error to obtain network weight parameters; the trained encoder and decoder are respectively used for encoding and reconstructing the embedded signal to obtain reconstructed multiple modal signals; S4, introducing disturbance to each reconstructed modal signal respectively, measuring the response change degree of other modal signals, calculating the disturbance influence value between modal signals, and constructing a modal disturbance response relationship matrix; S5, according to the modal disturbance response relationship matrix, constructing a disturbance direction sensitivity matrix, introducing a modal self-inhibition factor, calculating a fusion weight coefficient, and performing fusion processing on the multiple reconstructed modal signals to obtain the fusion processed building internal communication electric signal; S6, constructing a building internal communication electric signal processing model, inputting the building internal communication electric signal data set, sequentially passing through steps S2 to S5, and through iterative training until convergence, completing the optimization processing of the building internal communication electric signal.

2. The method of claim 1, wherein, The specific method for constructing the data set containing the propagation path, the frequency band parameter, the building material property and the channel response is: Obtain the communication electric signal original data in the building internal environment, including the transmitted signals, the reflected signals and the penetrated signals, obtain the signal sequence through the time-sharing sampling mode; at the same time, obtain the building structure information, including the building material type, the wall thickness, the floor structure, the door and window distribution information and the space partition structure; further obtain the frequency band parameter, including the transmission frequency, the carrier frequency and the channel bandwidth, as well as the historical communication data and the building environment parameter; According to the communication electric signal original data and the building structure information, mark the starting point, the ending point and the intermediate transmission medium of the propagation path, and construct the propagation path topology structure; According to the communication electric signal original data, the building structure information, the frequency band parameter, the historical communication data and the propagation path topology structure, create a communication electric signal data set containing the propagation path, the frequency band parameter, the building material property and the channel response information.

3. A method of processing electrical signals for communication within a building according to claim 2, wherein, The specific process of establishing the path propagation loss strategy includes: S21, coupling the environmental variable with the propagation path depth of the signal in the building structure, modeling the joint effect by using a nonlinear mapping mechanism, generating building response parameters that can dynamically reflect the response characteristics in the building interior, and the calculation formula is: where E is the architectural response parameter, θ env is the environmental variable, and D is the depth of the signal propagation path within the architecture. S22, define the frequency modulation loss term based on the quadratic function relationship between the frequency adjustment coefficient of the building internal communication telecommunication signal and the frequency deviation of the center frequency relative to the actual frequency collected by the communication path node, and the mathematical model is: Wherein, F is the frequency modulation loss term, β F is the frequency adjustment coefficient of the building internal communication signal, f0 is the center frequency, f c is the actual frequency collected by each path node in the communication process; S23, construct the path propagation loss strategy according to the frequency modulation loss term, combined with the propagation path depth of the signal in the building and the building response parameter relationship.

4. The method of claim 3, wherein, The specific method for obtaining the intermediate propagation state result describing the signal modulation characteristics is: S201、According to the path propagation loss strategy, a propagation power mechanism is constructed on the basis of considering path complexity constraints: where P eff (t) is a propagation power mechanism, A(t) is the transmit power of the communication signal at time t, R is a path complexity factor, and L(D) is a path propagation loss policy. S202, obtain the intermediate propagation state result for the spatial node pair by constructing a dynamic mapping mechanism that integrates the path loss weight adjustment factor and the propagation power mechanism, and the mathematical model is: w ij = exp(-λ·L(D ij )) S ij = P eff (t) · w ij ; In the formula, D ij represents the propagation path depth between node i and node j, w ij is the path loss weight, λ is the path loss weight adjustment factor, L(D ij ) is the path propagation loss strategy of node i and node j, S ij is the inter-node propagation state matrix value.

5. A method of processing electrical signals for communication within a building according to claim 4, wherein, The specific steps of constructing the non-Euclidean convolution strategy include: S31, construct an embedding mapping function for mapping the input signal to a non-Euclidean space, and establish a geometric metric model based on the structural mapping relationship between the embedded samples, so as to obtain the geodesic relationship between the samples in the embedded space, and the mathematical model is: z i = φ(x i ) = σ emb (W emb x i +b emb ), where x i is the ith input signal, z i is the embedded vector representation of the ith input signal, φ(·) is the embedding mapping function, σ emb is the nonlinear embedding mapping function, W emb is the weight of the nonlinear embedding mapping function, b emb is the bias of the nonlinear embedding mapping function, and n is the embedding dimension, is the n-dimensional Euclidean space, and → is the function mapping symbol, is the target non-Euclidean manifold space; S32, construct an adjacency weight function for describing the local structure relationship of the embedded space by introducing a tensor quantization mechanism reflecting the logarithmic mapping relationship between the embedded points, and combining a geometric distance metric method with a local curvature adjustment coefficient: where K(z i , z j ) is the adjacency weight function, denotes the logarithmic mapping tensor from z j to z i , κ i is the curvature adjustment coefficient at point z i , and τ is the bandwidth. S33, generate a local perception graph structure based on the adjacency weight function, and construct a non-Euclidean convolution strategy for the embedded space structure, and the mathematical model is: wherein, is the neighborhood of node i, σ is a differentiable processing module of a non-linear transformation, l is the l-th layer in a non-Euclidean convolution strategy, W (l) is the l-th layer weight matrix, is the neighborhood is the original feature map representation of node i at the l-th layer, is the convolution output feature.

6. A method of processing electrical signals for communication within a building according to claim 5, wherein, The specific steps of obtaining the reconstructed multiple modal signals include: S301, construct a non-Euclidean convolution encoder and decoder structure that adapts to the embedded manifold space by introducing a non-Euclidean space convolution strategy in the encoder and decoder structure, and train the non-Euclidean convolution encoder and decoder structure by using the reconstruction error as the target loss function, and the mathematical model is: where N is the number of training samples, Enc(x i ) is an encoder function, Dec(·) is a decoder function, Dec(Enc(·)) is a non-Euclidean convolutional encoder and decoder function, L rec is a target loss function; S302, encode and reconstruct the input signal based on the trained non-Euclidean convolution encoder and decoder structure to obtain the reconstructed multiple modal signals, and the mathematical model is: z i = Enc(x i ); In the formula, is the reconstruction result of the modal signal, z m denotes the mthreconstructed modal signal.

7. A method of processing electrical signals for communication within a building according to claim 6, wherein, The process of constructing the modal disturbance response relationship matrix includes: S41, construct a multi-modal signal set according to the reconstructed multiple modal signals, add a disturbance to each modal signal in the set respectively, and obtain a multi-modal signal set with added disturbances, and the mathematical model is: z' m = z m + δ m ; where z m represents the original signal of the mth modality, δ m is the introduced disturbance, z′ m is the disturbed modality signal; S42, modal input z before disturbance m and modal input z' after disturbance m , the response difference mapping mechanism of modal pairs is established, and the disturbance influence value between the modes is accurately quantified, and the mathematical model is: ΔR m，n = R n (z m ) - R n (z m ); wherein R n (·) denotes the response function of mode n, ΔR m，n denotes the disturbance influence value of mode m on mode n; S43, matrix integrate the disturbance influence values among the modes according to the index position to construct a unique modal disturbance response relationship matrix, and the mathematical model is: R = [AR m，n ] m×M ; In the formula, R represents the modal disturbance response matrix, and M represents the total number of modes.

8. The method of claim 7, wherein the method further comprises: The specific steps of the building internal communication telecommunication signal fusion processing include: S51, construct a direction sensitivity matrix for representing the direction interference distribution characteristics between the modes by normalizing the relative intensity proportion of the modal disturbance response value in a specific modal direction, and the mathematical model is: In the formula, S m,n denotes the directional disturbance sensitivity coefficient of mode m to mode n, ΔR m，n is the disturbance influence value; S52, construct a modal self-inhibition factor for measuring the adaptability of the mode to external disturbance by inversely measuring the response intensity of each mode in the direction sensitivity matrix, and the mathematical model is: wherein α m denotes the self-inhibition factor of the mode m; S53, construct a fusion weight coefficient for measuring the relative contribution degree of multi-modal information by normalizing the self-inhibition factor of each mode, and the mathematical model is: In the formula, w m denotes the fusion weight coefficient of the mth modality; S54, based on each reconstruction modal signal and its corresponding fusion weight factor, weighted combination is carried out to obtain the fusion processed building internal communication signal, and the mathematical model is: In the formula, Z represents the fusion processed building internal communication signal.

Images