Intelligent anti-interference method for multi-physics coupling and polarization diversity

Abstract

The invention relates to the technical field of wireless communication, in particular to an intelligent anti-interference method for multi-physics coupling and polarization diversity, which integrates a material shielding effect, a dust scattering characteristic and a dynamic polarization optimization mechanism so as to improve the reliability and the stability of wireless communication in a deep well environment. The method comprises the following steps: firstly, establishing a path loss correction model based on an eddy current shielding effect of a metal support material and a Mie scattering theory of dust concentration; secondly, providing an intelligent anti-interference layered framework, combining a direct sequence spread spectrum (DSSS) technology and a wavelet adaptive filtering method, and introducing a game theory to optimize multi-source node power distribution; and further designing a polarization-time delay joint compensation strategy, and suppressing multipath fading and inter-symbol interference in a non-line-of-sight scene by adopting a circularly polarized antenna and a channel equalizer based on a convolutional neural network and a long-short term memory network. Compared with a traditional method, the method has the advantages that the bit error rate can be reduced by 42% in a high-dust-concentration (200 g / m) environment, and the signal to interference plus noise ratio can be increased by 15 dB in a strong electromagnetic interference area.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of wireless communication, and specifically relates to an intelligent anti-interference method of multi-physical field coupling and polarization diversity. BACKGROUND

[0002] With the surge in global demand for mineral resources and the advancement of intelligent mine construction, wireless communication technology in deep well environments has become the core support for ensuring safety in production and improving mining efficiency. Deep well space has characteristics such as narrow roadway, complex support structure, high dust concentration, and dense electromagnetic interference sources (such as high-voltage equipment, frequency converters, and mechanical and electrical switches), which result in problems such as multipath fading, aggravated path loss, and significant time delay expansion for wireless signals. For example, in emergency situations such as coal mine collapse, traditional wired communication systems are prone to failure due to cable damage, while existing wireless technologies (such as WiFi and ZigBee) have obvious shortcomings in terms of penetration, anti-interference, and dynamic adaptability, making it difficult to achieve reliable two-way communication and real-time rescue command. According to statistics, the average mining depth of large and medium-sized coal mines in China exceeds 500 meters, and some mines have depths exceeding 1,000 meters, further exacerbating the stability challenges of signal transmission in complex electromagnetic environments.

[0003] Existing mine wireless communication technologies mainly include short-range wireless communication, acoustic wave transmission, electromagnetic wave transmission, and 5G technology. Among them, ZigBee is applied to underground sensor networks due to its low power consumption and network capacity advantages, but its anti-interference ability is limited; acoustic wave transmission propagates through oil pipes and uses single-carrier frequency domain equalization (SC-FDE) technology to suppress multipath interference, but the hardware complexity is high and is significantly affected by dust concentration; electromagnetic wave transmission (such as MWD system) has strong penetration, but is affected by metal support shielding effect and high-frequency noise interference, resulting in high bit error rate; 5G technology reduces roadway attenuation by optimizing 700-900MHz frequency band and introduces reconfigurable intelligent surface (RIS) to enhance coverage, but still faces the problem of signal quality fluctuations caused by dynamic electromagnetic interference (such as transient noise of mechanical and electrical equipment). In addition, existing anti-interference technologies such as adaptive filtering and blind source separation (BSS) have high computational complexity or residual interference problems, while signal detection methods based on deep learning (such as dual-path network DPN) can reduce bit error rate, but have insufficient joint optimization capability for complex multi-source interference.

[0004] Overall, the existing mine wireless communication technology generally has the following core problems: first, the environmental coupling effect modeling is insufficient, the existing research mostly uses a simplified path loss model, and fails to systematically integrate the electromagnetic shielding effect of supporting materials and the scattering absorption of dust particles and other multi-physical field coupling effects, resulting in limited channel prediction accuracy; second, the dynamic interference suppression capability is weak, the traditional anti-interference technology is difficult to adapt to the mixed characteristics of pulse noise and Gaussian noise in deep well environment, and lacks real-time perception and compensation mechanism for time-varying channel state; third, the polarization and multipath effect cooperative optimization is missing, the existing research mostly focuses on the independent performance analysis of a single polarization mode, ignores the multipath suppression potential of circularly polarized antennas in non-line-of-sight (NLOS) dense reflection scenarios, and lacks joint optimization strategies of dynamic balancing algorithm and polarization diversity technology, making it difficult to achieve cooperative suppression of complex electromagnetic interference and multipath fading. SUMMARY

[0005] The present application aims to solve the technical problems of insufficient path loss prediction accuracy, limited dynamic interference suppression capability, and lack of cooperative optimization of multipath fading and polarization effect in wireless communication under complex electromagnetic environment in deep well, and proposes an intelligent anti-interference method of multi-physical field coupling and polarization diversity, which realizes high-reliability and low-bit-error-rate wireless communication through the combination of multi-physical mechanism modeling and intelligent anti-interference strategy.

[0006] To achieve the above purpose, the present application adopts the following technical scheme: an intelligent anti-interference method of multi-physical field coupling and polarization diversity, comprising the following steps: constructing a multi-physical field coupling path loss model based on the eddy current shielding effect of metal support materials and the Mie scattering theory of dust concentration; designing an intelligent anti-interference layered architecture, including spread spectrum anti-narrowband interference, wavelet adaptive filtering to suppress impulse noise, and distributed power control based on game theory; using polarization diversity and deep learning channel equalization strategy, suppressing multipath effect through circularly polarized antennas, and combining CNN-LSTM network for dynamic channel equalization.

[0007] Further, the construction of the multi-physical field coupling path loss model includes: establishing a metal support shielding effect model, calculating the shielding attenuation factor based on eddy current loss and reflection loss; establishing a dust scattering effect model, calculating the extinction efficiency factor and total attenuation factor based on the Mie scattering theory; integrating the shielding attenuation factor and the total attenuation factor into the path loss formula to form a modified path loss model.

[0008] Further, the intelligent anti-interference layered architecture comprises: the first layer adopts a direct sequence spread spectrum (DSSS) technology, uses a Gold code sequence for spread spectrum modulation to suppress narrowband interference; the second layer adopts an adaptive threshold denoising method based on a discrete wavelet transform (DWT) to filter impulse noise; and the third layer constructs a distributed power control model based on non-cooperative game theory, and iteratively optimizes the transmission power of each node to maximize the system utility function.

[0009] Further, the processing gain of the DSSS technology is not less than 20 dB, the code length of the Gold code is 63, and the maximum cross-correlation value is not greater than 7.

[0010] Further, the adaptive threshold denoising method adopts a mixed soft and hard threshold strategy, and the threshold is dynamically determined based on a median absolute deviation (MAD).

[0011] Further, the utility function of the distributed power control model considers channel gain, noise power, interference power and energy consumption weight, and iteratively solves the Nash equilibrium point through a gradient projection algorithm.

[0012] Further, the polarization diversity and deep learning channel equalization strategy comprises: using a circularly polarized antenna to receive a signal to reduce the correlation between multipath components and achieve polarization diversity; and constructing a CNN-LSTM dynamic equalizer, including a one-dimensional convolution layer, a maximum pooling layer, an LSTM layer and a full connection layer, for time series modeling and symbol recovery of the received signal.

[0013] Further, the input of the CNN-LSTM equalizer is a time domain signal sequence with a length of 256, and the output is a symbol estimation value after equalization.

[0014] Further, a wireless communication system comprises: a transmitting module for generating and transmitting a DSSS modulated signal; a receiving module comprising a circularly polarized antenna and a CNN-LSTM equalizer; and a processing module for performing wavelet denoising, power control and channel equalization operations.

[0015] Further, the system is suitable for deep well safety monitoring, emergency communication or intelligent mine wireless communication scenarios.

[0016] The beneficial effects of the present application are: (1) the systematic fusion of metal shielding effect, dust scattering characteristics and geometric structure influence significantly improves the channel prediction accuracy; (2) the cooperative optimization of polarization diversity and dynamic equalization effectively reduces the bit error rate in NLOS scenarios; and (3) the construction of a low computational overhead layered anti-interference system greatly improves the SINR in a mixed noise environment. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1A three-level hierarchical anti-interference architecture diagram of the application.

[0018] Figure 2 Two polarization mode delay spectrum comparison and cumulative distribution curve of the application.

[0019] Figure 3 CNN-LSTM dynamic equalizer structure diagram.

[0020] Figure 4 Three-dimensional mine tunnel model.

[0021] Figure 5 Path loss result under high concentration dust, non-conductor support.

[0022] Figure 6 Path loss result under high concentration dust, high-conductive metal support.

[0023] Figure 7 Convergence and power balance in game theory power control.

[0024] Figure 8 Comparison of equalizer convergence performance in static channel.

[0025] Figure 9 Comparison of equalizer bit error rate performance in dynamic channel. DETAILED DESCRIPTION

[0026] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, other embodiments obtained by those skilled in the art without creative work are within the protection scope of the application. Improvements and modifications on the basis of the application are within the protection scope of the application.

[0027] The intelligent anti-interference method of multi-physical field coupling and polarization diversity provided by the application realizes reliability and stability in deep well complex environment by systematically integrating material shielding effect, dust scattering characteristics and dynamic anti-interference mechanism. Specifically, the eddy current shielding effect of metal support material is quantified based on finite element simulation (FEM), the nonlinear relationship between dust concentration and signal attenuation is established combining with Mie scattering theory, and the modified path loss model is proposed , breaks through the simplified assumption of traditional models on the coupling of multiple factors; on this basis, an intelligent anti-interference layered architecture is designed, the direct sequence spread spectrum (DSSS) technology is used to expand the signal bandwidth to suppress narrowband interference, the wavelet adaptive threshold denoising algorithm is used to eliminate impulse noise, and the game theory is introduced to optimize the power distribution of multiple source nodes, so that the signal-to-interference-and-noise ratio (SINR) is improved by 15dB in a mixed noise environment; at the same time, combined with polarization diversity and deep learning technology, a collaborative architecture of circularly polarized antenna and CNN-LSTM dynamic equalizer is proposed, which uses circularly polarized wave to suppress the phase inversion of multipath reflection, predicts the time-varying response based on historical channel state information (CSI), and reduces the bit error rate (BER) by 42% in a non-line-of-sight (NLOS) scenario. The model is verified by simulation and coal mine measurement, and shows significant robustness in a high-dust (200g / m³) and strong interference (85dBμV / m) environment, providing a theoretical and technical breakthrough for deep well communication system design.

[0028] The specific steps of the application are: 1) constructing a multi-physical field coupling path loss model; 2) designing an intelligent anti-interference layered architecture; 3) using polarization diversity and deep learning channel equalization strategy.

[0029] In step 1), the wireless signal propagation in deep well environment is affected by multiple physical field coupling, and the deep well communication environment not only has NLOS multipath, but also has strong nonlinear coupling characteristics due to the support vortex effect and the time-varying nature of dust concentration. The dense distribution of metal support structures (such as steel arches and metal nets) in the roadway forms an electromagnetic shielding layer through the vortex effect, significantly changing the propagation path and energy distribution of electromagnetic waves; at the same time, high concentration of dust (mainly composed of coal dust and rock debris) causes Mie scattering and absorption effect on electromagnetic waves, further aggravating signal attenuation. In addition, the multiple reflection and diffraction phenomenon caused by the long and narrow roadway geometry leads to serious multipath fading and time delay spread, especially in non-line-of-sight (NLOS) areas, the signal energy distribution presents a high degree of non-uniformity. However, existing channel models (such as COST231 and logarithmic distance model) are based on simplified assumptions, considering the above factors as independent variables, ignoring the nonlinear superposition effect of multiple physical field coupling on signal propagation. Therefore, the application proposes a multi-physical field coupling path loss model to optimize the above problems. In order to quantify the above multiple physical field coupling effects, the application respectively establishes a metal support shielding effect model and a dust scattering effect model, and finally integrates a multi-physical field coupling path loss model.

[0030] Metal support shielding effect model: the attenuation of electromagnetic waves by metal support structure is caused by the double action of eddy current loss and reflection loss. When electromagnetic waves penetrate metal mesh, alternating magnetic field induces eddy current on the surface of the conductor, and its power loss density can be represented as: where σ is the metal conductivity, ω is the electromagnetic wave angular frequency, B is the magnetic induction effective value, d is the thickness of the support material, and p is the resistivity.

[0031] In addition, the present application defines a shielding attenuation factor as: wherein, B is the magnetic induction effective value, d is the thickness of the support material, and p is the resistivity.

[0032] Dust scattering effect model: the attenuation of electromagnetic waves by deep well dust particles (particle size r = 1-100 μm) is contributed by scattering ( ) and absorption ( ). Based on Mie scattering theory, the extinction efficiency factor can be expressed as: wherein, is the size parameter; , is the Mie coefficient, n is the order index of the series expansion, represents the real part of a complex number.

[0033] Assuming that the dust particles are uniformly distributed, the particle number density per unit volume wherein is the mass concentration (g / m³), = 2.2 g / cm 3 is the SiC density, r eff = 10 μm is the equivalent particle size.

[0034] The total attenuation factor can be expressed as: wherein N is the particle number density per unit volume, is the extinction cross section of a single particle, is the extinction efficiency factor, is the mass concentration, is the SiC density, r eff is the equivalent particle size.

[0035] Multi-physical field coupling path loss model: based on the above single factor model, a multi-physical field coupling path loss formula is proposed. Specifically, by comprehensively considering the support shielding, dust scattering and roadway geometric effect, the path loss model is corrected as: wherein, represents the free space reference loss (d0 = 1 m, λ = 0.125 m@2.4 GHz), n represents the path loss index, represents the shadow fading component, is the shielding attenuation factor, is the total attenuation factor.

[0036] In step 2), the electromagnetic interference in the deep well environment has the characteristics of mixed pulse noise (electromechanical switch transient interference) and Gaussian noise, time-varying multipath fading. For example,Figure 1 As shown, the present application proposes a three-level hierarchical anti-jamming architecture of "spread spectrum-filtering-game optimization", which realizes the collaborative optimization of interference suppression and signal recovery through phased processing.

[0037] Spread spectrum anti-narrowband interference: In the first layer, to suppress narrowband interference in the deep well electromagnetic environment, direct sequence spread spectrum (DSSS) technology is used to expand the bandwidth of the original signal to 20MHz to improve the anti-interference ability. This technology uses the low cross-correlation characteristics of pseudo-random code (PN code) to realize the despreading suppression of interference signals at the receiving end.

[0038] The system constructs a Gold code sequence based on the IEEE802.15.4 standard, with a code length of 63 and a bit rate of 10Mcps, and the processing gain is: The autocorrelation peak N of this code is 63 and the maximum cross-correlation value is ≤9, which can effectively suppress co-frequency interference signals and achieve a balance between spread spectrum modulation and signal fidelity.

[0039] Wavelet adaptive threshold filtering: The second layer is aimed at transient pulse interference and proposes an adaptive denoising method based on discrete wavelet transform (DWT). This method performs multi-scale decomposition of the signal in the wavelet domain and applies dynamic thresholds to different frequency bands to enhance the suppression ability of burst non-stationary noise.

[0040] First, the original signal is decomposed into three layers by the wavelet basis function to extract the detail coefficients; then the hybrid soft and hard threshold strategy (Hybrid Shrinkage) is used for high-frequency coefficients, that is, the hard threshold is used to suppress sharp peaks in high-intensity pulse segments, and the soft threshold is used to retain signal details in low-intensity segments.

[0041] Finally, the threshold is determined by an adaptive formula based on the median absolute deviation (MAD): , is the adaptive threshold, is the median absolute deviation of the high-frequency wavelet coefficients, which is used to robustly estimate the noise amplitude; the constant 0.6745 is a correction factor under the standard normal distribution, which is used to convert to the noise standard deviation estimate; N is the number of high-frequency coefficients after wavelet decomposition, and the term 2logN comes from the maximum likelihood threshold theory, which is used to ensure the statistical optimality of the threshold. Finally, the signal is reconstructed by inverse wavelet transform to realize effective filtering of pulse interference.

[0042] Game theory driven distributed power allocation: In deep well communication systems, multiple source nodes (such as sensors, device controllers) autonomously adjust the transmission power, which can easily cause co-channel interference and energy waste. Traditional centralized power control methods rely on a central controller, which has a single point of failure risk and poor adaptability in dynamic environments.

[0043] To this end, the present application models the multi-node power allocation problem based on non-cooperative game theory, giving each node the ability to make autonomous decisions. In this model, each communication node in the network is considered a game player, whose goal is to maximize its utility function by considering channel quality and energy consumption. Through distributed optimization, each node eventually tends to Nash equilibrium, thus achieving global performance optimization without the need for centralized coordination.

[0044] The utility function of the ith node is defined as follows: wherein, Pi is the transmission power of node i, Gi is the corresponding channel gain, N is the noise power, Ii represents interference from other nodes, Ci is the energy consumption weight coefficient.

[0045] To solve the Nash equilibrium point, the present application uses a gradient projection algorithm to iteratively update the node power. The main steps are as follows: Iterative update The formula is as follows: wherein, a is the step size coefficient, Pi represents the non-negative projection operation.

[0046] Through multiple iterations, all nodes can converge to a stable power configuration without the need for communication and negotiation, ensuring communication quality while reducing interference and energy consumption.

[0047] In step 3), in a complex deep well communication environment, signal propagation is severely affected by multipath effects and time-varying interference. To improve the channel quality and demodulation reliability at the receiving end, the present application designs a joint strategy of "physical layer anti-interference + deep learning equalizer". First, a circularly polarized (CP) antenna is used as the front-end receiving device, which effectively reduces the correlation between multipath components by utilizing the polarization orthogonality of reflected and scattered waves, thereby compressing the RMS time delay spread and alleviating inter-symbol interference (ISI).

[0048] On this basis, in order to further compensate for the residual ISI effect, the application constructs a dynamic channel equalizer based on the CNN-LSTM architecture. The network combines the local disturbance perception ability of CNN and the time series modeling advantage of LSTM, which can adaptively adjust the equalization coefficient under different dust concentrations and material reflection conditions, realize the modeling of dynamic channel state information (CSI) and symbol recovery. Finally, while maintaining high reception reliability, the joint strategy significantly improves the convergence speed and equalization accuracy.

[0049] Circular polarization antenna multipath suppression mechanism: In the NLOS scene of mobile communication, the signal passes through multiple paths to reach the receiving end, the time delay spread increases, and the inter-symbol interference (ISI) is serious. Circular polarization (CP) antenna can produce orthogonal polarization components due to its equivalent radiation to left and right polarization components, which can significantly reduce the correlation between each path component, and achieve polarization diversity effect.

[0050] To quantitatively evaluate the performance of CP antenna in suppressing multipath time delay spread and improving diversity gain, the application uses the ray tracing simulation tool WirelessInSite to compare the channel time delay characteristics of CP and traditional linear polarization (LP) antennas in the same NLOS urban street environment. The simulation results show that the RMS time delay spread of CP antenna is 47% of that of LP antenna on average, i.e. reduced by about 53%; under the same link conditions, CP antenna can achieve a diversity gain of about 8.3dB.

[0051] Figure 2 The time delay spectrum comparison and cumulative distribution curve of the two polarization modes in the simulation scene are shown. As can be seen, the circular polarization antenna significantly compresses the multipath time delay distribution by orthogonalizing the polarization of the multipath components, effectively reducing the ISI, and providing a more "flat" input channel response with smaller time delay spread for the subsequent deep learning channel equalization module.

[0052] As shown in Figure 3 , the CNN-LSTM dynamic equalizer design, the input time domain received signal can be represented as a sequence: , where T=256, indicating the input sample length. The sequence is first processed by a one-dimensional convolution layer, and the convolution kernel scans the signal sequence by sliding window to effectively extract the local disturbance features between adjacent symbols.

[0053] The convolution process can be formally represented as: , where represents the output of the cth channel at time step t, is the cth convolution kernel parameter, is the bias term, is the ReLU activation function, k denotes the convolution kernel index, and the input sequence takes the value at time step .

[0054] After the convolution operation, the output features are down-sampled by 2 1 max-pooling to compress the feature dimension, expressed as , where denote the output before and after pooling, respectively. The pooling operation reduces the computational complexity while preserving the key features, which helps to improve the generalization performance of the network.

[0055] Next, the sequence features extracted by the CNN module are sent to the LSTM for time series modeling. The internal update mechanism of the LSTM is as follows: , where is the current input, is the output, is the hidden state of the previous time step, is the memory state, is the memory state of the previous time step, is the forget gate vector, is the input gate vector, is the output gate vector, is the input weight matrix, is the hidden layer weight matrix, is the bias vector, denotes the Sigmoid function, and denotes the element-wise multiplication.

[0056] Through the above gating mechanism, the evolution characteristics of the channel state information (CSI) in the time dimension are captured, and the hidden state hth_tht is output as the basis for the equalization output.

[0057] Finally, the network sends the output features of the LSTM to the fully connected layer to predict the equalization compensation coefficient: , where fc denotes the fully connected layer, is the equalized symbol estimate value, is the weight matrix of the fc layer, is the bias term of fc, is the hidden state output vector of the LSTM at time step t.

[0058] After multi-physical field coupling modeling, intelligent anti-interference layered processing, and polarization diversity dynamic equalization three-stage optimization, the method proposed in the present application can realize stable and efficient wireless communication in deep well complex electromagnetic environment, and the obtained transmission performance result has important reference value in mine safety monitoring and emergency command.

[0059] In this application, in order to evaluate the performance of the proposed "multi-physical field coupling path loss model + CP + CNN-LSTM joint anti-interference strategy" in complex deep well communication environment, the following three-layer simulation platform is built, and the process is executed in turn.

[0060] Firstly, a 100m×4m×3m three-dimensional mine roadway model is constructed in COMSOL Multiphysics, Figure 4 which is embedded with steel (thickness 5mm, electrical conductivity ) and glass steel (thickness 10mm, ) support structure respectively to simulate the shielding effect of different materials on 2.4GHz electromagnetic waves. The domain boundary is set as a perfect matched layer (PML) to absorb the outgoing wave and avoid boundary reflection; a dipole antenna with power of 20dBm is placed at the transmitting end, and a probe is placed every 10m along the center line of the roadway at the receiving end. Subsequently, through the joint simulation of "electromagnetic wave-frequency domain" module and "particle tracking" module, the multi-physical field coupling calculation of Mie scattering and absorption of 10μm dust (concentration 50-300g / m³) is realized, so as to obtain the spatial field distribution and power delay spectrum.

[0061] Then, on the MATLAB / Simulink platform, a baseband signal of 1Msps, QPSK modulation, and frame length of 256 symbols is generated, and AWGN (SNR 5-15dB) and pulse noise with duty cycle of 10% and peak power of-10dBm are injected into the channel to simulate the complex mixed interference in deep well environment. The received waveform and the corresponding COMSOL simulation impulse response (100000 groups in total) are used to form a training set, and a CNN-LSTM dynamic equalizer is built in the PyTorch framework and trained end-to-end, with the weighted sum of MSE and cross entropy as the optimization target.

[0062] Finally, in order to verify the multipath suppression effect and diversity gain of circularly polarized (CP) antenna, ray tracing simulation is performed in WirelessInSite with the same geometry and material parameters in COMSOL. The wall roughness is set to 0.1m to simulate irregular reflection, and the transmission link of linearly polarized (LP) and circularly polarized (CP) antennas is simulated respectively, and the RMS time delay spread and diversity gain of each receiving point are calculated.

[0063] To further highlight the modeling advantages of multi-physical field coupling channel model in deep well environment, the path loss of three representative scenarios is systematically compared and analyzed.

[0064] In scenario ① (empty tunnel), since there is no dust and the supporting material has very low conductivity, electromagnetic wave propagation mainly follows free space geometric attenuation, and both the traditional Log-Distance Path Loss Model and the coupling model of the application can well reproduce the COMSOL multi-physics simulation results in this scenario.

[0065] In scenario ② (high concentration of dust, non-conductor support), the Mie scattering and absorption effects of high-concentration particles begin to significantly enhance the non-linear growth of path loss Figure 5 ), resulting in an average error of about 4.2dB for the traditional Log-Distance model due to only considering geometric attenuation; in contrast, the dust attenuation term introduced in the coupling model accurately describes the scattering and absorption contributions, reducing the average error to within 1.5dB (about 64% reduction)

[0066] To more clearly present the comparison results of the most complex scenario ③ (high concentration of dust + high-conductive metal support), Figure 6 The path loss curve under this working condition is specially selected. The three broken lines in the figure represent the COMSOL multi-physics simulation results, the traditional Log-Distance model prediction value, and the multi-physics coupling model output proposed by the application, respectively, so that the comparison can be made directly: in the most complex environment, the traditional model severely underestimates the attenuation due to the neglect of dust scattering and metal eddy current shielding two physical mechanisms; while the coupling model realizes high-precision fitting of the COMSOL simulation points by superimposing the Mie scattering and eddy current loss terms in the Log-Distance formula, and its curve almost coincides with the simulation result, with an average fitting error of only 2.1dB.

[0067] As shown in Table 1, comparing the average errors of the traditional model and the coupling model under three scenarios, the coupling model is superior to the traditional model in all working conditions, especially in complex coupling conditions.

[0068] Table 1. Comparison of average path loss model errors under different scenarios

[0069] In scenario ①, due to the simple environment and no additional physical coupling, the coupling model can naturally degrade to the level of the traditional model.

[0070] In scenario ②, the dust scattering effect is prominent, and the coupling model reduces the error from 4.2dB to 1.5dB by adding the dust attenuation term.

[0071] In scenario ③, the most complex physical mechanism coexists, and the coupling model still maintains accurate reproduction of real attenuation, significantly better than the traditional model. The results fully prove that by incorporating the physical loss mechanisms of metal shielding and dust scattering into the path loss model, more accurate channel modeling can be achieved in deep well environments with multiple physical field interlacing, providing a reliable theoretical basis for subsequent channel simulation and equalization algorithm design.

[0072] To fully verify the effectiveness of the proposed "spread spectrum anti-jamming-wavelet filtering-game power control" three-level intelligent anti-jamming architecture, the application carries out experiments from three aspects of anti-jamming performance, signal recovery quality and power consumption control, covering typical deep well communication environment common narrowband interference, impulse noise and multi-node co-channel interference scenarios.

[0073] (1) Spread spectrum anti-jamming performance comparison and analysis: Under the test condition of SIR (Signal-to-Interference Ratio) of -5dB, the bit error rate performance of traditional baseband communication and the adopted DSSS (Direct Sequence Spread Spectrum) system under spectral interference condition is compared. As shown in Table 2, the DSSS system relies on Gold code spread spectrum (processing gain up to 20dB) to significantly reduce the bit error rate, and the BER remains at the order of 10 -3 under strong interference, far better than the 10 -1 order performance of the non-spread spectrum system.

[0074] Table 2. BER comparison under narrowband interference (SIR = -5dB)

[0075] (2) Wavelet filtering anti-impulse interference performance comparison and analysis: For typical non-Gaussian impulse interference in deep wells, a burst pulse signal interference scenario with 10% duty cycle is designed, and three wavelet denoising strategies are adopted: no filtering, fixed threshold denoising, and hybrid threshold (hard + soft) denoising. The mean square error (MSE) performance of 1000 samples is compared. As shown in Table 3, the proposed hybrid threshold method effectively suppresses sharp peak interference while maintaining low distortion, with an average MSE of 1.2x10 -3 , better than other methods, showing stronger robustness and adaptability.

[0076] Table 3. Comparison of mean square error of different wavelet denoising methods under pulse noise (10% duty cycle)

[0077] (3) Game theory driven distributed power control performance analysis: In the typical multi-node co-channel environment, a distributed network system containing 6 nodes is constructed, the transmission power of each node is randomly initialized, and the power strategy is iteratively solved based on the non-cooperative game model. Based on the comparison of centralized optimal power control and traditional fixed power method, the convergence, total energy consumption and SINR level of the proposed algorithm are analyzed.

[0078] As shown in Figure 7 , the game control strategy realizes rapid convergence within 20 iterations, and the power of each node tends to be balanced and stable. As shown in Table 4, the proposed algorithm realizes about 23.5% optimization in total power consumption control, while effectively improving the average SINR, verifying its adaptability and high efficiency in complex interference scenarios.

[0079] Table 4. Comparison and analysis of power control performance in multi-node co-channel interference scenarios

[0080] To evaluate the effectiveness of the proposed "circular polarization + CNN-LSTM" joint anti-interference strategy in typical deep well environment, a set of system comparison experiments are designed on the MATLAB communication link simulation platform, covering static channel (fixed dust concentration and support type) and dynamic channel (interference intensity and environmental parameter change) two test scenarios.

[0081] In the static channel test, the convergence performance and recovery accuracy of the traditional LMS, MMSE and the proposed CNN-LSTM equalizer under the condition of SNR=10dB and pulse noise duty cycle 10% are compared. As shown in Figure 8 , the NMSE of LMS and MMSE equalizers converges slowly, respectively tending to 0.035 and 0.024 after 50 Epochs, while the CNN-LSTM model can converge to 0.014 after about 20 Epochs, which is 42.5% lower than MMSE and 60.1% lower than LMS. As shown in Table 5, the comparison of the final NMSE and static average BER of the three further verifies the advantage of CNN-LSTM in static deep well channel: its BER is controlled at 4.7×10 -3 The following is significantly better than the traditional method.

[0082] Table 5. Comparison of final NMSE and BER performance of three equalizers under static test conditions

[0083] In the dynamic channel test, the application simulates the sudden scene of dust concentration from 50g / m³ to 300g / m³, and the interference intensity from 85dBμV / m to 95dBμV / m. AsFigure 9 As shown, when the frame number exceeds 50, the BER of the traditional equalizer rises obviously and is almost impossible to recover in the subsequent frames, while the CNN-LSTM equalizer only needs 3 frame updates to complete adaptive adjustment, and the BER fluctuation is stably controlled within 0.01, showing excellent real-time interference suppression ability and environmental adaptability.

[0084] In summary, the experiment fully verifies the robustness and generalization ability of the intelligent anti-interference architecture in the variable deep well environment, especially in the coupling noise source, nonlinear interference and time-varying channel conditions, which has significant performance advantages.

[0085] The above only describes one embodiment of the present application and does not limit the present application. Any modification, equivalent replacement or improvement within the technical scheme and principle of the present application, especially the principle application and design method, such as the principle and specific implementation method of the present application design, should be included in the protection scope of the present application.

Claims

1. An intelligent anti-jamming method of multi-physical field coupling and polarization diversity, characterized in that, The method comprises the following steps: A multi-physics coupling path loss model is constructed, based on the eddy current shielding effect of metal support materials and the Mie scattering theory of dust concentration; An intelligent anti-interference layered architecture is designed, including spread spectrum anti-narrowband interference, wavelet adaptive filtering to suppress impulse noise, and distributed power control based on game theory; Polarization diversity and deep learning channel equalization strategies are used to suppress multipath effects through circularly polarized antennas, and dynamic channel equalization is performed using a CNN-LSTM network.

2. The intelligent anti-jamming method of multi-physics coupling and polarization diversity according to claim 1, characterized in that, The construction of the multi-physics coupling path loss model comprises: A metal support shielding effect model is established, based on eddy current loss and reflection loss to calculate a shielding attenuation factor; A dust scattering effect model is established, based on the Mie scattering theory to calculate an extinction efficiency factor and a total attenuation factor; The shielding attenuation factor and the total attenuation factor are integrated into the path loss formula to form a revised path loss model. 3.The intelligent anti-interference method of multi-physics field coupling and polarization diversity according to claim 1, characterized in that, The intelligent anti-interference layered architecture comprises: The first layer uses direct sequence spread spectrum technology, using Gold code sequences for spread spectrum modulation to suppress narrowband interference; The second layer uses a discrete wavelet transform-based adaptive threshold denoising method to filter impulse noise; The third layer constructs a distributed power control model based on non-cooperative game theory, iteratively optimizing the transmit power of each node to maximize the system utility function.

4. The intelligent anti-jamming method of multi-physics field coupling and polarization diversity according to claim 3, characterized in that, The processing gain of the direct sequence spread spectrum technology is not less than 20 dB, the Gold code length is 63, and the maximum cross-correlation value is not greater than 7.

5. The intelligent anti-jamming method of multi-physics field coupling and polarization diversity according to claim 3, characterized in that, The adaptive threshold denoising method uses a hybrid soft and hard threshold strategy, with the threshold dynamically determined based on the median absolute deviation.

6. The intelligent anti-jamming method of multi-physics field coupling and polarization diversity according to claim 3, characterized in that, The utility function of the distributed power control model considers channel gain, noise power, interference power, and energy consumption weight, and iteratively solves the Nash equilibrium point using the gradient projection algorithm.

7. The intelligent anti-jamming method of multi-physics field coupling and polarization diversity according to claim 1, characterized in that, The polarization diversity and deep learning channel equalization strategy comprises: A circularly polarized antenna is used to receive signals to reduce the correlation between multipath components and achieve polarization diversity; A CNN-LSTM dynamic equalizer is constructed, including a one-dimensional convolution layer, a max pooling layer, an LSTM layer, and a fully connected layer, for time series modeling and symbol recovery of received signals.

8. The intelligent anti-jamming method of multi-physics field coupling and polarization diversity according to claim 7, characterized in that, The input of the CNN-LSTM dynamic equalizer is a time-domain signal sequence with a length of 256, and the output is the equalized symbol estimate value.

9. A wireless communication system configured to implement the method of any one of claims 1 to 8, characterized by The system comprises: A transmitting module for generating and transmitting DSSS modulated signals; A receiving module comprising a circularly polarized antenna and a CNN-LSTM dynamic equalizer; A processing module for performing wavelet denoising, power control, and channel equalization operations.

10. The wireless communication system of claim 9, wherein, The system is suitable for deep well safety monitoring, emergency communication, or intelligent mine wireless communication scenarios.

Images