A low voltage communication demodulation platform

Abstract

The present invention provides a low-voltage communication demodulation platform, which relates to the field of communication technology and comprises a demodulation main system. The demodulation main system comprises a signal acquisition module, a data storage module, a clock management module, an interface module, a fault detection and recovery module, a protocol conversion module, a signal processing module, a spectrum management module, a safety communication module, a power grid environment adaptation module and an intelligent demodulation module. The content of the intelligent demodulation module is a convolutional neural network (CNN) and a support vector machine (SVM) to realize adaptive modulation recognition and demodulation. The CNN is used to extract the characteristics of the carrier signal, and the SVM performs modulation mode classification. The intelligent demodulation module is combined with a Kalman filter through a generative adversarial network (GAN) to realize channel noise processing and enhancement. The GAN is used to generate a low-noise signal, and the Kalman filter further optimizes the signal quality.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a low-voltage communication demodulation platform. Background Art

[0002] According to a demodulation method and demodulation system of a spectrum domain communication signal with a publication number of WO2012075952A1, the method comprises: step 1, a cyclic autocorrelation function estimation value processing unit calculates the estimated value of the corresponding cyclic autocorrelation function according to a received continuous CDM-OFDM cyclic delay modulation signal of L OFDM cycles; step 2, a cyclic delay detection processing unit calculates the corresponding likelihood function for different cyclic delay values, and obtains the corresponding cyclic delay deficiency according to the maximum likelihood decision criterion; step 3, a cyclic modulation vector demapping processing unit maps the cyclic delay vector to obtain the transmitted cyclic modulation signal. The present invention can be applied to spectrum domain channel multiplexing transmission, and increases the system capacity while not affecting the transmission of ordinary OFDM signals; a block calculation method is adopted, which greatly reduces the calculation complexity and saves processing time and resources.

[0003] The above patent documents and prior art have the following technical problems when used:

[0004] Problem 1: The existing technology increases the capacity of the system by adding one-dimensional spectrum domain communication signal transmission on the basis of ordinary OFDM signal transmission, but its spectrum allocation strategy is relatively static and lacks sufficient flexibility and adaptability. Especially in multi-user scenarios or complex communication environments, the spectrum allocation of existing technologies is difficult to dynamically adjust according to real-time communication needs, resulting in low utilization efficiency of spectrum resources and limiting the scalability of the system in high-density user environments;

[0005] Problem 2: Although the existing technology has reduced some computational complexity by optimizing the cyclic delay modulation algorithm, the computational requirements are still high for communication systems that handle complex channel environments, especially in real-time applications. This computational overhead may significantly affect the response speed and overall performance of the system. The existing demodulation algorithm still consumes a lot of hardware resources and computing power when processing large-scale and diverse signals, which reduces the actual application efficiency of the system;

[0006] Problem three: Existing technologies mainly focus on optimizing signal modulation and demodulation technologies, but do not provide effective solutions for the security and anti-interference capabilities of communication systems. In complex communication environments, security threats such as electromagnetic interference and signal interception are becoming increasingly serious. Existing technologies lack effective encryption mechanisms and anti-interference measures, which means that the security of communication data cannot be fully guaranteed.

[0007] Question 4: In low-voltage DC carrier communication systems, fluctuations in the power grid environment and noise interference often have a significant impact on communication quality. Existing technologies lack effective channel equalization and adaptive compensation mechanisms when facing changes in the power grid environment, resulting in poor stability of communication signals when there is significant power grid interference, affecting system reliability, especially when applied in actual power grid scenarios. Summary of the invention

[0008] Technical issues solved

[0009] In view of the deficiencies in the prior art, the present invention provides a low-voltage communication demodulation platform to solve the problems in the above background technology.

[0010] Technical Solution

[0011] To achieve the above objectives, the present invention is implemented through the following technical solutions: a low-voltage communication demodulation platform, including a demodulation main system, the demodulation main system includes a signal acquisition module, a data storage module, a clock management module, an interface module, a fault detection and recovery module, a protocol conversion module, a signal processing module, a spectrum management module, a safety communication module, a power grid environment adaptation module and an intelligent demodulation module;

[0012] Sp1, the signal acquisition module uses the Kalman filter algorithm to perform real-time dynamic prediction and correction on the collected signal, remove noise interference and estimate the true value of the signal, and uses the wavelet transform algorithm to perform multi-scale decomposition on the signal to extract the signal characteristics of different frequency components, so as to facilitate the detection of mutation points and abnormal conditions in the signal;

[0013] Sp2, the data storage module combines the hash algorithm, the tree index algorithm and the data compression algorithm to store data, quickly locates and searches for data through the hash algorithm, improves the efficiency of data retrieval, and uses the tree index algorithm to organize the stored data in an orderly manner;

[0014] Sp3, the clock management module integrates the phase-locked loop (PLL) algorithm, the delay-locked loop (DLL) algorithm and the clock synchronization protocol algorithm, generates a stable high-frequency clock signal through the phase-locked loop (PLL) algorithm, and phase-locks it with the reference clock;

[0015] Sp4, the interface module identifies and parses different communication protocols through a protocol parsing algorithm, uses a data cache algorithm to temporarily store data to be sent or received to adapt to the differences between different data transmission rates and prevent data loss, and adjusts the data transmission rate on the interface through a flow control algorithm to avoid data congestion and transmission errors;

[0016] Sp5, the fault detection and recovery module combines the failure mode and effect analysis (FMEA) algorithm, the neural network fault diagnosis algorithm and the fault-tolerant redundancy algorithm to identify and evaluate potential failure modes through various components and functions of the FMEA algorithm system and establish a failure mode library;

[0017] Sp6, the protocol conversion module adopts a syntax analysis algorithm, a semantic mapping algorithm and an intermediate code generation algorithm. The syntax analysis algorithm parses the syntax structure of the source protocol and the target protocol to understand the organization form of the data. The semantic mapping algorithm converts the semantic information in the source protocol into semantics that the target protocol can understand, ensuring that the meaning of the data is not lost during the conversion process. The intermediate code generation algorithm generates code in the intermediate representation form based on syntax analysis and semantic mapping;

[0018] Sp7, the intelligent demodulation module realizes adaptive modulation recognition and demodulation through convolutional neural network (CNN) and support vector machine (SVM), introduces deep Q network (DQN) in reinforcement learning, uses CNN to extract features of received signals, inputs the extracted features into SVM for preliminary modulation type classification, and then DQN optimizes the demodulation strategy according to the output of SVM and historical demodulation experience, so as to realize accurate adaptive modulation recognition and demodulation;

[0019] Sp8, the spectrum management module dynamically allocates and manages the spectrum through the combination of Q learning and genetic algorithm, and introduces the simulated annealing algorithm at the same time, explores the state and action space of spectrum use through Q learning, obtains the preliminary strategy of spectrum allocation, crosses and mutates different spectrum allocation schemes through the genetic algorithm, finds the global optimal solution, and performs local search near the solution space searched by the genetic algorithm through the simulated annealing algorithm;

[0020] Sp9, the secure communication module achieves high security and anti-interference capability for dual-mode communication signals by combining chaos theory with low-density parity check (LDPC) error correction coding. At the same time, combined with the quantum encryption algorithm, the chaos theory encrypts the communication signal, and the LDPC error correction coding detects and corrects possible errors during signal transmission;

[0021] Sp10. The power grid environment adaptation module adopts a combination of convolutional neural network (CNN) and long short-term memory network (LSTM) to achieve channel equalization and compensation, and introduces particle swarm optimization algorithm (PSO). The CNN extracts the spatial characteristics of the power grid channel, and the LSTM processes the channel change information on the time series to achieve channel equalization and compensation. PSO optimizes the network parameters of CNN and LSTM, finds the best model structure and parameter configuration, and improves the performance of channel equalization and compensation. Under different power grid load and noise conditions, PSO is based on the real-time monitoring data of the channel.

[0022] Preferably, the signal acquisition module continuously updates the estimate of the signal state through Kalman filtering based on the prior knowledge and observation data of the signal, so that the collected signal is smoother and more accurate, decomposes the signal into different frequency sub-bands through wavelet transform, facilitates the analysis of the characteristics of the signal at different scales, detects transient interference in the signal, and uses an adaptive threshold algorithm to analyze the statistical information of the amplitude change of the signal, determines the threshold, removes the noise components below the threshold, and retains the useful signal part.

[0023] Preferably, when the clock management module is applied, a phase-locked loop (PLL) algorithm is used to compare the phase difference between the output clock and the reference clock, and the output frequency of the voltage-controlled oscillator is adjusted so that the output clock and the reference clock are phase-locked. A delay-locked loop (DLL) algorithm is used to adjust the delay of the clock signal so that the phase of the local clock is aligned with the input clock. The clock synchronization protocol algorithm exchanges time information through network communication and calculates the clock deviation and adjustment amount according to the algorithm specified in the protocol.

[0024] Preferably, the intelligent demodulation module combines a generative adversarial network (GAN) and a Kalman filter to enhance the processing of channel noise. The generative adversarial network (GAN) generates a low-noise version of the signal, and the Kalman filter further optimizes the quality of the demodulated signal. The fuzzy logic in the signal processing module is combined with a Bayesian network.

[0025] Preferably, the spectrum management module monitors spectrum resources in real time through Q learning, dynamically adjusts communication channels, and uses genetic algorithms to optimize spectrum allocation strategies to avoid spectrum interference and congestion.

[0026] Preferably, the secure communication module combines quantum key distribution (QKD) and elliptic curve cryptography (ECC) for communication encryption. QKD provides secure key generation and distribution, and ECC is used to efficiently encrypt data to ensure the security and efficiency of data transmission.

[0027] Preferably, the power grid environment adaptation module combines adaptive filtering (RLS) with a genetic algorithm to enhance the anti-interference capability of the system. The RLS filter adjusts the filtering parameters in real time to cope with sudden interference in the power grid, and the genetic algorithm optimizes the configuration of the filter through global search.

[0028] Preferably, the convolutional neural network (CNN) in the intelligent demodulation module is pre-trained and can adaptively adjust network parameters according to historical data in the power grid to improve the accuracy and efficiency of modulation recognition.

[0029] Preferably, the signal processing module combines a weighted average filter with a minimum mean square error (LMS) adaptive filter to further optimize the demodulated signal. The weighted average filter is used to suppress noise, and the LMS filter dynamically adjusts the filter weight according to the error signal to ensure signal stability. During the spectrum allocation process, the spectrum management module combines an adaptive scheduling mechanism based on reinforcement learning with genetic algorithm optimization to further improve spectrum utilization efficiency and communication stability.

[0030] Beneficial Effects

[0031] The present invention provides a low-voltage communication demodulation platform. It has the following beneficial effects:

[0032] 1. The present invention can adaptively perform spectrum dynamic allocation and optimization through the joint application of intelligent demodulation module, signal processing module and spectrum management module, especially combining Q learning and genetic algorithm in spectrum management. This improved solution can manage spectrum resources more flexibly and intelligently, especially in a multi-user communication environment, further improving the system capacity and spectrum utilization efficiency.

[0033] 2. The present invention introduces a combination of convolutional neural network (CNN) and support vector machine (SVM) in the demodulation process, adaptively identifies and demodulates the modulated signal through an intelligent algorithm, and combines FFT and Viterbi decoding for efficient signal processing. This technical solution greatly reduces the complexity of the demodulation algorithm, makes signal processing more efficient in a low-voltage DC carrier environment, and effectively improves the response speed of the system in real-time communication. In addition, the algorithm of the present invention reduces the demand for hardware resources through adaptive adjustment, further optimizes the system's computational overhead, and ensures its suitability for the actual deployment of large-scale communication systems.

[0034] 3. In terms of the security of the communication system, the present invention adopts chaos theory combined with low-density parity check (LDPC) error correction coding to effectively enhance the system's anti-interference ability and data security. In a complex power grid environment, the present invention further combines convolutional neural networks (CNN) and long short-term memory networks (LSTM) to develop a power grid environment adaptation module that can monitor and compensate for changes in power grid channels in real time, ensuring stable transmission of signals in high-noise and multi-interference environments. This makes the present invention have obvious advantages over existing technologies in dealing with power grid fluctuations and improving communication reliability and stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a diagram of the main demodulation system of the present invention;

[0036] Figure 2 It is an application logic diagram of the signal acquisition module, data storage module and clock management module of the present invention;

[0037] Figure 3 The interface module, fault detection and recovery module, and protocol conversion module of the present invention are applied to the logic diagram;

[0038] Figure 4 It is an application logic diagram of the intelligent demodulation module, spectrum management module and secure communication module of the present invention;

[0039] Figure 5 It is a hardware composition diagram of each module of the system of the present invention;

[0040] Figure 6 This is a spectrum simulation diagram of the dual-mode modulation signal of the present invention. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Specific embodiment one:

[0043] like Figures 1 to 6 As shown, a low-voltage communication demodulation platform includes a demodulation main system, which includes a signal acquisition module, a data storage module, a clock management module, an interface module, a fault detection and recovery module, a protocol conversion module, a signal processing module, a spectrum management module, a safety communication module, a power grid environment adaptation module and an intelligent demodulation module;

[0044] Sp1, the signal acquisition module uses the Kalman filter algorithm to perform real-time dynamic prediction and correction on the collected signals, remove noise interference and estimate the true value of the signal, use the wavelet transform algorithm to perform multi-scale decomposition on the signal, extract the signal characteristics of different frequency components, and facilitate the detection of mutation points and abnormal conditions in the signal. At the same time, the adaptive threshold algorithm is used to dynamically adjust the threshold according to the statistical characteristics of the signal to identify and remove the noise and interference components in the signal;

[0045] Sp2, the data storage module combines hash algorithm, tree index algorithm and data compression algorithm to store data, quickly locates and searches data through hash algorithm, improves the efficiency of data retrieval, uses tree index algorithm to organize the stored data in order, facilitates range query and fast traversal, and uses data compression algorithm to reduce the space required for data storage and improve the utilization rate of storage resources;

[0046] Sp3, the clock management module integrates the phase-locked loop (PLL) algorithm, the delay-locked loop (DLL) algorithm and the clock synchronization protocol algorithm. The phase-locked loop (PLL) algorithm generates a stable high-frequency clock signal and phase-locks it with the reference clock to ensure the accuracy and stability of the clock. The delay-locked loop (DLL) algorithm adjusts the clock phase through the delay line, accurately controls the delay of the clock signal, and compensates for the clock offset caused by factors such as the internal wiring of the chip. The clock synchronization protocol algorithm achieves constant synchronization in the distributed system to ensure the consistency of the system clock;

[0047] Sp4, the interface module identifies and parses different communication protocols through the protocol parsing algorithm, uses the data cache algorithm to temporarily store the data to be sent or received to adapt to the differences between different data transmission rates and prevent data loss, and adjusts the data transmission rate on the interface through the flow control algorithm to avoid data congestion and transmission errors;

[0048] Sp5, the fault detection and recovery module combines the failure mode and effect analysis (FMEA) algorithm, the neural network fault diagnosis algorithm and the fault-tolerant redundancy algorithm to identify and evaluate potential failure modes through the various components and functions of the FMEA algorithm system, and establish a failure mode library. During the operation of the system, the neural network fault diagnosis algorithm uses the system operation data collected by the sensor, including temperature, voltage and current, for real-time analysis and diagnosis. When a fault is detected, the fault-tolerant redundancy algorithm switches to the redundant hardware module or uses the backup data according to the fault type and severity;

[0049] Sp6. The protocol conversion module adopts a syntax analysis algorithm, a semantic mapping algorithm and an intermediate code generation algorithm. The syntax analysis algorithm parses the syntax structure of the source protocol and the target protocol and understands the organizational form of the data. The semantic mapping algorithm converts the semantic information in the source protocol into semantics that the target protocol can understand, ensuring that the meaning of the data is not lost during the conversion process. The intermediate code generation algorithm generates code in an intermediate representation form based on syntax analysis and semantic mapping to facilitate subsequent conversion and processing. The syntax analysis algorithm in the protocol conversion module parses the data format of the source protocol, builds a syntax tree, understands the hierarchical structure and field meaning of the data, and uses the semantic mapping algorithm to convert the data semantics in the source protocol into the semantics corresponding to the target protocol according to the pre-defined semantic mapping rules. The intermediate code generation algorithm converts the semantically mapped data into an intermediate code form. Finally, according to the syntax rules of the target protocol, the intermediate code is converted into a data format that meets the requirements of the target protocol for transmission or processing;

[0050] Sp7, the intelligent demodulation module realizes adaptive modulation recognition and demodulation through convolutional neural network (CNN) and support vector machine (SVM), introduces deep Q network (DQN) in reinforcement learning, uses CNN to extract features of received signals, inputs the extracted features into SVM for preliminary classification and judgment of modulation type, and then DQN further optimizes the demodulation strategy based on the output of SVM and historical demodulation experience to realize accurate adaptive modulation recognition and demodulation. CNN is responsible for automatically learning and extracting effective feature representations from input signals, and SVM makes classification decisions based on these features to judge the modulation type of the signal. On this basis, DQN adjusts the demodulation strategy based on reward feedback through continuous interaction with the demodulation process;

[0051] Sp8, spectrum management module dynamically allocates and manages spectrum through the combination of Q learning and genetic algorithm, and introduces simulated annealing algorithm at the same time. Through Q learning, the state and action space of spectrum usage are explored to obtain the preliminary strategy of spectrum allocation. Through genetic algorithm, different spectrum allocation schemes are crossover and mutation operations are performed to find the global optimal solution. Through simulated annealing algorithm, local search is performed near the solution space searched by genetic algorithm to avoid falling into the local optimal solution, and the dynamic allocation and management of spectrum is optimized. During the application process, Q learning interacts with the spectrum environment and continuously updates the Q value table to determine what actions to take under different spectrum states to obtain the maximum long-term return. Genetic algorithm uses its population evolution characteristics to search and optimize multiple spectrum allocation schemes in parallel, and generates better offspring schemes through selection, crossover and mutation operations.

[0052] Sp9, the secure communication module achieves high security and anti-interference capability for dual-mode communication signals by combining chaos theory with low-density parity check (LDPC) error correction coding. At the same time, combined with the quantum encryption algorithm, chaos theory encrypts the communication signal, and LDPC error correction coding detects and corrects possible errors during signal transmission. The quantum encryption algorithm uses the characteristics of quantum mechanics to provide an absolutely secure key distribution mechanism for communication. When the two communicating parties exchange keys, the quantum encryption algorithm is used to generate a secure key, and then the communication data is encrypted using chaos theory, and finally transmitted through LDPC coding. The chaotic sequence generated by the chaotic system is used as the encryption key to encrypt the original communication signal. LDPC coding encodes the encrypted signal at the sending end and adds redundant information to facilitate the detection and correction of errors introduced during the transmission process at the receiving end. The quantum encryption algorithm realizes secure key distribution through the quantum channel to ensure the security and randomness of the key. The receiving end first uses the quantum encryption algorithm to obtain the correct key, and then uses LDPC decoding to restore the encrypted signal, and finally restores the original communication data through the chaos decryption algorithm;

[0053] Sp10, the power grid environment adaptation module uses a combination of convolutional neural network (CNN) and long short-term memory network (LSTM) to achieve channel equalization and compensation, and introduces particle swarm optimization algorithm (PSO). CNN extracts the spatial characteristics of the power grid channel, and LSTM processes the channel change information on the time series to achieve channel equalization and compensation. PSO optimizes the network parameters of CNN and LSTM, finds the best model structure and parameter configuration, and improves the performance of channel equalization and compensation. Under different power grid load and noise conditions, PSO automatically adjusts the parameters of CNN and LSTM according to the real-time monitoring data of the channel to improve the communication quality. LSTM uses its memory unit and gating mechanism to model and predict the channel change information on the time series, and balances and compensates the current signal more accurately according to the historical channel status. The PSO algorithm searches for the optimal combination of CNN and LSTM network parameters in the parameter space by simulating the foraging behavior of bird flocks. Each particle represents a set of parameters. The speed and position of the particle are adjusted according to the fitness of the particle, that is, the bit error rate index after channel equalization.

[0054] The signal acquisition module uses Kalman filtering based on the prior knowledge and observation data of the signal to continuously update the estimation of the signal state, making the collected signal smoother and more accurate. The signal is decomposed into different frequency sub-bands through wavelet transform, which is convenient for analyzing the characteristics of the signal at different scales and detecting transient interference in the signal. The adaptive threshold algorithm is used to analyze the statistical information of the amplitude change of the signal, determine the threshold, and remove the noise components below the threshold while retaining the useful signal part. When writing data, the data storage module first calculates the hash value of the data through the hash algorithm and stores it as an index in the hash table for fast search. At the same time, the data is stored in the B+ tree structure according to the rules. The leaf nodes of the B+ tree contain the actual data records, and the non-leaf nodes are used for indexing and fast positioning. The data compression algorithm is used to compress the data before storing it to reduce the storage space occupied, and then decompress the data when reading it.

[0055] When the clock management module is applied, the phase-locked loop (PLL) algorithm compares the phase difference between the output clock and the reference clock, adjusts the output frequency of the voltage-controlled oscillator, and keeps the output clock and the reference clock in phase lock. The delay-locked loop (DLL) algorithm is used to delay and adjust the clock signal so that the local clock is aligned with the input clock. The clock synchronization protocol algorithm exchanges time information through network communication, calculates the clock deviation and adjustment amount according to the algorithm specified in the protocol, and keeps the clocks of each node synchronized. When the interface module is applied, when data is transmitted through the interface, the protocol parsing algorithm first identifies the protocol format used by the data, parses and encapsulates the data according to the protocol specification, and uses the data caching algorithm to temporarily store the data in the cache when the data transmission rate does not match, waiting for the right time to send and process it. At the same time, the flow control algorithm monitors the bandwidth usage and buffer status of the interface, and uses the token bucket algorithm and sliding window protocol mechanism to control the data transmission rate to ensure the stability and reliability of data transmission.

[0056] The intelligent demodulation module combines the generative adversarial network (GAN) and the Kalman filter to enhance the processing of channel noise. The generative adversarial network (GAN) generates a low-noise version of the signal, and the Kalman filter further optimizes the quality of the demodulated signal. The fuzzy logic in the signal processing module is combined with the Bayesian network to achieve intelligent switching between ASK and FSK dual-mode demodulation methods. The fuzzy logic handles the uncertainty of the channel state, and the Bayesian network is used to make inferences and decisions based on signal characteristics.

[0057] The spectrum management module uses Q learning to monitor spectrum resources in real time and dynamically adjust communication channels. Genetic algorithms are used to optimize spectrum allocation strategies to avoid spectrum interference and congestion.

[0058] The secure communication module combines quantum key distribution (QKD) and elliptic curve cryptography (ECC) for communication encryption. QKD provides secure key generation and distribution, and ECC is used to efficiently encrypt data to ensure the security and efficiency of data transmission.

[0059] The grid environment adaptation module combines adaptive filtering (RLS) with genetic algorithm to enhance the anti-interference ability of the system. The RLS filter adjusts the filtering parameters in real time to cope with sudden interference in the grid, and the genetic algorithm optimizes the configuration of the filter through global search.

[0060] The convolutional neural network (CNN) in the intelligent demodulation module is pre-trained and can adaptively adjust network parameters according to historical data in the power grid to improve the accuracy and efficiency of modulation recognition.

[0061] The signal processing module combines a weighted average filter with a least mean square error (LMS) adaptive filter to further optimize the demodulated signal. The weighted average filter is used to suppress noise, and the LMS filter dynamically adjusts the filter weight according to the error signal to ensure signal stability. During the spectrum allocation process, the spectrum management module combines an adaptive scheduling mechanism based on reinforcement learning with genetic algorithm optimization to further improve spectrum utilization efficiency and communication stability.

[0062] In this low-voltage DC carrier communication dual-mode demodulation system, each module works closely together to complete the processing, transmission and guarantee of communication signals. The signal acquisition module obtains the original signal from the low-voltage DC power line and performs preliminary processing, and then passes the signal to the intelligent demodulation module, signal processing module, spectrum management module, security communication module and power grid environment adaptation module for corresponding processing. In the whole process, the data storage module is used to store the relevant data generated by each module, the clock management module provides the system with an accurate clock signal, the interface module is responsible for data interaction with external devices, the fault detection and recovery module monitors the system operation status in real time and recovers when a fault occurs, and the protocol conversion module handles the conversion between different protocols. Specific embodiment 2:

[0064] like Figures 1 to 6 As shown, the key algorithm mentioned in Example 1 is analyzed in detail below, including its core mathematical formula and explanation:

[0065] 1.Signal acquisition module

[0066] Kalman Filter Formula

[0067] Status prediction:

[0068]

[0069] Update steps:

[0070]

[0071]

[0072] The status of the forecast; State transition matrix; : Control input matrix; Control input; Forecast error covariance; Kalman gain; Observation value; Observation matrix; Observation noise covariance.

[0073] Wavelet Transform:

[0074] Continuous wavelet transform formula:

[0075]

[0076] : Wavelet transform coefficients; Original signal; .

[0077] 2. Data storage module

[0078] Hash Algorithm:

[0079] Hash function:

[0080]

[0081] Hash value; constant; Hash table size; Enter your data.

[0082] 3. Clock management module

[0083] Phase-locked loop (PLL) formula:

[0084] Phase difference calculation:

[0085]

[0086] : output phase; Input phase; Gain constant; : Phase error.

[0087] 4. Intelligent demodulation module

[0088] Convolutional Neural Networks (CNN):

[0089] Convolution operation:

[0090]

[0091] in is the input signal or feature map, is the convolution kernel, is the size of the convolution kernel.

[0092] Activation function:

[0093]

[0094] The convolutional feature map is transformed nonlinearly through the activation function, so that the network can represent complex features.

[0095] Pooling operation:

[0096]

[0097] Pooling is used to reduce the size of feature maps and retain the main information, usually maximum pooling (MaxPooling) or average pooling.

[0098] SVM searches for the best hyperplane to separate samples of different categories. Through the kernel function, SVM can handle the situation of linear inseparability, and then find the optimal classification hyperplane after mapping to high-dimensional space. Combined with CNN feature extraction, SVM improves the accuracy of signal classification and modulation mode recognition. Specific embodiment three:

[0100] like Figures 1 to 6 As shown, the following is the specific application logic of each algorithm and each module of Example 1 and the detailed application steps in the low-voltage DC carrier communication dual-mode demodulation system.

[0101] 1.Signal acquisition module

[0102] Responsible for obtaining the original signal from the low-voltage DC power line and performing preliminary processing to remove noise and interference to ensure the data quality for subsequent processing.

[0103] Detailed application steps:

[0104] (1) Signal acquisition: Get the original signal from the power line; (2) Kalman filter: Set the initial state estimate and covariance matrix, perform state prediction, update the estimate based on the new observations, and output a smoothed signal; (3) Wavelet transform: Decompose the smoothed signal into multiple scales, extract low-frequency and high-frequency components, and identify mutation points and abnormal conditions; (4) Adaptive threshold: Calculate the mean and standard deviation of the signal, set a dynamic threshold, and remove noise below the threshold.

[0105] 2. Data storage module

[0106] Responsible for efficiently storing and managing signal processed data for quick retrieval and subsequent processing.

[0107] Detailed application steps: (1) Hash algorithm: Calculate the hash value of the data before data storage, use the hash value as an index, and store it in the hash table for fast search; (2) B+ tree index: Insert data into the B+ tree structure according to the rules to maintain its order. When querying, quickly locate data records through the B+ tree structure; (3) Data compression: Compress the data before storage to reduce space occupancy, and decompress it when reading the data.

[0108] 3. Clock management module

[0109] Ensure the accuracy and stability of the system's internal clock and ensure synchronization between modules.

[0110] Detailed application steps: (1) Phase-locked loop (PLL): compares the phase difference between the system output clock and the reference clock, and adjusts the output frequency of the voltage-controlled oscillator to maintain phase synchronization; (2) Delay locked loop (DLL):

[0111] Monitor the phase of the input clock and make delay adjustments to ensure that the local clock is aligned with the input clock phase; (3) Clock synchronization protocol: In a distributed system, time information is exchanged through the network and the clock deviation of each node is adjusted according to the protocol algorithm.

[0112] 4. Interface module

[0113] Handles data interaction with external devices to ensure reliable data transmission between different rates.

[0114] Detailed application steps: (1) Protocol parsing: Identify the received data protocol format, parse the data according to the protocol specification, and prepare to send or process; (2) Data cache: When the data transmission rate does not match, the data to be sent is temporarily stored in the cache area, waiting for the right time to transmit the data; (3) Flow control:

[0115] Monitor the bandwidth usage of the interface and use the token bucket mechanism and sliding window protocol to adjust the sending rate.

[0116] 5. Fault detection and recovery module

[0117] Application logic: monitor system operating status, identify faults and take recovery measures to ensure system reliability.

[0118] Detailed application steps:

[0119] (1) Failure Mode and Effects Analysis (FMEA): Identify and evaluate potential failure modes, establish a failure mode library, and update failure mode information in real time during system operation; (2) Neural Network Fault Diagnosis: Collect system operation data such as temperature, voltage, and current, use the trained neural network model to analyze the data, and make fault predictions; (3) Fault-Tolerant Redundancy: When a failure occurs, switch to redundant hardware modules or use backup data for recovery.

[0120] 6. Protocol conversion module

[0121] Application logic: implements data conversion between different communication protocols to ensure that the semantics of the data is not lost.

[0122] Detailed application steps: (1) Syntax analysis: parse the syntax structure of the source protocol and the target protocol, build a syntax tree, and understand the data hierarchy and field meaning; (2) Semantic mapping: convert the data semantics in the source protocol into the semantics corresponding to the target protocol according to predefined rules, and generate code in the intermediate representation form; (3) Data format conversion: convert the intermediate code into the target data format for transmission or processing according to the syntax rules of the target protocol.

[0123] 7. Intelligent demodulation module

[0124] Application logic: Perform modulation identification and demodulation of the signal to ensure that the received data can be accurately interpreted.

[0125] Detailed application steps: (1) Convolutional Neural Network (CNN): extract features of received signals, automatically learn effective feature representation of signals, and input the extracted features into Support Vector Machine (SVM) for classification; (2) Support Vector Machine (SVM): determine the modulation type based on the features output by CNN, output the modulation type, and pass it to Deep Q Network (DQN) for further optimization; (3) Reinforcement Learning (DQN):

[0126] According to the output of SVM and historical demodulation experience, the demodulation strategy is optimized and continuously updated during the demodulation process to improve the recognition accuracy.

[0127] 8. Spectrum Management Module

[0128] Application logic: Implement dynamic allocation and management of spectrum to optimize communication quality and resource utilization.

[0129] Detailed application steps:

[0130] (1) Q learning: real-time monitoring of spectrum resources, exploration of the state and action space of spectrum usage, and updating of the Q value table to determine the optimal spectrum allocation strategy; (2) Genetic algorithm: evaluate different spectrum allocation schemes, find the optimal solution through selection, crossover and mutation operations, and perform local search to avoid local optimal solutions based on the global optimal solution; (3) Simulated annealing algorithm: perform local search near the solution space of the genetic algorithm to optimize spectrum allocation.

[0131] 9. Secure communication module

[0132] Application logic: Provides a highly secure and anti-interference communication mechanism to ensure the security of data transmission.

[0133] Detailed application steps: (1) Chaos theory: Generate a chaotic sequence as an encryption key to encrypt the communication signal to ensure the randomness and security of the key; (2) LDPC coding: LDPC encode the encrypted signal, perform error detection and correction during signal transmission, and decode and restore the original signal at the receiving end; (3) Quantum encryption: Secure key generation and distribution is achieved through quantum key distribution, and the quantum channel is used to ensure the security of the key.

[0134] 10. Grid environment adaptation module

[0135] Application logic: Enhance the system's adaptability to changes in the power grid environment and optimize signal transmission quality.

[0136] Detailed application steps: (1) Convolutional neural network (CNN): extract the spatial characteristics of the power grid channel, analyze the channel status, and generate adaptive parameters based on the historical data of the power grid; (2) Long short-term memory network (LSTM): process time series channel change information, model and predict channel status, and optimize the current signal balance and compensation based on historical data; (3) Particle swarm optimization (PSO): optimize the network parameter combination of CNN and LSTM, improve the performance of channel balance and compensation, and find the optimal parameter combination by simulating the foraging behavior of bird flocks.

[0137] The specific application logic and steps of each module ensure the efficient operation of the low-voltage DC carrier communication dual-mode demodulation system. Each module ensures high-quality signal acquisition, storage, processing, transmission and security through sophisticated algorithms and collaboration. Such a design helps to improve the overall performance and reliability of the system and can cope with complex power grid environments and communication requirements. Specific embodiment four:

[0139] like Figures 1 to 6 As shown, the following is a detailed hardware composition and hardware description of each module in Example 1:

[0140] 1.Signal acquisition module

[0141] Hardware composition: Signal conditioning circuit: including amplifiers and filters, used to improve signal quality and remove high-frequency noise.

[0142] Analog-to-digital converter (ADC): converts analog signals into digital signals to ensure accuracy of subsequent processing. Microcontroller or FPGA: implements Kalman filtering and wavelet transform algorithms for signal processing.

[0143] Hardware Description:

[0144] Signal conditioning circuit: The signal amplitude is increased through the amplifier and the unnecessary frequency components are removed by the filter to enhance the usability of the signal.

[0145] Analog-to-digital converter (ADC): Provides high resolution (such as 12 bits or higher) to ensure the accuracy of signal acquisition.

[0146] Microcontroller / FPGA: Real-time computing capabilities required by processing algorithms and supporting flexible signal processing. 2. Data storage module

[0147] Hardware composition:

[0148] Flash memory or EEPROM: used to store signal data and related parameters.

[0149] Data processing unit: responsible for data hash processing and maintenance of B+ tree index structure.

[0150] Hardware Description:

[0151] Flash / EEPROM: Provides persistent storage, can quickly read and write data, and supports fast data retrieval. Data processing unit: Can implement data compression and index structure based on microcontroller or dedicated hardware to optimize storage efficiency.

[0152] 3. Clock management module

[0153] Hardware composition:

[0154] Phase-locked loop (PLL) chip: Generates a high-stability clock signal.

[0155] Clock synchronization chip: realizes network clock synchronization and supports various protocols.

[0156] Hardware Description:

[0157] PLL chip: provides low jitter and high stability to ensure the accuracy of the clocks of various parts of the system.

[0158] Clock synchronization chip: ensures the clock consistency of the system in a distributed environment and supports standard synchronization protocols.

[0159] 4. Interface module

[0160] Hardware composition:

[0161] Interface controller: supports parsing and conversion of different communication protocols.

[0162] Data buffer: stores data to be sent and received.

[0163] Hardware Description:

[0164] Interface controller: It can support multiple communication interfaces such as UART, SPI, I2C, etc. as needed to achieve flexible connection with external devices.

[0165] Data buffer: Provides sufficient storage space to prevent data loss and ensure smooth data transmission between different rates.

[0166] 5. Fault detection and recovery module

[0167] Hardware composition:

[0168] Sensor unit: monitors system operating status, such as temperature, voltage, and current.

[0169] Processing unit: implements failure mode and effects analysis (FMEA) and neural network algorithms.

[0170] Hardware Description:

[0171] Sensor unit: monitors system parameters in real time through various sensors and provides data to the processing unit. Processing unit: can use microcontroller or FPGA to realize real-time data processing and execute fault detection algorithm.

[0172] 6. Protocol conversion module

[0173] Hardware composition:

[0174] Protocol parser: used to parse and convert data formats of different protocols.

[0175] Microprocessor: Executes syntax analysis and semantic mapping algorithms.

[0176] Hardware Description: Protocol parser: designed as a programmable device, supports multiple communication protocols and ensures seamless data conversion.

[0177] Microprocessor: Provides powerful computing power to execute complex protocol conversion algorithms.

[0178] 7. Intelligent demodulation module

[0179] Hardware composition:

[0180] High-performance processor: supports processing of deep learning algorithms such as CNN and SVM.

[0181] Signal demodulation chip: a dedicated chip used for fast signal demodulation.

[0182] Hardware Description:

[0183] High-performance processor: GPU or dedicated AI accelerator can be selected to ensure the efficiency of algorithm operation. Signal demodulation chip: supports high-speed signal processing to ensure real-time demodulation performance.

[0184] 8. Spectrum Management Module

[0185] Hardware composition:

[0186] Spectrum Analyzer: Real-time monitoring of spectrum usage.

[0187] Processing unit: performs Q learning and genetic algorithm calculations.

[0188] Hardware Description:

[0189] Spectrum Analyzer: Provides high frequency response to ensure accurate analysis of the spectrum environment.

[0190] Processing unit: A microcontroller or FPGA is used to execute complex optimization algorithms.

[0191] 9. Secure communication module

[0192] Hardware composition:

[0193] Encryption processing unit: implements chaotic encryption and LDPC coding.

[0194] Quantum Key Distribution Module: Generates and distributes secure encryption keys.

[0195] Hardware Description:

[0196] Encryption processing unit: supports efficient encryption algorithms to ensure the security of data during transmission. Quantum key distribution module: uses the principles of quantum physics to provide secure key generation and exchange.

[0197] 10. Grid environment adaptation module hardware composition:

[0198] Signal processing unit: implements CNN and LSTM algorithms.

[0199] Monitoring sensors: monitor the load and noise status of the power grid.

[0200] Hardware Description: Signal processing unit: Uses high-performance processors to support real-time data analysis and model optimization.

[0201] Monitoring sensors: provide real-time grid status data to ensure the system's adaptability.

[0202] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "comprising a reference structure" do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0203] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A low voltage communication demodulation platform, including a demodulation main system, characterized in that: The demodulation main system includes a signal acquisition module, a data storage module, a clock management module, an interface module, a fault detection and recovery module, a protocol conversion module, a signal processing module, a spectrum management module, a safety communication module, a power grid environment adaptation module and an intelligent demodulation module. The working steps of the signal acquisition module, the data storage module, the clock management module, the interface module, the fault detection and recovery module, the protocol conversion module, the signal processing module, the spectrum management module, the safety communication module, the power grid environment adaptation module and the intelligent demodulation module in the demodulation main system are as follows: Sp1, the signal acquisition module performs real-time dynamic prediction and correction on the collected signals through the Kalman filter algorithm; Sp2, the data storage module stores data by combining a hash algorithm, a tree index algorithm and a data compression algorithm; Sp3, the clock management module integrates the phase-locked loop algorithm, the delay-locked loop algorithm and the clock synchronization protocol algorithm, generates a stable high-frequency clock signal through the phase-locked loop algorithm, and phase-locks it with the reference clock to ensure the accuracy and stability of the clock. The delay-locked loop algorithm adjusts the clock phase through the delay line, accurately controls the delay of the clock signal, and compensates for the clock offset caused by the internal wiring of the chip. The clock synchronization protocol algorithm realizes constant synchronization in the distributed system to ensure the consistency of the system clock; Sp4, the interface module identifies and parses different communication protocols through a protocol parsing algorithm, uses a data cache algorithm to temporarily store data to be sent or received to adapt to the differences between different data transmission rates and prevent data loss, and adjusts the data transmission rate on the interface through a flow control algorithm to avoid data congestion and transmission errors; Sp5, the fault detection and recovery module combines the fault mode and effect analysis algorithm, the neural network fault diagnosis algorithm and the fault-tolerant redundancy algorithm to identify and evaluate potential fault modes through the various components and functions of the FMEA algorithm system, and establish a fault mode library. During the operation of the system, the neural network fault diagnosis algorithm uses the system operation data collected by the sensor, including temperature, voltage and current, to perform real-time analysis and diagnosis. When a fault is detected, the fault-tolerant redundancy algorithm switches to a redundant hardware module or uses backup data according to the type and severity of the fault; Sp6, the protocol conversion module adopts a syntax analysis algorithm, a semantic mapping algorithm and an intermediate code generation algorithm; Sp7, the intelligent demodulation module realizes adaptive modulation recognition and demodulation through convolutional neural network and support vector machine, introduces deep Q network in reinforcement learning, extracts features of received signals, inputs the extracted features into SVM for preliminary classification and judgment of modulation type, and then DQN optimizes the demodulation strategy according to the output of SVM and historical demodulation experience to realize accurate adaptive modulation recognition and demodulation. CNN is responsible for automatically learning and extracting effective feature representation from input signals, SVM makes classification decisions based on these features and judges the modulation type of the signal. On this basis, DQN adjusts the demodulation strategy according to reward feedback through continuous interaction with the demodulation process; Sp8. The spectrum management module dynamically allocates and manages the spectrum by combining Q learning with a genetic algorithm, and introduces a simulated annealing algorithm. Through Q learning, the state and action space of spectrum use are explored to obtain a preliminary strategy for spectrum allocation. Through a genetic algorithm, different spectrum allocation schemes are crossover and mutation operations are performed to find the global optimal solution. Through a simulated annealing algorithm, a local search is performed near the solution space searched by the genetic algorithm to avoid falling into a local optimal solution, and the dynamic allocation and management of the spectrum is optimized. During the application process, Q learning interacts with the spectrum environment and continuously updates the Q value table to determine what actions to take under different spectrum states to obtain the maximum long-term return. The genetic algorithm uses its population evolution characteristics to search and optimize multiple spectrum allocation schemes in parallel, and generates a better offspring scheme through selection, crossover and mutation operations; Sp9, the secure communication module achieves high security and anti-interference capability for dual-mode communication signals by combining chaos theory with low-density parity check error correction coding; Sp10. The power grid environment adaptation module adopts a combination of convolutional neural network and long short-term memory network to achieve channel equalization and compensation, and introduces particle swarm optimization algorithm. The CNN extracts the spatial characteristics of the power grid channel, and the LSTM processes the channel change information on the time series to achieve channel equalization and compensation. The PSO optimizes the network parameters of the CNN and LSTM, finds the best model structure and parameter configuration, and improves the performance of channel equalization and compensation. Under different power grid load and noise conditions, the PSO automatically adjusts the parameters of the CNN and LSTM according to the real-time monitoring data of the channel to improve the communication quality. The LSTM uses its memory unit and gating mechanism to model and predict the channel change information on the time series, and balances and compensates the current signal more accurately according to the historical channel status. The PSO algorithm searches for the optimal combination of CNN and LSTM network parameters in the parameter space by simulating the foraging behavior of bird flocks. Each particle represents a set of parameters, and the speed and position of the particle are adjusted according to the fitness of the particle, that is, the bit error rate index after channel equalization.

2. A low voltage communication demodulation platform according to claim 1, characterized in that: The signal acquisition module continuously updates the estimation of the signal state based on the prior knowledge and observation data of the signal through Kalman filtering, so that the collected signal is smoother and more accurate. The signal is decomposed into different frequency sub-bands through wavelet transform, which is convenient for analyzing the characteristics of the signal at different scales and detecting transient interference in the signal. The adaptive threshold algorithm is used to analyze the statistical information of the amplitude change of the signal, determine the threshold, remove the noise components below the threshold, and retain the useful signal part. When writing data, the data storage module first calculates the hash value of the data through the hash algorithm, and stores it as an index in the hash table for fast search, removes noise interference and estimates the true value of the signal, and uses the wavelet transform algorithm to perform multi-scale signal analysis. Decompose and extract signal features of different frequency components to facilitate detection of mutation points and abnormal conditions in the signal. At the same time, the adaptive threshold algorithm dynamically adjusts the threshold according to the statistical characteristics of the signal to identify and remove noise and interference components in the signal. The hash algorithm is used to quickly locate and find data to improve the efficiency of data retrieval. The tree index algorithm is used to organize the stored data in an orderly manner to facilitate range queries and fast traversal. The data compression algorithm is used to reduce the space required for data storage and improve the utilization of storage resources. The syntax analysis algorithm parses the syntax structure of the source protocol and the target protocol to understand the organization of the data. The semantic mapping algorithm converts the semantic information in the source protocol into semantics that the target protocol can understand, ensuring that the meaning of the data is in the conversion process. No loss, the intermediate code generation algorithm generates the code in the form of intermediate representation based on the syntax analysis and semantic mapping, which is convenient for subsequent conversion and processing. The syntax analysis algorithm in the protocol conversion module parses the data format of the source protocol, builds a syntax tree, understands the hierarchical structure and field meaning of the data, and uses the semantic mapping algorithm to convert the data semantics in the source protocol into the semantics corresponding to the target protocol according to the pre-defined semantic mapping rules. The intermediate code generation algorithm converts the semantically mapped data into an intermediate code form. Finally, according to the syntax rules of the target protocol, the intermediate code is converted into a data format that meets the requirements of the target protocol for transmission or processing. At the same time, combined with the quantum encryption algorithm, the chaos theory performs on the communication signal Encryption, LDPC error correction coding detects and corrects possible errors during signal transmission. Quantum encryption algorithm uses the characteristics of quantum mechanics to provide an absolutely secure key distribution mechanism for communication. When the two communicating parties exchange keys, the quantum encryption algorithm is used to generate a secure key, and then the communication data is encrypted using chaos theory. Finally, it is transmitted through LDPC coding. The chaotic sequence generated by the chaotic system is used as the encryption key to encrypt the original communication signal. LDPC coding encodes the encrypted signal at the sending end and adds redundant information to facilitate the detection and correction of errors introduced during the transmission process at the receiving end. The quantum encryption algorithm realizes secure key distribution through the quantum channel to ensure the security and randomness of the key.The receiving end first uses the quantum encryption algorithm to obtain the correct key, then uses LDPC decoding to recover the encrypted signal, and finally restores the original communication data through the chaotic decryption algorithm.

3. A low voltage communication demodulation platform according to claim 1, characterized in that: When the clock management module is applied, the phase-locked loop algorithm compares the phase difference between the output clock and the reference clock, adjusts the output frequency of the voltage-controlled oscillator, so that the output clock and the reference clock are phase-locked, and uses the delay-locked loop algorithm to delay and adjust the clock signal so that the local clock is phase-aligned with the input clock. The clock synchronization protocol algorithm exchanges time information through network communication, calculates the clock deviation and adjustment amount according to the algorithm specified in the protocol, so that the clocks of each node are synchronized. When the interface module is applied, when data is transmitted through the interface, the protocol parsing algorithm first identifies the protocol format used by the data, parses and encapsulates the data according to the protocol specification, and uses the data caching algorithm to temporarily store the data in the cache area when the data transmission rate does not match, waiting for the appropriate time to send and process it. At the same time, the bandwidth usage and buffer status of the interface are monitored through the flow control algorithm, and the token bucket algorithm and sliding window protocol mechanism are used to control the data transmission rate to ensure the stability and reliability of data transmission.

4. A low voltage communication demodulation platform according to claim 1, characterized in that: The intelligent demodulation module combines a generative adversarial network and a Kalman filter to enhance the processing of channel noise. The generative adversarial network generates a low-noise version of the signal, and the Kalman filter further optimizes the quality of the demodulated signal. The fuzzy logic in the signal processing module is combined with the Bayesian network to achieve intelligent switching between ASK and FSK dual-mode demodulation methods. The fuzzy logic handles the uncertainty of the channel state, and the Bayesian network is used to make inferences and decisions based on signal characteristics.

5. A low voltage communication demodulation platform according to claim 1, characterized in that: The spectrum management module monitors spectrum resources in real time through Q learning and dynamically adjusts communication channels. Genetic algorithms are used to optimize spectrum allocation strategies to avoid spectrum interference and congestion.

6. A low voltage communication demodulation platform according to claim 1, characterized in that: The secure communication module combines quantum key distribution and elliptic curve encryption for communication encryption. QKD provides secure key generation and distribution, and ECC is used to efficiently encrypt data to ensure the security and efficiency of data transmission.

7. A low voltage communication demodulation platform according to claim 1, characterized in that: The grid environment adaptation module combines adaptive filtering with genetic algorithms to enhance the anti-interference capability of the system. The RLS filter adjusts the filtering parameters in real time to cope with sudden interference in the grid, and the genetic algorithm optimizes the configuration of the filter through global search.

8. A low voltage communication demodulation platform according to claim 1, characterized in that: The convolutional neural network in the intelligent demodulation module is pre-trained and can adaptively adjust network parameters according to historical data in the power grid to improve the accuracy and efficiency of modulation recognition.

9. A low voltage communication demodulation platform according to claim 1, characterized in that: The signal processing module combines a weighted average filter with a minimum mean square error adaptive filter to further optimize the demodulated signal. The weighted average filter is used to suppress noise, and the LMS filter dynamically adjusts the filter weight according to the error signal to ensure signal stability. During the spectrum allocation process, the spectrum management module combines an adaptive scheduling mechanism based on reinforcement learning with genetic algorithm optimization to further improve spectrum utilization efficiency and communication stability.

Images