Dual-mode network communication method and system based on HPLC + HRF

Abstract

The invention provides a dual-mode network communication method and system based on HPLC + HRF, and the method comprises the steps: carrying out the network initialization, and carrying out the initialization configuration of an HPLC module and an HRF module; data acquisition and feature extraction; training and predicting an LSTM neural network; selecting and optimizing a communication mode; dynamically adjusting the data transmission rate based on the prediction result of the LSTM neural network in combination with the current actual monitoring data; performing data transmission in the data transmission process according to the selected communication mode and data transmission rate; and data fusion and optimization: carrying out fusion processing, error correction and optimization sorting on the data from the HPLC and the HRF at a receiving end. By fusing the LSTM neural network prediction technology and the adaptive rate adjustment technology, accurate prediction of network communication link quality and intelligent optimization of the data transmission rate are realized, and the defects of the traditional network communication technology in a complex environment are effectively overcome.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to a dual-mode network communication method and system based on HPLC+HRF. Background Art

[0002] With the rapid development of Internet of Things technology, many fields such as smart home and smart grid have put forward more stringent requirements on the stability, reliability and intelligence of network communication. Single network communication technology, such as HPLC or HRF, often exposes many limitations in complex and changeable practical application scenarios.

[0003] High-speed power line carrier communication (HPLC) is a broadband power line carrier technology that transmits data through low-voltage power lines. It has the advantages of high speed, real-time performance, high anti-interference, high transmission, and inter-chip interconnection, which can meet the current needs of household and other electrical equipment for information collection. However, with the widespread use of power electronic devices and inverters, the electromagnetic environment of the power grid has become increasingly complex, making the environment of power line carrier communication increasingly harsh, resulting in a low success rate of fault reporting, a short single-hop communication distance, and the existence of "island / isolated point" phenomena in the network. For example, inside large building structures, although the power line layout is extensive, it is easily interfered by electrical equipment, causing the HPLC signal quality to fluctuate; and during the propagation of wireless signals, they are affected by factors such as obstacles and multipath effects, and there may be coverage blind spots or sharp signal attenuation.

[0004] High-speed wireless communication HRF refers to long-distance transmission communication between multiple nodes without the transmission of conductors or cables. Due to its advantages of low power consumption and flexible networking, it has gradually become a major information transmission mode in the construction of field networks. It can avoid various interferences and noises in power lines, overcome the shortcomings of isolated islands / isolated points, low success rate of power outage reports, and short single-hop communication distance. It is an effective supplement to HPLC communication. However, HRF communication is also susceptible to interference from the external environment, especially in complex environments such as buildings and underground parking lots, where there are problems such as severe signal attenuation and signal blind spots.

[0005] In addition, the traditional dual-mode network communication switching mechanism is mostly based on real-time link quality detection results for passive switching, lacking forward-looking prediction of link quality change trends. Therefore, it is urgent to develop a dual-mode network communication solution that integrates multiple advanced technologies and has intelligent prediction capabilities. Summary of the invention

[0006] The technical problem to be solved by the present invention is to provide a dual-mode network communication method and system based on HPLC+HRF. By integrating LSTM neural network prediction technology and adaptive rate adjustment technology, accurate prediction of network communication link quality and intelligent optimization of data transmission rate can be achieved, effectively overcoming the shortcomings of traditional network communication technology in complex environments, ensuring the efficiency, stability and reliability of network communication, so as to meet the growing demand for network communication in modern Internet of Things applications.

[0007] To solve the above technical problems, the embodiments of the present invention provide the following technical solutions: a dual-mode network communication method based on HPLC+HRF, comprising the following steps:

[0008] Network initialization: initializing and configuring the HPLC module and HRF module respectively, including setting the communication baud rate, channel coding mode, transmission power, and performing channel scanning operations to determine available channel resources;

[0009] Data collection and feature extraction: Continuously collect multi-dimensional data related to the quality of HPLC and HRF links at a specific sampling frequency. For the HPLC link, the voltage, current, signal frequency, phase and other parameters on the power line are collected for feature extraction. For the HRF link, the strength, frequency, bandwidth, bit error rate and other indicators of the wireless signal are collected for feature extraction.

[0010] LSTM neural network training and prediction: construct an LSTM neural network model with multiple hidden layers. The number of input layer nodes is determined according to the dimension of the extracted feature vector, and the number of output layer nodes corresponds to the number of predicted link quality indicators. The training is performed using the stochastic gradient descent algorithm with adaptive learning rate adjustment, and the early stopping method is used to prevent overfitting. After the training is completed, the real-time feature data is used to predict the signal strength change trend, bit error rate fluctuation range and other key indicators of the dual-mode network communication link quality;

[0011] Communication mode selection and optimization: select the appropriate communication mode based on the LSTM neural network prediction results and the current actual link quality detection data. When it is predicted that the HPLC link will meet the preset communication requirements in the short term in the future, the HPLC communication mode is selected first and the relevant communication parameters are adjusted in advance. When it is predicted that the HPLC link is about to fail and the HRF link quality is acceptable, switch to the HRF communication mode and adjust the transmission power and channel selection;

[0012] Adaptive rate adjustment: monitor the dynamic changes of the network environment in real time during data transmission, including the real-time fluctuation of link quality and the change of network load. Based on the prediction results of the LSTM neural network and the current actual monitoring data, the data transmission rate is dynamically adjusted. When the link quality improves and the network load decreases, the data transmission rate is increased, and vice versa. At the same time, a rate adjustment feedback mechanism is established to optimize and correct the rate adjustment strategy according to the actual effect of data transmission.

[0013] Data transmission and switching control: during the data transmission process, data transmission is performed according to the selected communication mode and data transmission rate, the real-time link quality under the current communication mode is monitored, and the communication mode is automatically switched if the link quality deteriorates and the link quality of another communication mode is available, data is segmented, numbered, and error-corrected, and relevant information including but not limited to the switching identifier and rate adjustment information is added to the data packet header when the communication mode is switched or the rate is adjusted, and data is reorganized and processed at the receiving end;

[0014] Data fusion and optimization: When dual-mode data is transmitted simultaneously, the data from HPLC and HRF are fused, error corrected, and optimized at the receiving end. Data is reorganized according to the identification information and numbering sequence in the data packet header. A hybrid error correction scheme based on CRC and FEC is used for error detection and correction. A priority queue algorithm is used for sorting and optimization integration based on the importance level and real-time requirements of the data.

[0015] Furthermore, in the data collection and feature extraction steps, for the HPLC link, fast Fourier transform is used to calculate the signal frequency components and principal component analysis is used to extract feature vectors, and for the HRF link, wavelet transform is used to extract the time-frequency domain features of the wireless signal.

[0016] Furthermore, in the LSTM neural network training and prediction steps, the number of hidden layers of the constructed LSTM neural network model is 3-5 layers, the number of input layer nodes is 10-20, the number of output layer nodes corresponds to the signal strength and bit error rate of 2, and the mean square error is used as the loss function to evaluate the model prediction performance.

[0017] Furthermore, in the communication mode selection and optimization step, the HPLC communication parameters adjusted in advance include optimizing the power line carrier frequency allocation and adjusting the encoding method, and the adjustment under the HRF communication mode includes dynamically adjusting the transmission power and channel selection according to the wireless signal propagation characteristics.

[0018] Furthermore, in the adaptive rate adjustment step, the rate adjustment is performed according to a preset rate adjustment strategy, and the strategy is optimized according to actual effects such as the bit error rate, accuracy and real-time performance of data transmission fed back by the receiving end. The adjustment methods include adjusting the encoding method and adjusting the transmission time interval.

[0019] Furthermore, in the data transmission and switching control step, the length of the data segment processing is a preset fixed length, and the switching identifier includes the switching reason and the number range of the data segments that have been transmitted in the original communication mode.

[0020] The present invention also proposes a dual-mode network communication system based on HPLC+HRF, comprising:

[0021] HPLC communication module, used to realize high-speed power line communication functions, including but not limited to power line signal modulation and demodulation, encoding and decoding, data transmission and reception operations;

[0022] HRF communication module, used to realize high-frequency wireless communication function, with the ability of transmitting and receiving wireless signals, frequency modulation, and channel coding;

[0023] A data acquisition and feature extraction unit, used to acquire data related to link quality and extract features, including a power line data acquisition and feature extraction module for acquiring data related to HPLC links and extracting features, and a wireless data acquisition and feature extraction module for acquiring data related to HRF links and extracting features;

[0024] The LSTM neural network prediction unit is used to predict the link quality based on the LSTM model, including the LSTM model construction module, training module and prediction calculation module, which are responsible for model construction, model training and link quality prediction calculation respectively;

[0025] A communication mode selection and optimization unit, which is used to select and optimize the communication mode according to the prediction results and the actual link quality, and includes a prediction result analysis module, a communication mode judgment module and a parameter optimization module. The prediction result analysis module is used to analyze the LSTM neural network prediction results, the communication mode judgment module is used to judge the communication mode according to the analysis results and the actual link quality, and the parameter optimization module is used to optimize the parameters of the selected communication mode;

[0026] The adaptive rate adjustment unit is used to monitor the dynamic changes of the network environment in real time and adjust the data transmission rate. It includes a rate monitoring module, a rate calculation module and a feedback optimization module. The rate monitoring module is used to monitor the link quality and network load. The rate calculation module calculates the rate based on the LSTM prediction results and monitoring data. The feedback optimization module optimizes the rate adjustment strategy according to the actual effect of data transmission;

[0027] A data processing and switching control unit, used for segmenting, numbering, error correction coding, communication mode switching and rate adjustment control of data, including a data segmentation module, a number adding module, an error correction coding module, a cache switching module and a rate adjustment execution module, wherein the data segmentation module is used for segmenting data, the number adding module is used for adding numbers to the segmented data, the error correction coding module is used for adding error correction coding information, the cache switching module is used for caching untransmitted data when the communication mode is switched and starting another communication mode for continued transmission, and the rate adjustment execution module is used for performing rate adjustment operations;

[0028] The data fusion and optimization unit is used to perform fusion processing, error correction and optimization sorting of dual-mode transmitted data at the receiving end.

[0029] Furthermore, the rate monitoring module, the rate calculation module and the feedback optimization module are specifically:

[0030] (1) Rate monitoring module, including link quality monitoring submodule and network load monitoring submodule

[0031] The link quality monitoring submodule is used to obtain the link quality related parameters in real time through interaction with the HPLC communication module and the HRF communication module;

[0032] The network load monitoring submodule is used to monitor the data traffic in the network, including but not limited to the number of currently connected devices, the frequency of data transmission requests from each device, the size of data packets transmitted, and the transmission rate;

[0033] (2) The rate calculation module includes a link quality rate adjustment submodule and a network load rate adjustment submodule.

[0034] The link quality rate adjustment submodule calculates the data transmission rate using a preset rate adjustment algorithm based on the link quality parameters obtained by the link quality monitoring submodule;

[0035] The network load rate adjustment submodule optimizes the data transmission rate in combination with the network load information provided by the network load monitoring submodule;

[0036] (3) Feedback optimization module, including transmission effect evaluation submodule and strategy adjustment submodule

[0037] The transmission effect evaluation submodule is used to collect the information fed back by the receiving end in real time during the data transmission process and to perform multi-dimensional evaluation on the data transmission effect;

[0038] The strategy adjustment submodule is used to optimize and modify the rate adjustment strategy according to the evaluation result of the transmission effect evaluation submodule.

[0039] Furthermore, the data segmentation module, the numbering module, the error correction coding module, the cache switching module and the rate adjustment execution module include:

[0040] (1) Data segmentation module, including segmentation strategy formulation submodule and segmentation execution submodule

[0041] The segmentation strategy formulation submodule is used to formulate data segmentation strategies based on the network communication protocol and the characteristics of data transmission;

[0042] The segmentation execution submodule is used to segment the data to be transmitted according to the segmentation rules determined by the segmentation strategy formulation submodule;

[0043] (2) Numbering adding module, including numbering rule setting submodule and numbering adding execution submodule

[0044] The numbering rule setting submodule is used to formulate a numbering rule system to add numbers to the segmented data. The numbering rule should cover the important dimensions of the data source, transmission sequence information, communication session identifier, and data version information;

[0045] The number adding execution submodule adds corresponding number information to each piece of data according to the numbering rules set by the numbering rule setting submodule;

[0046] (3) Error correction coding module, including coding mode selection submodule and coding execution submodule

[0047] The coding mode selection submodule is used to dynamically select the appropriate error correction coding mode according to the current link quality and network load;

[0048] The coding execution submodule is used to perform coding operations on the segmented and numbered data according to the error correction coding mode determined by the coding mode selection submodule;

[0049] (4) Cache switching module, including cache management submodule and switching control submodule

[0050] The cache management submodule is used to store the data that has not been fully transmitted during the communication mode switching or rate adjustment process;

[0051] The switching control submodule coordinates data caching and switching operations when communication mode switching or rate adjustment is required;

[0052] (5) Rate adjustment execution module, including rate adjustment instruction receiving submodule and rate adjustment operation execution submodule

[0053] The rate adjustment instruction receiving submodule is used to receive the rate adjustment instruction from the adaptive rate adjustment unit, the instruction including the target rate value, the adjustment reason and the adjustment time requirement information;

[0054] The rate adjustment operation execution submodule is used to adjust the data transmission rate according to the instruction information parsed by the rate adjustment instruction receiving submodule.

[0055] Furthermore, the data fusion and optimization unit includes a data fusion processing module, an error correction module, and an optimization sorting module, wherein:

[0056] (1) Data fusion processing module, including data synchronization and alignment submodule, data merging and deduplication submodule

[0057] The data synchronization and alignment submodule uses the timestamp information carried in the data packet header to align the data arriving at the receiving end at different times according to their original sending order;

[0058] Data merging and deduplication submodule, after data synchronization and alignment, the data merging and deduplication submodule merges the data from the two modes;

[0059] (2) Error correction module, including error detection submodule and error correction execution submodule

[0060] The error detection submodule is used to perform data integrity check based on the cyclic redundancy check mechanism and the CRC check code added by the data sender, calculate the CRC value of the received data, and compare it with the CRC check code sent by the sender. If the two do not match, it indicates that an error may have occurred in the data transmission process;

[0061] The error correction execution submodule is used for performing corresponding error correction operations according to the error correction coding method adopted by the data sending end when the error detection submodule detects that there is an error in the data;

[0062] (3) Optimize the sorting module, including the data evaluation submodule and the sorting execution submodule

[0063] The data importance assessment submodule determines the importance level of data based on factors including but not limited to the type and source of the data and the degree of impact on the upper-layer application system;

[0064] The sorting execution submodule is used to optimize the sorting of the error-corrected data according to the importance level and weight determined by the data importance evaluation submodule.

[0065] The beneficial effects of the above technical solution of the present invention are as follows:

[0066] 1. The present invention accurately predicts the quality of the dual-mode network communication link through the LSTM neural network. The system can know the link quality change trend in advance and take proactive measures before the link quality deteriorates, such as switching the communication mode in advance or adjusting the data transmission rate, effectively avoiding data transmission interruptions or errors caused by sudden link failures.

[0067] 2. The adaptive rate adjustment technology of the present invention dynamically adjusts the data transmission rate according to the link quality and network load. When the link quality is poor or the network is congested, the rate is reduced to ensure the accuracy of data transmission, and when the link quality is good and the network is idle, the rate is increased to improve the transmission efficiency, thereby maintaining reliable data transmission in various complex network environments.

[0068] 3. The error detection and correction mechanism in the data fusion processing of the present invention can identify and repair the erroneous data generated in the dual-mode transmission process. Whether it is HPLC data errors caused by power line interference or HRF data errors caused by the multipath effect of wireless signals, they can be effectively processed, further enhancing the reliability of data transmission and ensuring the integrity and accuracy of the data obtained by the receiving end. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 It is a flow chart of the dual-mode network communication method based on HPLC+HRF of the present invention;

[0070] Figure 2 It is a principle block diagram of the dual-mode network communication system based on HPLC+HRF of the present invention;

[0071] Figure 3 The schematic diagram of the application of HPLC and HRF dual-mode communication in electric energy data communication of the present invention;

[0072] Figure 4 It is a node topology diagram of HPLC and HRF dual-mode communication of the dual-mode network communication method and system based on HPLC+HRF of the present invention. DETAILED DESCRIPTION

[0073] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0074] Example 1

[0075] like Figure 1 As shown, the present invention proposes a dual-mode network communication method based on HPLC+HRF, comprising the following steps:

[0076] S1, network initialization, initialization configuration of HPLC module and HRF module respectively, including setting communication baud rate, channel coding mode, transmission power, and performing channel scanning operation to determine available channel resources;

[0077] S2. Data collection and feature extraction: Continuously collect multi-dimensional data related to the quality of HPLC link and HRF link at a specific sampling frequency. For HPLC link, collect parameters such as voltage, current, signal frequency, phase, etc. on the power line for feature extraction. For HRF link, collect wireless signal strength, frequency, bandwidth, bit error rate, etc. for feature extraction.

[0078] S3, LSTM neural network training and prediction, build an LSTM neural network model with multiple hidden layers. The number of input layer nodes is determined according to the dimension of the extracted feature vector, and the number of output layer nodes corresponds to the number of predicted link quality indicators. The stochastic gradient descent algorithm with adaptive learning rate adjustment is used for training, and the early stopping method is used to prevent overfitting. After the training is completed, the real-time feature data is used to predict the signal strength change trend, bit error rate fluctuation range and other key indicators of the dual-mode network communication link quality;

[0079] S4, communication mode selection and optimization, select the appropriate communication mode according to the LSTM neural network prediction results and the current actual link quality detection data. When it is predicted that the HPLC link will meet the preset communication requirements in the short term in the future, the HPLC communication mode is preferred and the relevant communication parameters are adjusted in advance. When it is predicted that the HPLC link is about to fail and the HRF link quality is acceptable, switch to the HRF communication mode and adjust the transmission power and channel selection;

[0080] S5, adaptive rate adjustment, real-time monitoring of dynamic changes in the network environment during data transmission, including real-time fluctuations in link quality and changes in network load, dynamically adjusting the data transmission rate based on the LSTM neural network prediction results combined with the current actual monitoring data, increasing the data transmission rate when link quality improves and network load decreases, and decreasing it otherwise. At the same time, a rate adjustment feedback mechanism is established to optimize and correct the rate adjustment strategy based on the actual effect of data transmission;

[0081] S6, data transmission and switching control, during the data transmission process, data transmission is performed according to the selected communication mode and data transmission rate, the real-time link quality under the current communication mode is monitored, and the communication mode is automatically switched if the link quality deteriorates and the link quality of another communication mode is available, the data is segmented, numbered, and error-corrected, and relevant information including but not limited to the switching identifier and rate adjustment information is added to the data packet header when the communication mode is switched or the rate is adjusted, and the data is reorganized and processed at the receiving end;

[0082] S7. Data fusion and optimization. When dual-mode data is transmitted simultaneously, the data from HPLC and HRF are fused, error corrected, and optimized at the receiving end. Data is reorganized according to the identification information and numbering sequence in the packet header. A hybrid error correction scheme based on CRC and FEC is used for error detection and correction. A priority queue algorithm is used for sorting and optimization integration based on the importance level and real-time requirements of the data.

[0083] In this embodiment, in the data collection and feature extraction steps, for the HPLC link, fast Fourier transform is used to calculate the signal frequency components and principal component analysis is used to extract feature vectors, and for the HRF link, wavelet transform is used to extract the time-frequency domain features of the wireless signal.

[0084] In this embodiment, in the LSTM neural network training and prediction steps, the number of hidden layers of the constructed LSTM neural network model is 3-5 layers, the number of input layer nodes is 10-20, the number of output layer nodes corresponds to the signal strength and bit error rate of 2, and the mean square error is used as the loss function to evaluate the model prediction performance.

[0085] In this embodiment, in the communication mode selection and optimization step, the HPLC communication parameters adjusted in advance include optimizing the power line carrier frequency allocation and adjusting the encoding method, and the adjustment under the HRF communication mode includes dynamically adjusting the transmission power and channel selection according to the wireless signal propagation characteristics.

[0086] In this embodiment, in the adaptive rate adjustment step, the rate adjustment is performed according to a preset rate adjustment strategy, and the strategy is optimized according to actual effects such as the bit error rate fed back by the receiving end, the accuracy and real-time performance of data transmission, and the adjustment methods include adjusting the encoding method and adjusting the transmission time interval.

[0087] In this embodiment, in the data transmission and switching control step, the length of the data segment processing is a preset fixed length, and the switching identifier includes the switching reason and the number range of the data segments that have been transmitted in the original communication mode.

[0088] The present invention is based on the dual-mode network communication method of HPLC+HRF, and the specific implementation principle is:

[0089] Network initialization: When the dual-mode network communication system is started, the HPLC module and the HRF module are fully and carefully initialized and configured. This covers the precise setting of key communication parameters, such as the power line carrier frequency range (e.g. 2-12MHz), modulation method (e.g. orthogonal frequency division multiplexing OFDM), initial transmission power (e.g. 10dBm), etc. of the HPLC module, and the wireless frequency band (e.g. 2.4GHz or 5GHz), modulation method (e.g. QPSK), initial transmission power (e.g. 15dBm), etc. of the HRF module. At the same time, the channel scanning program is started to conduct a comprehensive detection of the power line network and wireless environment, identify and record the available channel resources and their related characteristic parameters, and lay a solid foundation for subsequent data transmission.

[0090] Data collection and feature extraction: The system continuously collects multi-dimensional data closely related to the quality of HPLC and HRF links at a specific sampling frequency (e.g., every 100 milliseconds). For HPLC links, the instantaneous values ​​of voltage and current on the power line are collected, and the fast Fourier transform (FFT) algorithm is used to accurately calculate the signal frequency components. The phase information of the signal is accurately obtained with the help of the phase detection circuit. Then, advanced feature extraction algorithms such as principal component analysis (PCA) are used to extract feature vectors that are key to link quality, such as the voltage amplitude change rate, current harmonic content, and signal frequency offset. For HRF links, the wireless signal strength indicator (RSSI) circuit is used to obtain the received signal strength in real time, the signal bandwidth is accurately measured by the spectrum analyzer, the bit error rate data is statistically analyzed, and the time-frequency domain features of the wireless signal are deeply extracted using methods such as wavelet transform, such as the signal's multipath delay spread, Doppler frequency shift and other feature vectors. These rich and accurate feature data will provide high-quality data input for subsequent LSTM neural network predictions.

[0091] LSTM neural network training and prediction: Carefully construct an LSTM neural network model specifically for dual-mode network communication link quality prediction. The model contains multiple hidden layers (such as 3-5 layers), the number of input layer nodes is determined according to the extracted feature vector dimension (such as 10-20 nodes), and the number of output layer nodes corresponds to the number of predicted link quality indicators (such as 2 nodes for signal strength and bit error rate). The link quality data collected historically and processed after feature extraction is scientifically divided into training set and test set in a ratio of 80:20, and the stochastic gradient descent (SGD) algorithm with adaptive learning rate adjustment is used to deeply train the LSTM model. During the training process, the model prediction performance is strictly evaluated through loss functions such as mean square error (MSE), and the early stopping method is used to effectively prevent the occurrence of overfitting. After the training is completed, the latest link quality feature data collected in real time is input into the trained LSTM model to achieve accurate prediction of the quality of HPLC links and HRF links in the future (such as the next 5 seconds, 10 seconds, etc.), including key indicators such as signal strength change trend and bit error rate fluctuation range, providing forward-looking decision-making basis for subsequent communication mode selection and rate adjustment.

[0092] Communication mode selection and optimization: Based on the prediction results of the LSTM neural network and combined with the current actual link quality detection data, an intelligent communication mode selection and optimization strategy is formulated. When it is predicted that the HPLC link will maintain good quality in the short term and meet the preset communication requirements, the HPLC communication mode is selected for data transmission first, and the relevant communication parameters are adjusted in advance according to the prediction information, such as optimizing the power line carrier frequency allocation and adjusting the encoding method to improve the transmission efficiency; if the prediction shows that the HPLC link is about to fail or the quality is significantly reduced, and the HRF link quality can be maintained at an acceptable level during the prediction period, it will be actively switched to the HRF communication mode in advance, and the transmission power and channel selection will be dynamically adjusted according to the wireless signal propagation characteristics to ensure the continuity and stability of communication. In some special scenarios, such as when there are extremely high requirements for data transmission rate and the prediction shows that both dual-mode links can provide stable services within a certain period of time, the data flow can be reasonably allocated to the HPLC and HRF dual channels based on the prediction results to achieve coordinated high-speed transmission and give full play to the advantages of the dual-mode network. In addition, while determining the communication mode, combined with the adaptive rate adjustment technology, the initial data transmission rate is intelligently determined according to the link quality prediction results and the current network load. For example, when it is predicted that the link quality is good and the network load is low, a higher data transmission rate is selected (such as 100Mbps in HPLC mode and 50Mbps in HRF mode); when there is a certain fluctuation in link quality or the network load is high, the data transmission rate is appropriately reduced to ensure the reliability and stability of data transmission.

[0093] Adaptive rate adjustment: During data transmission, the dynamic changes of the network environment are monitored in real time, including real-time fluctuations in link quality (such as changes in signal strength and bit error rate), changes in network load (such as the number of connected devices and data traffic size), and other factors. Based on the prediction ability of the LSTM neural network, the changing trend of link quality is predicted in advance, and the data transmission rate is dynamically adjusted in combination with the current actual monitoring data. When the link quality improves and the network load decreases, the data transmission rate is gradually increased; conversely, when the link quality decreases or the network load increases, the data transmission rate is promptly reduced to maintain the stability and reliability of data transmission. At the same time, a rate adjustment feedback mechanism is established to optimize and correct the rate adjustment strategy based on the actual effect of data transmission (such as the accuracy and real-time nature of data transmission and feedback information from the receiving end) to ensure that the adaptive rate adjustment mechanism can accurately adapt to changes in the network environment.

[0094] Data transmission and switching control: In the process of data transmission, efficient data transmission is carried out strictly in accordance with the selected communication mode and the adaptively determined data transmission rate. On the one hand, the real-time link quality under the current communication mode is closely monitored. If the link quality of the communication mode currently in use suddenly deteriorates to the point where it cannot meet normal communication needs, and the link quality of another communication mode can be restored to a usable state in a short period of time based on the LSTM prediction results, the data is quickly and smoothly switched to another communication mode to continue data transmission, and the communication mode switching information is recorded in detail, including the switching time, switching reason, and the link quality parameters before and after, so as to facilitate in-depth system performance analysis and optimization. On the other hand, the data is scientifically segmented and assigned a unique number before data transmission. Combined with error correction coding technology, it ensures that the data can be accurately reorganized, checked, and transmitted during the switching of different communication modes and rate adjustment, effectively avoiding data loss or errors.

[0095] Data fusion and optimization (receiving end): When dual-mode data is transmitted simultaneously, an intelligent data fusion and optimization processing mechanism is built at the receiving end. First, the data from HPLC and HRF are accurately reorganized based on the identification information and number sequence in the data packet header to restore the original data sequence. Then, advanced error correction algorithms, such as a hybrid error correction scheme based on a combination of cyclic redundancy check (CRC) and forward error correction (FEC), are used to perform comprehensive error detection and correction on the reorganized data, effectively eliminating various errors introduced during data transmission. Finally, according to the importance level of the data (such as control instruction data has the highest priority, followed by real-time monitoring data, and then ordinary business data) and real-time requirements (such as priority processing of time-sensitive video stream data), the data is intelligently sorted and optimized and integrated to ensure that the receiving end can quickly, accurately and orderly obtain complete and high-quality data information, providing reliable data support for the upper-level application system.

[0096] Example 2

[0097] like Figure 2 As shown, the present invention also proposes a dual-mode network communication system based on HPLC+HRF, comprising:

[0098] HPLC communication module 101, used to implement high-speed power line communication functions, including but not limited to power line signal modulation and demodulation, encoding and decoding, data transmission and reception operations;

[0099] HRF communication module 102, used to realize high-frequency wireless communication function, with the ability of transmitting and receiving wireless signals, frequency modulation, and channel coding;

[0100] A data collection and feature extraction unit 103, used to collect data related to link quality and extract features, including a power line data collection and feature extraction module for collecting HPLC link related data and extracting features, and a wireless data collection and feature extraction module for collecting HRF link related data and extracting features;

[0101] The LSTM neural network prediction unit 104 is used to predict the link quality based on the LSTM model, including an LSTM model construction module, a training module and a prediction calculation module, which are respectively responsible for building the model, training the model and performing link quality prediction calculation;

[0102] The communication mode selection and optimization unit 105 is used to select and optimize the communication mode according to the prediction result and the actual link quality, and includes a prediction result analysis module, a communication mode judgment module and a parameter optimization module. The prediction result analysis module is used to analyze the prediction result of the LSTM neural network, the communication mode judgment module is used to judge the communication mode according to the analysis result and the actual link quality, and the parameter optimization module is used to optimize the parameters of the selected communication mode;

[0103] The adaptive rate adjustment unit 106 is used to monitor the dynamic changes of the network environment in real time and adjust the data transmission rate, including a rate monitoring module, a rate calculation module and a feedback optimization module. The rate monitoring module is used to monitor the link quality and network load. The rate calculation module calculates the rate based on the LSTM prediction result and the monitoring data. The feedback optimization module optimizes the rate adjustment strategy according to the actual effect of data transmission;

[0104] The data processing and switching control unit 107 is used for segmenting, numbering, error correction coding, communication mode switching and rate adjustment control of data, and includes a data segmentation module, a number adding module, an error correction coding module, a cache switching module and a rate adjustment execution module. The data segmentation module is used to segment the data, the number adding module is used to add numbers to the segmented data, the error correction coding module is used to add error correction coding information, the cache switching module is used to cache the untransmitted data when the communication mode is switched and start another communication mode to continue transmission, and the rate adjustment execution module is used to perform the rate adjustment operation;

[0105] The data fusion and optimization unit 108 is used to perform fusion processing, error correction and optimization sorting on the dual-mode transmitted data at the receiving end.

[0106] In this embodiment, the rate monitoring module, the rate calculation module and the feedback optimization module are specifically:

[0107] (1) Rate monitoring module, including link quality monitoring submodule and network load monitoring submodule

[0108] Link quality monitoring submodule: Through deep interaction with the HPLC communication module and the HRF communication module, link quality related parameters are obtained in real time. For the HPLC link, the signal strength, bit error rate, noise level and other indicators on the power line are continuously monitored; for the HRF link, close attention is paid to the reception strength of the wireless signal, the signal-to-noise ratio, the signal fluctuation caused by the multipath effect and other information. High-precision sensors and signal detection circuits are used to ensure that the collected data is highly accurate and real-time. For example, for the monitoring of signal strength, the resolution can be accurate to 0.1dBm, and the statistical error of the bit error rate is controlled within a very small range.

[0109] Network load monitoring submodule: Comprehensively monitor the data traffic in the network, including the number of currently connected devices, the frequency of data transmission requests of each device, the size of the data packet of the data transmission, the transmission rate and other information. By deploying traffic monitoring software or hardware probes at key nodes of the network, network traffic data can be captured in real time and analyzed in depth. For example, the proportion and change trend of different types of data traffic (such as control instruction traffic, real-time monitoring data traffic, and ordinary business data traffic) can be distinguished, providing a comprehensive network load information basis for rate calculation and adjustment.

[0110] (2) Rate calculation module, including link quality rate adjustment submodule and network load rate adjustment submodule

[0111] Rate adjustment submodule based on link quality: Based on the link quality parameters obtained by the link quality monitoring submodule, the data transmission rate is calculated using a pre-set rate adjustment algorithm. When the HPLC link signal strength is high and the bit error rate is low, a basic rate increase ratio is determined based on the comparison between the signal strength and the preset signal strength threshold, and the bit error rate and the bit error rate threshold. For example, if the signal strength is 10dBm higher than the good threshold and the bit error rate is less than 0.1%, the basic rate can be increased by 50%. For HRF links, a similar calculation method is used based on parameters such as reception strength and signal-to-noise ratio, while considering the propagation characteristics of wireless signals and channel state information, such as the degree of multipath fading, channel busyness and other factors that affect the rate. For example, when the channel is relatively idle and the reception strength is stable at a high level, the modulation order of data transmission is appropriately increased to increase the transmission rate.

[0112] Rate adjustment submodule based on network load: Combined with the network load information provided by the network load monitoring submodule, the data transmission rate is further optimized. When the number of connected devices in the network is small and the data flow is low, the rate is appropriately increased to make full use of network bandwidth resources; conversely, when the network load is high, such as when the number of connected devices exceeds the preset device number threshold and the total data flow is close to the upper limit of the network bandwidth, the data transmission rate is proportionally reduced according to the load level. For example, if the network load reaches 80% bandwidth utilization, the rate is reduced by 30% to avoid data transmission errors or delays caused by network congestion. At the same time, this submodule also considers the priority of different types of data traffic to ensure that high-priority data (such as control instructions) can still maintain a relatively stable transmission rate when the network load is high, while low-priority data is subject to more rate adjustments.

[0113] (3) Feedback optimization module, including transmission effect evaluation submodule, strategy adjustment submodule,

[0114] Transmission effect evaluation submodule: During the data transmission process, the information fed back by the receiving end is collected in real time, and the data transmission effect is evaluated in multiple dimensions. By analyzing factors such as the change in the bit error rate of the data received by the receiving end, the delay time of data transmission, data integrity, and the processing feedback of the upper-layer application system on the data (such as whether the data triggers the corresponding control operation or business processing in a timely and accurate manner), the effectiveness of the current rate adjustment strategy is comprehensively evaluated. For example, if the bit error rate fed back by the receiving end continues to be higher than the preset acceptable bit error rate upper limit, and the data transmission delay causes abnormal operation of the upper-layer application system, it is determined that there is a problem with the current rate adjustment strategy.

[0115] Strategy adjustment submodule: optimize and modify the rate adjustment strategy according to the evaluation results of the transmission effect evaluation submodule. When it is found that a certain rate adjustment strategy leads to poor data transmission effect, try different adjustment methods. For example, if the rate increase based on link quality leads to an increase in the bit error rate, you may consider adjusting the encoding method, switching from efficient but weak error correction coding to stronger error correction but slightly lower efficiency coding, and adjusting the rate calculation parameters in the rate calculation module accordingly; or adjust the transmission time interval to reduce the bit error rate by extending the time interval for data transmission, but at the same time it will slightly reduce the transmission rate. In addition, the strategy adjustment submodule can also automatically adjust the parameters and thresholds of the rate adjustment strategy according to the long-term change trend of the network environment (such as the periodic change law of the network load within a day, the stability difference of the link quality in different time periods, etc.), so that the entire adaptive rate adjustment mechanism can better adapt to the dynamic changes of the network and continuously improve the performance and reliability of data transmission.

[0116] In this embodiment, the data segmentation module, the number adding module, the error correction coding module, the cache switching module and the rate adjustment execution module are specifically:

[0117] (1) Data segmentation module

[0118] Segmentation strategy formulation submodule: formulate scientific and reasonable data segmentation strategies based on network communication protocols and the characteristics of data transmission. Taking into account the transmission requirements and processing efficiency of different types of data (such as control instruction data, real-time monitoring data, large-capacity file data, etc.), different segment length ranges are set for each type of data. For example, for control instruction data, since its data volume is usually small and the real-time requirements are extremely high, its segment length is set to a shorter fixed value, such as 64 bytes, to ensure fast transmission and processing; for real-time monitoring data, according to the frequency of data generation and the size of data volume, the segment length is set between 256-512 bytes; and for large-capacity file data, a dynamic segmentation method is used, and the segment length is adjusted between 1024-4096 bytes according to the current bandwidth status and transmission load of the network, so as to improve transmission efficiency while ensuring data integrity.

[0119] Segmentation execution submodule: According to the segmentation rules determined by the segmentation strategy formulation submodule, the data to be transmitted is accurately segmented. In the segmentation process, the logical structure and integrity of the data are fully considered to avoid dividing closely related data into different segments. For example, for a data packet containing multiple data fields and there is a dependency relationship between the fields, by analyzing the structure and protocol specifications of the data packet, it is ensured that the related fields are completely divided into the same segment during segmentation. At the same time, after the segmentation operation is completed, the corresponding segment identification information is generated for each segment of data. The identification information contains key information such as the segment number, the data packet identifier, and the data type, so that each segment of data can be accurately identified and processed in the subsequent data transmission, switching, and reorganization process.

[0120] (2) Numbering module

[0121] Numbering rule setting submodule: Develop a comprehensive and unique numbering rule system for adding numbers to segmented data. The numbering rule should cover important dimensions such as the source of the data (i.e., from the HPLC or HRF communication module), transmission sequence information, the communication session identifier, and data version information. For example, a hierarchical numbering structure is adopted, in which the high-order part indicates the source of the data (such as 0 for HPLC, 1 for HRF), the middle part indicates the transmission sequence number, and the numbering is incremented in the order in which the data is generated or enters the transmission queue. The low-order part indicates the unique identifier of the communication session and the data version number, so as to ensure that in a complex network communication environment, data from different sources, different orders, and different versions can be accurately distinguished.

[0122] Numbering Adding Execution Submodule: According to the numbering rules set by the Numbering Rules Setting Submodule, add corresponding numbering information to each segment of data. In the process of adding numbers, ensure the accuracy and completeness of the numbering information, and tightly bind the numbering information to the data segment, so that during the data transmission process, whether in the same communication mode or in the process of switching communication modes, the receiving end can quickly and accurately reorganize and process the data according to the numbering information. At the same time, the numbering information is verified and encrypted to prevent the number from being tampered with or damaged during the transmission process, thereby ensuring the reliability and security of data transmission.

[0123] (3) Error Correction Coding Module

[0124] Coding mode selection submodule: Dynamically select the appropriate error correction coding mode according to the reliability requirements of network communication, the current link quality and network load. When the link quality is good and the network load is low, give priority to coding modes with high coding efficiency but relatively weak error correction capabilities, such as convolutional coding, to increase the data transmission rate; when there is a certain fluctuation in link quality or the network load is high, switch to coding modes with strong error correction capabilities but slightly lower coding efficiency, such as Turbo coding or low-density parity check code (LDPC). For example, by real-time monitoring of the link's bit error rate and the network's bandwidth utilization, when the bit error rate is less than 0.1% and the bandwidth utilization is less than 50%, select convolutional coding; when the bit error rate is higher than 0.5% or the bandwidth utilization is higher than 80%, switch to Turbo coding or LDPC coding.

[0125] Coding execution submodule: According to the error correction coding method determined by the coding method selection submodule, the segmented and numbered data is coded. During the coding process, the data is coded accordingly according to the algorithm requirements of the selected coding method to generate a coded data block containing error correction information. For example, for Turbo coding, the data is interleaved according to a specific interleaving rule, and then encoded by two recursive system convolution encoders respectively, and then the encoded results are combined and processed to generate the final coded data block. At the same time, the coding method identification information is added to the coded data block so that the receiving end can accurately identify the coding method used during decoding, thereby performing the correct decoding operation.

[0126] (4) Cache switching module

[0127] Cache management submodule: Build an efficient cache management system to store data that has not been fully transmitted during communication mode switching or rate adjustment. The cache management system adopts a dynamic memory allocation strategy to reasonably allocate cache space based on the type, size, and expected cache time of the data. For example, a small but extremely fast-response cache area is allocated for control instruction data and real-time monitoring data to ensure that these data with high real-time requirements can be quickly read and written in the cache; for large-capacity file data, a larger cache space is allocated, and a hierarchical cache structure is adopted to store data that may be transmitted in the near future in the first-level cache close to the processor, and data that will be transmitted in the distant future is stored in the second or third-level cache to improve cache utilization and data read and write efficiency.

[0128] Switching control submodule: When communication mode switching or rate adjustment is required, it is responsible for coordinating data caching and switching operations. Before the communication mode is switched, the switching control submodule first suspends data transmission in the current communication mode, quickly transfers the untransmitted data to the corresponding cache area of ​​the cache management submodule, and records the location information of the current transmission (such as the range of transmitted data segment numbers, the storage location of the data in the cache, etc.). Then, according to the switched communication mode or the adjusted rate requirements, read the data from the cache and start the new communication mode or rate for data transmission. During the switching process, ensure the continuity and integrity of the data to avoid data loss or repeated transmission. For example, when switching from HPLC communication mode to HRF communication mode, the switching control submodule converts and processes the data in the HPLC cache according to the transmission requirements of the HRF communication module, and then transmits it to the HRF communication module for transmission, while updating the relevant transmission status information and cache management information.

[0129] (5) Rate Adjustment Execution Module

[0130] Rate adjustment instruction receiving submodule: Receives rate adjustment instructions from the adaptive rate adjustment unit. The instructions contain key information such as the target rate value, adjustment reason (such as link quality change, network load change, etc.) and adjustment time requirements. The rate adjustment instruction receiving submodule parses and verifies the received instructions to ensure the integrity and accuracy of the instructions. For example, check whether the target rate value in the instruction is within the rate range allowed by the system, whether the adjustment reason is reasonable and consistent with the current network status, etc. If the instruction is found to be erroneous or incomplete, promptly feedback the error information to the adaptive rate adjustment unit and request to resend the instruction.

[0131] Rate adjustment operation execution submodule: According to the instruction information parsed by the rate adjustment instruction receiving submodule, the actual adjustment operation is performed on the data transmission rate. When executing the rate adjustment, it works in coordination with relevant modules such as the communication mode selection and optimization unit, the data processing and switching control unit, etc. For example, when the rate needs to be reduced, the rate adjustment operation execution submodule may adjust the time interval for data transmission, reduce the modulation order, or reduce the number of data channels for parallel transmission; when the rate needs to be increased, the opposite operation is taken, such as shortening the transmission time interval, increasing the modulation order, or increasing the number of parallel transmission channels. At the same time, during the rate adjustment process, the status and performance indicators of data transmission (such as bit error rate, transmission delay, etc.) are monitored in real time to ensure that the rate adjustment operation does not have a negative impact on the stability and reliability of data transmission. If an abnormality is found in the data transmission during the adjustment process, the rate adjustment operation is suspended in time, and the abnormality is reported to the adaptive rate adjustment unit for further analysis and processing.

[0132] In this embodiment, the data fusion and optimization unit includes a data fusion processing module, an error correction module, and an optimization sorting module, wherein:

[0133] (1) Data fusion processing module

[0134] Data synchronization and alignment submodule: Since the data transmitted by the dual modes may differ in transmission time, transmission order, etc., the data synchronization and alignment submodule first performs time synchronization processing on the data from HPLC and HRF. Through the timestamp information carried in the data packet header, the data arriving at the receiving end at different times are aligned according to their original transmission order. For example, if it is found that the timestamp of a certain data segment from HPLC is earlier than that of the adjacent data segment from HRF, but it arrives at the receiving end later due to transmission delays, etc., it will be adjusted to the correct sequential position to ensure the consistency of the data sequence on the time axis. At the same time, according to the numbering information added to each segment of data by the data segmentation module at the sending end, the data is further numbered and aligned to ensure the logical coherence and integrity of the data, laying the foundation for subsequent fusion processing.

[0135] Data merging and deduplication submodule: After data synchronization and alignment, the data merging and deduplication submodule merges the data from the two modes. For repeatedly transmitted data segments (which may be due to redundant transmission or network jitter during the communication mode switching process, resulting in repeated reception of some data), duplicate data is identified and removed by comparing the content, number, and related feature information of the data segment. For example, a hash algorithm is used to perform fast hash calculations on the data segment, and the calculated hash value is used as the unique identifier of the data segment. The hash value is compared to determine whether the data segment is repeated. After removing duplicate data, the data from different modes are merged according to the data packets or data sequences to which they originally belonged to form a complete, non-duplicate data set, ready for error correction processing.

[0136] (2) Error Correction Module

[0137] Error detection submodule: Use a variety of error detection technologies to perform comprehensive detection on the fused data. First, based on the cyclic redundancy check (CRC) mechanism, the data integrity is checked according to the CRC check code added by the data sender. Calculate the CRC value of the received data and compare it with the CRC check code sent by the sender. If the two do not match, it indicates that an error may have occurred in the data transmission process. At the same time, combined with other error detection methods, such as parity check, checksum and other methods, multi-dimensional error detection is performed on the data. For example, for some control instruction data with extremely high requirements for data accuracy, in addition to CRC check, multiple parity checks are also used to detect whether there are error bits in the data from different angles, thereby improving the accuracy and reliability of error detection.

[0138] Error correction execution submodule: When the error detection submodule detects that there is an error in the data, the error correction execution submodule performs the corresponding error correction operation according to the error correction coding method adopted by the data sender. If the sender adopts forward error correction (FEC) coding, such as Turbo coding or low-density parity check code (LDPC), the erroneous data is corrected by using the redundant information of the coding and the error correction algorithm. For example, for LDPC-coded data, the error bits are gradually repaired and the original data information is restored through the iterative decoding algorithm according to the check matrix and the received coded data. During the error correction process, multiple iterative calculations may be required to ensure that the error is completely corrected, depending on the severity of the error and the performance of the error correction algorithm. If there are still uncorrectable errors after the error correction algorithm is processed, the erroneous data will be marked and the relevant error information will be recorded, and an error prompt will be sent to the upper-layer application system so that further processing measures can be taken, such as requesting retransmission of data.

[0139] (3) Optimize the sorting module

[0140] Data importance assessment submodule: Evaluate and classify the importance of data. Determine the importance level of data based on factors such as the type and source of data and the degree of impact on the upper-level application system. For example, control instruction data (such as device switch instructions in smart home systems, power dispatch instructions in smart grids, etc.) are set to the highest importance level because these data directly affect the operating status and control functions of the system; real-time monitoring data (such as environmental sensor data, equipment operating status data, etc.) are set to a higher importance level, and these data have important reference value for real-time monitoring and decision-making of the system; and ordinary business data (such as user entertainment video data, file transfer data, etc.) are set to a lower importance level. By assigning different importance weights to different types of data, a basis is provided for subsequent optimization sorting.

[0141] Sorting execution submodule: According to the importance level and weight determined by the data importance assessment submodule, the corrected data is optimized and sorted. Data structures such as priority queues are used to prioritize data with high importance levels at the front of the queue to ensure that these data can be processed by the upper-layer application system first. For example, at the receiving end of the smart home system, when device control instruction data and ordinary entertainment data are received at the same time, the control instruction data is prioritized into the processing queue so that it can be executed immediately, while the entertainment data is placed at the back of the queue, waiting for the system to process it in idle time. At the same time, in the sorting process, the real-time requirements of the data are also considered. For data with high real-time requirements (such as key frame data in real-time video stream data), even if its importance level is relatively low, its position in the queue will be appropriately advanced to ensure that the data can be processed in time and meet the real-time requirements of the upper-layer application system, thereby improving the data processing efficiency and quality of the entire dual-mode network communication system at the receiving end, and providing more reliable and efficient data services for the upper-layer application system.

[0142] The specific implementation principle of the dual-mode network communication system based on HPLC+HRF of the present invention is:

[0143] HPLC communication module: undertakes the core functions of high-speed power line communication, including efficient modulation and demodulation of power line signals, accurate encoding and decoding, stable data transmission and reception, etc. Through close connection with the power line network, it realizes reliable transmission and interaction of data on the power line medium.

[0144] HRF communication module: responsible for the realization of high-frequency wireless communication functions, with the ability of strong transmission and high-sensitivity reception of wireless signals, flexible frequency modulation, efficient channel coding, etc., with the help of wireless antennas and peripheral equipment to build wireless data communication links, to ensure the smooth transmission of data in wireless channels.

[0145] Data acquisition and feature extraction unit: It is deeply connected with the HPLC communication module and the HRF communication module to collect various raw data reflecting the link quality in real time, and use advanced data processing algorithms and feature engineering techniques to extract key feature vectors closely related to the link quality changes, providing a rich and accurate data basis for subsequent LSTM neural network predictions.

[0146] LSTM neural network prediction unit: Based on the constructed LSTM neural network model, it receives real-time feature data from the data acquisition and feature extraction unit, uses the trained model to perform link quality prediction calculations, and outputs the prediction results of the dual-mode network communication link quality in the future, including the change trends of key indicators such as signal strength and bit error rate, providing forward-looking decision-making basis for the communication mode selection and optimization unit and the adaptive rate adjustment unit.

[0147] Communication mode selection and optimization unit: Comprehensively consider the prediction results provided by the LSTM neural network prediction unit and the current actual link quality information fed back by the link quality detection unit, determine the most suitable data transmission communication mode according to the preset intelligent communication mode selection strategy, and dynamically optimize and adjust the communication parameters under the selected mode. At the same time, control the precise transmission path of data on the corresponding communication module to achieve intelligent switching and optimized configuration of communication modes.

[0148] Adaptive rate adjustment unit: monitors the dynamic changes of the network environment in real time, including factors such as link quality and network load, and dynamically calculates and adjusts the data transmission rate based on the prediction results of the LSTM neural network and the current actual monitoring data. At the same time, a rate adjustment feedback mechanism is established to optimize and correct the rate adjustment strategy according to the actual effect of data transmission, ensuring the effectiveness and accuracy of the adaptive rate adjustment mechanism.

[0149] Data processing and switching control unit: During the data transmission process, it is responsible for scientific segmentation processing of data, unique number marking, and fine control operations during communication mode switching and rate adjustment. Through carefully designed data segmentation strategies and numbering rules, combined with efficient cache management mechanisms, it ensures the integrity, accuracy and continuity of data during the switching of different communication modes and rate adjustment, effectively avoiding data transmission interruption or confusion.

[0150] Data fusion and optimization unit (receiving end): build an intelligent data fusion and optimization processing engine at the receiving end to accurately reorganize, deeply detect and correct errors, and intelligently sort and optimize the data from dual-mode transmission. By integrating a variety of advanced data processing technologies and algorithms, it ensures that the receiving end can quickly and accurately obtain complete and high-quality data information, providing solid data guarantee for the stable operation of the upper-level application system.

[0151] In summary, the present invention has the following technical effects:

[0152] (1) Improve communication reliability:

[0153] Through the LSTM neural network's accurate prediction of the quality of the dual-mode network communication link, the system can know the trend of link quality changes in advance and take proactive measures before the link quality deteriorates, such as switching the communication mode in advance or adjusting the data transmission rate, effectively avoiding data transmission interruptions or errors caused by sudden link failures. For example, in a smart factory environment, for the transmission of control instructions for production equipment, this reliability guarantee can prevent production accidents caused by communication failures and ensure the continuity and stability of the production process.

[0154] Adaptive rate adjustment technology dynamically adjusts the data transmission rate according to the link quality and network load. When the link quality is poor or the network is congested, the rate is reduced to ensure the accuracy of data transmission. When the link quality is good and the network is idle, the rate is increased to improve transmission efficiency, thereby maintaining reliable data transmission in various complex network environments.

[0155] The error detection and correction mechanism in data fusion processing can identify and repair erroneous data generated during dual-mode transmission. Whether it is HPLC data errors caused by power line interference or HRF data errors caused by multipath effects of wireless signals, they can be effectively processed, further enhancing the reliability of data transmission and ensuring the integrity and accuracy of data obtained by the receiving end.

[0156] (2) Enhance network adaptability:

[0157] Combining HPLC and HRF dual-mode communications, it fully leverages the advantages of wide coverage of power line networks and flexible deployment of wireless networks, and can adapt to communication needs in different scenarios, such as Figure 4 As shown in Figure 1, it is a node topology diagram of HPLC and HRF dual-mode communication. Figure 3 As shown, it is a schematic diagram of the application of HPLC and HRF dual-mode communication in the present invention in the occasion of electric energy data communication. In the figure, except for the water / heat / gas meter and the intelligent leakage protection switch which are HRF single-mode communication, the rest are HPLC+HRF dual-mode communication.

[0158] LSTM neural network prediction combined with adaptive rate adjustment enables the system to automatically optimize communication parameters according to changes in network environment at different times and in different areas. For example, during peak network usage in the office area, the data transmission rate is automatically reduced and dual-mode traffic is reasonably allocated according to the increase in network load; and at night when the network is idle, the rate is increased to speed up the execution of tasks such as data backup or system updates, greatly improving the network's adaptability to different usage scenarios and load changes.

[0159] The data acquisition and feature extraction unit collects and analyzes multi-dimensional data, and can perceive subtle changes in the network environment in real time, including changes in power line noise, dynamics of wireless signal interference sources, etc., providing rich information basis for the system to adjust communication strategies in a timely manner, allowing the network to quickly adapt to complex and changeable actual application environments.

[0160] (3) Optimize data transmission efficiency:

[0161] The communication mode selection and optimization strategy based on the prediction results of the LSTM neural network can select the optimal communication mode or enable dual-mode collaborative transmission at the right time. For example, when it is predicted that the HPLC link quality will be good for a period of time and large data files will be transmitted, the HPLC mode will be selected first and its parameters will be optimized to make full use of the high bandwidth characteristics of the power line network and increase the data transmission rate; when mobility is required or the wireless signal in the local area is good, it will switch to the HRF mode, avoiding the performance bottleneck of a single mode in a specific scenario and maximizing the data transmission efficiency.

[0162] Adaptive rate adjustment technology ensures data transmission reliability while flexibly adjusting the rate according to link quality and network load, avoiding resource waste or transmission delay caused by fixed rate settings. For example, when transmitting high-definition video streams, dynamically adjusting the rate according to network conditions can ensure smooth video playback, without causing network congestion and video freezes due to too high a rate, or causing serious degradation of video quality due to too low a rate, effectively improving user experience and actual data transmission efficiency.

[0163] (4) Intelligent decision-making and management:

[0164] The introduction of LSTM neural network enables the system to have the ability to intelligently predict changes in link quality. Operations such as communication mode selection and optimization, and adaptive rate adjustment are all based on the results of such intelligent predictions, reducing the limitations of human intervention and traditional empirical decision-making. For example, in multiple IoT application scenarios in smart cities, the system can automatically make the best communication decisions based on the network conditions in different regions and time periods, without the need for manual configuration and adjustment one by one, greatly improving the intelligence level and efficiency of network management.

[0165] The data fusion and optimization unit intelligently processes data at the receiving end, including not only error correction and sorting, but also dynamic management based on the importance and real-time nature of the data. For example, for some urgent security alert data, it can directly skip the regular data queuing and waiting process and trigger the corresponding security response mechanism first, which reflects the system's intelligent priority management capability in the data processing link and improves the entire network communication system's ability to respond to emergencies and its overall intelligence.

[0166] The entire dual-mode network communication system realizes intelligent monitoring and management of the entire network communication process through the collaborative work and data interaction between various modules. From data collection and feature extraction to link quality prediction, communication mode selection, data transmission control, and data processing at the receiving end, each link is interconnected and optimized based on intelligent algorithms, forming a closed-loop intelligent communication management system, which effectively improves the stability, reliability and overall performance of network communication and meets the high requirements of modern Internet of Things applications for intelligent network communication.

[0167] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A dual-mode network communication method based on HPLC+HRF, characterized in that: The following steps are involved: Network initialization: initializing and configuring the HPLC module and the HRF module respectively, and performing channel scanning operations to determine available channel resources; Data collection and feature extraction: continuously collect multi-dimensional data related to the quality of HPLC link and HRF link at a specific sampling frequency, and perform data feature extraction; LSTM neural network training and prediction: building and training an LSTM neural network model with multiple hidden layers. After training, use real-time feature data to predict the quality indicators of the dual-mode network communication link. Communication mode selection and optimization: switching communication modes based on LSTM neural network prediction results and current actual link quality detection data; Adaptive rate adjustment: monitor the dynamic changes of the network environment in real time during data transmission, dynamically adjust the data transmission rate based on the LSTM neural network prediction results combined with the current actual monitoring data, establish a rate adjustment feedback mechanism, and optimize and correct the rate adjustment strategy according to the actual effect of data transmission; Data transmission and switching control, during the data transmission process, data transmission is performed according to the selected communication mode and data transmission rate, real-time link quality in the current communication mode is monitored, and the communication mode is automatically switched if the link quality deteriorates and the link quality of another communication mode is available; Data fusion and optimization, when dual mode is used to transmit data simultaneously, the data from HPLC and HRF are fused, error corrected and optimized sorted at the receiving end, and data are reorganized according to the identification information and numbering sequence in the data packet header. A hybrid error correction scheme based on CRC and FEC is used for error detection and correction, and a priority queue algorithm is used for sorting and optimization integration.

2. The dual-mode network communication method based on HPLC+HRF according to claim 1, characterized in that: In the data collection and feature extraction steps, for the HPLC link, fast Fourier transform is used to calculate the signal frequency components and principal component analysis is used to extract feature vectors. For the HRF link, wavelet transform is used to extract the time-frequency domain features of the wireless signal.

3. The dual-mode network communication method based on HPLC+HRF according to claim 1, characterized in that: In the LSTM neural network training and prediction steps, the number of hidden layers of the constructed LSTM neural network model is 3-5 layers, the number of input layer nodes is 10-20, the number of output layer nodes corresponds to the signal strength and the bit error rate is 2, and the mean square error is used as the loss function to evaluate the model prediction performance.

4. The dual-mode network communication method based on HPLC+HRF according to claim 1, characterized in that: In the communication mode selection and optimization step, the HPLC communication parameters adjusted in advance include optimizing the power line carrier frequency allocation and adjusting the encoding method, and the adjustment under the HRF communication mode includes dynamically adjusting the transmission power and channel selection according to the wireless signal propagation characteristics.

5. The dual-mode network communication method based on HPLC+HRF according to claim 1, characterized in that: In the adaptive rate adjustment step, the rate adjustment is performed according to a preset rate adjustment strategy, and the strategy is optimized. The adjustment methods include adjusting the encoding method and adjusting the transmission time interval.

6. The dual-mode network communication method based on HPLC+HRF according to claim 1, characterized in that: In the data transmission and switching control step, the length of the data segment processing is a preset fixed length, and the switching identifier includes the switching reason and the number range of the data segments that have been transmitted in the original communication mode.

7. The dual-mode network communication system based on HPLC+HRF is characterized by: include: HPLC communication module, used to realize high-speed power line communication functions, including but not limited to power line signal modulation and demodulation, encoding and decoding, data transmission and reception operations; HRF communication module, used to realize high-frequency wireless communication function, with the ability of transmitting and receiving wireless signals, frequency modulation, and channel coding; A data acquisition and feature extraction unit, used to acquire data related to link quality and extract features, including a power line data acquisition and feature extraction module for acquiring data related to HPLC links and extracting features, and a wireless data acquisition and feature extraction module for acquiring data related to HRF links and extracting features; The LSTM neural network prediction unit is used to predict the link quality based on the LSTM model, including the LSTM model construction module, training module and prediction calculation module, which are responsible for model construction, model training and link quality prediction calculation respectively; A communication mode selection and optimization unit, which is used to select and optimize the communication mode according to the prediction results and the actual link quality, and includes a prediction result analysis module, a communication mode judgment module and a parameter optimization module. The prediction result analysis module is used to analyze the LSTM neural network prediction results, the communication mode judgment module is used to judge the communication mode according to the analysis results and the actual link quality, and the parameter optimization module is used to optimize the parameters of the selected communication mode; The adaptive rate adjustment unit is used to monitor the dynamic changes of the network environment in real time and adjust the data transmission rate, including a rate monitoring module, a rate calculation module and a feedback optimization module; the rate monitoring module is used to monitor the link quality and network load, the rate calculation module calculates the rate based on the LSTM prediction results and monitoring data, and the feedback optimization module optimizes the rate adjustment strategy according to the actual effect of data transmission; A data processing and switching control unit, used for segmenting, numbering, error correction coding, communication mode switching and rate adjustment control of data, including a data segmentation module, a number adding module, an error correction coding module, a cache switching module and a rate adjustment execution module; the data segmentation module is used for segmenting data, the number adding module is used for adding numbers to the segmented data, the error correction coding module is used for adding error correction coding information, the cache switching module is used for caching untransmitted data when the communication mode is switched and starting another communication mode for continued transmission, and the rate adjustment execution module is used for performing rate adjustment operations; The data fusion and optimization unit is used to perform fusion processing, error correction and optimization sorting of dual-mode transmitted data at the receiving end.

8. The dual-mode network communication system based on HPLC+HRF according to claim 7, characterized in that: The rate monitoring module, rate calculation module and feedback optimization module are specifically: (1) Rate monitoring module, including link quality monitoring submodule and network load monitoring submodule The link quality monitoring submodule is used to obtain the link quality related parameters in real time through interaction with the HPLC communication module and the HRF communication module; The network load monitoring submodule is used to monitor the data traffic in the network, including but not limited to the number of currently connected devices, the frequency of data transmission requests from each device, the size of data packets transmitted, and the transmission rate; (2) The rate calculation module includes a link quality rate adjustment submodule and a network load rate adjustment submodule. The link quality rate adjustment submodule calculates the data transmission rate using a preset rate adjustment algorithm based on the link quality parameters obtained by the link quality monitoring submodule; The network load rate adjustment submodule optimizes the data transmission rate in combination with the network load information provided by the network load monitoring submodule; (3) Feedback optimization module, including transmission effect evaluation submodule and strategy adjustment submodule The transmission effect evaluation submodule is used to collect the information fed back by the receiving end in real time during the data transmission process and to perform multi-dimensional evaluation on the data transmission effect; The strategy adjustment submodule is used to optimize and modify the rate adjustment strategy according to the evaluation result of the transmission effect evaluation submodule.

9. The dual-mode network communication system based on HPLC+HRF according to claim 7, characterized in that: The data segmentation module, the numbering module, the error correction coding module, the cache switching module and the rate adjustment execution module include: (1) Data segmentation module, including segmentation strategy formulation submodule and segmentation execution submodule The segmentation strategy formulation submodule is used to formulate data segmentation strategies based on the network communication protocol and the characteristics of data transmission; The segmentation execution submodule is used to segment the data to be transmitted according to the segmentation rules determined by the segmentation strategy formulation submodule; (2) Numbering adding module, including numbering rule setting submodule and numbering adding execution submodule The numbering rule setting submodule is used to formulate a numbering rule system to add numbers to the segmented data. The numbering rule should cover the important dimensions of the data source, transmission sequence information, communication session identifier, and data version information; The number adding execution submodule adds corresponding number information to each piece of data according to the numbering rules set by the numbering rule setting submodule; (3) Error correction coding module, including coding mode selection submodule and coding execution submodule The coding mode selection submodule is used to dynamically select the appropriate error correction coding mode according to the current link quality and network load; The coding execution submodule is used to perform coding operations on the segmented and numbered data according to the error correction coding mode determined by the coding mode selection submodule; (4) Cache switching module, including cache management submodule and switching control submodule The cache management submodule is used to store the data that has not been fully transmitted during the communication mode switching or rate adjustment process; The switching control submodule coordinates data caching and switching operations when communication mode switching or rate adjustment is required; (5) Rate adjustment execution module, including rate adjustment instruction receiving submodule and rate adjustment operation execution submodule The rate adjustment instruction receiving submodule is used to receive the rate adjustment instruction from the adaptive rate adjustment unit, the instruction including the target rate value, the adjustment reason and the adjustment time requirement information; The rate adjustment operation execution submodule is used to adjust the data transmission rate according to the instruction information parsed by the rate adjustment instruction receiving submodule.

10. The dual-mode network communication system based on HPLC+HRF according to claim 7, characterized in that: The data fusion and optimization unit includes a data fusion processing module, an error correction module, and an optimization sorting module, wherein: (1) Data fusion processing module, including data synchronization and alignment submodule, data merging and deduplication submodule The data synchronization and alignment submodule uses the timestamp information carried in the data packet header to align the data arriving at the receiving end at different times according to their original sending order; Data merging and deduplication submodule, after data synchronization and alignment, the data merging and deduplication submodule merges the data from the two modes; (2) Error correction module, including error detection submodule and error correction execution submodule The error detection submodule is used to perform data integrity check based on the cyclic redundancy check mechanism and the CRC check code added by the data sender, calculate the CRC value of the received data, and compare it with the CRC check code sent by the sender. If the two do not match, it indicates that an error may have occurred in the data transmission process; The error correction execution submodule is used for performing corresponding error correction operations according to the error correction coding method adopted by the data sending end when the error detection submodule detects that there is an error in the data; (3) Optimize the sorting module, including the data evaluation submodule and the sorting execution submodule The data importance assessment submodule determines the importance level of data based on factors including but not limited to the type and source of the data and the degree of impact on the upper-layer application system; The sorting execution submodule is used to optimize the sorting of the error-corrected data according to the importance level and weight determined by the data importance evaluation submodule.

Images