Electronic building block type edge embedded data acquisition method based on LoRa communication technology

Abstract

The invention discloses an electronic building block type edge embedded data acquisition method based on a LoRa communication technology, and relates to the technical field of Internet of Things and edge computing, and the method comprises the following steps: in a LoRa communication process, through a data monitoring module deployed in a LoRa receiver (gateway) and a network server, carrying out edge embedded data acquisition on the LoRa; multi-dimensional communication parameters generated by the nodes in the data transmission process are collected and recorded in real time; the method comprises the following steps: preprocessing collected original data information, and constructing a multi-dimensional structured data set by taking'node identification + timestamp 'as a main index; according to the method, conflicts among codes are identified through communication parameter sensing, feature engineering and machine learning, spreading factors are dynamically adjusted based on conflict strength and node priorities, signal time-frequency decoupling is achieved, the data reliability and channel utilization efficiency of the LoRa network in a dense deployment scene are effectively improved, the method has self-adaptive regulation and control capacity, and the data loss risk is reduced.

Description

Technical Field

[0001] The present invention relates to the technical fields of the Internet of Things and edge computing, and particularly to an edge-embedded data acquisition method based on LoRa communication technology in the form of electronic building blocks. Background Art

[0002] "Edge-embedded data acquisition based on LoRa communication technology" refers to an innovative data acquisition method that combines low-power wide-area network communication (LoRa) with a modular hardware structure. This method utilizes the pluggable and combinable characteristics of electronic building blocks to integrate various environmental sensors (such as temperature and humidity, light, air pressure, etc.) into building block modules with independent functions. Each module is embedded with a low-power embedded processor and a LoRa communication unit, which can perform preliminary processing at the edge level after collecting data, such as data preprocessing, denoising, and screening, only retaining the valid information and remotely sending it to the central receiver or cloud platform using the LoRa protocol. This method realizes a data acquisition mechanism of "nearby processing and remote transmission", which not only reduces network bandwidth and energy consumption, but also greatly improves the flexibility and scalability of the system, and is particularly suitable for data acquisition environments with limited device deployment, power supply constraints, or the need for rapid setup, such as complex application scenarios like field monitoring, smart campuses, and environmental education experiments.

[0003] The existing technology has the following deficiencies: In the existing technology, LoRa communication generally uses spread spectrum modulation technology (Chirp Spread Spectrum, CSS) for data transmission, and realizes low-power, long-distance wireless communication capabilities by configuring different spreading factors (Spreading Factor, SF) and frequency channels. However, in an actual multi-node deployment environment, when multiple LoRa nodes send data using the same spreading factor and frequency channel within a similar time window, a hidden terminal collision problem is likely to occur. This problem refers to the fact that since nodes are mutually imperceptible (i.e., outside each other's wireless coverage range), the sending behavior cannot be coordinated, resulting in multiple data frames arriving at the receiving end simultaneously, causing signal overlap and inter-symbol interference collision (Chirp Collision), leading to data demodulation failure.

[0004] More seriously, the commonly used channel conflict detection mechanisms in the prior art, such as the ALOHA protocol and the Listen-Before-Talk (LBT) strategy, are difficult to effectively identify and avoid the above-mentioned hidden conflicts. Among them, the ALOHA protocol belongs to a non-listening transmission mechanism and does not have the ability to sense the channel state before transmission; while most LoRa terminals (especially Class A types) do not implement an efficient LBT function, and nodes usually blindly send data after a set interval, lacking the ability of dynamic conflict perception and response. Therefore, when a hidden conflict occurs, the sending node cannot identify the failure of data transmission, and the receiving end cannot feedback errors in time, resulting in a large amount of key data being quietly lost, seriously affecting the stability, reliability, and data integrity of the LoRa communication network in the edge data collection scenario.

[0005] The above information disclosed in the background art section is only used to strengthen the understanding of the background of the present disclosure, so it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0006] The object of the present invention is to provide an electronic building block type edge embedded data collection method based on LoRa communication technology, which can identify the inter-code conflict through communication parameter perception, feature engineering, and machine learning, and dynamically adjust the spreading factor based on the conflict intensity and node priority to achieve "time-frequency decoupling" of the signal, effectively improve the data reliability and channel utilization efficiency of the LoRa network in the dense deployment scenario, have the ability of adaptive regulation, and reduce the risk of data loss, so as to solve the problems in the above background art.

[0007] In order to achieve the above object, the present invention provides the following technical solutions: An electronic building block type edge embedded data collection method based on LoRa communication technology, including the following steps:

[0008] During the LoRa communication process, through the data monitoring module deployed in the LoRa receiver and the network server, the multi-dimensional communication parameters generated by the node during the data transmission process are collected and recorded in real time;

[0009] Preprocess the collected original data information, and construct a multi-dimensional structured data set with "node identifier + timestamp" as the main index;

[0010] Extract the key features related to the inter-code conflict from the data set by means of feature engineering, and comprehensively analyze the extracted key features to quantify the authenticity of the conflict;

[0011] Input the key features that have been analyzed and quantified into a machine learning model pre-trained based on historical data, and use the model to intelligently evaluate the communication process to judge whether there is an inter-code interference conflict in the current communication window;

[0012] When inter - symbol interference conflicts are detected among nodes, the spreading factors of each node are dynamically adjusted based on the conflict intensity and node priority to increase the "time - frequency" distance between symbols, so as to decouple the originally highly overlapping signals and avoid conflicts.

[0013] Preferably, during the LoRa communication process, the node communication parameters are collected and recorded in real - time through data monitoring modules deployed in the receiver and the network server, including the following steps:

[0014] When the gateway receives the data frame uploaded by the terminal node, it automatically captures its communication metadata;

[0015] The monitoring module preliminarily analyzes and classifies the collected original parameter data, removes invalid or duplicate information, and marks the communication status;

[0016] The parsed data is uploaded to the network server, and the data management module on the server side completes further structured archiving and time - series collation to ensure the traceability and timeliness of the parameter data.

[0017] Preferably, key features related to inter - symbol conflicts are extracted from the data set through feature engineering means. The extracted features include the change in the offset between the node reporting frequency and the actually received frequency at the gateway, and the time overlap degree of data frames of different nodes at the receiving end. The change in the offset between the node reporting frequency and the actually received frequency at the gateway and the time overlap degree of data frames of different nodes at the receiving end are comprehensively analyzed under the detection window to generate a frequency - boundary shift reference value and a frame - overlap window reference value respectively, and the conflict authenticity is quantified through the frequency - boundary shift reference value and the frame - overlap window reference value.

[0018] Preferably, the specific steps for comprehensively analyzing the change in the offset between the node reporting frequency and the actually received frequency at the gateway under the detection window to generate a frequency - boundary shift reference value are as follows:

[0019] The data monitoring module collects in real - time the transmission frequency reported by each node during the data - frame communication process and the actual frequency received at the gateway end, and calculates the frequency offset between the two. The calculation formula is as follows:

[0020] Δf i =|f rx，i -f tx，i |

[0021] where f rx，i is the actual frequency received at the gateway end, f tx，i is the transmission frequency reported by the node, and Δf i is the frequency offset;

[0022] To improve the perception ability of subtle frequency tails, a response function based on hyper-curve amplification is introduced to perform non-linear enhancement processing on the frequency offset amount to obtain a frequency offset enhanced response value. The calculation expression is as follows: E i = tanh(γ·Δf i ), where γ is the frequency offset response amplification coefficient, tanh(·) is the hyperbolic tangent function, and E i is the frequency offset enhanced response value;

[0023] Based on the obtained frequency offset enhanced response value E i , comprehensive processing is performed on all communication events within the entire detection window to construct a frequency boundary drift reference value, which is used to characterize the spectrum boundary drift trend caused by inter-symbol interference between nodes. The calculation expression is as follows:

[0024]

[0025] , where FBDRV is the frequency boundary drift reference value, δ is the drift trend amplification factor, N is the number of communication events, and κ is the boundary sensitive mapping factor.

[0026] Preferably, the specific steps for comprehensively analyzing the time overlap degree of different node data frames at the receiving end to generate a frame overlap window reference value are as follows:

[0027] Capture and time boundary determination are performed on the data frames sent from multiple LoRa terminal nodes to the receiving end. A time interval pair D is constructed based on the start receiving time and end receiving time of each data frame at the receiving end a = [s a , e a , where D a is the receiving time interval segment of the a-th data frame at the receiving end, s a is the start time point when the a-th data frame is detected at the receiving end, and e a is the end time point when the a-th data frame finishes receiving at the receiving end. An overlap determination function is introduced to determine whether any two data frames overlap in the time dimension. The determination logic is as follows:

[0028]

[0029] , where O(a, b) is the overlap determination function used to determine whether any two data frames overlap in the time dimension. D b is the receiving time interval segment of the b-th data frame at the receiving end, f a and f b are the frequency channels used by the a-th data frame and the b-th data frame respectively, and sf a and sf bThey are the spreading factors adopted by the a-th data frame and the b-th data frame respectively, and their common value range is from SF7 to SF12. It indicates that the time periods of the a-th data frame and the b-th data frame overlap at the receiving end;

[0030] After the frame overlap relationship matrix is constructed, a non-linear amplification and overlap ratio evaluation function is introduced to quantitatively model the conflict intensity and construct a reference value for the frame overlap window. The formula is as follows:

[0031]

[0032] , where FOWRV is the reference value of the frame overlap window, H is the total number of frames, λ is the conflict index amplification factor used to adjust the response amplitude of the conflict accumulation value to the final index value, tanh is the hyperbolic tangent function, and Φ ab is the overlap intensity factor, which represents the actual overlap time length of the a-th data frame and the b-th data frame on the receiving time axis.

[0033] Preferably, the frequency boundary shift reference value and the frame overlap window reference value after analysis and quantization are input into a machine learning model pre-trained based on historical data. The model generates the inter-symbol interference conflict risk coefficient, and the communication process is intelligently evaluated based on the inter-symbol interference conflict risk coefficient to determine whether there is an inter-symbol interference conflict in the current communication window.

[0034] Preferably, the inter-symbol interference conflict risk coefficient generated when the communication process is intelligently evaluated by a machine learning model pre-trained based on historical data is compared and analyzed with a pre-set reference threshold of the inter-symbol interference conflict risk coefficient to determine whether there is an inter-symbol interference conflict in the current communication window. The judgment logic is as follows:

[0035] If the inter-symbol interference conflict risk coefficient is greater than the pre-set reference threshold of the inter-symbol interference conflict risk coefficient, it is determined that there is an inter-symbol interference conflict in the current communication window; if the inter-symbol interference conflict risk coefficient is less than or equal to the pre-set reference threshold of the inter-symbol interference conflict risk coefficient, it is determined that there is no inter-symbol interference conflict in the current communication window.

[0036] Preferably, when it is recognized that there is an inter-symbol interference conflict between nodes, the spreading factors of each node are dynamically adjusted based on the conflict intensity and node priority to increase the "time-frequency" distance between codes, and the specific steps for decoupling and conflict avoidance of the originally highly overlapping signals are as follows:

[0037] The generated inter-symbol interference conflict risk coefficient CCRC is compared and analyzed with the pre-set reference threshold of the inter-symbol interference conflict risk coefficient to determine whether the current node is in a high-risk conflict state. The judgment logic is as follows:

[0038]

[0039] , where CCRC q is the inter-symbol interference conflict risk coefficient of the q-th node, and Θ is the reference threshold of the inter-symbol interference conflict risk coefficient, is the set of target nodes identified as needing spreading factor adjustment;

[0040] For each conflict node in the set of target nodes that need spreading factor adjustment, combining its conflict intensity and node priority weight, determine the spreading factor adjustment increment, and its calculation formula is as follows:

[0041]

[0042] , where ΔSF q is the spreading factor adjustment increment, indicating the increment of the spreading factor that node q needs to increase, c is the conflict-priority adjustment coefficient, P q is the priority weight of the q-th node, P max is the maximum value of the priority, represents the ceiling operation;

[0043] After obtaining the spreading adjustment amount of each node, dynamically correct the original spreading factor and ensure that the updated value is within the range allowed by the system. The correction formula is as follows:

[0044]

[0045] , where is the new spreading factor finally allocated to the q-th node, is the spreading factor currently used by the q-th node, SF max is the maximum allowed spreading factor value, SF min is the minimum allowed spreading factor value, min{SF max , (·)} is to prevent the spreading factor from exceeding the maximum value, is to prevent the spreading factor from being lower than the minimum value.

[0046] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0047] The present invention effectively realizes the intelligent identification and dynamic avoidance of inter - symbol interference conflicts in the LoRa multi - node communication process, significantly improving the data transmission reliability and channel utilization efficiency of the system in a densely deployed and asynchronous communication environment. This method is based on the real - time perception of communication parameters. By constructing a structured data set with the core index of "node identifier + timestamp", combined with feature engineering and machine learning models, it realizes the accurate assessment of conflict risks, and dynamically adjusts the spreading factor according to the conflict intensity and node service priority, achieving the "time - frequency decoupling" of conflict signals at the physical layer. Compared with the passive communication method of blindly sending and unperceivable conflicts in the traditional ALOHA protocol, this method has a high degree of self - adaptability and self - regulation ability. It can actively perceive the communication state and optimize parameter configuration, effectively reducing the risk of data loss caused by inter - symbol conflicts, ensuring the stable communication of key edge nodes, and providing an innovative technical path for constructing a highly robust data acquisition system in the low - power wide - area Internet of Things environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0049] Figure 1 It is a flowchart of the method for the electronic building block - type edge - embedded data acquisition method based on LoRa communication technology of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0050] Now, the example embodiments will be described more fully with reference to the accompanying drawings. However, the example embodiments can be implemented in various forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that this disclosure will be more complete and comprehensive, and will fully convey the concept of the example embodiments to those skilled in the art.

[0051] The present invention provides an Figure 1 electronic building block - type edge - embedded data acquisition method based on LoRa communication technology, including the following steps:

[0052] During the LoRa communication process, through the data monitoring module deployed in the LoRa receiver (gateway) and the network server, the multi - dimensional communication parameters generated by the node during data transmission are collected and recorded in real - time;

[0053] During the LoRa communication process, through the data monitoring modules integrated in the LoRa receiver (gateway) and the network server, multi-dimensional communication parameters generated during data transmission and reception by each terminal node can be continuously and real-time collected and recorded. These communication parameters include, but are not limited to, the receive timestamp of the data frame, frequency channel, spreading factor, bandwidth, received signal strength indicator (RSSI), signal-to-noise ratio (SNR), transmit power, CRC check status, frame reception result (success or failure), etc. The core role of this process is to provide basic data support for subsequent comprehensive analysis of the network operating status, evaluation of the communication quality between nodes, and identification of potential conflict or interference problems. Especially in the scenario of dense multi-node deployment, only through real-time and fine-grained parameter monitoring can the signal overlap behavior or inter-symbol interference risk that may occur between nodes be accurately grasped. In addition, this data acquisition mechanism also provides a basis for dynamically adjusting communication parameters (such as spreading factor, time slot arrangement, frequency hopping strategy, etc.), and constructs the prerequisite for intelligent optimization and conflict self-regulation of the LoRa network.

[0054] During the LoRa communication process, the real-time collection and recording of node communication parameters are carried out through the data monitoring modules deployed in the receiver (gateway) and the network server, including the following steps: First, when the gateway receives the data frame uploaded by the terminal node, it automatically captures its communication metadata, including frequency, spreading factor (SF), bandwidth, RSSI, SNR, timestamp, and CRC check result, etc.; Second, the monitoring module conducts preliminary parsing and classification on the collected raw parameter data, removes invalid or duplicate information, and marks the communication status (such as successful reception, conflict failure, etc.); Third, the parsed data is uploaded to the network server, and the data management module on the server side completes further structured archiving and time series sorting to ensure the traceability and timeliness of the parameter data. This process realizes the full-process perception of the LoRa network operating status and data closed-loop monitoring.

[0055] Preprocess the collected raw data information, and construct a multi-dimensional structured data set with "node identifier + timestamp" as the main index;

[0056] Preprocess the collected original data information, and construct a multi-dimensional structured data set with the "node identifier + timestamp" as the main index. This refers to, on the basis of cleaning, standardizing, aligning, and formatting the raw communication data collected in real time in the LoRa network, using the unique identifier of each terminal node (such as Node ID) and its corresponding data sending or receiving time (timestamp) as a joint index to uniformly organize communication parameters in different dimensions (such as frequency, spreading factor, RSSI, SNR, frame status, etc.) into structured record units. The role of this process is to standardize unstructured, sequentially disordered, and discretely distributed communication data into a data set that is convenient for analysis and modeling, so as to achieve dynamic tracking of node behavior on the time axis and improve the accuracy and efficiency of subsequent conflict identification, feature extraction, and model training. By constructing this structured data set, it can provide a unified data view for intelligent analysis algorithms and is an indispensable basic link in the entire LoRa network communication status evaluation and interference management process.

[0057] Extract key features related to inter-symbol interference from the data set through feature engineering means, and comprehensively analyze the extracted key features to quantify the authenticity of the interference.

[0058] Extract key features related to inter-symbol interference from the data set through feature engineering means. The extracted features include the change in the offset between the node-reported frequency and the frequency actually received by the gateway, and the time overlap degree of data frames of different nodes at the receiving end. The change in the offset between the node-reported frequency and the frequency actually received by the gateway and the time overlap degree of data frames of different nodes at the receiving end are comprehensively analyzed under the detection window to generate a frequency boundary drift reference value and a frame overlap window reference value respectively, and the authenticity of the interference is quantified through the frequency boundary drift reference value and the frame overlap window reference value.

[0059] When there is a significant deviation between the frequency reported by a node and the frequency actually received by the gateway, and the deviation amplitude shows a non-random increasing trend within a short period of time, it can usually be regarded as one of the important characteristic features of inter-symbol interference conflicts among nodes. This is because in LoRa communication, the frequency modulation of the node transmission is highly accurate, and under normal circumstances, the frequency deviation should be kept within the hardware tolerance range (such as within ±1kHz). However, when multiple nodes transmit simultaneously in the same frequency channel and similar time windows and use the same spreading factor, their modulation signals will physically superimpose in the air. Especially when there is a lack of orthogonality protection between signals with the same SF, the receiving end may be affected by aliasing interference, manifested as blurred or shifted spectral boundaries. At this time, the gateway will have a deviation in identifying the center frequency during the frequency-domain demodulation process, resulting in an obvious difference between the recorded received frequency and the actual transmission frequency of the node. The amplification of this deviation reflects the existence of complex multi-node interference superposition in the channel. It can be seen that the dynamic anomaly of the frequency deviation is not only a direct manifestation of physical interference but also provides a sensitive and indirect quantifiable indicator for the identification of inter-symbol conflicts.

[0060] The specific steps for comprehensively analyzing the change in the offset between the frequency reported by the node and the frequency actually received by the gateway under the detection window to generate a frequency boundary drift reference value are as follows:

[0061] The data monitoring module collects in real time the transmission frequency reported by each node during the data frame communication process and the actual frequency received by the gateway, and calculates the frequency offset between the two. The calculation expression is as follows:

[0062] Δf i =|f rx，i -f tx，i |

[0063] where, f rx，i is the actual frequency received by the gateway, which refers to the actual received frequency value detected and recorded by the LoRa receiver (gateway) when receiving the node data frame for the i-th time, and f tx，i is the transmission frequency reported by the node, representing the transmission frequency value declared by the LoRa node in its data frame during the i-th communication, and Δf i is the frequency offset, representing the absolute difference between the actual transmission frequency of the node and the frequency actually received by the receiver during the i-th communication event;

[0064] To enhance the perception ability of subtle frequency tails, a response function based on hyperbola amplification is introduced to perform non-linear enhancement processing on the frequency offset to obtain a frequency offset enhanced response value. The calculation expression is as follows: E i =tanh(γ·Δf i), where γ is the frequency deviation response magnification factor, which is used to adjust the amplification strength of the frequency deviation value before entering the tanh mapping function, and determines the sensitivity of the system to small frequency deviation changes. The value range is 0.1-10. The larger the value, the more obvious the amplification of low-amplitude frequency deviation. tanh(·) is the hyperbolic tangent function, with a domain of real numbers and a range of (-1,1). E i is the frequency offset enhanced response value, which indicates the frequency offset response intensity obtained after nonlinear mapping in the i-th node communication event. It is used as the core intermediate feature in the subsequent frequency boundary drag index construction process to measure the frequency stability risk of each communication.

[0065] The core purpose of introducing the hyperbolic tangent function tanh(·) is to perform nonlinear enhancement and controlled compression processing on the node frequency offset to improve the sensitive recognition ability of subtle spectrum disturbances, while suppressing the interference of extreme offset values on the overall judgment. Since the frequency drift in LoRa communication often manifests as small changes in the early stage, directly using linear function processing will result in unclear response to these changes, which is easy to cause misjudgment and omission. By inputting the amplified frequency offset as the independent variable into tanh(γ·Δf i ), it can be achieved that: on the one hand, when the offset value is small, the function grows approximately linearly, which plays an effective amplification role; on the other hand, when the offset value increases to a certain extent, the tanh function tends to the saturation interval, so that the output value remains within a limited range, avoiding the imbalance of the model response caused by individual high offset values. This feature enables this method to maintain high sensitivity to initial spectrum disturbances and maintain the judgment stability and robustness of the overall system when facing multi-node inter-code interference conflicts.

[0066] This processing method can significantly amplify the interference risk caused by frequency offset in the feature dimension, thereby forming a more distinctive spectrum disturbance feature.

[0067] Based on the acquired frequency shift enhanced response value E i ,All communication events are processed comprehensively in the entire detection window to construct the frequency boundary drag reference value to characterize the spectrum boundary drift trend caused by the inter-code interference between nodes. The calculation expression is as follows:

[0068]

[0069] , where FBDRV is the frequency boundary drag reference value, δ is the drag trend amplification factor, which is used to control the linear amplification ratio of the entire drag index to the frequency offset response value, with a value range of 1-10, N is the number of communication events, and k is the boundary sensitive mapping factor, which regulates the pull-up amplitude of the high frequency offset response in the nonlinear mapping. It is a key parameter for controlling whether the local drift is rapidly amplified, with a value range of 0.1-0.5.

[0070] Finally, the larger the frequency boundary drift reference value is, the stronger the frequency offset response and the higher the cumulative volatility in the current window are, indicating that the frequency resource conflict or inter-symbol interference between nodes is more obvious.

[0071] The offset change between the node-reported frequency and the actually received frequency by the gateway is comprehensively analyzed under the detection window to generate the frequency boundary drift reference value. When multiple nodes send data frames almost synchronously under the same frequency channel and spreading factor conditions, their signals will be superimposed in the air, resulting in spectrum boundary jitter or recognition error during frequency-domain demodulation at the receiving end of the gateway, thus causing a significant increase in the offset between the node-reported frequency and the actually received frequency. By continuously analyzing the change trend of this offset, if the frequency boundary drift reference value continues to increase, it indicates that the receiving end is being interfered by multi-node overlapping signals, suggesting the risk of inter-symbol conflict; while when the reference value is stable and the offset amplitude is maintained within the hardware error tolerance range, it indicates that the communication environment is relatively clean, the node signals do not overlap significantly, and it can be judged that there is no obvious inter-symbol interference conflict.

[0072] When the data frames sent by different LoRa nodes show severe overlap in the time dimension at the receiving end, it is very likely to indicate the existence of inter-symbol interference conflict (Chirp Collision) between nodes. This is because LoRa communication uses spread-spectrum modulation technology, and the demodulation of data from different nodes depends on their discrimination in time and frequency spectrum. If multiple nodes use the same frequency channel and spreading factor (SF), and their signals arrive at the receiver almost simultaneously, the degree of overlap of these data frames in the "time-frequency" space will increase sharply, and the receiving end will not be able to correctly separate the chirp structures of these signals, resulting in demodulation failure. Such overlap phenomena are often not recognized by ALOHA-like protocols and do not automatically trigger retransmission or conflict feedback, causing "silent" data loss. Therefore, the stronger the overlap of data frames in time at the receiving end, the more likely it is to occur phase and frequency modulation interference between signals, which becomes a key criterion for inter-symbol interference conflict and can be used as a significant judgment feature in the conflict detection algorithm.

[0073] The specific steps for comprehensively analyzing the time overlap degree of data frames from different nodes at the receiving end under the detection window to generate the frame overlap window reference value are as follows:

[0074] Capture and determine the time boundary of the data frames sent from multiple LoRa terminal nodes to the receiving end, and construct the time interval pair Dx = [s a , e a based on the start receiving time and end receiving time of each data frame at the receiving end, where D a is the receiving time interval segment of the a-th data frame at the receiving end, and s ais the start time point when the a-th data frame is detected at the receiving end, e a is the end time point when the a-th data frame finishes receiving at the receiving end. An overlapping determination function is introduced to determine whether there is an overlap between any two data frames in the time dimension. The determination logic is as follows:

[0075]

[0076] , where O(a, b) is the overlapping determination function used to determine whether there is an overlap between any two data frames in the time dimension. Its output result has only two cases: O(a, b) = 1 indicates that the a-th data frame and the b-th data frame overlap in time at the receiving end, and they use the same frequency channel and the same spreading factor (SF), so there is a possibility of interference conflict; O(a, b) = 0 indicates that the two frames do not overlap in time or use different frequencies / SFs, and there is no effective conflict relationship, D b is the receiving time interval of the b-th data frame at the receiving end, f a and f b are the frequency channels used by the a-th data frame and the b-th data frame respectively, sf a and sf b are the spreading factors adopted by the a-th data frame and the b-th data frame respectively, and the common value range is from SF7 to SF12, indicates that the a-th data frame and the b-th data frame overlap in the time period at the receiving end;

[0077] After completing the construction of the frame - to - frame overlap relationship matrix, a non - linear amplification and overlap ratio evaluation function is introduced to quantitatively model the conflict intensity and construct a reference value for the frame overlap window. The formula is as follows:

[0078]

[0079] , where FOWRV is the reference value of the frame overlap window, H is the total number of frames, λ is the conflict index amplification factor used to adjust the response amplitude of the conflict accumulation value to the final index value, tanh is the hyperbolic tangent function used to normalize the accumulation result to a limited range to avoid system output distortion caused by exponential growth, Φ ab is the overlap intensity factor, indicating the actual overlap time length of the a-th data frame and the b-th data frame on the receiving time axis.

[0080] The hyperbolic tangent function, tanh, is introduced. Its core role is to perform non-linear boundary restriction and numerical smoothing on the accumulated value of the conflict intensity. Specifically, when multiple data frames overlap in time and the channel parameters are the same, the conflict accumulation value may rise rapidly. If a linear mapping is directly used, it may lead to an infinite increase in the output value, affecting the system stability and model interpretability. The tanh function has a natural S-shaped curve characteristic, which can map the input value to a finite interval to avoid the amplification of extreme values. In this scenario, the input of the tanh function is always a non-negative real number (i.e., the conflict intensity ≥ 0), so its output range is [0, 1), that is, it always approaches 1 but is not equal to 1. In this way, when the system faces high-density frame overlaps, the FOWRV value will quickly approach 1, indicating a serious conflict; while in the case of low overlaps, the change of FOWRV is relatively smooth, which helps the system to gradually identify the conflict evolution process, thus improving the sensitivity and stability of the control strategy.

[0081] The FOWRV obtained through the above calculation can be used as a metric for the frame-level conflict density within the current monitoring window. The higher the FOWRV, the stronger the temporal overlap of the data frames sent by different nodes under the same frequency and spreading factor conditions, indicating a higher probability of inter-symbol interference conflict, and vice versa indicating a lower conflict risk.

[0082] The larger the frame overlap window reference value generated by comprehensively analyzing the temporal overlap degree of different node data frames at the receiving end under the detection window, the more data frames overlap partially or completely in the same time period. Especially when the frequency channels and spreading factors are the same, this overlap will significantly increase the risk of demodulation interference between signals, and is extremely likely to trigger inter-symbol interference conflicts (Chirp Collision). Since the LoRa receiving mechanism does not have a strong decoupling ability for concurrent frames with the same frequency and SF, an increase in the frame overlap window reference value can be directly regarded as a sign of an increased conflict risk. On the contrary, when this reference value is small, it means that the frames of each node are relatively staggered in time, the signal overlap degree is low, and the receiver can keep the signals separated during the demodulation process, thus determining that the system is in a conflict-free state.

[0083] Input the key features that have been analyzed and quantified into a machine learning model pre-trained based on historical data, and use the model to intelligently evaluate the communication process to determine whether there is an inter-symbol interference conflict within the current communication window;

[0084] Input the frequency boundary shift reference value and the frame overlap window reference value that have been analyzed and quantified into a machine learning model pre-trained based on historical data, generate an inter-symbol interference conflict risk coefficient through the model, and based on the inter-symbol interference conflict risk coefficient, intelligently evaluate the communication process to determine whether there is an inter-symbol interference conflict within the current communication window.

[0085] The "machine learning model pre-trained based on historical data" refers to an intelligent algorithm model with discriminative ability constructed by collecting and organizing a large number of sample data of actual communication behaviors that occur during LoRa communication (including typical cases of normal communication and inter-symbol interference conflicts), and annotating, feature extracting, and model training these sample data. The construction process of this model usually includes: First, extract multi-dimensional key parameter features from the original communication logs, such as frequency boundary drift reference value, frame overlap window reference value, RSSI change amplitude, SNR mutation times, frame demodulation failure rate, spreading factor conflict density, etc.; Then, label the data as "conflict" or "normal" according to whether there is actually an inter-symbol conflict phenomenon in each sample; Finally, use appropriate machine learning algorithms (such as random forest, gradient boosting tree, support vector machine, LSTM time series model, or lightweight deep neural network) to learn and train these labeled data. The training goal is to enable the model to automatically learn the implicit relationship between different feature combination patterns and inter-symbol interference conflicts, and establish the mapping ability between "input features → conflict risk".

[0086] After being deployed on the LoRa receiver or edge network server, this model can receive the actual parameter inputs from the current communication process in real time, such as key indicators like the analyzed frequency boundary drift reference value and frame overlap window reference value, and quickly calculate and output a quantified "inter-symbol interference conflict risk coefficient". The risk coefficient is the confidence estimate of whether there is overlapping interference in the current node group's communication behavior, usually a floating-point value between 0 and 1, and the higher the value, the stronger the conflict risk. Based on this output result, the system can intelligently evaluate the security of the current communication window. For example, when the risk coefficient exceeds the set threshold, intervention measures such as spreading factor adjustment, time staggering scheduling, or channel switching are triggered. Compared with traditional methods based on rule matching or threshold judgment, the solution based on the machine learning model has stronger generalization ability and dynamic adaptability, can mine complex non-linear relationships, is applicable to LoRa network scenarios with multiple nodes, multiple environments, and multiple conflict types, and greatly improves the self-awareness, self-discrimination, and self-regulation abilities during the communication process.

[0087] The machine learning model is not limited here. Any machine learning model that can comprehensively analyze the frequency boundary drift reference value FBDRV and the frame overlap window reference value FOWRV to generate the inter-symbol interference conflict risk coefficient CCRC can be used. To implement the technical solution of the present invention, the present invention provides a specific implementation method:

[0088] The formula for generating the inter-code interference conflict risk coefficient CCRC is as follows: CCRC=f1·FBDRV+f2·FOWRV, where f1 and f2 are the preset proportional coefficients of the frequency boundary drag reference value FBDRV and the frame overlap window reference value FOWRV, respectively, and both f1 and f2 are greater than 0.

[0089] "Preset proportionality coefficient" refers to: in the weighted calculation process of generating the inter-code interference conflict risk coefficient CCRC, in order to balance the influence of the frequency boundary drag reference value FBDRV and the frame overlap window reference value FOWRV on the final risk assessment result, two preset constant weights, namely f1 and f2. These two coefficients reflect the sensitivity tendency of the model between different features.

[0090] Specifically, f1 is the weight assigned to FBDRV, which indicates the contribution ratio of this indicator to the conflict risk; f2 is the corresponding weight of FOWRV. Since the conflict characteristics represented by the frequency boundary drag reference value FBDRV and the frame overlap window reference value FOWRV are different in nature, one is biased towards the fluctuation trend of the spectrum layer, and the other is biased towards the frame overlap in the time dimension, the settings of the two coefficients can be adjusted according to historical statistics, model training or expert experience. The role of the preset proportional coefficient is to give the feature a reasonable initial influence in the absence of a real-time training update mechanism or in the model initialization stage, thereby ensuring the physical meaning and stability of the CCRC output results. The figure clearly states that both f1 and f2 are greater than 0, which means that both features have positive contributions in risk assessment and will not be offset or ignored.

[0091] It can be seen from the inter-code interference conflict risk coefficient that the larger the frequency boundary drag reference value generated after comprehensive analysis of the offset change between the node reported frequency and the frequency actually received by the gateway under the detection window, and the larger the frame overlap window reference value generated after comprehensive analysis of the time overlap degree of data frames of different nodes at the receiving end under the detection window, the larger the inter-code interference conflict risk coefficient generated when the communication process is intelligently evaluated by the machine learning model pre-trained based on historical data, indicating that the probability of inter-code interference conflicts between nodes is greater, and vice versa, the probability of inter-code interference conflicts between nodes is smaller.

[0092] The ISI conflict risk coefficient generated by the intelligent evaluation of the communication process through the machine learning model pre-trained based on historical data is compared and analyzed with the pre-set ISI conflict risk coefficient reference threshold to determine whether there is an ISI conflict in the current communication window. The judgment logic is as follows:

[0093] If the inter-symbol interference conflict risk coefficient is greater than the pre-set reference threshold of the inter-symbol interference conflict risk coefficient, it is determined that there is an inter-symbol interference conflict within the current communication window; if the inter-symbol interference conflict risk coefficient is less than or equal to the pre-set reference threshold of the inter-symbol interference conflict risk coefficient, it is determined that there is no inter-symbol interference conflict within the current communication window.

[0094] When it is recognized that there is an inter-symbol interference conflict among nodes, the spreading factors of each node are dynamically adjusted based on the conflict intensity and node priority to widen the "time-frequency" distance between symbols, so as to decouple and avoid conflicts for the originally highly overlapping signals.

[0095] When it is recognized that there is an inter-symbol interference conflict among nodes, the spreading factors (Spreading Factor, SF) of each node are dynamically adjusted based on the conflict intensity and node priority to widen the "time-frequency" distance between symbols. Its core function is to achieve the differentiation of signals in the time dimension and modulation form, so as to reduce or avoid the signal overlapping interference caused when multiple nodes send data simultaneously on the same frequency channel and under the same SF. LoRa communication uses spread-spectrum modulation. Different SFs will result in different durations of each data frame, data transmission rates, and modulation chirp characteristics. Therefore, even if multiple signals overlap in time, as long as the SFs are different, the receiving end (gateway) also has a certain concurrent demodulation ability. By quantitatively evaluating the conflict intensity, the system can identify which nodes have a high risk of interference; at the same time, combined with the node priority (such as critical data nodes, nodes with sufficient power, or nodes with low tolerance for delay), a lower SF is dynamically allocated to improve its transmission rate and resource acquisition priority, while a high SF is allocated to nodes with low priority or high conflict interference degree to extend their transmission time and stagger their signal arrival windows to achieve off-peak demodulation. The fundamental goal of this mechanism is to improve the channel utilization rate and communication stability of the entire LoRa network in high-density deployment scenarios through "code distance widening" and "transmission timing separation", significantly reduce the packet loss rate caused by conflicts, ensure the reliable demodulability of data when multiple nodes communicate simultaneously, and is an important technical means to realize the adaptive and intelligent operation of the LoRa network.

[0096] When it is recognized that there is an inter-symbol interference conflict among nodes, the spreading factors of each node are dynamically adjusted based on the conflict intensity and node priority to widen the "time-frequency" distance between symbols, and the specific steps to decouple and avoid conflicts for the originally highly overlapping signals are as follows:

[0097] Compare and analyze the generated inter-symbol interference conflict risk coefficient CCRC with the pre-set reference threshold of the inter-symbol interference conflict risk coefficient to determine whether the current node is in a high-risk conflict state. The judgment logic is as follows:

[0098]

[0099] , where CCRC q is the inter-symbol interference conflict risk coefficient of the q-th node, and Θ is the reference threshold of the inter-symbol interference conflict risk coefficient. is the set of target nodes identified as requiring spreading factor adjustment;

[0100] This step is used to screen out intervention objects with significant conflict risks from all communication nodes, providing a target basis for subsequent parameter optimization.

[0101] For each conflict node in the set of target nodes that requires spreading factor adjustment, combining its conflict intensity and node priority weight, determine the spreading factor adjustment increment. The calculation formula is as follows:

[0102]

[0103] , where ΔSF q is the spreading factor adjustment increment, representing the increment of the spreading factor that node q needs to increase. c is the conflict-priority adjustment coefficient, used to adjust the sensitivity of the overall SF adjustment range, with a value range of 0 < κ ≤ 1, and P q is the priority weight of the q-th node, and P max is the maximum value of the priority, represents the ceiling operation;

[0104] This step can dynamically determine the spreading adjustment range according to the severity of the conflict and the importance of the node task, ensuring that high-priority nodes have higher communication priority and low-priority nodes perform conflict avoidance behaviors.

[0105] After obtaining the spreading adjustment amount for each node, dynamically correct the original spreading factor and ensure that the updated value is within the system allowable range. The correction formula is as follows:

[0106]

[0107] , where is the new spreading factor finally assigned to the q-th node, is the spreading factor currently used by the q-th node, and SF max is the maximum allowable spreading factor value, and SF min is the minimum allowable spreading factor value, min{SF max , (·)} is to prevent the spreading factor from exceeding the maximum value, is to prevent the spreading factor from being lower than the minimum value.

[0108] By dynamically updating the node spreading factor and ensuring the adjustment result is always within the legal range allowed by the system through the maximum and minimum constraint mechanism. This step not only ensures the effectiveness of conflict intervention but also prevents the increase in communication delay or abnormal energy consumption caused by excessive adjustment, thus achieving a dual balance between steady-state control and communication performance.

[0109] The present invention effectively realizes the intelligent identification and dynamic avoidance of inter-symbol interference conflicts in the LoRa multi-node communication process, significantly improving the data transmission reliability and channel utilization efficiency of the system in a densely deployed and asynchronous communication environment. This method is based on the real-time perception of communication parameters. By constructing a structured data set with "node identifier + timestamp" as the core index, combined with feature engineering and machine learning models, it realizes the accurate assessment of conflict risks, and dynamically adjusts the spreading factor according to the conflict intensity and node service priority, achieving the "time-frequency decoupling" of conflict signals at the physical layer. Compared with the passive communication method of blindly sending and unperceivable conflicts in the traditional ALOHA protocol, this method has a high degree of self-adaptability and self-regulation ability, can actively sense the communication state and optimize parameter configuration, effectively reducing the risk of data loss caused by inter-symbol conflicts, ensuring the stable communication of key edge nodes, and providing an innovative technical path for building a highly robust data acquisition system in the low-power wide-area Internet of Things environment.

[0110] The above formulas are all dimensionless and take their numerical calculations. The formula is obtained by collecting a large amount of data for software simulation to get a formula closest to the real situation. The preset parameters in the formula are set by those skilled in the art according to the actual situation.

[0111] Only some exemplary embodiments of the present invention have been described by way of illustration. Undoubtedly, for those of ordinary skill in the art, without departing from the spirit and scope of the present invention, the described embodiments can be modified in various different ways. Therefore, the above drawings and description are illustrative in nature and should not be construed as limiting the protection scope of the claims of the present invention.

[0112] It should be noted that in this article, if there are relational terms such as first and second, they are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of another identical element in the process, method, article or device including the element.

[0113] It should be understood that in various embodiments of the present application, the magnitudes of the serial numbers of the above processes do not imply the order of execution, and the order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0114] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Skilled professionals can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0115] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be repeated herein.

[0116] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0117] In addition, the functional units in various embodiments of the present application can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.

[0118] As described above, only the specific implementation manners of the present application are provided, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present application, and all of them should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claimed rights.

[0119] Only some exemplary embodiments of the present invention have been described above by way of illustration. Undoubtedly, for those of ordinary skill in the art, without departing from the spirit and scope of the present invention, the described embodiments can be modified in various different ways. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the protection scope of the claims of the present invention.

Claims

1. An edge-embedded data acquisition method based on LoRa communication technology, characterized in that It includes the following steps: During the LoRa communication process, the multi-dimensional communication parameters generated by nodes during data transmission are collected and recorded in real time through data monitoring modules deployed in LoRa receivers and network servers; The collected raw data information is preprocessed, and a multi-dimensional structured data set is constructed with "node identifier + timestamp" as the main index; Key features related to inter-symbol conflict are extracted from the data set through feature engineering means, and the extracted key features are comprehensively analyzed to quantify the authenticity of the conflict; The key features that have been analyzed and quantified are input into a machine learning model pre-trained based on historical data, and the communication process is intelligently evaluated through the model to determine whether there is inter-symbol interference conflict within the current communication window; When it is identified that there is inter-symbol interference conflict among nodes, the spreading factors of each node are dynamically adjusted based on the conflict intensity and node priority to widen the "time-frequency" distance between symbols, so as to decouple the originally highly overlapping signals and avoid conflicts; 2. The method for edge-embedded data acquisition of electronic building blocks based on LoRa communication technology according to claim 1, wherein During the LoRa communication process, the communication parameters of nodes are collected and recorded in real time through data monitoring modules deployed in receivers and network servers, including the following steps: When the gateway receives the data frame uploaded by the terminal node, it automatically captures its communication metadata; The monitoring module makes a preliminary analysis and classification of the collected raw parameter data, removes invalid or duplicate information, and marks the communication status; The parsed data is uploaded to the network server, and the data management module on the server side completes further structured archiving and time series arrangement to ensure the traceability and timeliness of the parameter data; 3. The method for edge-embedded data acquisition of electronic building blocks based on LoRa communication technology according to claim 1, wherein Key features related to inter-symbol conflict are extracted from the data set through feature engineering means. The extracted features include the change in the offset between the node reporting frequency and the actual received frequency at the gateway, and the time overlap degree of data frames of different nodes at the receiving end. The change in the offset between the node reporting frequency and the actual received frequency at the gateway and the time overlap degree of data frames of different nodes at the receiving end are comprehensively analyzed under the detection window to generate a frequency boundary shift reference value and a frame overlap window reference value respectively, and the authenticity of the conflict is quantified through the frequency boundary shift reference value and the frame overlap window reference value; 4. The method for edge-embedded data acquisition of electronic building blocks based on LoRa communication technology according to claim 3, wherein The specific steps for comprehensively analyzing the change in the offset between the node reporting frequency and the actual received frequency at the gateway under the detection window to generate a frequency boundary shift reference value are as follows: The transmission frequency reported by each node during data frame communication and the actual frequency received at the gateway end are collected in real time through the data monitoring module, and the frequency offset between the two is calculated. The calculation formula is as follows: Δf i = |f rx，i - f tx，i | where f rx，i is the actual frequency received by the gateway side, f tx，u is the transmission frequency reported by the node, and Δf i is the frequency offset; To improve the perception ability of subtle frequency tails, a response function based on hyper-curve amplification is introduced to perform non-linear enhancement processing on the frequency offset amount, and a frequency offset enhanced response value is obtained. The calculation expression is as follows: E i = tanh(γ·Δf i ), where γ is the frequency offset response amplification coefficient, tanh(·) is the hyperbolic tangent function, and E i is the frequency offset enhanced response value; Based on the obtained enhanced response value E of the frequency offset i , all communication events are comprehensively processed within the entire detection window to construct a frequency boundary drift reference value, which is used to characterize the spectrum boundary drift trend caused by inter-symbol interference between nodes. The calculation expression is as follows: In the formula, FBDRV is the frequency boundary shift reference value, δ is the shift trend amplification factor, N is the number of communication events, and κ is the boundary sensitive mapping factor; 5. The method for edge-embedded data acquisition of electronic building blocks based on LoRa communication technology according to claim 3, wherein The specific steps for comprehensively analyzing the time overlap degree of data frames of different nodes at the receiving end under the detection window to generate a frame overlap window reference value are as follows: Capture the data frames sent from multiple LoRa terminal nodes to the receiver and determine the time boundaries, and construct a time interval pair D based on the start reception time and end reception time of each data frame at the receiver a =[s a , e a , where D a is the reception time interval segment of the a-th data frame at the receiver, s a is the start time point when the a-th data frame is detected at the receiver, and e a is the end time point when the reception of the a-th data frame is completed at the receiver. Introduce an overlap determination function to determine whether there is an overlap between any two data frames in the time dimension. The determination logic is as follows: where O(a, b) is an overlap determination function used to determine whether there is an overlap in the time dimension between any two data frames, D b is the reception time interval segment of the b-th data frame at the receiving end, f a and f b are the frequency channels used by the a-th data frame and the b-th data frame respectively, sf a and sf a are the spreading factors adopted by the a-th data frame and the b-th data frame respectively, and the common value range is from SF7 to SF12, indicates that there is an overlap in the time periods of the a-th data frame and the b-th data frame at the receiving end; After the frame overlap relationship matrix is constructed, a non-linear amplification and overlap ratio evaluation function is introduced to quantitatively model the conflict intensity and construct a frame overlap window reference value. The formula is as follows: Wherein, FOWRV is the frame overlap window reference value, H is the total number of frames, λ is the conflict index amplification factor, which is used to adjust the response amplitude of the conflict accumulation value to the final index value, tanh is the hyperbolic tangent function, and Φ ab is the overlap intensity factor, indicating the actual overlap time length of the a-th data frame and the b-th data frame on the reception time axis.

6. The method for electronically building block type edge embedded data acquisition based on LoRa communication technology according to claim 3, wherein Input the frequency boundary shift reference value and frame overlap window reference value after analysis and quantization into a machine learning model pre-trained based on historical data. Generate an inter-symbol interference conflict risk coefficient through the model, and conduct an intelligent evaluation of the communication process based on the inter-symbol interference conflict risk coefficient to determine whether there is an inter-symbol interference conflict within the current communication window.

7. The method for edge-embedded data acquisition of electronic building blocks based on LoRa communication technology according to claim 6, characterized in that Compare and analyze the inter-symbol interference conflict risk coefficient generated during the intelligent evaluation of the communication process by a machine learning model pre-trained based on historical data with a pre-set reference threshold of the inter-symbol interference conflict risk coefficient to determine whether there is an inter-symbol interference conflict within the current communication window. The judgment logic is as follows: If the inter-symbol interference conflict risk coefficient is greater than the pre-set reference threshold of the inter-symbol interference conflict risk coefficient, it is determined that there is an inter-symbol interference conflict within the current communication window; if the inter-symbol interference conflict risk coefficient is less than or equal to the pre-set reference threshold of the inter-symbol interference conflict risk coefficient, it is determined that there is no inter-symbol interference conflict within the current communication window.

8. The method for edge-embedded data acquisition of electronic building blocks based on LoRa communication technology according to claim 7, characterized in that, When it is recognized that there is an inter-symbol interference conflict between nodes, dynamically adjust the spreading factor of each node based on the conflict intensity and node priority to widen the "time-frequency" distance between inter-symbols and achieve the decoupling of originally highly overlapping signals and conflict avoidance. The specific steps are as follows: Compare and analyze the generated inter-symbol interference conflict risk coefficient CCRC with the pre-set reference threshold of the inter-symbol interference conflict risk coefficient to determine whether the current node is in a high-risk conflict state. The judgment logic is as follows: where CCRC q is the interference risk coefficient of the q-th node, and Θ is the reference threshold of the interference risk coefficient; is the set of target nodes identified as needing to adjust the spreading factor; For each conflict node in the set of target nodes that need to adjust the spreading factor, combine its conflict intensity and node priority weight to determine the spreading factor adjustment increment. The calculation formula is as follows: where, ΔSF q is the spreading factor adjustment increment, representing the increment of the spreading factor that node q needs to increase, c is the conflict-priority adjustment coefficient, P q is the priority weight of the q-th node, P max is the maximum value of the priority, represents the ceiling operation; After obtaining the spreading adjustment amount of each node, dynamically correct the original spreading factor and ensure that the updated value is within the range allowed by the system. The correction formula is as follows: wherein, is the newly allocated spreading factor finally assigned to the q-th node, is the spreading factor currently used by the q-th node, SF max is the maximum allowed spreading factor value, SF min is the minimum allowed spreading factor value, min{SF max , (·)} is to prevent the spreading factor from exceeding the maximum value, is to prevent the spreading factor from being lower than the minimum value.

Images