Packet loss compensation optimization method and system for data transmission of communication equipment of Internet of Things

Abstract

The invention relates to the technical field of Internet of Things communication, and particularly discloses an Internet of Things communication equipment data transmission packet loss compensation optimization method and system, and the method comprises the steps: obtaining a data packet transmission sequence, and obtaining the receiving state identifiers and transmission timestamps of a plurality of data packets according to the data packet transmission sequence; and acquiring network transmission quality parameters, and acquiring a received signal strength value sequence, a network jitter value sequence and a data packet delay fluctuation value sequence in a continuous time window according to the network transmission quality parameters. According to the method, by dynamically obtaining the multi-dimensional network transmission quality parameters in the continuous time window and calculating the actual loss probability value, misjudgment caused by a fixed threshold value is avoided, reliable recovery of the data flow is achieved in combination with a prediction compensation mechanism and consistency verification, and the problem of packet loss caused by network fluctuation in the environment of the Internet of Things is effectively solved.

Description

Technical Field

[0001] This invention relates to the field of Internet of Things (IoT) communication technology, and in particular to a method and system for optimizing packet loss compensation in IoT communication devices. Background Technology

[0002] With the widespread application of IoT technology, IoT communication devices play a crucial role in real-time data transmission in key scenarios such as industrial automation, smart cities, and environmental monitoring. The reliability and real-time performance of data transmission directly affect the stability of system operation; however, the complex characteristics of wireless channels make data packet loss an increasingly prominent problem. In the industrial IoT environment, uncertainties such as router buffer overflows due to network congestion, physical layer interruptions caused by wireless signal attenuation and multipath effects, signal fluctuations caused by device mobility, electromagnetic interference, and equipment failures collectively exacerbate the risk of data packet loss. These factors not only cause data flow interruptions but may also trigger system-level failures, such as equipment malfunctions or production stoppages in industrial control scenarios.

[0003] In existing technologies, lost data packets are retransmitted to restore transmission, but their inherent round-trip delay makes them unsuitable for high-real-time applications, especially under dynamic network conditions where they significantly increase end-to-end latency. Forward error correction (FEC) can recover from errors by adding redundant information, but it inevitably consumes additional bandwidth, reducing overall transmission efficiency, which is particularly detrimental to resource-constrained IoT devices. State estimation compensation methods that have emerged in recent years, such as Kalman filter-based prediction models, attempt to infer lost content from historical data, but they have significant drawbacks: system noise characteristics are simplified to static assumptions, failing to cope with the dynamic changes in noise statistics in real-world networks; the compensation process relies solely on observation data from a single path, failing to effectively integrate redundant information from multi-path transmissions to improve robustness; and the lack of real-time monitoring and adaptive adjustment mechanisms for compensation results in continuous degradation of compensation accuracy over long periods. Furthermore, traditional packet loss detection mechanisms use fixed time thresholds to determine packet status, making it difficult to distinguish between actual loss and temporary delays in scenarios with network jitter or severe signal fluctuations, easily leading to misjudgments. For example, in industrial environments with sudden communication interference, fixed thresholds may incorrectly mark delayed data packets as lost, triggering unnecessary compensation operations and introducing data distortion. These limitations make it difficult for existing solutions to maintain stable and reliable transmission performance in complex and ever-changing IoT environments, necessitating a packet loss compensation mechanism that can dynamically sense network status, integrate multi-source information, and possess continuous optimization capabilities. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for optimizing packet loss compensation in data transmission of Internet of Things (IoT) communication devices, so as to solve the technical problems mentioned in the background.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A method for optimizing packet loss compensation in Internet of Things (IoT) communication devices includes: Obtain the data packet transmission sequence, and based on the data packet transmission sequence, obtain the reception status identifier and transmission timestamp of multiple data packets; Obtain network transmission quality parameters, and based on the network transmission quality parameters, obtain the received signal strength numerical sequence, network jitter numerical sequence, and data packet delay fluctuation numerical sequence within a continuous time window; The arrival status sequence of data packets is obtained based on the received status identifier and the numerical sequence of received signal strength, network jitter, and data packet delay fluctuation within a continuous time window. The actual loss probability value of each data packet is obtained based on the arrival status sequence of data packets and network transmission quality parameters. Determine whether the actual loss probability value exceeds a preset threshold; If the time limit is not exceeded, the data packet is marked as not lost. If the value exceeds the limit, the data packet is marked as confirmed lost, and a predicted compensation value for the data packet content is generated based on the data packet in the confirmed lost state. The predicted compensation value is inserted into the data transmission sequence, and the consistency of the data packets before and after the insertion position is checked.

[0006] Preferably, the step of obtaining the data packet arrival status sequence based on the received status identifier and the received signal strength numerical sequence, network jitter numerical sequence, and data packet delay fluctuation numerical sequence within a continuous time window includes: The received status identifier is parsed to extract the arrival status type and corresponding timestamp of each data packet, wherein the arrival status type includes successful reception, delayed reception, and loss; The received signal strength numerical sequence is divided into time windows, the moving average and variance of the received signal strength values ​​in each window are obtained, and a received signal strength statistic is generated based on the moving average and variance of the received signal strength values ​​in each window. The network jitter numerical sequence is divided into time windows, and the maximum, minimum and standard deviation of the network jitter values ​​in each window are obtained. The network jitter statistics are obtained based on the maximum, minimum and standard deviation of the network jitter values ​​in each window. The data packet delay fluctuation numerical sequence is divided into time windows to obtain the fluctuation range and rate of change of the data packet delay value within each window, and the data packet delay fluctuation statistics are obtained based on the fluctuation range and rate of change of the data packet delay value within each window. Aligned data is obtained by aligning and associating the received status identifier with the received signal strength statistics, network jitter statistics, and data packet delay fluctuation statistics according to timestamps. Based on the alignment data, an arrival status label is assigned to each data packet by comparing the relative magnitudes of the received signal strength value, network jitter value, and data packet delay fluctuation value with a preset threshold. Combine the arrival status tags of all data packets in chronological order to generate a data packet arrival status sequence.

[0007] Preferably, the step of obtaining the actual loss probability value of each data packet based on the data packet arrival status sequence and network transmission quality parameters includes: Obtain the arrival status label and historical arrival status sequence of each data packet from the data packet arrival status sequence; Based on the network transmission quality parameters, obtain the numerical sequences of received signal strength, network jitter, and data packet delay fluctuation for the corresponding time window; Obtain historical packet loss rate statistics based on historical arrival status sequences; The numerical sequences of received signal strength, network jitter, and data packet delay fluctuation are normalized and converted into numerical values ​​with uniform dimensions. The initial loss probability value of each data packet is obtained based on historical packet loss rate statistics and uniform unit values; The initial loss probability value is adjusted by time decay to generate the actual loss probability value.

[0008] Preferably, the step of generating a predicted compensation value for the packet content based on the packet whose loss status has been confirmed includes: The state estimation parameters are obtained based on the preset state vector and the historical data packet content sequence. The preset state vector includes data packet feature parameters and transmission timing parameters, and the state estimation parameters include state transition matrix coefficients, observation matrix coefficients, process noise covariance, and observation noise covariance. The predicted state value at the current moment is obtained based on the coefficients of the state transition matrix. The prediction error covariance matrix at the current moment is obtained based on the state prediction value and the process noise covariance. Generate the Kalman gain matrix based on the prediction error covariance matrix and the observation noise covariance matrix; The updated state estimate and error covariance matrix are obtained based on the Kalman gain matrix and the difference between the actual and predicted observations. Obtain a preliminary estimate of the packet content based on the updated state estimate; Obtain the confidence interval of the preliminary estimate based on the error covariance matrix; The data packet content prediction compensation value is obtained based on the confidence interval and the preliminary estimate.

[0009] Preferably, the step of inserting the predicted compensation value into the data transmission sequence includes: Based on the path identification information in the data packet transmission sequence, monitoring data of multiple transmission paths are obtained, and a set of multipath estimates for the same data packet is obtained based on the monitoring data of the multiple transmission paths. Historical estimation accuracy data is obtained based on the historical transmission records of each transmission path, and path reliability weighting coefficients are obtained based on the historical estimation accuracy data. The correlation coefficient of the estimated values ​​is obtained based on the numerical differences between the estimated values ​​in the multipath estimated value set. The fusion weight allocation scheme is obtained based on the path reliability weight coefficient and the estimated value correlation coefficient. The weighted fusion estimate is obtained based on the fusion weight allocation scheme and the multi-path estimate set, and the overall error covariance after fusion is obtained based on the error statistical characteristics of each path estimate. The spatiotemporal consistency test result is obtained based on the numerical relationship between the weighted fusion estimate and the content of adjacent data packets, and the final confirmed prediction compensation value is obtained based on the spatiotemporal consistency test result. The insertion position coordinates are obtained based on the sequence number information in the data packet transmission sequence, and the updated data transmission sequence is obtained based on the insertion position coordinates and the finally confirmed prediction compensation value.

[0010] Preferably, the step of performing a consistency check on the data packets before and after the insertion position includes: The numerical difference dataset is obtained based on the actual received data packets and the corresponding predicted compensation values. Based on the numerical difference dataset, difference statistical features are obtained, wherein the difference statistical features include mean, variance, skewness coefficient and kurtosis coefficient; The error distribution model parameters are obtained based on the statistical characteristics of the differences, and the prediction compensation accuracy index is obtained based on the numerical difference dataset and the error distribution model parameters. The systematic deviation component and random error component are obtained based on the accuracy index and error distribution model parameters, and the deviation correction parameters are obtained based on the systematic deviation component. Obtain the corrected compensation result based on the deviation correction parameters and the predicted compensation value, and obtain the compensation verification report based on the corrected compensation result and the accuracy index. If the compensation verification report meets the preset conditions, the compensation verification report is considered normal, and the verification is completed. If the verification report does not meet the preset conditions, the verification report is judged to be abnormal, the verification is not completed, and the process is returned to obtain the difference statistical features based on the numerical difference dataset.

[0011] This invention also discloses a data transmission packet loss compensation optimization system for Internet of Things (IoT) communication devices, comprising: The data packet transmission sequence processing module is used to obtain the data packet transmission sequence and obtain the reception status identifier and transmission timestamp of multiple data packets based on the data packet transmission sequence. The network transmission quality parameter processing module is used to acquire network transmission quality parameters and, based on the network transmission quality parameters, acquire a sequence of received signal strength values, a sequence of network jitter values, and a sequence of data packet delay fluctuation values ​​within a continuous time window. The actual loss probability value acquisition module is used to obtain the data packet arrival status sequence based on the received status identifier and the received signal strength value sequence, network jitter value sequence and data packet delay fluctuation value sequence within a continuous time window, and to obtain the actual loss probability value of each data packet based on the data packet arrival status sequence and network transmission quality parameters. The judgment module is used to determine whether the actual loss probability value exceeds a preset threshold; If the time limit is not exceeded, the data packet is marked as not lost. If the value exceeds the limit, the data packet is marked as confirmed lost, and a predicted compensation value for the data packet content is generated based on the data packet in the confirmed lost state. The verification module is used to insert the predicted compensation value into the data transmission sequence and to perform consistency verification on the data packets before and after the insertion position.

[0012] Preferably, the actual loss probability value acquisition module includes: The historical arrival status sequence acquisition unit is used to obtain the arrival status tag and historical arrival status sequence of each data packet based on the data packet arrival status sequence. The network transmission quality parameter analysis unit is used to obtain the received signal strength numerical sequence, network jitter numerical sequence, and data packet delay fluctuation numerical sequence for the corresponding time window based on the network transmission quality parameters. The historical packet loss rate statistics acquisition unit is used to obtain historical packet loss rate statistics based on the historical arrival status sequence. The numerical unification unit is used to normalize the numerical sequences of received signal strength, network jitter, and data packet delay fluctuation, converting them into numerical values ​​with uniform dimensions. The initial loss probability value acquisition unit is used to acquire the initial loss probability value of each data packet based on historical packet loss rate statistics and uniform unit values. The actual loss probability value generation unit is used to adjust the initial loss probability value by time decay to generate the actual loss probability value.

[0013] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of a data transmission packet loss compensation optimization method for Internet of Things communication devices.

[0014] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of a method for optimizing data transmission packet loss compensation in Internet of Things communication devices.

[0015] The beneficial effects of this application are as follows: This invention dynamically acquires multi-dimensional network transmission quality parameters within a continuous time window and calculates the actual loss probability value, avoiding misjudgments caused by fixed thresholds. It combines a predictive compensation mechanism with consistency verification to achieve reliable data stream recovery, effectively solving the packet loss problem caused by network fluctuations in the Internet of Things environment. It has the advantages of being able to dynamically sense changes in network status to accurately distinguish between real packet loss and delayed data packets and reduce misjudgments, improving packet loss detection accuracy by fusing multi-source network parameters, reducing transmission delay without retransmission by using a predictive compensation mechanism, and ensuring the reliability of compensation data through consistency verification. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of a method flow according to an embodiment of this application.

[0017] Figure 2 This is a schematic diagram of the system structure according to an embodiment of this application.

[0018] Figure 3 This is a schematic diagram of the internal structure of a computer device according to an embodiment of this application.

[0019] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0020] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0021] like Figure 1 As shown, this application provides a method for optimizing packet loss compensation in data transmission for IoT communication devices, applied to data transmission between an energy storage power supply monitoring module and a monitoring center, including: S1. Obtain the data packet transmission sequence, and obtain the reception status identifier and transmission timestamp of multiple data packets based on the data packet transmission sequence; S2. Obtain network transmission quality parameters, and obtain the received signal strength numerical sequence, network jitter numerical sequence, and data packet delay fluctuation numerical sequence within a continuous time window based on the network transmission quality parameters; S3. Obtain the data packet arrival status sequence based on the received status identifier and the received signal strength value sequence, network jitter value sequence and data packet delay fluctuation value sequence within the continuous time window, and obtain the actual loss probability value of each data packet based on the data packet arrival status sequence and network transmission quality parameters. S4. Determine whether the actual loss probability value exceeds a preset threshold; If the time limit is not exceeded, the data packet is marked as not lost. If the value exceeds the limit, the data packet is marked as confirmed lost, and a predicted compensation value for the data packet content is generated based on the data packet in the confirmed lost state. S5. Insert the predicted compensation value into the data transmission sequence, and at the same time perform consistency verification on the data packets before and after the insertion position.

[0022] As described in steps S1-S5 above, obtaining the data packet transmission sequence refers to extracting the time sequence record of data packets from the communication device. This can be achieved by network interface monitoring or log file parsing, such as capturing data streams or reading device buffer contents using packet capture tools. The received signal strength numerical sequence within a continuous time window refers to the sequence of signal strength measurements collected within a preset time range. In practical applications, this sequence can be generated by segmenting the original signal data using a sliding time window, such as using a fixed-width window for data slicing or dynamically adjusting the window size based on network events. This is mainly to reflect the time-varying characteristics of signal strength. The process of obtaining the data packet arrival status sequence based on the received status identifier and the received signal strength numerical sequence, network jitter numerical sequence, and data packet delay fluctuation numerical sequence within the continuous time window can be understood as comprehensively judging the data packet status by integrating multi-source information. This can be done using a decision tree model or rule engine for status classification, such as based on threshold comparison or pattern matching. This is mainly to reduce misjudgments in status identification. Obtaining the actual loss probability value refers to calculating the probability of actual data packet loss, such as using a logistic regression model or support vector machine. Its input includes historical status and network parameters. This is mainly to achieve dynamically adaptive packet loss detection. When the actual packet loss probability exceeds a preset threshold, the data packet is marked as confirmed lost, and a predicted compensation value for the packet content is generated. This predicted compensation value can be generated using time series prediction techniques, such as autoregressive models or neural network predictions, primarily to recover data content without increasing bandwidth overhead. Finally, the predicted compensation value is inserted into the data transmission sequence for consistency verification. This consistency verification verifies the logical consistency between the compensation data and adjacent data packets, which can be obtained using difference analysis or correlation checks, such as calculating the absolute value or relative rate of change of numerical differences. This is primarily to ensure the integrity and reliability of the data stream. This application dynamically assesses packet loss probability by fusing multi-dimensional network parameters, avoiding misjudgments caused by fixed time thresholds and overcoming the poor adaptability to time-varying noise characteristics. Furthermore, it optimizes compensation accuracy through multi-source information fusion and ensures the continuous stability of the compensation effect through a closed-loop verification mechanism.

[0023] By acquiring the data packet transmission sequence and extracting the receive status identifier and transmission timestamp, a temporal correlation basis for data packets is established, thereby solving the problem that fixed time thresholds cannot distinguish between actual loss and temporary delay. After the network transmission quality parameters are acquired, numerical sequences of received signal strength, network jitter, and data packet delay fluctuations within a continuous time window are generated. Windowed statistical processing is applied to calculate statistics such as moving average, variance, maximum value, and standard deviation, effectively capturing the time-varying characteristics of network parameters and overcoming the limitation of fixed noise characteristics in traditional methods. The receive status identifier and the three types of network parameter sequences are aligned and correlated by timestamp to form a data packet arrival status sequence. The actual loss probability value is calculated based on this sequence and network transmission quality parameters. A time decay adjustment mechanism is introduced to focus on the recent network state, ensuring that the probability assessment is synchronized with the current environment. The actual loss probability value is compared with a preset threshold. If it does not exceed the threshold, the data packet is marked as not lost; if it exceeds the threshold, it is marked as confirmed lost and the generation of a predicted compensation value is triggered. The predicted compensation value is inserted into the data transmission sequence. At the same time, numerical difference analysis and error distribution modeling are performed on the data packets before and after the insertion position. After the consistency check is completed, a complete data stream is formed.

[0024] In one embodiment, the step of obtaining the data packet arrival status sequence based on the reception status identifier and the received signal strength numerical sequence, network jitter numerical sequence, and data packet delay fluctuation numerical sequence within a continuous time window includes: S301. Parse the received status identifier and extract the arrival status type and corresponding timestamp of each data packet, wherein the arrival status type includes successful reception, delayed reception and loss; S302. Divide the received signal strength numerical sequence into time windows, obtain the moving average and variance of the received signal strength values ​​in each window, and generate a received signal strength statistic based on the moving average and variance of the received signal strength values ​​in each window. S303. Divide the network jitter value sequence into time windows, obtain the maximum value, minimum value and standard deviation of the network jitter value in each window, and obtain the network jitter statistics based on the maximum value, minimum value and standard deviation of the network jitter value in each window; S304. Divide the data packet delay fluctuation value sequence into time windows, obtain the fluctuation range and rate of change of the data packet delay value in each window, and obtain the data packet delay fluctuation statistics based on the fluctuation range and rate of change of the data packet delay value in each window. S305. Align and associate the received status identifier with the received signal strength statistics, network jitter statistics and data packet delay fluctuation statistics according to the timestamp to obtain aligned data; S306. Based on the alignment data, by comparing the relative magnitudes of the received signal strength value, network jitter value, and data packet delay fluctuation value with a preset threshold, an arrival status label is assigned to each data packet. S307. Combine the arrival status tags of all data packets in chronological order to generate a data packet arrival status sequence.

[0025] As described in steps S301-S307 above, the receive status identifier refers to a set of metadata used to record status information during data packet transmission. It can be implemented by adding a status flag field to the data packet header or by using an independent status feedback message, aiming to provide a basic basis for classifying the status of data packets in the transmission link. The received signal strength statistics are quantitative indicators reflecting the quality of the wireless channel. They can be implemented using a sliding window algorithm combined with an exponentially weighted moving average method, aiming to eliminate instantaneous signal fluctuation interference and accurately characterize channel stability. The network jitter statistics are statistical parameters describing the characteristics of network delay changes. They can be implemented using a combination of segmented extreme value detection and dynamic variance threshold adjustment, aiming to distinguish between normal network fluctuations and abnormal interference events. The data packet delay fluctuation statistics are characteristic quantities that characterize the dynamic characteristics of data packet transmission delay. They can be implemented using delay differential sequence analysis and trend change point detection techniques, aiming to identify the persistence and suddenness of delay changes.

[0026] The process involves parsing the received status identifiers to extract the arrival status type and corresponding timestamp for each data packet, thus establishing a basic status classification framework. Based on this, statistical features are extracted from the received signal strength, network jitter, and data packet delay fluctuation sequences within a continuous time window. Multi-dimensional statistics are generated by calculating parameters such as moving average, variance, and extreme range. Subsequently, the received status identifiers are spatiotemporally correlated with each statistic based on a timestamp alignment mechanism, forming a unified analysis benchmark. Furthermore, by comprehensively comparing the relative relationships between each statistic and preset thresholds, data packet arrival status labels are dynamically assigned. Finally, all status labels are combined in chronological order to generate a structured data packet arrival status sequence. This process, through the temporal alignment and joint analysis of multi-source parameters, avoids the limitations of single-threshold judgment, enabling the status identification process to adapt to dynamic changes in the network environment and effectively distinguish between temporary delays and actual data loss.

[0027] In one embodiment, the step of obtaining the actual loss probability value of each data packet based on the data packet arrival status sequence and network transmission quality parameters includes: S308. Obtain the arrival status tag and historical arrival status sequence of each data packet from the data packet arrival status sequence; S309. Obtain the received signal strength numerical sequence, network jitter numerical sequence, and data packet delay fluctuation numerical sequence for the corresponding time window based on the network transmission quality parameters. S310. Obtain historical packet loss rate statistics based on historical arrival status sequences; S311. Normalize the received signal strength numerical sequence, network jitter numerical sequence, and data packet delay fluctuation numerical sequence to convert them into uniform dimension values. S312. Obtain the initial loss probability value of each data packet based on historical packet loss rate statistics and uniform unit values; S313. Adjust the initial loss probability value by time decay to generate the actual loss probability value.

[0028] As described in steps S308-S313 above, the arrival status label refers to the status type marked during data packet transmission. It can be implemented using discrete status values ​​such as successful reception, delayed reception, or loss, with the aim of distinguishing data packets with different transmission results. The historical arrival status sequence can be understood as a continuous record of data packet status arranged in chronological order. It can be implemented using a status sequence of multiple data packets within a sliding time window, with the aim of capturing the temporal correlation characteristics of transmission behavior. The historical packet loss rate statistic refers to a quantitative indicator of packet loss frequency calculated based on historical data. It can be implemented using the ratio of the number of lost data packets to the total number of transmitted data packets within a time window, with the aim of reflecting long-term transmission stability. The unified dimension value refers to the dimensionless parameter after standardization. It can be implemented by using min-max normalization or Z-score standardization methods to convert different dimension parameters to the same numerical range, with the aim of eliminating the dimensional differences of received signal strength, network jitter, and data packet delay fluctuations. The time decay adjustment refers to the processing mechanism of applying time-sensitive weight decay to the probability value. It can be implemented by using an exponential decay function to reduce the weight of historical data, with the aim of enabling the probability estimate to respond quickly to sudden changes in network status.

[0029] First, by extracting the arrival status label and historical arrival status sequence of each data packet based on the arrival status sequence, a temporal context for data packet transmission behavior is established. Simultaneously, based on network transmission quality parameters, the numerical sequences of received signal strength, network jitter, and data packet delay fluctuation within the corresponding time window are obtained to comprehensively characterize the current dynamic characteristics of the network. Then, historical packet loss rate statistics are calculated based on the historical arrival status sequence to quantify long-term transmission trends. On this basis, the numerical sequences of received signal strength, network jitter, and data packet delay fluctuation are normalized to eliminate dimensional barriers and generate uniform dimensional values. Next, the initial loss probability value is calculated by fusing the historical packet loss rate statistics and the uniform dimensional values, allowing the probability estimate to simultaneously reflect historical trends and real-time network conditions. Finally, time decay adjustment is applied to the initial loss probability value, and a timeliness weighting mechanism is used to reduce the influence of historical data, enabling the actual loss probability value to dynamically adapt to changes in the network environment, thereby accurately distinguishing between actual data packet loss and temporary delays.

[0030] In one embodiment, the step of generating a predicted compensation value for the packet content based on the packet whose loss status has been confirmed includes: S401. Obtain state estimation parameters based on the preset state vector and the historical data packet content sequence. The preset state vector includes data packet feature parameters and transmission timing parameters. The state estimation parameters include state transition matrix coefficients, observation matrix coefficients, process noise covariance, and observation noise covariance. S402. Obtain the predicted state value at the current moment based on the coefficients of the state transition matrix; S403. Obtain the prediction error covariance matrix at the current moment based on the state prediction value and the process noise covariance. S404. Generate the Kalman gain matrix based on the prediction error covariance matrix and the observation noise covariance matrix; S405. Obtain the updated state estimate and error covariance matrix based on the Kalman gain matrix and the difference between the actual and predicted observations. S406. Obtain a preliminary estimate of the data packet content based on the updated state estimate; S407. Obtain the confidence interval of the preliminary estimate based on the error covariance matrix; S408. Obtain the data packet content prediction compensation value based on the confidence interval and the preliminary estimate.

[0031] As described in steps S401-S408 above, the preset state vector refers to a mathematical vector describing the dynamic behavior characteristics of data packets. It can be implemented using a combination of parameters including data packet size, type identifier, and timestamp sequence. Its purpose is to integrate multi-dimensional transmission characteristics to characterize the network state evolution law. The state estimation parameters refer to the core parameter set characterizing the dynamic model characteristics of the system. Among them, the state transition matrix coefficients can be understood as a coefficient matrix describing the temporal dependence of data packet states, the observation matrix coefficients can be understood as the mapping relationship between associated state variables and actual observation values, and the process noise covariance and observation noise covariance can be understood as statistics that quantify the uncertainty of the system model and measurement error. Its purpose is to dynamically characterize the random fluctuation characteristics of the network environment. The Kalman gain matrix refers to a matrix that dynamically adjusts the fusion weights of prediction and observation information. It can be implemented using an adaptive weight matrix calculated based on the current error covariance and observation noise covariance. Its purpose is to optimize the information fusion ratio to cope with noise mutations. The confidence interval refers to a statistical interval that quantifies the reliable range of the estimation results. It can be implemented using the probability distribution boundary calculated based on the error covariance matrix. Its purpose is to provide a statistical reliability basis for the compensation value.

[0032] State estimation parameters are obtained through correlation analysis between a preset state vector and historical data packet content sequences, enabling the model to capture the dynamic behavior characteristics of data transmission and update in real time with network status. Based on this, the state prediction value at the current moment is derived using the coefficients of the state transition matrix, reflecting the continuity and historical dependence of data packet transmission. The prediction error covariance matrix is ​​calculated by combining the process noise covariance to dynamically evaluate prediction reliability. Subsequently, a Kalman gain matrix is ​​generated based on the prediction error covariance matrix and the observation noise covariance, adaptively adjusting the fusion weights of prediction and observation information. Then, the difference between actual and predicted observations is fused using the Kalman gain matrix to update the state estimate and error covariance matrix, promptly correcting for time-varying noise. Finally, a preliminary estimate is generated from the updated state estimate, and a statistically reliable prediction compensation value is output by combining the confidence interval constraints determined by the error covariance matrix. This process forms a closed-loop adaptive mechanism from historical data input to compensation value output, effectively overcoming the error accumulation problem caused by a fixed noise model by dynamically updating the noise statistical characteristics and state estimation parameters.

[0033] In one embodiment, the step of inserting the predicted compensation value into the data transmission sequence includes: S501. Obtain monitoring data of multiple transmission paths based on path identification information in the data packet transmission sequence, and obtain a set of multipath estimates for the same data packet based on the monitoring data of the multiple transmission paths. S502. Obtain historical estimation accuracy data based on the historical transmission records of each transmission path, and obtain the path reliability weight coefficient based on the historical estimation accuracy data; S503. Obtain the correlation coefficient of the estimated values ​​based on the numerical differences between the estimated values ​​in the multipath estimated value set; S504. Obtain the fusion weight allocation scheme based on the path reliability weight coefficient and the estimated value correlation coefficient; S505. Obtain the weighted fusion estimate based on the fusion weight allocation scheme and the multi-path estimate set, and obtain the overall error covariance after fusion based on the error statistical characteristics of each path estimate. S506. Obtain the spatiotemporal consistency test result based on the numerical relationship between the weighted fusion estimate and the content of adjacent data packets, and obtain the final confirmed prediction compensation value based on the spatiotemporal consistency test result. S507. Obtain the insertion position coordinates based on the sequence number information in the data packet transmission sequence, and obtain the updated data transmission sequence based on the insertion position coordinates and the finally confirmed prediction compensation value.

[0034] As described in steps S501-S507 above, the path reliability weight coefficient refers to a dynamic parameter that quantifies the long-term stability of the transmission path. It can be implemented using an exponentially weighted moving average algorithm based on historical transmission accuracy or a sliding window statistical method. The purpose is to ensure that high-reliability paths dominate the fusion process, thereby improving the quality of compensation value generation. The estimated value correlation coefficient can be understood as an indicator that characterizes the degree of statistical dependence between estimated values ​​of different paths. It can be calculated using the Pearson correlation coefficient or the Spearman rank correlation coefficient. The purpose is to identify and reduce the redundant influence of highly correlated paths and avoid overfitting in weight allocation. The fusion weight allocation scheme specifically refers to the mechanism that determines the contribution ratio by combining path reliability and estimated value independence. It can be implemented using linear combination optimization or a Bayesian inference framework. The purpose is to reduce the interference of noise and outliers on the final result.

[0035] By identifying the real-time status of different transmission paths using path identification information, multi-path monitoring data is acquired and a multi-path estimation set is formed, thereby aggregating redundant information to enhance the system's robustness to network fluctuations. Secondly, historical estimation accuracy data is extracted based on historical transmission records to dynamically generate path reliability weight coefficients. Simultaneously, the numerical differences between estimates are analyzed to calculate the correlation coefficient of the estimates; these two factors together constitute a fusion weight allocation scheme. Subsequently, the multi-path estimates are weighted and fused according to this scheme to generate a comprehensive estimate and quantify its uncertainty. Then, spatiotemporal consistency is checked by comparing the timing and numerical logic of adjacent data packets, and only verified compensation values ​​are adopted. Finally, the insertion position coordinates are precisely located based on the sequence number, completing the seamless integration of compensation values. This process organically connects with the predicted compensation values ​​generated in the preceding scheme. Through multi-source information fusion and adaptive verification mechanisms, the limitations of single-path compensation are effectively overcome, ensuring high-precision and high-reliability data flow reconstruction in complex network environments.

[0036] In one embodiment, the step of performing a consistency check on the data packets before and after the insertion position includes: S508. Obtain the numerical difference dataset based on the actual received data packets and the corresponding predicted compensation values; S509. Obtain statistical features of the differences based on the numerical difference dataset, wherein the statistical features of the differences include mean, variance, skewness coefficient and kurtosis coefficient; S510. Obtain the error distribution model parameters based on the statistical characteristics of the differences, and obtain the prediction compensation accuracy index based on the numerical difference dataset and the error distribution model parameters. S511. Obtain systematic deviation components and random error components based on accuracy indicators and error distribution model parameters, and obtain deviation correction parameters based on systematic deviation components; S512. Obtain the corrected compensation result based on the deviation correction parameters and predicted compensation value, and obtain the compensation verification report based on the corrected compensation result and accuracy index. If the compensation verification report meets the preset conditions, the compensation verification report is considered normal, and the verification is completed. If the verification report does not meet the preset conditions, the verification report is judged to be abnormal, the verification is not completed, and the process is returned to obtain the difference statistical features based on the numerical difference dataset.

[0037] As described in steps S508-S512 above, the numerical difference dataset refers to the set of numerical differences between the actual received data packets and the predicted compensation values. It can be stored and managed in the form of an array or database table. Its purpose is to quantify the compensation error and provide basic data for subsequent statistical analysis. The statistical characteristics of the differences include the mean, variance, skewness coefficient, and kurtosis coefficient. The mean reflects the central tendency of the error, the variance reflects the degree of dispersion, the skewness coefficient reflects the asymmetry of the distribution, and the kurtosis coefficient reflects the sharpness of the distribution. These can be calculated using statistical software libraries or custom algorithms. Their purpose is to comprehensively capture the multidimensional characteristics of the error distribution and avoid the limitations of a single indicator in adapting to the dynamic network environment. The error distribution model parameters refer to the parameters describing the characteristics of the error probability distribution. The set of data, which can be estimated using Gaussian mixture models or Bayesian inference methods, aims to transform the original discrepancy data into a probabilistic model, enabling the verification criteria to be dynamically adjusted based on the current network state. Systematic bias components refer to the error trend attributable to fixed factors, while random error components refer to unpredictable fluctuations. Bias correction parameters are adjustment coefficients used to correct systematic biases; these can be generated based on least squares or gradient descent algorithms to accurately locate correctable error sources and achieve targeted optimization. The compensation verification report is a comprehensive evaluation document containing accuracy indicators and error component analysis; it can be generated using structured data formats to provide multi-dimensional verification evidence and enhance the interpretability of the verification results.

[0038] A numerical difference dataset is constructed based on the actual received data packets and predicted compensation values, serving as the initial basis for error analysis. Subsequently, statistical characteristics of the differences, including mean, variance, skewness coefficient, and kurtosis coefficient, are calculated to comprehensively capture the central tendency, dispersion, asymmetry, and sharpness of the error distribution. Based on accuracy indicators and model parameters, systematic deviation components and random error components are separated, and correctable deviation components are extracted to generate deviation correction parameters. Then, the predicted compensation values ​​are adjusted using these deviation correction parameters to obtain the correction results, generating a compensation verification report containing multi-dimensional indicators. Finally, the verification is completed or returned to iteration based on whether the verification report meets preset conditions, forming a closed-loop feedback mechanism. This step sequence design ensures that the verification process can dynamically adapt to changes in network status, accurately identify correctable errors, and effectively avoid accidental misjudgments in single verifications, thereby guaranteeing the integrity of the data stream.

[0039] like Figure 2 As shown, the present invention also provides a data transmission packet loss compensation and optimization system for Internet of Things (IoT) communication devices, comprising: The data packet transmission sequence processing module 1 is used to acquire the data packet transmission sequence and obtain the reception status identifier and transmission timestamp of multiple data packets based on the data packet transmission sequence. The network transmission quality parameter processing module 2 is used to acquire network transmission quality parameters and, based on the network transmission quality parameters, acquire a sequence of received signal strength values, a sequence of network jitter values, and a sequence of data packet delay fluctuation values ​​within a continuous time window. The actual loss probability value acquisition module 3 is used to obtain the data packet arrival status sequence based on the receiving status identifier and the received signal strength value sequence, network jitter value sequence and data packet delay fluctuation value sequence within a continuous time window, and to obtain the actual loss probability value of each data packet based on the data packet arrival status sequence and network transmission quality parameters. Module 4 is used to determine whether the actual loss probability value exceeds a preset threshold. If the time limit is not exceeded, the data packet is marked as not lost. If the value exceeds the limit, the data packet is marked as confirmed lost, and a predicted compensation value for the data packet content is generated based on the data packet in the confirmed lost state. The verification module 5 is used to insert the predicted compensation value into the data transmission sequence and to perform consistency verification on the data packets before and after the insertion position.

[0040] Preferably, the actual loss probability value acquisition module 3 includes: The historical arrival status sequence acquisition unit is used to obtain the arrival status tag and historical arrival status sequence of each data packet based on the data packet arrival status sequence. The network transmission quality parameter analysis unit is used to obtain the received signal strength numerical sequence, network jitter numerical sequence, and data packet delay fluctuation numerical sequence for the corresponding time window based on the network transmission quality parameters. The historical packet loss rate statistics acquisition unit is used to obtain historical packet loss rate statistics based on the historical arrival status sequence. The numerical unification unit is used to normalize the numerical sequences of received signal strength, network jitter, and data packet delay fluctuation, converting them into numerical values ​​with uniform dimensions. The initial loss probability value acquisition unit is used to acquire the initial loss probability value of each data packet based on historical packet loss rate statistics and uniform unit values. The actual loss probability value generation unit is used to adjust the initial loss probability value by time decay to generate the actual loss probability value.

[0041] like Figure 3 As shown, the present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of a data transmission packet loss compensation optimization method for Internet of Things communication devices.

[0042] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of a method for optimizing data transmission packet loss compensation in Internet of Things communication devices.

[0043] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in this application and in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual-speed SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0044] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0045] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent results or equivalent process transformations made based on the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for optimizing packet loss compensation in Internet of Things (IoT) communication devices, characterized in that, include: Obtain the data packet transmission sequence, and based on the data packet transmission sequence, obtain the reception status identifier and transmission timestamp of multiple data packets; Obtain network transmission quality parameters, and based on the network transmission quality parameters, obtain the received signal strength numerical sequence, network jitter numerical sequence, and data packet delay fluctuation numerical sequence within a continuous time window; The arrival status sequence of data packets is obtained based on the received status identifier and the numerical sequence of received signal strength, network jitter, and data packet delay fluctuation within a continuous time window. The actual loss probability value of each data packet is obtained based on the arrival status sequence of data packets and network transmission quality parameters. Determine whether the actual loss probability value exceeds a preset threshold; If the time limit is not exceeded, the data packet is marked as not lost. If the value exceeds the limit, the data packet is marked as confirmed lost, and a predicted compensation value for the data packet content is generated based on the data packet in the confirmed lost state. The predicted compensation value is inserted into the data transmission sequence, and the consistency of the data packets before and after the insertion position is checked.

2. The method for optimizing packet loss compensation in IoT communication devices according to claim 1, characterized in that, The step of obtaining the data packet arrival status sequence based on the received status identifier and the received signal strength numerical sequence, network jitter numerical sequence, and data packet delay fluctuation numerical sequence within a continuous time window includes: The received status identifier is parsed to extract the arrival status type and corresponding timestamp of each data packet, wherein the arrival status type includes successful reception, delayed reception, and loss; The received signal strength numerical sequence is divided into time windows, the moving average and variance of the received signal strength values ​​in each window are obtained, and a received signal strength statistic is generated based on the moving average and variance of the received signal strength values ​​in each window. The network jitter numerical sequence is divided into time windows, and the maximum, minimum and standard deviation of the network jitter values ​​in each window are obtained. The network jitter statistics are obtained based on the maximum, minimum and standard deviation of the network jitter values ​​in each window. The data packet delay fluctuation numerical sequence is divided into time windows to obtain the fluctuation range and rate of change of the data packet delay value within each window, and the data packet delay fluctuation statistics are obtained based on the fluctuation range and rate of change of the data packet delay value within each window. Aligned data is obtained by aligning and associating the received status identifier with the received signal strength statistics, network jitter statistics, and data packet delay fluctuation statistics according to timestamps. Based on the alignment data, an arrival status label is assigned to each data packet by comparing the relative magnitudes of the received signal strength value, network jitter value, and data packet delay fluctuation value with a preset threshold. Combine the arrival status tags of all data packets in chronological order to generate a data packet arrival status sequence.

3. The method for optimizing packet loss compensation in IoT communication devices according to claim 1, characterized in that, The steps for obtaining the actual loss probability value of each data packet based on the data packet arrival status sequence and network transmission quality parameters include: Obtain the arrival status label and historical arrival status sequence of each data packet from the data packet arrival status sequence; Based on the network transmission quality parameters, obtain the numerical sequences of received signal strength, network jitter, and data packet delay fluctuation for the corresponding time window; Obtain historical packet loss rate statistics based on historical arrival status sequences; The numerical sequences of received signal strength, network jitter, and data packet delay fluctuation are normalized and converted into numerical values ​​with uniform dimensions. The initial loss probability value of each data packet is obtained based on historical packet loss rate statistics and uniform unit values; The initial loss probability value is adjusted by time decay to generate the actual loss probability value.

4. The method for optimizing packet loss compensation in IoT communication devices according to claim 1, characterized in that, The steps of generating predicted compensation values ​​for packet content based on packets whose loss status has been confirmed include: The state estimation parameters are obtained based on the preset state vector and the historical data packet content sequence. The preset state vector includes data packet feature parameters and transmission timing parameters, and the state estimation parameters include state transition matrix coefficients, observation matrix coefficients, process noise covariance, and observation noise covariance. The predicted state value at the current moment is obtained based on the coefficients of the state transition matrix. The prediction error covariance matrix at the current moment is obtained based on the state prediction value and the process noise covariance. Generate the Kalman gain matrix based on the prediction error covariance matrix and the observation noise covariance matrix; The updated state estimate and error covariance matrix are obtained based on the Kalman gain matrix and the difference between the actual and predicted observations. Obtain a preliminary estimate of the packet content based on the updated state estimate; Obtain the confidence interval of the preliminary estimate based on the error covariance matrix; The data packet content prediction compensation value is obtained based on the confidence interval and the preliminary estimate.

5. The method for optimizing packet loss compensation in IoT communication devices according to claim 1, characterized in that, The step of inserting the predicted compensation value into the data transmission sequence includes: Based on the path identification information in the data packet transmission sequence, monitoring data of multiple transmission paths are obtained, and a set of multipath estimates for the same data packet is obtained based on the monitoring data of the multiple transmission paths. Historical estimation accuracy data is obtained based on the historical transmission records of each transmission path, and path reliability weighting coefficients are obtained based on the historical estimation accuracy data. The correlation coefficient of the estimated values ​​is obtained based on the numerical differences between the estimated values ​​in the multipath estimated value set. The fusion weight allocation scheme is obtained based on the path reliability weight coefficient and the estimated value correlation coefficient. The weighted fusion estimate is obtained based on the fusion weight allocation scheme and the multi-path estimate set, and the overall error covariance after fusion is obtained based on the error statistical characteristics of each path estimate. The spatiotemporal consistency test result is obtained based on the numerical relationship between the weighted fusion estimate and the content of adjacent data packets, and the final confirmed prediction compensation value is obtained based on the spatiotemporal consistency test result. The insertion position coordinates are obtained based on the sequence number information in the data packet transmission sequence, and the updated data transmission sequence is obtained based on the insertion position coordinates and the finally confirmed prediction compensation value.

6. The method for optimizing data transmission packet loss compensation in IoT communication devices according to claim 1, characterized in that, The steps for verifying the consistency of data packets before and after the insertion position include: The numerical difference dataset is obtained based on the actual received data packets and the corresponding predicted compensation values. Based on the numerical difference dataset, difference statistical features are obtained, wherein the difference statistical features include mean, variance, skewness coefficient and kurtosis coefficient; The error distribution model parameters are obtained based on the statistical characteristics of the differences, and the prediction compensation accuracy index is obtained based on the numerical difference dataset and the error distribution model parameters. The systematic deviation component and random error component are obtained based on the accuracy index and error distribution model parameters, and the deviation correction parameters are obtained based on the systematic deviation component. Obtain the corrected compensation result based on the deviation correction parameters and the predicted compensation value, and obtain the compensation verification report based on the corrected compensation result and the accuracy index. If the compensation verification report meets the preset conditions, the compensation verification report is considered normal, and the verification is completed. If the compensation verification report does not meet the preset conditions, the compensation verification report is judged to be abnormal, the verification is not completed, and the process returns to obtaining the difference statistical features based on the numerical difference dataset.

7. A data transmission packet loss compensation and optimization system for Internet of Things (IoT) communication devices, characterized in that, include: The data packet transmission sequence processing module is used to obtain the data packet transmission sequence and obtain the reception status identifier and transmission timestamp of multiple data packets based on the data packet transmission sequence. The network transmission quality parameter processing module is used to acquire network transmission quality parameters and, based on the network transmission quality parameters, acquire a sequence of received signal strength values, a sequence of network jitter values, and a sequence of data packet delay fluctuation values ​​within a continuous time window. The actual loss probability value acquisition module is used to obtain the data packet arrival status sequence based on the received status identifier and the received signal strength value sequence, network jitter value sequence and data packet delay fluctuation value sequence within a continuous time window, and to obtain the actual loss probability value of each data packet based on the data packet arrival status sequence and network transmission quality parameters. The judgment module is used to determine whether the actual loss probability value exceeds a preset threshold; If the time limit is not exceeded, the data packet is marked as not lost. If the value exceeds the limit, the data packet is marked as confirmed lost, and a predicted compensation value for the data packet content is generated based on the data packet in the confirmed lost state. The verification module is used to insert the predicted compensation value into the data transmission sequence and to perform consistency verification on the data packets before and after the insertion position.

8. The PID algorithm-based closed-loop temperature control system for a 3D printer's heated bed according to claim 7, characterized in that, The actual loss probability value acquisition module includes: The historical arrival status sequence acquisition unit is used to obtain the arrival status tag and historical arrival status sequence of each data packet based on the data packet arrival status sequence. The network transmission quality parameter analysis unit is used to obtain the received signal strength numerical sequence, network jitter numerical sequence, and data packet delay fluctuation numerical sequence for the corresponding time window based on the network transmission quality parameters. The historical packet loss rate statistics acquisition unit is used to obtain historical packet loss rate statistics based on the historical arrival status sequence. The numerical unification unit is used to normalize the numerical sequences of received signal strength, network jitter, and data packet delay fluctuation, converting them into numerical values ​​with uniform dimensions. The initial loss probability value acquisition unit is used to acquire the initial loss probability value of each data packet based on historical packet loss rate statistics and uniform unit values. The actual loss probability value generation unit is used to adjust the initial loss probability value by time decay to generate the actual loss probability value.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images