A transmission control method and system for Internet of Things terminal data

Abstract

The present invention discloses a transmission control method and system for Internet of Things terminal data, and specifically relates to the technical field of Internet of Things terminal control, and is used to solve the problem of inaccurate transmission control of Internet of Things terminals; the present invention reasonably determines the sensor type, distribution and data characteristics, formulates acquisition targets and transmission requirements, sets sampling frequencies and acquisition windows, and adopts asynchronous or synchronous acquisition strategies through Internet of Things terminal application scenarios to achieve efficient data acquisition and scheduling, ensures data integrity through noise filtering, outlier detection and data normalization processing, quantifies data priority according to entropy shift index, identifies important events in combination with dynamic threshold clustering, dynamically adjusts bandwidth resources, ensures that high-priority data is transmitted first, and adopts an adaptive scheduling mechanism and a time slice polling algorithm to manage queues of data of different priorities, optimizes data transmission efficiency, and improves the reliability of Internet of Things terminal equipment.

Description

Technical Field

[0001] The present invention relates to the field of Internet of Things terminal control technology, and more specifically, to a method and system for controlling transmission of Internet of Things terminal data. Background Art

[0002] In the Internet of Things (IoT) system, terminal devices collect various environmental or equipment status data in real time through sensors, and transmit this data to the central server or cloud through the network for analysis and processing. In actual deployment, IoT terminal devices may be distributed in different application scenarios such as factories, transportation, agriculture, and medical care. In each scenario, the type, importance, and urgency of the data of terminal devices are different. For example, in industrial monitoring scenarios, equipment status data (such as temperature and vibration) needs to be transmitted first to detect equipment anomalies in time, while in smart agriculture, sensor information such as soil moisture and meteorological data can be processed with a delay and does not need to have the same priority as equipment failure data.

[0003] Deficiencies in existing technologies: IoT terminal devices may be deployed in environments with poor network conditions. Network fluctuations can cause data loss or delays, affecting the timeliness and integrity of data, resulting in situations such as limited bandwidth, frequent network fluctuations, and poor device computing power. Under these restrictions, ensuring the priority transmission of key data is the most important part of the IoT terminal data transmission process. However, the transmission priority scheduling of IoT terminal devices is not accurate enough, resulting in delays or omissions in the transmission of key data, affecting subsequent device control. Summary of the invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides a transmission control method and system for IoT terminal data to solve the problem of inaccurate transmission control of IoT terminals in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A method for controlling transmission of data from an Internet of Things terminal comprises the following steps:

[0007] Determine the application scenarios of IoT terminals based on the type, distribution and physical characteristics of the sensor data collected, formulate the data collection goals and transmission demand assessment, and clarify the deployment location and functions of various sensors;

[0008] According to the type of sensor and the data change rate, set the sampling frequency and data acquisition window, and use asynchronous or synchronous acquisition strategies for data acquisition and transmission scheduling, perform noise filtering and outlier detection on the collected data, and normalize different types of data;

[0009] Quantify data priority based on entropy shift index, identify important event changes based on dynamic threshold clustering algorithm, and dynamically adjust bandwidth resources to determine data transmission priority;

[0010] Based on the priority allocation results, independent queues are allocated to data flows of different priorities. The queue capacity is adjusted through an adaptive scheduling mechanism, and a time slice polling algorithm is used for priority polling to optimize data transmission efficiency and queue management.

[0011] In a preferred embodiment, based on the type, distribution and physical characteristics of the sensor data collected, the IoT terminal application scenario is determined, the data collection target and transmission demand assessment are formulated, and the deployment location and function of each type of sensor are clarified. The specific process is as follows:

[0012] Determine the distribution density and location of sensors, clarify the types of sensors used by terminal devices, including temperature, humidity, vibration, pressure, and gas concentration sensors, and analyze the deployment of all sensors in IoT terminals;

[0013] Determine the data type and characteristics of the collection, including numeric, discrete, and symbolic data types;

[0014] Determine the environmental characteristics, which include the dynamics of the scene and the abnormalities within the scene. The dynamics of the scene include the ambient temperature, humidity, and frequency of equipment use. The abnormalities within the scene include the impact of equipment failure or emergency conditions on data.

[0015] The scenario of the IoT terminal is obtained by combining the distribution density and location of the sensors, the data types and characteristics collected by the sensors, and the environmental characteristics;

[0016] Environmental sensors collect environmental data, crowd detectors, noise sensors, infrared sensors, and smoke alarms collect data as sensor data, and cameras collect data as video data;

[0017] Determine the target data and core data content to be collected and the data transmission requirements, and determine the physical deployment location and function of the sensor based on the application scenario.

[0018] In a preferred embodiment, the sampling frequency and data acquisition window are set according to the type of sensor and the data change rate, and an asynchronous or synchronous acquisition strategy is used for data acquisition and transmission scheduling, noise filtering and outlier detection are performed on the collected data, and different types of data are normalized. The specific process is as follows:

[0019] Determine the data collection frequency of each sensor based on the sensor type, data change rate, and bandwidth resources, and set a data collection window for each sensor;

[0020] A synchronous collection mechanism is used for time-series correlation data, and an asynchronous collection mechanism is used for independent data sources;

[0021] After data collection, the collected raw data is preprocessed, and the high-frequency noise signals in the sensor data are detected and filtered out using the noise filtering method based on wavelet transform; the distribution density estimation technology is used to detect the abnormal values ​​appearing in the high-dimensional data space, mark them, and smooth them through the interpolation algorithm;

[0022] The data were processed using a normalization method based on Laplace regularization.

[0023] In a preferred embodiment, data priority is quantified according to the entropy shift index, important event changes are identified based on a dynamic threshold clustering algorithm, and bandwidth resources are dynamically adjusted to determine data transmission priority. The specific process is as follows:

[0024] Use kernel-based support vector regression to extract features from the collected data, extract the main features from each data stream, and form feature vectors from the main features through nonlinear mapping;

[0025] From the sensor data of IoT terminal devices, construct the data vector at each time point and calculate the high-dimensional data entropy at each time point;

[0026] Dynamically track and increment the entropy offset, recalculate the entropy value of the data, and calculate the entropy difference between the previous and next time points through the entropy offset formula. The entropy difference indicates the change in the entropy value of the data between adjacent time points.

[0027] Compare the data entropy of adjacent time points to dynamically detect the uncertainty increment of the data stream. The uncertainty increment indicates the uncertainty change of the data over time.

[0028] Based on the calculated entropy offset, a dynamic threshold clustering algorithm is used to identify important event changes and determine whether the change requires adjustment of data priority. The offset represents the trend of entropy value changes at adjacent time points.

[0029] For the identified large entropy shift data points, the importance of the data is quantified and a priority is assigned using the entropy shift index, which represents the degree of fluctuation in the data stream.

[0030] Bandwidth resources are allocated to data points according to the calculated entropy shift index.

[0031] In a preferred embodiment, based on the priority allocation result, independent queues are allocated to data flows of different priorities, the queue capacity is adjusted through an adaptive scheduling mechanism, and a time slice polling algorithm is used for priority polling to optimize data transmission efficiency and queue management. The specific process is as follows:

[0032] Perform queue management and scheduling control on data according to the assigned data priority;

[0033] Assign independent queues to data flows of different priorities, and initialize each queue according to the priority weight of the data;

[0034] The queue capacity is adjusted based on the adaptive scheduling mechanism. After the queue is initialized, each queue is monitored in real time to analyze the queue occupancy and data flow status. The capacity and transmission order of each priority queue are dynamically adjusted according to the changes in network bandwidth and data flow through the adaptive scheduling mechanism. When the occupancy rate of the high-priority queue exceeds the preset threshold, the high-priority queue capacity is automatically expanded, or the transmission frequency of the low-priority data flow is limited.

[0035] Perform queue priority adjustment based on delay perception. Introduce a queue priority adjustment mechanism based on delay perception. By real-time monitoring of the transmission delay of each data flow, determine whether the queue priority needs to be adjusted. When the transmission delay of the queue exceeds the preset delay threshold, the queue priority is automatically increased.

[0036] Introducing bandwidth optimization and queue clearing mechanisms, clearing low-priority data that has not been processed for a long time through bandwidth optimization algorithms;

[0037] After completing the adaptive queue management and priority polling scheduling, transmission reliability enhancement, packet retransmission mechanism, bandwidth optimization and flow regulation, congestion control and fault recovery processing are carried out.

[0038] In a preferred embodiment, after completing the adaptive queue management and priority polling scheduling, transmission reliability enhancement, data packet retransmission mechanism, bandwidth optimization and flow control, congestion control and fault recovery processing are performed. The specific process is as follows:

[0039] Use the data packet confirmation mechanism. After each transmission, the receiving end sends a confirmation message to the sending end to confirm that the data has been successfully received. If the receiving end does not receive the data packet within the specified time, or the data packet is damaged, the data packet retransmission mechanism is started. In case of network congestion, the redundancy control mechanism is enabled.

[0040] If the data packet retransmission fails, the fault recovery mechanism is started, and the data packet transmission is abandoned according to the set timeout period, and the relevant information is recorded, and the untransmitted priority data continues to be processed.

[0041] A transmission control system for IoT terminal data, used to implement the above-mentioned transmission control method for IoT terminal data, comprising:

[0042] The sensor deployment module is used to determine the application scenarios of IoT terminals based on the type, distribution and physical characteristics of the sensor data collected, formulate the target of data collection and transmission demand assessment, and clarify the deployment location and function of various sensors;

[0043] The data transmission specification module is used to set the sampling frequency and data acquisition window according to the type of sensor and the data change rate, and adopt asynchronous or synchronous acquisition strategies to perform data acquisition and transmission scheduling, perform noise filtering and outlier detection on the collected data, and perform standardization processing on different types of data;

[0044] The priority quantification module quantifies data priority according to the entropy shift index, identifies important event changes based on the dynamic threshold clustering algorithm, and dynamically adjusts bandwidth resources to determine data transmission priority;

[0045] The transmission control module is used to allocate independent queues to data flows of different priorities based on the priority allocation results, adjust the queue capacity through an adaptive scheduling mechanism, and use a time slice polling algorithm for priority polling to optimize data transmission efficiency and queue management.

[0046] Technical effects and advantages of the present invention:

[0047] The present invention determines the type, distribution and physical characteristics of sensor data collected through the application scenarios of IoT terminals, formulates the target of data collection and transmission demand assessment, and realizes efficient data collection and scheduling by setting the sampling frequency and acquisition window, adopts asynchronous or synchronous strategies, and performs noise filtering, outlier detection and normalization processing on the data to ensure the integrity and accuracy of the data, thereby ensuring the efficient collection of multi-type sensor data, adapting to the data requirements in different application scenarios, and optimizing the data management and transmission performance of IoT terminal devices;

[0048] The data priority is quantified through the entropy shift index, combined with the dynamic threshold clustering algorithm to identify changes in important events, and the bandwidth resources are dynamically adjusted to ensure the timely transmission of high-priority data. The queue management and scheduling of data streams of different priorities are performed through the adaptive scheduling mechanism and time slice polling algorithm, which optimizes the data transmission efficiency, avoids queue blocking, and improves the data transmission reliability and overall performance in complex network environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 The present invention is a flow chart of a method for controlling transmission of Internet of Things terminal data.

[0050] Figure 2 The present invention is a structural schematic diagram of a transmission control system for Internet of Things terminal data. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] Example 1: Figure 1 As shown, a transmission control method for IoT terminal data includes the following steps:

[0053] Determine the application scenarios of IoT terminals based on the type, distribution and physical characteristics of the sensor data collected, formulate the data collection goals and transmission demand assessment, and clarify the deployment location and functions of various sensors;

[0054] According to the type of sensor and the data change rate, set the sampling frequency and data acquisition window, and use asynchronous or synchronous acquisition strategies for data acquisition and transmission scheduling, perform noise filtering and outlier detection on the collected data, and normalize different types of data;

[0055] Quantify data priority based on entropy shift index, identify important event changes based on dynamic threshold clustering algorithm, and dynamically adjust bandwidth resources to determine data transmission priority;

[0056] Based on the priority allocation results, independent queues are allocated to data flows of different priorities. The queue capacity is adjusted through an adaptive scheduling mechanism, and a time slice polling algorithm is used for priority polling to optimize data transmission efficiency and queue management.

[0057] Based on the priority allocation of dynamic event recognition, determine the data collection scenario of the IoT terminal and ensure the accuracy and adaptability of the data collection steps. Scenario determination is the basis for data collection. It is necessary to clarify the specific environment of the IoT terminal application and the core goal of data collection. This step requires a comprehensive assessment of the physical environment, equipment layout and data transmission requirements of the collection scenario. Scenario determination includes the following key parts:

[0058] Clarify the sensor network type and distribution, that is, determine the distribution density and location of sensors, clarify the types of sensors used by terminal devices (such as temperature, humidity, vibration, pressure, gas concentration, etc.), and analyze the deployment of these sensors in IoT terminals;

[0059] Determine the type and characteristics of the collected data, that is, clarify the type of data collected (such as numerical, discrete or symbolic), and determine the actual physical meaning of each type of data and its scope of influence according to the application scenario of the terminal device. The data characteristics in the scenario may be heterogeneous, so it is necessary to determine the comparability and relevance of the data in advance, and determine the environmental characteristics, analyze the dynamics of the scenario (such as ambient temperature, humidity, frequency of equipment use) and the potential impact of abnormal conditions in the scenario (such as equipment failure or emergency state) on the data, and determine the scenario of the IoT terminal based on the above;

[0060] In the IoT terminal application scenario, setting data collection goals and evaluating transmission requirements are important steps to ensure efficient operation of the system, including the following considerations:

[0061] Determine the goal of data collection. The goal of data collection is to solve specific application needs. Clarifying the target data and core data content of the collection will help define the performance indicators, response time and storage requirements of the entire data collection;

[0062] It should be noted that core data content refers to data that is critical to achieving specific application requirements in IoT terminal applications. These data directly reflect key parameters or states and can effectively support system decision-making, analysis or control. For example, in an environmental monitoring system, core data content may include temperature, humidity, gas concentration, etc.; in smart manufacturing, core data may be vibration, pressure or operating status of equipment. Core data content usually has a higher collection frequency and priority to ensure timely and accurate capture of key information in application scenarios.

[0063] Transmission demand assessment: The purpose of assessing data transmission demand is to ensure that data can be accurately and real-time transmitted to the data center for processing or decision-making. The assessment of transmission demand can consider data transmission frequency, bandwidth and network conditions, data flow, and data real-time performance.

[0064] In addition to determining the sensor type and distribution, it is also necessary to clarify the physical deployment location and function of the sensor according to the specific application scenario to ensure the accuracy and coverage of data collection, as follows:

[0065] The deployment location should be determined based on the monitoring target of the sensor, the physical characteristics of the application scenario, and the functional requirements of the sensor. It is necessary to ensure that the sensor is placed in a location where accurate data can be obtained and to avoid data interference and repeated collection. Different types of sensors have different functions in different scenarios, and deployment must ensure that they play the best role.

[0066] For example, in smart farmland, soil moisture sensors need to be deployed in the root area of ​​crops. The depth should be determined according to the growth needs of crops. They are usually arranged in shallow, middle and deep layers to comprehensively monitor moisture changes at different depths.

[0067] For example, in the intelligent farmland environment monitoring system, in the farmland scenario, sensors are arranged in different areas of the farmland, including multiple layers in the soil (shallow, middle, and deep layers), and at different heights in the air. The distribution density of sensors is adjusted according to the area of ​​the farmland and the type of crops. In high-density areas (such as areas with dense root systems), the distribution interval of soil moisture sensors is 5 meters, while in low-density areas (such as near irrigation facilities), the interval between sensors is 10 meters. In addition, air temperature and humidity sensors are arranged at different heights of the field at intervals of 20 meters to monitor the entire crop growth environment.

[0068] Sensor data collection data types and characteristics: Soil moisture sensor: collects soil moisture content. The data is a continuous analog signal. Since soil moisture changes slowly, the collection frequency is once an hour. Soil temperature sensor: monitors soil temperature. The data is numerical and the collection frequency is once every half an hour to ensure that the crop root temperature is within the optimal range.

[0069] Environmental characteristics analysis, seasonal changes: Ambient temperature and humidity, soil moisture content will fluctuate significantly with the seasons. The sensor system needs to adjust the acquisition frequency in different seasons. For example, the acquisition frequency of soil moisture sensors needs to be increased in the dry season to monitor moisture changes in a timely manner. Terrain changes: Due to the topographic differences of farmland, low-lying areas may have problems with excessive moisture or water accumulation. Therefore, sensors in these areas need to be distributed at a higher density and the acquisition frequency needs to be increased.

[0070] In the intelligent farmland environment monitoring of this IoT terminal, the distribution density and location of sensors are determined according to the crop growth requirements and the terrain characteristics of the farmland. The data types and collection frequencies of different types of sensors are dynamically adjusted according to environmental changes and abnormal conditions to ensure the accuracy and real-time nature of data collection. Such a scenario can achieve precise agricultural management, including crop growth status monitoring, irrigation optimization, and prevention of the impact of environmental changes on crops.

[0071] Scenario determination of IoT terminals helps define the core environmental parameters for collecting data, and by sorting out sensor types and distribution, clarifies the importance of various types of data in practical applications, providing background basis for subsequent priority allocation.

[0072] After the scenario is determined, the preparation stage for actual data collection begins. During the actual data collection process, the sensors operate according to the set rules, and the data collection frequency and scheduling mechanism are designed according to the characteristics of different sensors, including the following aspects:

[0073] Determine the data collection frequency of each sensor based on the sensor type, data change rate, and bandwidth resources. Important sensors (such as vibration sensors) can collect data at high frequencies, while non-critical sensors (such as temperature and humidity) can moderately reduce the frequency. Set a data collection window for each sensor to ensure that data is collected and transmitted within a reasonable time frame. That is, by configuring the collection window length, avoid transmission congestion caused by excessive data traffic.

[0074] In a multi-sensor environment, choose between asynchronous and synchronous acquisition strategies. For data with strong time-series correlation (such as temperature and pressure), a synchronous acquisition mechanism is used; for independent data sources (such as gas concentration), asynchronous acquisition can be used to reduce bandwidth burden.

[0075] By determining the acquisition frequency, timing and data window length, the accuracy and efficiency of sensor data acquisition are guaranteed. The reasonable design of the scheduling mechanism lays the foundation for the smooth transmission of data streams and ensures that the sensor data acquisition scheduling under low bandwidth conditions meets the requirements.

[0076] After data collection, the collected raw data is preprocessed to ensure the integrity and accuracy of the data, including the detection and processing of noise data and outliers. The preprocessing process is divided into the following parts: Use a noise filtering method based on wavelet transform to detect and filter out high-frequency noise signals in sensor data to ensure that data quality is not affected by environmental noise; Use distribution density estimation technology to detect outliers that appear in high-dimensional data space. For data points with excessively large outliers, the system will mark them and smooth them through an interpolation algorithm to ensure that they do not affect the overall data trend; Normalize the collected multidimensional data to ensure that different types of data can be analyzed on the same dimension. A normalization method based on Laplace regularization can be used to ensure that each sensor data has the same metric.

[0077] On the basis of preprocessing data, it is necessary to structure the collected multi-dimensional data stream into a model to form a matrix expression that can be used for subsequent high-dimensional entropy measurement. The original data stream is mapped to a high-dimensional space through feature extraction and dimensionality reduction to form a dynamic time series data matrix. The specific steps are as follows:

[0078] Use kernel-based support vector regression (SVR) to extract features from the collected data, extract the main features (such as temperature trends, pressure fluctuations, etc.) from each data stream, and form feature vectors through nonlinear mapping;

[0079] Map all sensor data to a high-dimensional space, that is, set the collected values ​​of different sensors at different times. At this time, the data already has dynamic characteristics, that is, a multi-dimensional time series data stream is formed;

[0080] Construct the data vector at each time point t from the sensor data of the IoT terminal device ,in, represents the collected value of the i-th sensor at time t, n represents the total number of sensors, and the high-dimensional data entropy is calculated at each time point , through the following formula: ,in, It is a probability distribution based on the collected data. It is a regulating factor used to control the data entropy fluctuations in extreme situations. By calculating the data entropy value at each time point, the uncertainty of the data collected by the IoT terminal can be measured. The entropy value formula introduces a regulating factor on the basis of traditional entropy to avoid excessive fluctuations in data entropy under extreme probability distributions.

[0081] Dynamic tracking and incremental determination of entropy offset is performed to recalculate the entropy value of high-dimensional data at time t+1, and the entropy difference between the previous and next time points t and t+1 is calculated through the following entropy offset formula. The entropy difference represents the magnitude of the entropy change of the data between adjacent time points. The formula is as follows: ,in, is to prevent small values ​​of zero probability from occurring, , Respectively represent the probability distribution of data collected by i sensors at time t and time t+1. Through this formula, the entropy offset in the high-dimensional data stream is dynamically tracked, and the uncertainty change of the data stream between adjacent time points is evaluated;

[0082] By comparing the data entropy at adjacent time points, the uncertainty increment of the data stream is dynamically detected. The uncertainty increment indicates the uncertainty change of the data over time. The larger the entropy offset, the greater the change in the data stream over time. The entropy offset indicates the change trend of the entropy value at adjacent time points.

[0083] Based on the entropy offset calculated previously, the dynamic threshold clustering algorithm is used to identify important event changes and determine whether these changes require adjustment of data priority. The settings are as follows: ,in, is the adjustment coefficient; is the smoothing factor, Z is the time window length, is the adjustment index, k is the length of the observation period, and n is the total number of time windows. The dynamic threshold is used to dynamically adjust the setting of the cluster center point to ensure that the system can identify important change events in the data in real time. By dynamically adjusting the cluster threshold, the transmission control system can flexibly respond to data changes in different time periods, especially in rapidly changing data scenarios. Through real-time threshold adjustment, abnormal events can be effectively captured to avoid misjudgment problems that may be caused by static thresholds.

[0084] It should be noted that the adjustment coefficient , used to scale and adjust the dynamic threshold as a whole, controlling the sensitivity and response speed to the entropy offset, affecting the size and flexibility of the threshold. The size of the adjustment coefficient determines the overall scale of the threshold. When increases, the threshold increases accordingly, the sensitivity to entropy shift decreases, and the priority adjustment is triggered only when significant events occur. When it decreases, the system threshold is lowered, it is more sensitive to small fluctuations, and changes in the data can be identified more quickly; smoothing factor It is used to avoid the situation where the denominator is zero or the entropy offset fluctuates too much during the calculation process. The smoothing factor ensures that even if the entropy changes close to zero or extreme data occurs, the system can still calculate the threshold normally and will not cause calculation errors due to the denominator being zero; adjust the index , which is used to control the nonlinear change amplitude of the dynamic threshold and the responsiveness to historical entropy changes, the sensitivity to changes in entropy offset, and determines how the system accumulates and weights entropy changes of historical data.

[0085] For the identified large entropy shift data points, the entropy shift index (EDI) is further used to quantify the importance of the data and assign priorities. The entropy shift index represents the degree of fluctuation in the data stream. The entropy shift index calculation formula is as follows: ,in, represents the entropy offset of the ith sensor, is the exponential tuning parameter, is the weight coefficient. The entropy offset index is used to determine which data points need to be transmitted first, ensuring that data points with larger entropy offsets receive higher priorities. The entropy offset index combines the entropy offset with the dynamic threshold to evaluate the importance of each data point in the entire data stream. The combination of exponential regulation and logarithmic function ensures that priority allocation under different weights is smooth and flexible, and determines the transmission strategy under low bandwidth conditions.

[0086] According to the entropy shift index calculated above, appropriate bandwidth resources are allocated to each data point through the bandwidth allocation formula: ,in, is the entropy offset index of the ith sensor, j represents the entropy offset index of the jth sensor, is the bandwidth of the ith data point, n is the number of sensors, is the total available bandwidth, The amplification factor is used to increase the proportion of high-priority data in bandwidth allocation. This formula ensures that critical data streams are given priority in transmission bandwidth, while low-priority data is appropriately delayed in transmission, and dynamic resource allocation based on priority is achieved to ensure that high-priority data obtains sufficient bandwidth and low-priority data does not block transmission. The use of the amplification factor ensures the flexibility and dynamism of bandwidth allocation and meets the data transmission needs in different scenarios.

[0087] The high-dimensional data entropy deviation measurement method based on dynamic threshold clustering can reasonably assign priorities to IoT terminal data and realize dynamic data transmission scheduling through bandwidth management.

[0088] After completing the high-dimensional data entropy offset measurement priority allocation processing steps based on dynamic threshold clustering, the data is reasonably queued and scheduled according to the assigned priority. The main goal of adaptive queue management and priority polling scheduling is to ensure that high-priority data is processed first, while also ensuring that low-priority data can be transmitted within a certain period of time to avoid long-term data delay or loss. The specific steps are as follows:

[0089] Based on the priority calculated by the entropy deviation index (EDI), firstly, independent queues are allocated to data flows of different priorities. Each queue is initialized according to the priority weight of the data, and the initial queue length and data buffer space are set. The queue allocation formula is as follows: ,in, is the capacity of the i-th priority queue, is the capacity of the jth priority queue, n is the total capacity of the priority queues, is the total queue capacity of the system, The amplification factor is used to dynamically allocate queue space according to the weights of different priorities, ensuring that high-priority data obtains more buffer resources.

[0090] It should be noted that a data point refers to a single data value collected by the sensor of an IoT terminal at a specific time. For example, the temperature value collected by a temperature sensor at a certain point in time, the pressure value recorded by a pressure sensor, etc. Each sensor may continuously generate multiple data points, which represent different time periods, different sensor locations or different types of measurement values. The importance and changes of each data point are analyzed through the entropy shift index. The system allocates bandwidth resources according to the priority of the data point to ensure that important data points are transmitted first.

[0091] By dynamically allocating queue capacity, we ensure that high-priority data has sufficient queue space to avoid queue congestion caused by data backlog. At the same time, low-priority data can also be processed within a certain range to avoid excessive delays.

[0092] Based on adaptive scheduling, queue monitoring and adjustment, after the queue is initialized, each queue is monitored in real time, and the queue occupancy and data flow status are analyzed. Through the adaptive scheduling mechanism, the capacity and transmission order of each priority queue are dynamically adjusted according to the changes in network bandwidth and data flow. When the occupancy rate of the high priority queue exceeds the preset threshold, When the system automatically expands the capacity of the high priority queue or limits the transmission frequency of the low priority data flow. The queue monitoring algorithm is as follows: ,in, is the queue capacity adjustment coefficient, which is used to control the queue capacity; n is the total capacity of the priority queue, is the smoothing factor, is the adjustment index, which is used to control the sensitivity of dynamic expansion of the queue;

[0093] It should be noted that the smoothing factor is usually used to prevent the denominator from being zero or to avoid extreme data fluctuations from having too great an impact on the calculation results. In order to ensure the robustness of the system, the smoothing factor is usually set to a very small positive number. The specific value of the smoothing factor is usually determined based on the magnitude of the data or the physical limitations of the system; assuming that the temperature data collected by the sensor fluctuates from 0.01 to 1 degree, through the data standardization process, it is observed that some sensor values ​​are close to 0 in extreme cases. In order to avoid the denominator being zero in the entropy offset calculation, the smoothing factor can be set to 0.001 to ensure the computability of the formula in extreme cases without significantly affecting the overall calculation of the data;

[0094] The queue capacity adjustment coefficient determines the sensitivity of queue expansion or reduction. Its value needs to be dynamically adjusted according to the real-time load of the system and the change of network bandwidth. The queue capacity adjustment coefficient can be estimated by the bandwidth usage and queue backlog in historical data, or adjusted through a feedback-based adaptive mechanism. The feedback control method is to monitor the queue occupancy rate in real time. When the queue occupancy rate exceeds a certain threshold, the queue capacity adjustment coefficient is used to expand the queue capacity. Usually, the system will first set an initial value and then adjust it according to feedback. The initial value can be determined through simulation tests. For example, according to historical traffic data and system load characteristics, a medium sensitivity value that can ensure system stability can be selected.

[0095] The regulation index controls the sensitivity and amplitude of adaptive adjustment, affecting the degree of change in queue capacity and priority polling scheduling. The regulation index is usually determined through simulation testing and model optimization. The purpose is to avoid excessive fluctuations while ensuring system sensitivity. By simulating changes in data traffic, the regulation index is adjusted to observe the response of the system under different loads. The value of the regulation index is usually between 1 and 3. When the regulation index increases, the sensitivity of queue expansion and scheduling will decrease, the system will be more stable, but the response speed will be slower; when the regulation index is small, the system will be more sensitive to traffic changes, but the fluctuation will also be greater. It can also be adjusted with the load of the system. When the load is light, the index can be increased to improve system stability. When the load is heavy, the index can be reduced to increase sensitivity.

[0096] For example, in an IoT terminal system with frequent data fluctuations, setting the adjustment index to 2 may be a reasonable initial value, which means that when the queue occupancy exceeds the threshold, the queue capacity is adjusted more slowly, avoiding the waste of resources caused by over-expansion and not completely ignoring low-priority queues. In actual operation, if the data flow fluctuates too much, the adjustment index can be gradually reduced to 1.2 to improve system sensitivity.

[0097] Based on queue management, the system adopts a priority polling scheduling mechanism to regularly check the data flow status in each priority queue and schedule it through a time slice polling algorithm, that is, high-priority data occupies more transmission resources in scheduling, while low-priority data can also get processing opportunities regularly;

[0098] Perform queue priority adjustment based on delay perception, that is, to prevent high-priority data from being delayed due to queue blocking, introduce a queue priority adjustment mechanism based on delay perception, and determine whether the queue priority needs to be adjusted by real-time monitoring of the transmission delay of each data flow. When the preset delay threshold is exceeded, the priority of the queue is automatically increased and more bandwidth resources and queue space are allocated. The delay monitoring algorithm is as follows: ,in, represents the round trip time of the ith queue at time t, is the base round trip time, represents the monitoring time window length of the ith queue or data flow, Determines the time range in which the system observes and calculates the delay of the queue or data flow. By collecting statistics and analyzing the delay changes within this time period, it can be determined whether the queue priority or bandwidth resources need to be adjusted. It is the delay smoothing factor. The delay smoothing factor is determined according to actual needs. Through the delay perception mechanism, the priority of the high-delay queue can be dynamically improved to prevent the key data from affecting the overall transmission efficiency due to delay accumulation. The delay perception algorithm ensures that the key data in the queue can be processed in a timely manner, especially when the network load is heavy;

[0099] To prevent low-priority queues from occupying system resources for a long time, a bandwidth optimization and queue cleaning mechanism is introduced. In this step, the low-priority data that has not been processed for a long time is cleaned up through the bandwidth optimization algorithm to release system resources. The basis for queue cleaning is the entropy offset index and delay value of the data. The cleaning algorithm is as follows: ,when When the value is too large, some low-priority data in the queue will be cleared first to release resources. That is, through bandwidth optimization and queue clearing mechanism, low-priority data that occupies too many resources can be reasonably cleared to ensure that high-priority data obtains sufficient transmission resources and queue space;

[0100] After completing the adaptive queue management and priority polling scheduling, the transmission control of IoT terminal data needs to be further optimized to ensure the reliability, real-time and efficiency of data transmission. In addition, transmission reliability enhancement, packet retransmission mechanism, bandwidth optimization and traffic regulation, congestion control and fault recovery processing can be performed in the future.

[0101] After the priority data transmission is completed, the system strengthens the reliability of the transmission, especially in the environment of network fluctuation. In order to ensure that the key data can be transmitted completely and accurately, the system will use the data packet confirmation mechanism. After each transmission, the receiving end sends a confirmation message (ACK) to the sending end to confirm that the data has been successfully received. If the receiving end does not receive the data packet within the specified time, or the data packet is damaged, the system will trigger the retransmission mechanism to ensure that the data will not be lost;

[0102] When data packet loss or timeout without confirmation is detected, the data packet retransmission mechanism is started. That is, in some cases (such as network congestion or high packet loss rate), the redundancy control mechanism is enabled, such as by adding redundant information or forward error correction technology to improve the data transmission success rate. Critical data will have more retransmission opportunities and redundant protection to ensure its successful transmission even under poor network conditions. During the transmission process, bandwidth is a key resource, especially in concurrent transmission between IoT terminal devices. Bandwidth allocation is dynamically optimized according to the priority of different data and the status of the queue.

[0103] High-priority data will be given priority in bandwidth allocation during transmission, while low-priority data will be transmitted without affecting high-priority data. Bandwidth optimization will be adjusted dynamically according to real-time network conditions and loads to ensure that critical data obtains sufficient bandwidth resources without causing transmission delays due to low-priority data occupying bandwidth. Bandwidth optimization ensures that critical data can still be transmitted first when network resources are limited. At the same time, bandwidth allocation is dynamically adjusted to enable low-priority data to be transmitted without affecting critical data, maximizing the use of existing network resources.

[0104] If a data packet cannot be successfully transmitted after multiple retransmissions, the system will start a fault recovery mechanism. This mechanism will abandon the transmission of the data packet after a certain timeout period and record relevant information for subsequent fault diagnosis and processing. At the same time, it will continue to process other untransmitted priority data to avoid the entire system from stagnating due to the failure of individual data packets to be transmitted. In special circumstances, network failures may be circumvented by trying other transmission paths or methods.

[0105] It should be noted that the thresholds involved in the embodiments may be determined according to specific scenarios and requirements, and are usually adjusted and optimized based on factors such as historical data and real-time data.

[0106] The present invention determines the type, distribution and physical characteristics of sensor data collected through the application scenarios of IoT terminals, formulates the target of data collection and transmission demand assessment, and realizes efficient data collection and scheduling by setting the sampling frequency and acquisition window, adopts asynchronous or synchronous strategies, and performs noise filtering, outlier detection and normalization processing on the data to ensure the integrity and accuracy of the data, thereby ensuring the efficient collection of multi-type sensor data, adapting to the data requirements in different application scenarios, and optimizing the data management and transmission performance of IoT terminal devices;

[0107] The data priority is quantified through the entropy shift index, combined with the dynamic threshold clustering algorithm to identify changes in important events, and the bandwidth resources are dynamically adjusted to ensure the timely transmission of high-priority data. The queue management and scheduling of data streams of different priorities are performed through the adaptive scheduling mechanism and time slice polling algorithm, which optimizes the data transmission efficiency, avoids queue blocking, and improves the data transmission reliability and overall performance in complex network environments.

[0108] Embodiment 2: This embodiment is a system embodiment of embodiment 1, which is used to implement a transmission control method for IoT terminal data introduced in embodiment 1, such as Figure 2 As shown, specifically including:

[0109] The sensor deployment module determines the application scenarios of IoT terminals based on the type, distribution and physical characteristics of the sensor data collected, formulates the target of data collection and the evaluation of transmission requirements, and clarifies the deployment location and function of various sensors;

[0110] The data transmission specification module is used to set the sampling frequency and data acquisition window according to the type of sensor and the data change rate, and adopt asynchronous or synchronous acquisition strategies to perform data acquisition and transmission scheduling, perform noise filtering and outlier detection on the collected data, and perform standardization processing on different types of data;

[0111] The priority quantification module quantifies data priority according to the entropy shift index, identifies important event changes based on the dynamic threshold clustering algorithm, and dynamically adjusts bandwidth resources to determine data transmission priority;

[0112] The transmission control module is used to allocate independent queues to data flows of different priorities based on the priority allocation results, adjust the queue capacity through an adaptive scheduling mechanism, and use a time slice polling algorithm for priority polling to optimize data transmission efficiency and queue management.

[0113] The above formulas are all dimensionless and calculated numerically. Specific dimension removal can be achieved by various means such as standardization, which will not be elaborated here. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0114] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium may be a magnetic medium (e.g., a floppy disk, an ATA hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium may be a solid-state ATA hard disk.

[0115] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0116] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0117] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0118] 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, and may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0119] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0120] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for controlling the transmission of data of an Internet of Things terminal, characterized in that: The steps include: Determine the application scenarios of IoT terminals based on the type, distribution and physical characteristics of the sensor data collected, formulate the data collection goals and transmission demand assessment, and clarify the deployment location and functions of various sensors; According to the type of sensor and the data change rate, set the sampling frequency and data acquisition window, and use asynchronous or synchronous acquisition strategies for data acquisition and transmission scheduling, perform noise filtering and outlier detection on the collected data, and normalize different types of data; Quantify data priority based on entropy shift index, identify important event changes based on dynamic threshold clustering algorithm, and dynamically adjust bandwidth resources to determine data transmission priority; Based on the priority allocation results, independent queues are allocated to data flows of different priorities. The queue capacity is adjusted through an adaptive scheduling mechanism, and a time slice polling algorithm is used for priority polling to optimize data transmission efficiency and queue management. The data priority is quantified according to the entropy shift index, and the changes of important events are identified based on the dynamic threshold clustering algorithm. The bandwidth resources are dynamically adjusted to determine the data transmission priority. The specific process is as follows: Use kernel-based support vector regression to extract features from the collected data, extract the main features from each data stream, and form feature vectors from the main features through nonlinear mapping; Construct the data vector at each time point t ,in, represents the collected value of the i-th sensor at time t, n represents the total number of sensors, and the high-dimensional data entropy is calculated at each time point , the formula is: ,in, It is a probability distribution based on the collected data. It is a regulating factor that controls the fluctuation of data entropy in extreme cases; From the sensor data of IoT terminal devices, construct the data vector at each time point and calculate the high-dimensional data entropy at each time point; Dynamically track and increment the entropy offset, calculate the entropy value of the data, and use the entropy offset formula to calculate the entropy difference between the previous and next time points, and the entropy difference between the previous and next time points t and t+1. The entropy difference represents the change in the entropy value of the data between adjacent time points. The formula is as follows: ,in, is to prevent small values ​​of zero probability from occurring, , They represent the probability distribution of data collected by the i sensors at time t and time t+1 respectively, and the entropy difference represents the change amplitude of the entropy value of the data between adjacent time points; Compare the data entropy of adjacent time points to dynamically detect the uncertainty increment of the data stream. The uncertainty increment indicates the uncertainty change of the data over time. Based on the calculated entropy offset, a dynamic threshold clustering algorithm is used to identify important event changes and determine whether the change requires adjustment of data priority. The entropy offset represents the trend of entropy value changes at adjacent time points. For the identified large entropy shift data points, the importance of the data is quantified and a priority is assigned using the entropy shift index. The entropy shift index represents the degree of fluctuation in the data stream. The entropy shift index calculation formula is as follows: ,in, represents the entropy offset of the ith sensor, is the exponential tuning parameter, is the weight coefficient; Bandwidth resources are allocated to data points according to the calculated entropy shift index. The bandwidth allocation formula is: ,in, is the entropy offset index of the ith sensor, j represents the entropy offset index of the jth sensor, is the bandwidth of the ith data point, n is the number of sensors, is the total available bandwidth, is the magnification factor.

2. The method for controlling the transmission of Internet of Things terminal data according to claim 1, characterized in that: According to the type, distribution and physical characteristics of the sensor data collected, determine the application scenario of the IoT terminal, formulate the target of data collection and transmission demand assessment, and clarify the deployment location and function of various sensors. The specific process is as follows: Determine the distribution density and location of sensors, clarify the types of sensors used by terminal devices, including temperature, humidity, vibration, pressure, and gas concentration sensors, and analyze the deployment of all sensors in IoT terminals; Determine the data type and characteristics of the collection, including numeric, discrete, and symbolic data types; Determine the environmental characteristics, which include the dynamics of the scene and the abnormalities within the scene. The dynamics of the scene include the ambient temperature, humidity, and frequency of equipment use. The abnormalities within the scene include the impact of equipment failure or emergency conditions on data. The application scenarios of IoT terminals are obtained by combining the distribution density and location of sensors, the data types and characteristics collected by sensors, and environmental characteristics; Determine the target data and core data content to be collected and the data transmission requirements.

3. The method for controlling the transmission of Internet of Things terminal data according to claim 2, characterized in that: According to the type of sensor and the data change rate, set the sampling frequency and data acquisition window, and use asynchronous or synchronous acquisition strategies for data acquisition and transmission scheduling, perform noise filtering and outlier detection on the collected data, and normalize different types of data. The specific process is as follows: Determine the data collection frequency of each sensor based on the sensor type, data change rate, and bandwidth resources, and set a data collection window for each sensor; A synchronous collection mechanism is used for time-series correlation data, and an asynchronous collection mechanism is used for independent data sources; After data collection, the collected raw data is preprocessed, and the high-frequency noise signal in the sensor data is detected and filtered out using a noise filtering method based on wavelet transform; Use distribution density estimation technology to detect outliers in high-dimensional data space, mark them, and smooth them through interpolation algorithms; The data were processed using a normalization method based on Laplace regularization.

4. The method for controlling the transmission of Internet of Things terminal data according to claim 3 is characterized in that: Based on the priority allocation results, independent queues are allocated to data flows of different priorities. The queue capacity is adjusted through the adaptive scheduling mechanism, and the time slice polling algorithm is used for priority polling to optimize data transmission efficiency and queue management. The specific process is as follows: Perform queue management and scheduling control on data according to the assigned data priority; Assign independent queues to data flows of different priorities, and initialize each queue according to the priority weight of the data; The queue capacity is adjusted based on the adaptive scheduling mechanism. After the queue is initialized, each queue is monitored in real time to analyze the queue occupancy and data flow status. The capacity and transmission order of each priority queue are dynamically adjusted according to the changes in network bandwidth and data flow through the adaptive scheduling mechanism. When the occupancy rate of the high-priority queue exceeds the preset threshold, the high-priority queue capacity is automatically expanded, or the transmission frequency of the low-priority data flow is limited. Perform queue priority adjustment based on delay perception. Introduce a queue priority adjustment mechanism based on delay perception. By real-time monitoring of the transmission delay of each data flow, determine whether the queue priority needs to be adjusted. When the transmission delay of the queue exceeds the preset delay threshold, the queue priority is automatically increased. Introducing bandwidth optimization and queue clearing mechanisms, clearing low-priority data that has not been processed for a long time through bandwidth optimization algorithms; After completing the adaptive queue management and priority polling scheduling, transmission reliability enhancement, packet retransmission mechanism, bandwidth optimization and flow regulation, congestion control and fault recovery processing are carried out.

5. The method for controlling the transmission of Internet of Things terminal data according to claim 4, characterized in that: After completing the adaptive queue management and priority polling scheduling, the transmission reliability enhancement, packet retransmission mechanism, bandwidth optimization and traffic regulation, congestion control and fault recovery processing are carried out. The specific process is as follows: Use the data packet confirmation mechanism. After each transmission, the receiving end sends a confirmation message to the sending end to confirm that the data has been successfully received. If the receiving end does not receive the data packet within the specified time, or the data packet is damaged, the data packet retransmission mechanism is started. In case of network congestion, the redundancy control mechanism is enabled. If the data packet retransmission fails, the fault recovery mechanism is started, and the data packet transmission is abandoned according to the set timeout period, and the information is recorded.

6. A transmission control system for Internet of Things terminal data, used to implement a transmission control method for Internet of Things terminal data according to any one of claims 1 to 5, characterized in that: include: The sensor deployment module is used to determine the application scenarios of IoT terminals based on the type, distribution and physical characteristics of the sensor data collected, formulate the target of data collection and transmission demand assessment, and clarify the deployment location and function of various sensors; The data transmission specification module is used to set the sampling frequency and data acquisition window according to the type of sensor and the data change rate, and adopt asynchronous or synchronous acquisition strategies to perform data acquisition and transmission scheduling, perform noise filtering and outlier detection on the collected data, and perform standardization processing on different types of data; The priority quantification module quantifies data priority according to the entropy shift index, identifies important event changes based on the dynamic threshold clustering algorithm, and dynamically adjusts bandwidth resources to determine data transmission priority; The transmission control module is used to allocate independent queues to data flows of different priorities based on the priority allocation results, adjust the queue capacity through an adaptive scheduling mechanism, and use a time slice polling algorithm for priority polling to optimize data transmission efficiency and queue management.

Images