A timing data management method of an edge computing gateway

Abstract

The application discloses a kind of edge computing gateway's time series data management method, it is related to edge computing and industrial internet of things technical field, including the following steps: in the preset fixed time period, the communication bandwidth occupancy of edge computing gateway, buffer remaining rate and scheduling thread occupancy are recorded respectively, and resource use original data matrix covering all time points in fixed time period is constructed.The application generates scheduling resource abnormal occupancy index dynamically by periodically monitoring resource use state and fusing high-priority task scheduling performance, realizes intelligent perception and scheduling optimization of resource pressure of edge computing gateway.When abnormal index rises, buffer redistribution and low task data compression are automatically triggered by system, data writing and scheduling real-time performance of critical task are preferentially guaranteed, task starvation and data loss problem are effectively avoided, and stability and response capability of system in high-pressure environment are improved.

Description

A method for managing time-series data in an edge computing gateway Technical Field

[0001] This invention relates to the fields of edge computing and industrial Internet of Things (IIoT) technology, and specifically to a time-series data management method for an edge computing gateway. Background Technology

[0002] Time-series data management of edge computing gateways refers to a comprehensive management method within an edge computing architecture that efficiently stores, organizes, processes, and schedules continuous data (i.e., time-series data) with timestamp characteristics collected from various sensors, control devices, or terminal systems locally on the gateway side. This method not only supports large-scale, high-frequency data writing and low-latency data reading, but also leverages edge computing capabilities for real-time preprocessing, cleaning, compression, and aggregation of data to improve data utilization efficiency. Furthermore, it dynamically determines when to upload data to the cloud or central system based on network status, business strategies, or computing load, thereby achieving data autonomy, bandwidth savings, and accelerated response at the edge. This approach is widely used in scenarios such as industrial sites, smart manufacturing, energy monitoring, and traffic sensing.

[0003] Existing technologies have the following shortcomings: In scenarios involving multi-task concurrency and mixed priority queue scheduling, existing edge computing gateways suffer from rigid scheduling policies and a lack of dynamic awareness in resource allocation mechanisms. Specifically, when low-priority tasks (such as periodic data uploads and background data cleaning) continuously occupy core scheduling resources (including but not limited to communication bandwidth, buffer space, and scheduling threads) for a certain period, high-priority critical time-series tasks (such as security alarm information and device anomaly events) may not be able to obtain timely scheduling permissions, resulting in scheduling delays or even "task starvation." This problem is difficult to expose during routine system testing and functional verification phases because test environments often have sufficient bandwidth resources and low task density, making it difficult to simulate the sudden concurrency of multi-source data or bandwidth-limited conditions in actual operation. Once the aforementioned scheduling resources are continuously occupied, critical event data will fail to be uploaded to the platform in a timely manner, leading to the failure of early warning mechanisms, delayed security control responses, and potentially irreversible consequences such as serious production interruptions, equipment damage, and even personal injury.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a time-series data management method for edge computing gateways to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a time-series data management method for an edge computing gateway, comprising the following steps:

[0007] Within a preset fixed time period, the communication bandwidth utilization rate, buffer remaining rate, and scheduling thread utilization rate of the edge computing gateway are recorded respectively, and a raw data matrix of resource usage covering all time points within the fixed time period is constructed.

[0008] Based on the original data matrix of resource usage, the usage fluctuation slope and resource usage density of communication bandwidth utilization, buffer remaining rate and scheduling thread utilization are calculated in a fixed time period. Weighting factors are configured according to the task priority level corresponding to each resource indicator, and the resource indicators are merged to generate a single resource anomaly characterization value.

[0009] Within a fixed time period, extract the execution status information of all tasks set as high priority, identify the number of tasks whose waiting time exceeds the preset response threshold, and calculate their cumulative blocking time to generate a high priority task blocking coefficient.

[0010] The abnormal resource representation value and the high-priority task obstruction coefficient are mapped and fused into a nonlinear function to obtain the abnormal resource occupancy index of the current scheduling cycle, which serves as the abnormal perception index of the dynamic scheduling system.

[0011] When the abnormal occupancy index of scheduling resources exceeds the set threshold, the buffer resource reallocation mechanism is triggered to dynamically adjust the buffer quota ratio corresponding to different priority tasks to ensure the continuity of data writing and uploading operations for critical tasks. At the same time, data fragments of low-priority tasks are compressed to retain only their statistical summary information, thereby reducing buffer resource pressure.

[0012] Preferably, within a preset fixed time period, the steps for recording the communication bandwidth utilization rate, buffer remaining rate, and scheduling thread utilization rate of the edge computing gateway, and constructing the original resource usage data matrix are as follows:

[0013] To analyze the communication behavior of the edge computing gateway during operation, a fixed sampling time interval is set, and the real-time data transmission rate of the communication channel is continuously sampled. The communication bandwidth utilization sequence is calculated by the ratio of the sampling points to the theoretical maximum bandwidth.

[0014] At the time point synchronized with communication sampling, the usage status of the buffers allocated to each task type in the edge computing gateway is periodically read, and the ratio of the remaining buffer capacity to the total capacity at each sampling time is calculated to form a buffer remaining rate sequence.

[0015] The ratio between the active number of task scheduling threads inside the edge computing gateway and the total thread pool capacity is statistically analyzed to obtain the scheduling thread occupancy rate sequence. The communication bandwidth occupancy rate sequence, buffer remaining rate sequence and scheduling thread occupancy rate sequence are time-aligned and a resource usage raw data matrix is ​​constructed according to a unified timestamp.

[0016] Preferably, the specific steps for fusing resource indicators to generate a single resource anomaly representation value based on the original resource usage data matrix are as follows:

[0017] The time-series occupancy data corresponding to the communication bandwidth occupancy rate, buffer remaining rate, and scheduling thread occupancy rate in the original data matrix of resource usage are processed. A continuous variation model is constructed by curve fitting within a fixed time period. The mean value of the derivative is calculated by taking the derivative of the fitted curve at each sampling point to obtain the usage fluctuation slope of each resource type.

[0018] Based on the calculation of fluctuation slope, the time-series occupancy data within the same period is used to calculate the cumulative time of communication bandwidth occupancy, buffer remaining rate and scheduling thread occupancy in the region close to the resource saturation threshold. The cumulative time is then compared with the total sampling time to calculate the resource occupancy density.

[0019] Based on the task priority system preset by the edge computing gateway, corresponding weight factors are configured for communication bandwidth utilization, buffer remaining rate and scheduling thread utilization. The fluctuation slope and resource utilization density are normalized according to the weighted fusion rule, and finally synthesized into a single resource anomaly characterization value, which is used to quantitatively describe the severity of overall resource anomaly within a fixed time period.

[0020] Preferably, the specific steps for configuring weight factors based on the edge computing gateway's preset task priority level system and fusing them to generate resource anomaly representation values ​​are as follows:

[0021] Based on the predefined task priority system within the edge computing gateway, the sensitivity of various high-priority tasks to communication bandwidth utilization, buffer remaining rate, and scheduling thread utilization is quantitatively evaluated. Weight factor values ​​for each type of resource are set according to the sensitivity, and a mapping function relationship between task priority and resource usage indicators is established.

[0022] The usage fluctuation slope and resource usage density data, which are calculated separately for communication bandwidth utilization, buffer remaining rate, and scheduling thread utilization over a fixed time period, are uniformly normalized.

[0023] A one-to-one correspondence is established between the preset weighting factors and the normalized communication bandwidth utilization rate, buffer remaining rate, and scheduling thread utilization rate indicators. The normalization of various resources is calculated by weighting the fluctuation slope and the normalized resource utilization density in sequence, and finally a single resource anomaly characterization value is generated.

[0024] Preferably, the specific steps for extracting the execution status information of high-priority tasks and generating the obstruction coefficient of high-priority tasks within a fixed time period are as follows:

[0025] Within a fixed time period set by the edge computing gateway, based on the task scheduling queue log and task execution status record, all tasks marked as high priority are filtered out, their waiting time data from entering the scheduling queue to starting execution is extracted, and compared with the preset response time threshold. The set of high priority tasks whose waiting time exceeds the response time threshold is identified, and a high priority task response timeout list is established.

[0026] For the identified high-priority task response timeout list, iterate through the scheduling waiting interval of each task, count its actual waiting time and accumulate it to calculate the total time that all high-priority tasks are blocked in the scheduling process within a fixed time period, and at the same time record the number of blocked tasks.

[0027] The high-priority task blocking coefficient is generated by dividing the total duration of high-priority tasks blocked by the total number of high-priority tasks in the current period, and further combining this with the average waiting tolerance time of a unit task.

[0028] Preferably, the specific steps for mapping and fusing the resource anomaly representation value with the high-priority task obstruction coefficient into a nonlinear function to generate the scheduling resource anomaly occupancy index are as follows:

[0029] After a fixed time period ends, the scheduling system of the edge computing gateway obtains the calculated resource anomaly characterization value and the high-priority task obstruction coefficient obtained based on the high-priority task scheduling status analysis, and combines the two to form a joint index input vector.

[0030] The scheduling system calls a pre-trained nonlinear function model and uses the resource anomaly representation value and the high-priority task obstruction coefficient in the joint index input vector as the function independent variables for calculation.

[0031] The fusion result output by the nonlinear function model is defined as the scheduling resource anomaly occupancy index for the current scheduling cycle, and the scheduling resource anomaly index is registered as a dynamic anomaly perception indicator in the scheduling control module of the edge computing gateway.

[0032] Preferably, when the abnormal occupancy index of scheduling resources exceeds a set threshold, a buffer resource reallocation mechanism is triggered to dynamically adjust the buffer quota ratio corresponding to different priority tasks. At the same time, the specific steps for performing compression processing on data fragments of low-priority tasks are as follows:

[0033] Based on the generated abnormal scheduling resource occupancy index and the set reference threshold, the buffer pressure index for the current period is calculated using the following expression:

[0034]

[0035] In the formula, It is a buffer stress index. It is the index of abnormal resource occupancy during scheduling, θ ref It is a reference threshold for the abnormal resource occupancy index.

[0036] After completing the buffer pressure index After calculation, based on the buffer pressure index The target buffer quota ratios for high-priority and low-priority tasks are calculated using the following expression:

[0037]

[0038] In the formula, β high It is the target buffer quota ratio for high-priority tasks, β base It is the base quota ratio for high-priority tasks, δ is the quota adjustment range coefficient, tanh is the hyperbolic tangent function, and β is the base quota ratio for high-priority tasks. low It is the buffer quota ratio for low-priority tasks;

[0039] Based on the acquired quota ratio and the available buffer capacity, the write space that high-priority tasks can occupy and the write space that low-priority tasks can reserve are determined. The calculation expression is as follows:

[0040]

[0041] In the formula, C high This is the capacity of the high-priority task buffer, C avail η is the currently available buffer capacity, η is the compression retention factor, and C is the current available buffer capacity. low This is the buffer capacity after compression of low-priority tasks.

[0042] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0043] This invention achieves efficient linkage between resource pressure perception and task scheduling behavior by periodically monitoring and modeling the usage status of core resources such as communication bandwidth, buffers, and scheduling threads in an edge computing gateway, and dynamically generating a scheduling resource anomaly occupancy index based on the scheduling execution status of high-priority tasks. The method employs a nonlinear mapping mechanism to fuse resource anomaly representation values ​​with task obstruction coefficients, improving the system's sensitivity and accuracy in responding to sudden resource shortages and task blocking states. When the anomaly index rises, the system can trigger real-time buffer resource reallocation and data compression strategies to release resource bottlenecks, ensuring the continuity of high-priority task writes and real-time scheduling, thereby effectively preventing problems such as task starvation, data loss, and warning failures. This mechanism significantly enhances the scheduling stability, task reliability, and resource scheduling intelligence of the edge computing system under high concurrency and complex operating conditions, demonstrating good engineering adaptability and practical application value. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0045] Figure 1 is a flowchart of a time-series data management method for an edge computing gateway according to the present invention. Detailed Implementation

[0046] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0047] This invention provides a time-series data management method for an edge computing gateway, as shown in Figure 1, comprising the following steps:

[0048] Within a preset fixed time period, the communication bandwidth utilization rate, buffer remaining rate, and scheduling thread utilization rate of the edge computing gateway are recorded respectively. A resource usage raw data matrix covering all time points within the fixed time period is constructed. The resource usage raw data matrix is ​​used to represent the time-series occupancy status of various core scheduling resources within this period.

[0049] Within a preset fixed time period, the communication bandwidth utilization, buffer remaining rate, and scheduling thread utilization of the edge computing gateway are recorded, and a raw resource usage data matrix is ​​constructed. Its core function is to provide the edge computing system with a complete, detailed, and quantifiable dynamic resource profile, providing a data foundation for subsequent scheduling strategy optimization and anomaly identification. As a computing node executing multiple tasks concurrently, the edge computing gateway must simultaneously handle various operations such as upload tasks, processing tasks, and alarm responses. These tasks compete for limited resources such as communication, caching, and computing threads during execution. By continuously recording the status of these three key resources within a fixed period, a high-time-resolution resource usage time series can be formed. The raw data matrix, constructed after time alignment, reflects the actual fluctuation trajectory of resource utilization within that period. For example, the changing trend of communication bandwidth utilization can reveal whether sudden data uploads are crowding out normal upload channels, buffer remaining rate can warn whether cache space is nearing exhaustion, and scheduling thread utilization can reflect whether the system's processing capacity is saturated. This original data matrix not only enables a quantitative expression of the current resource status, but also provides an essential data foundation for subsequent construction of resource anomaly representation values, analysis of task blocking root causes, and triggering of dynamic scheduling control mechanisms. This enables the system to have intelligent decision-making capabilities based on objective resource status perception, and enhances the scheduling robustness and real-time assurance capabilities of multi-task collaborative processing in edge computing environments.

[0050] Within a preset fixed time period, the communication bandwidth utilization, buffer remaining rate, and scheduling thread utilization of the edge computing gateway are recorded respectively. The steps to construct the original data matrix of resource usage are as follows:

[0051] To analyze the communication behavior of the edge computing gateway during operation, a fixed sampling time interval is set, and the real-time data transmission rate of the communication channel is continuously sampled. The communication bandwidth utilization sequence is calculated by the ratio of the sampling points to the theoretical maximum bandwidth, which reflects the dynamic change process of communication resources within the fixed time period.

[0052] During the operation of the edge computing gateway, to accurately capture the dynamic status of communication resources, a fixed sampling time interval (e.g., once per second) is first set to continuously collect the real-time data transmission rate of each communication interface of the gateway. The ratio of the actual transmission rate at each moment to the theoretical maximum bandwidth corresponding to that interface is calculated, thus forming a time series of communication bandwidth utilization. This series not only accurately reflects the load change trend of the communication link within the period, but also provides data basis for subsequent identification of abnormal resource states such as sudden bandwidth congestion and bandwidth saturation, and is one of the key inputs in the scheduling resource awareness mechanism.

[0053] At the time point synchronized with communication sampling, the usage status of the buffers allocated to each task type in the edge computing gateway is periodically read, the ratio of the remaining capacity of the buffer to the total capacity at each sampling time is calculated, a buffer remaining rate sequence is formed, and smoothing is performed in combination with historical capacity change trends to improve the accuracy of abnormal change identification.

[0054] To achieve real-time monitoring of memory resource usage, the edge computing gateway periodically acquires buffer usage information for various tasks (such as uploading, processing, and cleaning) at the time point synchronized with communication data sampling. It calculates the ratio of remaining capacity to total capacity at each sampling point, generating a buffer availability sequence. This sequence reflects the dynamic availability of cache resources within the scheduling cycle. Combined with smoothing algorithms such as moving averages, it can identify trends of gradual buffer depletion or sudden, sharp declines, providing crucial early warning for preventing high-priority tasks from being unable to write data due to cache exhaustion.

[0055] The ratio between the active number of task scheduling threads inside the edge computing gateway and the total thread pool capacity is statistically analyzed to obtain the scheduling thread occupancy rate sequence. The communication bandwidth occupancy rate sequence, buffer remaining rate sequence and scheduling thread occupancy rate sequence are time-aligned and constructed according to a unified timestamp to form a resource usage raw data matrix. This forms a complete data infrastructure that reflects the temporal occupancy status of three key resources: communication bandwidth, buffer, and scheduling threads within the period.

[0056] To assess the availability of processing capacity, at the same time points as the previous two types of data sampling, the system statistically analyzes the ratio between the number of currently active scheduling threads and the total capacity of the thread pool, forming a time series of scheduling thread occupancy. This series reflects the thread resource pressure on the gateway when processing task scheduling. By aligning it with the communication bandwidth occupancy series and the buffer remaining rate series using a unified timestamp, a raw resource usage data matrix is ​​constructed. This matrix, as a complete temporal resource occupancy mapping structure, supports subsequent resource fusion modeling, anomaly detection, and scheduling control decisions, and is a fundamental core component for realizing dynamic resource regulation logic.

[0057] Based on the original data matrix of resource usage, the usage fluctuation slope and resource usage density of communication bandwidth utilization, buffer remaining rate and scheduling thread utilization are calculated in a fixed time period. Weighting factors are configured according to the task priority level corresponding to each resource indicator, and the resource indicators are integrated to generate a single resource anomaly characterization value, which is used to quantify the overall degree of abnormal resource usage in this period.

[0058] The core function of this step is to transform the multi-dimensional and heterogeneous resource usage status information of the edge computing gateway within a fixed time period into a unified and quantifiable comprehensive anomaly indicator, thereby providing the scheduling system with accurate and real-time resource anomaly perception capabilities. By calculating the usage fluctuation slope and resource occupancy density of three key resource indicators—communication bandwidth utilization, buffer remaining rate, and scheduling thread utilization—it can not only capture the dynamic fluctuation trend during resource usage but also reflect the persistence of high-pressure states. This dual-dimensional analysis mode breaks through the traditional static judgment method based on a single utilization rate or average value, and can more sensitively identify the risk of instantaneous or continuous resource congestion. Simultaneously, in the fusion phase, by introducing weighting factors set by the task priority level system, the model fully considers the actual impact weight of various resources in the system on the scheduling success rate of high-priority tasks. The performance of resource-sensitive tasks in the system will therefore be amplified during the fusion process, improving the overall scheduling model's responsiveness to critical task assurance capabilities. Ultimately, the single resource anomaly representation value generated through weighted normalization serves as the core quantitative indicator for the scheduling system to perceive resource status throughout the entire cycle. This value can be directly used to determine whether control strategies such as resource reallocation, task preemption scheduling, or cache adjustment are necessary, effectively improving the stability and real-time performance of the edge computing system in complex multi-task concurrent scenarios. This step establishes a logical closed loop from multi-source resource status perception to scheduling control response, forming a fundamental link in the system's intelligent and refined resource scheduling management.

[0059] The specific steps for fusing resource indicators to generate a single resource anomaly representation value based on the original resource usage data matrix are as follows:

[0060] The time-series occupancy data corresponding to the communication bandwidth occupancy rate, buffer remaining rate, and scheduling thread occupancy rate in the original resource usage data matrix are processed. Continuous variation models are constructed using curve fitting within a fixed time period. The mean value of the derivatives of the fitted curves at each sampling point is calculated to obtain the usage fluctuation slope of each resource type. This slope reflects the rate of change and the degree of fluctuation of the usage of communication bandwidth, buffer, and scheduling thread resources within the period, providing a quantitative basis for evaluating the dynamic stability of system resources.

[0061] In the raw resource usage data matrix collected by the edge computing gateway, communication bandwidth utilization, buffer remaining rate, and scheduling thread utilization are all presented in time series form. To obtain the usage fluctuation trend of various resources within a fixed time period, curve fitting processing is first performed on the time series data of each type of resource. For example, multinomial regression or spline interpolation is used to construct a continuous change curve of each resource over time. Subsequently, the curve is differentiated to extract the instantaneous rate of change at each time point, and the mean value of the derivative value within the period is calculated as the usage fluctuation slope of that type of resource. This fluctuation slope reflects the rate of change and fluctuation intensity of resource utilization within the period, and can be used to identify whether there are sudden, high-frequency dynamic fluctuations in resources, providing a basis for the system to judge scheduling stability.

[0062] Based on the calculation of fluctuation slope, the time-series occupancy data within the same period is further used to calculate the cumulative time of communication bandwidth occupancy, buffer remaining rate, and scheduling thread occupancy in the region close to the resource saturation threshold. The cumulative time is then compared with the total sampling time to calculate the resource occupancy density. This density is used to characterize the degree of persistence of communication resources, cache resources, and scheduling thread resources in a high-pressure operating state within the period, thereby supplementing the time stability characteristics that characterize abnormal resource states.

[0063] After obtaining the usage fluctuation slope, the system continues to calculate the cumulative duration for which communication bandwidth utilization, buffer availability, and scheduling thread utilization are near saturation based on the resource usage time series within the same time period. Specifically, reasonable high-load thresholds are set (e.g., communication utilization above 85%, buffer availability below 15%, and thread utilization above 90%). At each sampling point, it is determined whether a high-pressure state has been entered, and the duration of this state is accumulated. Finally, this is compared with the total period time to obtain a resource utilization density index. This index reflects whether resources are under high stress for an extended period within the period, continuously quantifying system operating pressure and serving as an important characteristic for assessing resource overload risk and scheduling bottlenecks.

[0064] Based on the task priority level system preset by the edge computing gateway, corresponding weight factors are configured for communication bandwidth utilization, buffer remaining rate and scheduling thread utilization. The fluctuation slope and resource utilization density are normalized according to the weighted fusion rule and finally synthesized into a single resource anomaly characterization value, which is used to quantitatively describe the severity of overall resource anomaly within a fixed time period. This serves as an important judgment indicator for the scheduling system to subsequently implement control strategies such as resource reallocation and task priority preemption.

[0065] After calculating the fluctuation slope and resource occupancy density, to enhance the accuracy of anomaly identification, it is necessary to adjust for differences in resource sensitivity based on task priority. According to the edge computing gateway's settings for different task priority levels, corresponding weighting factors are configured for communication bandwidth occupancy, buffer remaining rate, and scheduling thread occupancy. For example, critical tasks may rely more on communication and thread resources, thus receiving higher weighting values. After normalizing the fluctuation slope and occupancy density of each resource indicator, they are merged into a single resource anomaly representation value according to weighting rules. This representation value comprehensively considers the resource change rate and high-pressure persistence, and quantitatively reflects the differences in resource dependence based on task importance. It is a key input parameter for the scheduling system to determine whether resources are abnormally occupied in the current cycle, playing a decisive role in triggering dynamic adjustment strategies.

[0066] The specific steps for configuring weight factors and generating resource anomaly representation values ​​based on the edge computing gateway's preset task priority system are as follows:

[0067] Based on the predefined task priority system within the edge computing gateway, the sensitivity of various high-priority tasks to communication bandwidth utilization, buffer remaining rate, and scheduling thread utilization is quantitatively evaluated. A weight factor value is set for each type of resource according to the sensitivity. The higher the weight factor value, the more critical the resource is to the scheduling timeliness and task completion of high-priority tasks. A mapping function relationship between task priority and resource usage indicators is established to provide parameter basis for task demand guidance for subsequent fusion models.

[0068] In edge computing gateway systems, different types of tasks exhibit significant differences in their resource dependencies. For example, security alarm information upload tasks are highly sensitive to communication bandwidth and scheduling thread resources, while periodic log cleaning tasks primarily rely on buffer resources. To accurately reflect this difference, a task priority system must first be established, categorizing all system tasks into multiple priority levels based on business urgency and data timeliness. Subsequently, based on the resource call behavior of high-priority tasks during historical execution, statistical analysis is used to clarify the dependence of each type of task on communication bandwidth utilization, buffer remaining rate, and scheduling thread utilization, and corresponding weight factors are configured for each of the three resource types. This step ensures that subsequent resource anomaly detection reflects the actual urgency of task scheduling, transitioning from resource-centric to business-demand-oriented modeling, and laying the foundation for ultimately generating integrated indicators with scheduling decision significance.

[0069] A "predefined task priority system" refers to a set of hierarchical rules pre-defined by system designers or operations personnel in an edge computing gateway system. These rules are based on the importance, urgency, and timeliness of tasks within the business process and are used to classify and prioritize all tasks running in the system. This system typically includes several priority levels, such as "high priority," "medium priority," "low priority," or a more granular five-level classification system. Each task level corresponds to a different level of scheduling assurance. Its core function is to provide decision-making support for the scheduling system, enabling it to prioritize the allocation of scheduling resources for high-priority tasks when resources are scarce or concurrent tasks conflict. This ensures that critical business processes (such as device alarms, status synchronization, and fault reporting) have real-time processing capabilities, thereby improving the overall system's robustness and responsiveness.

[0070] When determining the sensitivity of high-priority tasks to three types of resources—communication bandwidth utilization, buffer availability, and scheduling thread utilization—a dual approach based on runtime data analysis and business logic correlation can be used for quantitative evaluation. First, from the perspective of runtime data, historical scheduling logs and task performance statistics can be used to quantify the changes in task completion time under different resource load conditions. For example, when communication bandwidth utilization exceeds 80%, the completion delay of a certain type of task increases significantly, indicating its sensitivity to communication resources. Similarly, by analyzing the correlation between task execution failure rate and buffer availability or thread blocking, a resource-task sensitivity model can be established.

[0071] Secondly, modeling is performed from a business logic perspective, combining the functional objectives of tasks to determine their dependencies on specific resources. For example, event alerting tasks typically require millisecond-level real-time performance and are extremely sensitive to scheduling thread occupancy; while periodic log archiving tasks rely more on the continuous writing capability of the buffer and are more dependent on the cache remaining rate. Through these two types of analysis, a sensitivity coefficient between 0 and 1 can be assigned to each type of resource, and then mapped to a weighting factor based on the task hierarchy structure. This factor is used in the subsequent weighted fusion model for resource anomaly representation, thereby ensuring that the impact of critical resource status changes on critical tasks is accurately quantified and responded to.

[0072] The usage fluctuation slope and resource occupancy density data obtained by calculating the communication bandwidth occupancy rate, buffer remaining rate, and scheduling thread occupancy rate within a fixed time period are uniformly normalized. The minimum-maximum linear transformation method is used to map all the original data to the same standard numerical range, eliminating the differences in numerical scale and dimension of each resource indicator. This ensures that all indicators have an equivalent basis for comparison during the weighted fusion process and avoids an inappropriate dominant effect in the calculation due to the original magnitude of a certain resource being too large.

[0073] Because communication bandwidth utilization, buffer availability, and scheduling thread utilization exhibit significant differences in their numerical range, dimensions, and distribution characteristics across the two dimensions of fluctuation slope and resource utilization density, directly weighting and fusing these metrics can lead to results biased towards metrics with larger absolute values, impacting model fairness. Therefore, it is necessary to normalize the fluctuation slope and utilization density for each resource type separately. Specifically, a minimum-maximum linear transformation is used, mapping the minimum value of each metric within the current period to 0, the maximum value to 1, and other values ​​to the corresponding interval in a linear proportion. Through this normalization operation, different resource metrics are unified to the same standardized numerical scale, ensuring consistency and comparability in the weighted fusion process. The core function of this step is to eliminate interference from the original metric numerical characteristics, ensuring that the fusion result truly reflects the actual degree of metric change and the abnormal characteristics that the scheduling system needs to focus on.

[0074] A one-to-one correspondence is established between the preset weighting factors and the normalized communication bandwidth utilization rate, buffer remaining rate, and scheduling thread utilization rate indicators. According to the weighted average fusion algorithm, the normalization of various resources is calculated by using the fluctuation slope and the normalized resource utilization density in sequence to generate a single resource anomaly characterization value. This resource anomaly characterization value is used to quantitatively characterize the severity of overall resource anomaly within the current fixed time period, and serves as an important basis for the edge computing gateway scheduling system to decide whether to execute scheduling control strategies such as resource reallocation, task preemption, or buffer expansion in real-time scenarios.

[0075] After normalizing resource metrics, the system establishes a one-to-one correspondence between the weighting factors set in the first step and the various metrics processed in the second step. Then, using a weighted average fusion method, the normalized fluctuation slopes of communication bandwidth utilization, buffer remaining rate, and scheduling thread utilization are weighted and superimposed with the normalized resource utilization density, ultimately fusing them into a single resource anomaly representation value. This anomaly representation value reflects the overall abnormal utilization level of the system's core resources within the current scheduling cycle as a continuous numerical value. The key function of this step is to compress multi-dimensional resource status data into a single indicator that can be accessed in real time and used for decision-making reference. This allows the scheduling system to quickly determine whether there are resource bottlenecks or scheduling obstacles, and trigger dynamic control strategies such as resource reallocation and task priority adjustment accordingly, effectively improving the real-time response capability and task assurance level of the edge computing system.

[0076] Within a fixed time period, extract the execution status information of all tasks set as high priority, identify the number of tasks whose waiting time exceeds the preset response threshold, and calculate their cumulative blocking time to generate a high priority task blocking coefficient, which is used to represent the degree of scheduling impact on high priority tasks due to abnormal resource occupation.

[0077] The purpose of this step is to dynamically evaluate the system's ability to guarantee critical tasks under resource constraints or concurrent conflicts by performing refined quantitative analysis of the scheduling and execution status of high-priority tasks in the edge computing gateway within a specific time period. This provides real-time and reliable performance feedback for optimizing resource scheduling strategies. In edge computing environments, critical tasks such as device alarm uploading and fault event reporting are typically set as high priority, and their scheduling timeliness directly affects the system's security and responsiveness. However, in scenarios with multiple concurrent tasks and intense resource contention, even with priority settings, if the resource allocation mechanism cannot respond in a timely manner, high-priority tasks may still experience long waiting times, affecting event processing efficiency. Through this step, the system first identifies which high-priority tasks in the current fixed period have actual waiting times exceeding preset response thresholds, then calculates their total blocking time, and generates a high-priority task blocking coefficient based on the number of blocked tasks. This coefficient, as a continuous and quantifiable scheduling performance indicator, objectively reflects the scheduling latency of critical tasks in the real-world operating environment and reveals the impact of abnormal resource consumption on core business tasks. This metric allows the scheduling system to determine in real time whether to adjust the current resource allocation strategy, trigger task preemption, or expand buffer resources, thereby enabling task scheduling to evolve from "rule-driven" to "feedback-driven" and improving the stability, response speed, and business continuity of the edge computing system under complex operating conditions.

[0078] The specific steps for extracting the execution status information of high-priority tasks and generating the obstruction coefficient of high-priority tasks within a fixed time period are as follows:

[0079] Within a fixed time period set by the edge computing gateway, based on the task scheduling queue log and task execution status record, all tasks marked as high priority are filtered out, their waiting time data from entering the scheduling queue to starting execution is extracted, and compared with the preset response time threshold. The set of high priority tasks whose waiting time exceeds the response time threshold is identified, and a high priority task response timeout list is established.

[0080] Within a fixed time period set by the edge computing gateway, the system first extracts scheduling logs and task execution status data from the task scheduling management module, filtering out a set of tasks marked as high priority. Then, for each high-priority task, it obtains the timestamp of its entry into the scheduling queue and the timestamp of its actual start execution, calculates the task's scheduling waiting time, and compares it with a preset response time threshold in the system. All tasks whose waiting time exceeds the response threshold are added to the high-priority task response timeout list.

[0081] The purpose of this step is to identify abnormal tasks from all high-priority tasks that have delayed response due to untimely resource scheduling, providing target samples for subsequent statistical analysis of the scheduling system's inadequacy in ensuring critical tasks.

[0082] For the identified high-priority task response timeout list, iterate through the scheduling waiting interval of each task, count its actual waiting time and accumulate it to calculate the total time that all high-priority tasks are blocked in the scheduling process within a fixed time period, and at the same time record the number of blocked tasks to reflect the scheduling system's resource response capability to high-priority tasks.

[0083] After obtaining the list of high-priority tasks that have timed out, the system enters the statistics phase. It accumulates the scheduling wait time for each task in the list to obtain the total blocking time experienced by all high-priority tasks that timed out within the current time period. Simultaneously, the system counts the number of timed-out tasks, which, together with the total blocking time, reflects the scheduling system's failure to meet the requirements for high-priority tasks within that period.

[0084] The purpose of this step is to assess the severity of scheduling bottlenecks from two dimensions: time and number of tasks. It reveals whether the current resource scheduling strategy of the system has problems with the untimely processing of high-priority tasks, and provides necessary data for generating scheduling performance indicators that reflect the task guarantee capability.

[0085] The total duration of high-priority tasks being blocked is divided by the total number of high-priority tasks in the current period, and further combined with the average waiting tolerance time of a unit task to generate a high-priority task blocking coefficient. This coefficient serves as a dynamic quantitative indicator of the edge computing gateway scheduling performance, used to measure the actual performance of the current resource scheduling strategy in ensuring the critical task guarantee under high pressure conditions, and as a reference for subsequent decision-making on whether to trigger resource reallocation or priority strategy adjustment.

[0086] After calculating the blocking duration and the number of blocked tasks, the system uses both as basic variables to calculate the high-priority task blocking coefficient. Specifically, the total blocking duration of high-priority tasks is divided by the total number of high-priority tasks within the period to obtain the average blocking duration per task. This average duration is then compared to the system-defined average response tolerance time per task to generate the high-priority task blocking coefficient. This coefficient, as a quantifiable scheduling performance indicator, dynamically reflects the processing delay of high-priority tasks by the current scheduling mechanism. It is a crucial basis for determining whether to adjust the scheduling strategy, perform resource reallocation, or trigger critical task preemption logic.

[0087] This step transforms the scheduling performance of high-priority tasks into a numerical model that can be used for decision feedback, which helps improve the agility of the scheduling system and the ability to ensure task response.

[0088] The abnormal resource representation value and the high-priority task obstruction coefficient are mapped and fused into a nonlinear function to obtain the abnormal resource occupancy index of the current scheduling cycle. This index serves as an abnormality perception indicator for the dynamic scheduling system, used to determine in real time whether resource reallocation or task preemption scheduling strategy needs to be executed, so as to ensure the real-time scheduling of high-priority tasks under resource shortage conditions.

[0089] The purpose of this step is to enable edge computing gateways to achieve real-time perception and early warning control of scheduling risks in multi-task concurrency and resource contention environments by constructing a highly integrated and responsive anomaly indicator system. Specifically, this step maps and fuses two key indicators—resource anomaly representation values ​​and high-priority task obstruction coefficients—into a nonlinear function model to generate a scheduling resource anomaly occupancy index. This index not only reflects the overall pressure status of core resources such as communication bandwidth, buffers, and scheduling threads within the current scheduling cycle but also comprehensively considers the degree of interference experienced by high-priority tasks during actual scheduling. Compared to traditional single-indicator monitoring methods, this integrated index uses nonlinear functions to model the complex coupling relationships between input variables, possessing enhanced response characteristics to extreme states. It can amplify scheduling risk signals under critical conditions such as sudden resource shortages and dense task obstruction, improving the system's sensitivity to identifying scheduling anomaly trends. By using this index as an anomaly perception indicator for the dynamic scheduling system, the scheduling engine can automatically trigger a series of control strategies, such as resource reallocation, task preemption, and buffer adjustment, when anomaly values ​​exceed preset thresholds, effectively preventing high-priority tasks from being unable to be processed in a timely manner due to resource saturation. This mechanism realizes a logical closed loop from "resource status monitoring" to "scheduling behavior driven", enabling the system to have intelligent scheduling capabilities with autonomous judgment and rapid response, significantly improving the real-time performance of task execution, the robustness of the scheduling system, and the overall reliability of edge computing services.

[0090] The specific steps for mapping and fusing the resource anomaly representation value with the high-priority task obstruction coefficient into a nonlinear function to generate the scheduling resource anomaly occupancy index are as follows:

[0091] After a fixed time period ends, the scheduling system of the edge computing gateway obtains the calculated resource anomaly characterization value and the high-priority task obstruction coefficient obtained based on the high-priority task scheduling status analysis. The two are combined to form a joint index input vector, which is used as an input variable to express the composite characteristics of the system's resource usage status and key task scheduling performance in the current period, providing multi-dimensional basic data for subsequent execution of nonlinear fusion mapping.

[0092] After a fixed time period ends, the edge computing gateway scheduling system first extracts key input variables from two sub-modules: one is a resource anomaly representation value calculated by fusing communication bandwidth utilization, buffer remaining rate, and scheduling thread utilization; the other is a high-priority task obstruction coefficient derived from the high-priority task execution status analysis. The system combines these two independent indicators into a joint input structure in the form of a two-dimensional vector, constructing input samples based on timestamp alignment. This step integrates and expresses the two key dimensions of "resource status" and "task performance" in the scheduling system, forming a multi-dimensional feature vector that simultaneously reflects resource load pressure and task execution anomalies within the current scheduling period. This provides a data support foundation for subsequent nonlinear model processing and improves the completeness and expressive power of the system's modeling.

[0093] The scheduling system calls a pre-trained nonlinear function model, which uses the resource anomaly representation value and the high-priority task obstruction coefficient in the joint index input vector as function independent variables for calculation. The nonlinear function model adopts an activation function structure with adjustable parameters, such as a sigmoid function, an exponential growth function, or a logistic regression function, to model the nonlinear coupling relationship between input variables and to amplify the response to critical states such as sudden resource anomalies or continuous obstruction of key tasks, thereby improving the scheduling system's ability to sensitively identify abnormal situations.

[0094] After the input vector is constructed, the system feeds it into a pre-trained nonlinear function model. This model can employ activation functions commonly used in neural networks, such as sigmoid, exponential, or ReLU functions. By introducing adjustable weight parameters and nonlinear factors, it models the complex coupling relationship between resource anomaly representations and high-priority task obstruction coefficients. The function calculation process not only considers the nonlinear amplification effect between variables but also enhances the response to the deterioration trend of scheduling performance under extreme conditions. This makes the output value more sensitive to the intensity of anomalies when the system is in a state of sudden resource occupancy or intensive obstruction of critical tasks. The purpose of this step is to construct a "discriminant function" for scheduling performance anomalies through refined modeling, significantly improving the system's ability to identify critical states and providing a dynamic decision-making basis for scheduling control strategies.

[0095] A pre-trained nonlinear function model refers to a mathematical function structure constructed and optimized using machine learning or fitting algorithms based on a large amount of historical operational data, task scheduling records, and resource usage status before the deployment of an edge computing gateway. This model captures the complex nonlinear relationship between resource status indicators and scheduling performance. Unlike traditional static rules, this model can not only characterize the interaction effect between abnormal resource values ​​and the obstruction coefficient of high-priority tasks, but also enhance the response to scheduling bottlenecks under critical conditions. For example, during model training, the system identifies the nonlinear coupling trend between sudden increases in resource occupancy and sharp increases in task waiting times, and amplifies the sensitivity to anomalies by adjusting function parameters. Its role is to enable the edge computing gateway's scheduling system to have highly flexible anomaly recognition capabilities. When input features are in complex intersection intervals, the model can output a more discriminative and judgmental anomaly index, significantly improving the system's perception accuracy and response speed to scheduling anomaly trends, and providing reliable data support for intelligent, real-time resource management strategies.

[0096] The fusion result output by the nonlinear function model is defined as the scheduling resource anomaly occupancy index for the current scheduling cycle. The scheduling resource anomaly index is registered as a dynamic anomaly perception indicator in the scheduling control module of the edge computing gateway. This indicator is used to determine in real time whether the system is in an abnormal state of resource scheduling performance. When the index exceeds a set threshold, the scheduling system can automatically trigger a resource reallocation mechanism, a task priority enhancement strategy, or a buffer dynamic adjustment logic. This realizes the transformation from a fixed rule-driven approach to a dynamic resource scheduling control mechanism based on intelligent perception feedback, effectively improving the real-time performance of critical task scheduling and the overall stability of the system operation.

[0097] After the nonlinear function model is calculated, the system defines its output value as the scheduling resource anomaly occupancy index and registers this index in the scheduling engine as a dynamic anomaly detection indicator. The scheduling system continuously monitors the changes in this index and compares it with the set anomaly threshold in real time. Once the index exceeds the threshold, it immediately triggers the built-in resource reallocation mechanism, high-priority task promotion strategy, or dynamic cache expansion logic to ensure the scheduling continuity of critical tasks and the integrity of data transmission. The purpose of this step is to compress the complex abnormal states of resources and tasks into an operable and controllable scheduling indicator, enabling the scheduling system to have a closed-loop capability from "state awareness" to "adaptive response," truly realizing the transformation from traditional static strategies to intelligent perception-driven scheduling mechanisms, and comprehensively improving the operational intelligence and stability of edge computing systems.

[0098] When the abnormal occupancy index of scheduling resources exceeds the set threshold, the buffer resource reallocation mechanism is triggered to dynamically adjust the buffer quota ratio corresponding to different priority tasks. By releasing the buffer space occupied by low-priority tasks and temporarily expanding the buffer capacity of high-priority tasks, the continuity of data writing and uploading operations of critical tasks is ensured. At the same time, data fragments of low-priority tasks are compressed to retain only their statistical summary information, reducing buffer resource pressure and preventing high-priority tasks from queuing or losing data due to insufficient buffer resources, thereby improving the scheduling efficiency and response reliability of the system.

[0099] The purpose of this step is to provide a dynamic buffer control mechanism with real-time response and resource awareness capabilities in the complex operating environment where edge computing gateways face high scheduling resource constraints. This ensures the continuity of data processing and system stability for high-priority critical tasks under unexpected circumstances. Specifically, when the abnormal occupancy index of scheduling resources exceeds the system's set threshold, the system automatically determines that the current resource scheduling status has encountered a potential bottleneck, which may prevent critical tasks from writing or uploading data in a timely manner, thus triggering a buffer resource reallocation mechanism. This mechanism dynamically adjusts the buffer quotas for high-priority and low-priority tasks based on the scheduling pressure intensity. By releasing the buffer space already occupied by low-priority tasks, it temporarily expands the cache write capacity of high-priority tasks, ensuring that their critical data will not encounter write interruptions or queuing blockages during periods of resource scarcity. At the same time, to prevent overall system memory overflow, a compression strategy is implemented for low-priority tasks, retaining only critical statistical summary information, thereby reducing resource consumption while preserving business integrity as much as possible. This mechanism not only improves the system's responsiveness to critical tasks, but also enhances the flexibility of resource scheduling and the hierarchical fault tolerance of task processing. It effectively avoids serious consequences such as data loss, upload delays, or alarm failures caused by resource scheduling imbalances, and ultimately achieves high reliability, high real-time performance, and business continuity when multiple tasks are processed concurrently in an edge computing environment.

[0100] When the abnormal occupancy index of scheduling resources exceeds a set threshold, a buffer resource reallocation mechanism is triggered to dynamically adjust the buffer quota ratios corresponding to tasks of different priorities. At the same time, the specific steps for compressing data fragments of low-priority tasks are as follows:

[0101] Based on the generated abnormal scheduling resource occupancy index and the set reference threshold, the buffer pressure index for the current period is calculated using the following expression:

[0102]

[0103] In the formula, This is the buffer pressure index, representing the level of buffer resource pressure caused by abnormal resource scheduling in the current scheduling cycle. It serves as a direct driving parameter for the subsequent dynamic reallocation mechanism of buffer resources. The larger the value, the greater the resource scheduling pressure on the buffer. This is the scheduling resource anomaly index, a fusion indicator that comprehensively reflects the status of key system resources (communication bandwidth utilization, buffer remaining rate, scheduling thread utilization) and the degree of obstruction of high-priority tasks within the current scheduling cycle. A higher value indicates a more severe scheduling anomaly, requiring stronger intervention for system resource reallocation. θ ref It is a reference threshold for the abnormal resource occupancy index.

[0104] This step enables the effective transformation of scheduling awareness indicators into buffer control signals.

[0105] After completing the buffer pressure index After calculation, based on the buffer pressure index The target buffer quota ratios for high-priority and low-priority tasks are calculated using the following expression:

[0106]

[0107] In the formula, β high This is the target buffer quota ratio for high-priority tasks, representing the proportion of buffer resources the system should allocate to high-priority tasks in the current cycle. It is one of the final output parameters of this step and directly affects the data write guarantee level for high-priority tasks. β base This is the base quota ratio for high-priority tasks, a preset buffer allocation baseline value for the system under normal scheduling conditions, typically between 0.5 and 0.7. It represents the default level of protection for high-priority tasks when there are no anomalies. δ is the quota adjustment range coefficient, representing the maximum dynamic increase in the buffer ratio for high-priority tasks allowed when an abnormal resource state is detected; the larger the value, the stronger the system's tilt capability under pressure, typically ranging from 0.1 to 0.4. tanh is the hyperbolic tangent function, and β... low This is the buffer quota ratio for low-priority tasks, representing the proportion of buffer space reserved by the system for low-priority tasks in the current cycle; its value is determined inversely by the quota for high-priority tasks, ensuring that the total quota sums to 1.

[0108] The hyperbolic tangent function (Tanh function) is a common sigmoid nonlinear function. Its output value ranges from -1 to 1 across its entire domain and is symmetric about the origin. As the input approaches 0, the tanh output approaches 0; as the input approaches positive infinity, the output approaches 1; and as it approaches negative infinity, the output approaches -1. This smooth, nonlinear growth characteristic makes it widely used in machine learning and dynamic scheduling systems to amplify the changing trends of input variables while suppressing drastic fluctuations caused by extreme values. The main function of this approach is to normalize and map the buffer pressure exponent (which can be an arbitrarily large positive value) to a finite interval (0 to 1), while maintaining a sensitive response to abnormal pressure changes. This allows the system to quickly increase the buffer quota ratio for high-priority tasks when it senses a sudden increase in buffer pressure; and when the pressure is low or stable, the scheduling system can maintain a stable state, preventing frequent resource fluctuations caused by minor changes. This achieves the triple objectives of smooth adjustment of the scheduling strategy, nonlinear amplification, and boundary control.

[0109] Through this step, the system can achieve dynamic, tilted scheduling and allocation of resources among tasks of different priorities without interrupting the original task logic.

[0110] Based on the acquired quota ratio and the available buffer capacity, the write space that high-priority tasks can occupy and the write space that low-priority tasks can reserve are determined. The calculation expression is as follows:

[0111]

[0112] In the formula, C high This refers to the high-priority task buffer capacity, representing the buffer capacity dynamically allocated to high-priority tasks during the current scheduling cycle, measured in bytes (Byte) or megabytes (MB). This value directly affects the write continuity and upload reliability of high-priority tasks under resource-constrained environments. avail η is the currently available buffer capacity, representing the physical or logical buffer space still available for allocation in the system. It is used to allocate write areas for tasks of different priorities within a scheduling cycle. This value is typically updated in real-time by the memory management subsystem or cache control module. η is the compression retention factor, representing the application strength of the compression mechanism. It controls the compression retention ratio of low-priority task data, and its value range is typically [0,1]. η=1 indicates no compression and full retention, while η=1 indicates only 20% of the digest information is retained. This parameter is the core control variable of the system's embedded compression strategy and can be dynamically configured based on task type and data characteristics. lowThis is the compressed buffer capacity for low-priority tasks, representing the effective write buffer capacity actually allocated to low-priority tasks in the current cycle. Under resource-constrained conditions, this capacity is usually limited to the minimum reserved space after compression, used to store task statistical summaries or partially simplified data to reduce system load.

[0113] This step ensures the continuity of writing to high-priority tasks and significantly reduces the resource consumption of the buffer without completely discarding low-priority data. It effectively avoids task queuing and blocking and the loss of critical data, further improving the real-time performance and stability of the scheduling system.

[0114] This invention achieves efficient linkage between resource pressure perception and task scheduling behavior by periodically monitoring and modeling the usage status of core resources such as communication bandwidth, buffers, and scheduling threads in an edge computing gateway, and dynamically generating a scheduling resource anomaly occupancy index based on the scheduling execution status of high-priority tasks. The method employs a nonlinear mapping mechanism to fuse resource anomaly representation values ​​with task obstruction coefficients, improving the system's sensitivity and accuracy in responding to sudden resource shortages and task blocking states. When the anomaly index rises, the system can trigger real-time buffer resource reallocation and data compression strategies to release resource bottlenecks, ensuring the continuity of high-priority task writes and real-time scheduling, thereby effectively preventing problems such as task starvation, data loss, and warning failures. This mechanism significantly enhances the scheduling stability, task reliability, and resource scheduling intelligence of the edge computing system under high concurrency and complex operating conditions, demonstrating good engineering adaptability and practical application value.

[0115] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0116] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0117] It should be noted that, in this document, the use of relational terms such as "first" and "second" is merely for distinguishing one entity or operation from another, and does not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0118] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply 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 this application.

[0119] Those skilled in the art will recognize that the units and algorithm steps of the various examples 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 implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0120] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0121] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0122] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0123] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0124] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for managing time-series data in an edge computing gateway, characterized in that, Includes the following steps: Within a preset fixed time period, the communication bandwidth utilization rate, buffer remaining rate, and scheduling thread utilization rate of the edge computing gateway are recorded respectively, and a raw data matrix of resource usage covering all time points within the fixed time period is constructed. Based on the raw data matrix of resource usage, the usage fluctuation slope and resource utilization density of communication bandwidth utilization rate, buffer remaining rate, and scheduling thread utilization rate within the fixed time period are calculated respectively. Weighting factors are configured according to the task priority level corresponding to each resource indicator, and the resource indicators are merged to generate a single resource anomaly characterization value. Within the fixed time period, the execution status information of all tasks set as high priority is extracted, the number of tasks whose waiting time exceeds the preset response threshold is identified, and their cumulative blocking time is counted to generate a high-priority task blocking coefficient. The abnormal resource representation value is mapped and fused with the obstruction coefficient of high-priority tasks into a nonlinear function to obtain the abnormal scheduling resource occupancy index of the current scheduling cycle, which serves as the abnormality detection indicator of the dynamic scheduling system. When the abnormal scheduling resource occupancy index exceeds a set threshold, a buffer resource reallocation mechanism is triggered to dynamically adjust the buffer quota ratio corresponding to different priority tasks to ensure the continuity of data writing and uploading operations for critical tasks. At the same time, data fragments of low-priority tasks are compressed to retain only their statistical summary information, thereby reducing buffer resource pressure.

2. The time-series data management method for an edge computing gateway according to claim 1, characterized in that, Within a preset fixed time period, the following steps are taken to record the communication bandwidth utilization rate, buffer remaining rate, and scheduling thread utilization rate of the edge computing gateway, and to construct the original data matrix of resource usage: For the communication behavior of the edge computing gateway during operation, a fixed sampling time interval is set, and the real-time data transmission rate of the communication channel is continuously sampled. The communication bandwidth utilization rate sequence is calculated by the ratio of the sampling points to the theoretical maximum bandwidth. At the time point synchronized with the communication sampling, the buffer usage status allocated to each task type in the edge computing gateway is periodically read, and the ratio of the remaining buffer capacity to the total capacity at each sampling time is calculated to form a buffer remaining rate sequence. The ratio between the active number of task scheduling threads inside the edge computing gateway and the total thread pool capacity is statistically analyzed to obtain the scheduling thread occupancy rate sequence. The communication bandwidth occupancy rate sequence, buffer remaining rate sequence and scheduling thread occupancy rate sequence are time-aligned and a resource usage raw data matrix is ​​constructed according to a unified timestamp.

3. The time-series data management method for an edge computing gateway according to claim 1, characterized in that, Based on the original resource usage data matrix, the following steps are taken to calculate the usage fluctuation slope and resource occupancy density of communication bandwidth utilization, buffer remaining rate, and scheduling thread utilization within a fixed time period, and to integrate the resource indicators to generate a single resource anomaly characterization value according to the task priority level corresponding to each resource indicator: The time-series occupancy data corresponding to communication bandwidth utilization, buffer remaining rate, and scheduling thread utilization in the original resource usage data matrix are processed. A continuous variation model is constructed using curve fitting within a fixed time period. The mean of the derivatives is calculated after taking the derivatives of the fitted curves at each sampling point to obtain the usage fluctuation slope of each resource type. Based on the fluctuation slope calculation, the time-series occupancy data within the same period is used to calculate the cumulative time of communication bandwidth utilization, buffer remaining rate, and scheduling thread utilization in the region close to the resource saturation threshold. The cumulative time is then compared with the total sampling time to calculate the resource occupancy density. Based on the task priority system preset by the edge computing gateway, corresponding weight factors are configured for communication bandwidth utilization, buffer remaining rate and scheduling thread utilization. The fluctuation slope and resource utilization density are normalized according to the weighted fusion rule, and finally a single resource anomaly characterization value is generated to quantitatively describe the severity of overall resource anomaly within a fixed time period.

4. The time-series data management method for an edge computing gateway according to claim 3, characterized in that, Based on the pre-defined task priority system of the edge computing gateway, corresponding weight factors are configured for communication bandwidth utilization, buffer remaining rate, and scheduling thread utilization. The usage fluctuation slope and resource utilization density are then normalized according to a weighted fusion rule to finally generate a single resource anomaly representation value. The specific steps are as follows: Based on the pre-defined task priority system within the edge computing gateway, the sensitivity of various high-priority tasks to communication bandwidth utilization, buffer remaining rate, and scheduling thread utilization is quantitatively evaluated. Weight factor values ​​for each type of resource are set according to the sensitivity, and a mapping function relationship between task priority and resource usage indicators is established. The usage fluctuation slope and resource utilization density data calculated for communication bandwidth utilization, buffer remaining rate, and scheduling thread utilization within a fixed time period are uniformly normalized. Based on the one-to-one correspondence established between the preset weighting factors and the normalized usage fluctuation slope and normalized resource occupancy density, the normalized usage fluctuation slope and normalized resource occupancy density of various resources are calculated and weighted sequentially to finally generate a single resource anomaly characterization value.

5. The time-series data management method for an edge computing gateway according to claim 1, characterized in that, The specific steps for extracting the execution status information of high-priority tasks and generating the high-priority task obstruction coefficient within a fixed time period are as follows: Within the fixed time period set by the edge computing gateway, based on the task scheduling queue log and task execution status record, all tasks marked as high priority are filtered out, and their waiting time data from entering the scheduling queue to starting execution is extracted and compared with a preset response time threshold. The set of high-priority tasks whose waiting time exceeds the response time threshold is identified, and a high-priority task response timeout list is established. For the identified high-priority task response timeout list, the scheduling waiting interval of each task is traversed, its actual waiting time is counted and accumulated, and the total time during which all high-priority tasks are blocked in the scheduling process within the fixed time period is calculated, and the number of blocked tasks is recorded at the same time. The total time during which high-priority tasks are blocked is divided by the total number of high-priority tasks in the current period, and further combined with the average waiting tolerance time of a unit task, the high-priority task obstruction coefficient is generated.

6. The time-series data management method for an edge computing gateway according to claim 1, characterized in that, The resource anomaly characterization value and the high-priority task obstruction coefficient are mapped and fused into a nonlinear function to obtain the scheduling resource anomaly occupancy index for the current scheduling period, which serves as the anomaly perception index of the dynamic scheduling system. The specific steps are as follows: After a fixed time period ends, the scheduling system of the edge computing gateway obtains the calculated resource anomaly characterization value and the high-priority task obstruction coefficient obtained based on the high-priority task scheduling status analysis, and combines the two to form a joint index input vector. The scheduling system calls a pre-trained nonlinear function model and uses the resource anomaly representation value and the high-priority task obstruction coefficient in the joint index input vector as the function independent variables for calculation. The fusion result output by the nonlinear function model is defined as the scheduling resource anomaly occupancy index for the current scheduling cycle, and the scheduling resource anomaly index is registered as a dynamic anomaly perception indicator in the scheduling control module of the edge computing gateway.

7. The time-series data management method for an edge computing gateway according to claim 1, characterized in that, When the abnormal scheduling resource occupancy index exceeds a set threshold, a buffer resource reallocation mechanism is triggered to dynamically adjust the buffer quota ratios for tasks of different priorities. Simultaneously, data fragments from low-priority tasks are compressed. The specific steps are as follows: Based on the generated abnormal scheduling resource occupancy index and the set reference threshold, the buffer pressure index for the current period is calculated using the following expression: In the formula, It is a buffer stress index. It is a resource utilization index for scheduling. It is a reference threshold for the abnormal resource occupancy index. After completing the buffer pressure index After calculation, based on the buffer pressure index The target buffer quota ratios for high-priority and low-priority tasks are calculated using the following expression: In the formula, It is the target buffer quota ratio for high-priority tasks. This is the base quota ratio for high-priority tasks. It is the quota adjustment range coefficient. It is the hyperbolic tangent function. This represents the buffer quota ratio for low-priority tasks. Based on the acquired quota ratio and the available buffer capacity, the write space that high-priority tasks can occupy and the write space that low-priority tasks can reserve are determined. The calculation expression is as follows: In the formula, This is the capacity of the high-priority task buffer. This is the currently available buffer capacity. It is the compression retention factor. This is the buffer capacity after compression of low-priority tasks.

Images