Low-delay cross-end interaction network construction method for disaster prevention and reduction emergency drill

Abstract

The application provides a low-delay cross-terminal interaction network construction method for disaster prevention and reduction emergency drills, and relates to the technical field of data processing. The method comprises the following steps: performing data type identification processing according to original drill data from each drill terminal; performing independent cache and timestamp recording processing to generate timestamp marked data with queue sequence information; performing delay sensitivity calculation processing according to the timestamp marked data to generate priority scheduling data for data processing sequence control; performing time slice allocation and delay offset verification processing according to the priority scheduling data to generate effective to-be-processed data corrected by time slices; performing instant transparent processing according to the effective to-be-processed data to generate instant transparent synchronization data; and performing cross-terminal pushing processing according to the instant transparent synchronization data to form a low-delay cross-terminal state update result. The application improves the delay of the cross-terminal interaction network construction method for disaster prevention and reduction emergency drills.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method for constructing a low-latency cross-terminal interactive network for disaster prevention and mitigation emergency drills. Background Technology

[0002] Currently, disaster prevention and mitigation emergency drill systems typically rely on a "central server-multiple terminals" architecture to achieve cross-terminal interaction. A typical process involves various drill terminals (such as mobile devices, PCs, and command screens) periodically reporting their status, such as location, action commands, and equipment readings, to the server at a fixed frame rate. The server then aggregates the data from each terminal and synchronously pushes the processing results to all participating terminals to form a unified virtual drill environment. To ensure data consistency across different terminals, existing technologies often employ WebSocket long connections or real-time transmission based on UDP hole punching. Server-side queue caching and timing correction modules sort, filter, and merge multi-source data to achieve cross-terminal interaction.

[0003] In multi-person collaborative emergency drills, such as simulating a "smoke spread in a subway station," the movement of drill participants, data from environmental sensors (such as smoke concentration and temperature), and dispatch instructions from the command center need to be synchronized to all terminals within hundreds of milliseconds. However, under the existing central server-side architecture, when a certain type of data (such as smoke sensor readings) suddenly increases in a short period, the server-side queue caching module experiences instantaneous backlog, forcing subsequent "personnel location" data to wait for the server to complete the sorting and merging of the previous batch of data. For example, in a sudden situation where smoke sensors increase from 1 to 20 readings per second, the server-side sorting module will prioritize processing the densely arriving sensor data, causing personnel location data to be unable to be broadcast in time within 300–500 ms. As a result, the location seen by the command center will deviate significantly from the actual actions of the drill participants, leading to misinterpretation of drill instructions and failing to meet the requirements of low-latency, cross-terminal synchronization in emergency drills. Summary of the Invention

[0004] The purpose of this invention is to provide a method for constructing a low-latency cross-terminal interactive network for disaster prevention and mitigation emergency drills, aiming to solve the problems mentioned in the background art.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows: A method for constructing a low-latency cross-platform interactive network for disaster prevention and mitigation emergency drills, the method comprising: Based on the raw exercise data from each exercise terminal, data type identification processing is performed. By identifying whether it belongs to personnel location data or environmental sensor node data, a classified data stream with type labels is generated. Independent caching and timestamp recording are performed based on the classified data stream. By establishing independent queues in their respective data cache units and recording their arrival time sequence, timestamp-marked data with queue sequence information is generated. Delay sensitivity calculation is performed based on timestamp-marked data. Priority scheduling data for data processing order control is generated by analyzing queue length changes and arrival time distribution. Based on priority scheduling data, time slice allocation and delay offset verification are performed. By allocating processing time slices to the corresponding data channels and deleting data that exceeds the preset offset threshold, valid data to be processed after time slice correction is generated. Real-time transparent transmission processing is performed on the valid unprocessed data. By bypassing the cross-type sorting or merging process for high-priority data and generating a synchronization package based on its independent timestamp record table, real-time transparent transmission synchronization data is generated. Cross-terminal push processing is performed based on real-time transparent transmission and synchronization data. By pushing the latest interaction status to each exercise terminal, a low-latency cross-terminal status update result is formed.

[0006] The above-described solution of the present invention has at least the following beneficial effects: First, by identifying the data types of the raw training data from each training terminal, and forming a categorized data stream with type labels by combining personnel location data and environmental sensor node data, different types of data are separated before entering the server-side processing link. This avoids the mutual crowding phenomenon caused by the mixing of multiple data sources into the same queue in existing technologies, and provides clear data boundaries for subsequent low-latency scheduling.

[0007] Secondly, by establishing independent cache queues and recording independent timestamps for each type of data stream, timestamp-marked data is formed, enabling the server to monitor the arrival rhythm and queue changes of each type of data in real time. Compared to existing technologies that only perform time-series correction during the unified sorting stage, this "first divide into queues, then mark the time sequence" approach allows for accurate location of sudden pressure surges in the corresponding channels, thus providing a basis for dynamically adjusting the processing order.

[0008] Furthermore, by generating priority scheduling data based on queue length changes and arrival distribution using timestamp-marked data, the server can identify the impact of short-term surges in environmental sensor node data on different channels and correspondingly increase the processing priority of the personnel location data channel. In this way, even in situations like a sudden increase in smoke sensor data from low frequency to high frequency, personnel location data will not be blocked in the sorting and merging process for an extended period, helping to shorten waiting times in critical states.

[0009] Furthermore, by using priority-driven time slice allocation and delay offset verification, old data exceeding the preset offset threshold is removed at the beginning of each processing time slice. This reduces the situation where outdated data occupies processing resources, ensuring that the data entering the current round of processing remains timely. It avoids the accumulation of invalid delays caused by "old data still being sorted and merged" after queue backlog in existing technologies, thereby maintaining the real-time nature of cross-end updates.

[0010] Finally, by performing real-time transparent transmission on high-priority data among the valid pending data, synchronization packets are directly generated based on their independent timestamps, bypassing cross-type sorting or merging processes. This allows critical data such as personnel locations to complete cross-terminal broadcasting via a shorter processing path. Combined with cross-terminal push processing, terminals can still receive timely updates on location and interaction status during periods of high-frequency environmental data reporting, thereby reducing the risk of discrepancies between the location seen by the command center and actual actions, and meeting the requirements of disaster prevention and mitigation emergency drills for low latency and cross-terminal consistency. Attached Figure Description

[0011] Figure 1 This is a flowchart of a method for constructing a low-latency cross-terminal interactive network for disaster prevention and mitigation emergency drills, provided by an embodiment of the present invention. Detailed Implementation

[0012] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0013] like Figure 1 As shown, embodiments of the present invention propose a method for constructing a low-latency cross-terminal interactive network for disaster prevention and mitigation emergency drills, the method comprising: Based on the raw exercise data from each exercise terminal, data type identification processing is performed. By identifying whether it belongs to personnel location data or environmental sensor node data, a classified data stream with type labels is generated. Independent caching and timestamp recording are performed based on the classified data stream. By establishing independent queues in their respective data cache units and recording their arrival time sequence, timestamp-marked data with queue sequence information is generated. Delay sensitivity calculation is performed based on timestamp-marked data. Priority scheduling data for data processing order control is generated by analyzing queue length changes and arrival time distribution. Based on priority scheduling data, time slice allocation and delay offset verification are performed. By allocating processing time slices to the corresponding data channels and deleting data that exceeds the preset offset threshold, valid data to be processed after time slice correction is generated. Real-time transparent transmission processing is performed on the valid unprocessed data. By bypassing the cross-type sorting or merging process for high-priority data and generating a synchronization package based on its independent timestamp record table, real-time transparent transmission synchronization data is generated. Cross-terminal push processing is performed based on real-time transparent transmission and synchronization data. By pushing the latest interaction status to each exercise terminal, a low-latency cross-terminal status update result is formed.

[0014] In this embodiment of the invention, by introducing the identification and separation of classified data streams into the training network, different types of data have clear attribute labels before entering the processing link, thereby avoiding the situation where multiple types of data are mixed and piled up at the entry point. Since personnel location data and environmental sensor node data have different characteristics, processing them separately enables subsequent processes to be independently scheduled according to their respective characteristics, which helps to reduce the possibility of data waiting at the entry stage.

[0015] The categorized data is stored in independent data cache units, each with its own timestamp, ensuring the data has an identifiable state in both the sequence and time dimensions. This caching method accurately tracks the cumulative quantity and arrival rhythm of each data category, providing a reliable foundation for subsequent urgency analysis. When a large influx of data occurs within a short period, the arrival time sequence information formed in the cache accurately reflects the queue's changing trends, facilitating the timely identification of potential delay risks.

[0016] By analyzing timestamped data, and identifying changes in queue length and data arrival distribution characteristics, scheduling results reflecting the urgency of various data types can be generated. These scheduling results express the appropriate processing order for different types of data within the current cycle, allowing subsequent processing flows to prioritize content more likely to affect the overall synchronization of the exercise, thereby reducing the possibility of interaction interference between different types of data due to shared processing resources.

[0017] When allocating data processing windows based on scheduling results, setting multiple processing time slices and cleaning up data exceeding the delay offset threshold at the beginning of each time slice can reduce the consumption of processing resources by outdated data. The corrected data better reflects the current exercise status, ensuring that the processed content is valuable and effective, thus improving overall data responsiveness.

[0018] The corrected data is then transmitted in real time, allowing data that meets priority processing criteria to directly generate synchronization packets for cross-device updates without waiting for the cross-type data sorting process. This reduces additional latency caused by internal merging processes, enabling terminals to obtain synchronized content in a shorter time, thereby enhancing the continuity of interaction among exercise participants.

[0019] The cross-device push process sends synchronization packets to all participating terminals, enabling each terminal to receive the update results promptly. This mechanism maintains synchronization across different terminals during the exercise, allowing changes in participant behavior to be reflected in the overall exercise scenario within a short time, thus improving the overall collaborative experience of the exercise.

[0020] For example, in a simulated subway station smoke diffusion scenario, environmental sensor nodes may continuously report changes in smoke concentration within a short period of time, while the timeliness of personnel location data is more crucial for dispatch instructions. Implementing this method identifies sudden increases in environmental data by monitoring queue changes, automatically prioritizing personnel location data, and directly generating synchronization packets and pushing them to the terminal when the data meets the pass-through conditions. This allows command personnel to promptly grasp the location of personnel, thereby reducing instruction delays caused by data accumulation.

[0021] In a preferred embodiment of the present invention, the method for setting the preset offset threshold specifically includes: First, during the deployment phase, collect the update cycle requirements of different data types in typical exercises. For example, personnel location data usually requires a shorter validity period, while environmental sensor node data allows for a relatively longer validity period. Secondly, based on the above update cycle requirements, determine the maximum acceptable lag time range for each type of data, record this range as a candidate threshold interval and store it as a threshold configuration relationship. The form of the configuration table can refer to the expiration time setting table of the existing message queue. Secondly, during trial runs or historical drill replays, observe the actual effects under the candidate threshold range, such as whether the drill situation can still be maintained after old data is removed, and whether key data is deleted prematurely, and make fine adjustments to the candidate threshold range accordingly. Finally, the fine-tuned threshold range is solidified as a preset offset threshold, and the corresponding threshold is obtained at runtime by matching the threshold configuration relationship according to the data type. This ensures that the offset threshold meets the real-time requirements while avoiding information loss caused by excessive culling.

[0022] In a preferred embodiment of the present invention, cross-terminal push processing is performed based on real-time transparent transmission synchronization data. By pushing the latest interaction status to each training terminal, a low-latency cross-terminal status update result is formed, specifically including: First, the real-time transparent transmission synchronization data is written to the push cache area on the server side, and a unified packet format for cross-terminal transmission is generated. This packet format is compatible with the existing formats used for multi-terminal message broadcasting, allowing the receiving terminal to read the internal fields without additional parsing logic.

[0023] Secondly, by calling existing long-connection push frameworks on the server side, such as bidirectional connections based on WebSocket or lightweight broadcasts based on UDP, the packet content is delivered to each training terminal with established connections. Each terminal has already established a basic channel for maintaining a connection with the server at the beginning of the training, so synchronization packets can be directly injected into the existing link without re-establishing a connection.

[0024] Secondly, upon receiving the synchronization packet, each training terminal reads the timestamp and data fields within it and refreshes the interface according to its own display logic. For example, mobile terminals can update personnel locations on the map, while large-screen terminals can refresh the status of key nodes in the situation map.

[0025] Finally, by setting a lightweight acknowledgment mechanism in the push link, the server can determine whether synchronization is complete based on the acknowledgment message returned by the terminal. In the event of significant terminal latency or failure to receive a push notification, the server can resend the synchronization packet according to retry strategies commonly used in existing technologies to ensure the continuity of the exercise process.

[0026] In a preferred embodiment of the present invention, delay sensitivity calculation is performed based on timestamp data, and priority scheduling data for data processing order control is generated by analyzing queue length changes and arrival time distribution, including: The queue growth identification process is performed on the timestamp-marked data. By comparing the changes in the number of arrivals over multiple consecutive time periods, growth analysis results representing the queue growth trend are generated. Based on the growth analysis results, arrival interval fluctuation analysis is performed. By identifying whether the arrival interval shows a concentrated shortening phenomenon in a short period, a surge identification result is generated to indicate the surge trend. Based on the results of the surge identification, sensitivity levels are classified. Multiple sensitivity levels are divided according to the strength of the growth trend and the degree of the surge, generating sensitivity level content to characterize the urgency of processing each data channel. Priority mapping is performed based on the sensitivity level content. By mapping higher sensitivity levels to higher processing priorities, priority scheduling data for controlling the order of data processing is generated.

[0027] In this embodiment of the invention, by performing queue growth identification, arrival interval fluctuation analysis, sensitivity level classification, and priority mapping on timestamped data, the data processing workflow can acquire a more refined ability to determine urgency. When data accumulates rapidly within a short period or the arrival interval continuously shortens, this method can promptly identify potential surges, thereby adjusting the processing order to allocate processing resources to data that requires immediate processing. By combining growth trends and the degree of surges during the sensitivity classification process, the scheduling results are more closely aligned with the actual load conditions, reducing the likelihood of misjudgments or delayed adjustments. The final priority scheduling results are more consistent with the current exercise rhythm, reducing processing deviations caused by differences in data types and making the overall processing actions more real-time.

[0028] In a preferred embodiment of the present invention, queue growth identification processing is performed on the timestamp-marked data. By comparing the changes in the number of arrivals over multiple consecutive time periods, growth analysis results characterizing the queue growth trend are generated, specifically including: First, the timestamp data is divided into multiple equal time segments according to its recording time. These time segments can be determined based on the actual amount of data in the exercise, for example, by dividing them into time slices at the level of hundreds of milliseconds or seconds. This step is consistent with the "time window counting method" used in existing technologies for traffic statistics.

[0029] Secondly, the number of data arrivals is counted within each time period to obtain a continuous sequence of arrivals over a period of time. Existing message processing middleware widely adopts similar statistical methods, such as message rate calculation based on sliding windows.

[0030] Secondly, by comparing the changes in quantity between adjacent time periods, we can identify whether the queue is showing an increasing trend. For example, when the number of arrivals in a certain time period is significantly higher than in the previous time period, we can determine that the queue is in an accumulation state. This judgment is described by the textual rule of "whether the change in quantity is continuously increasing," without the need to introduce mathematical formulas.

[0031] Finally, the aforementioned trends are recorded as growth analysis results. These results are stored as a separate field and can be used for subsequent sensitivity assessments, enabling timely adjustments to the processing order when environmental data experiences a sudden surge.

[0032] In a preferred embodiment of the present invention, priority mapping is performed based on the sensitivity level content. By mapping higher sensitivity levels to higher processing priorities, priority scheduling data for data processing order control is generated, specifically including: First, multiple levels are set for sensitivity levels, for example, gradually increasing from lower to higher levels, with each level corresponding to different data urgency. This approach can support multi-level scheduling.

[0033] Secondly, a corresponding processing priority is matched for each sensitivity level. This is done through a preset relationship table, meaning that the higher the sensitivity level, the higher its matching processing priority.

[0034] Furthermore, the generated priority information will be appended to the channel scheduling record corresponding to the data, and used as a sorting basis in subsequent time slice allocation processing.

[0035] Finally, by accurately converting sensitivity levels into priority values ​​that can be used for scheduling, the processing order can be dynamically adjusted according to the urgency of the data, thereby improving the timeliness of response during the exercise.

[0036] In a preferred embodiment of the present invention, time slice allocation and delay offset verification processing are performed according to priority scheduling data. This involves allocating processing time slices to the corresponding data channels and deleting data exceeding a preset offset threshold to generate valid data to be processed after time slice correction, including: Priority scheduling data is pre-arranged in time slices by arranging high-priority data channels in earlier processing intervals to generate initial time slice arrangement results for preliminary determination of processing time periods. Based on the initial time slice ranking results, the time slice is refined by further dividing the initial ranking interval into multiple continuous sub-intervals to generate a refined time slice structure that adapts to dynamic traffic. Offset determination is performed based on timestamp-marked data. By comparing the interval between the data arrival time and the current time, offset annotation content is generated to indicate the newness or oldness of the data. Data that has exceeded the time limit is removed based on the offset annotation. By deleting data that exceeds the preset offset threshold, a data set that only meets the timeliness requirements is generated. The dataset is processed within a time slice. By reordering the remaining data within the same time slice, valid data to be processed after time slice correction is generated.

[0037] In this embodiment of the invention, by performing time-slice pre-scheduling, time-slice refinement, offset determination, timeout culling, and time-slice consolidation on the scheduling results, the data processing flow can maintain a more stable rhythm under different load conditions. Refining the processing segments improves the adaptability of the processing window to traffic fluctuations, allowing for more precise processing order for different types of data. The offset determination process distinguishes between new and old data, accurately removing old data that is no longer relevant during the culling process, preventing resources from being consumed by low-value data. The consolidated effective data set maintains a reasonable processing order within the time slice, helping to reduce data latency accumulation and keeping the processing chain compact.

[0038] In a preferred embodiment of the present invention, priority scheduling data is pre-arranged in time slices by arranging high-priority data channels in earlier processing intervals to generate preliminary time slice arrangements for initially determining processing time periods. Specifically, this includes: First, based on the priority scheduling content generated in the previous step, the multiple data channels are initially sorted according to priority from high to low.

[0039] Secondly, the processing order is set according to the sorting results, allocating high-priority channels to the early part of the time slice and low-priority channels to the later part of the time slice, so that critical data can be processed at the beginning of the processing cycle. High-priority channels refer to data input channels that carry high-priority data.

[0040] Furthermore, by setting time slice boundaries, the entire processing cycle is divided into multiple consecutive intervals, and the corresponding channels are bound to these intervals.

[0041] Finally, the initial time slice arrangement will serve as the basis for subsequent time slice refinement and offset verification, ensuring a clear timing structure when handling load changes.

[0042] In a preferred embodiment of the present invention, offset determination processing is performed based on timestamp marker data. By comparing the interval between the data arrival time and the current time, offset annotation content indicating the newness or oldness of the data is generated, specifically including: First, the server reads the arrival time corresponding to each piece of data from the timestamp data. This arrival time is the standard time recorded when the data is written to the cache unit. Secondly, the server reads the current time when entering the current processing time slice, and both are under the same timing benchmark, ensuring comparability; Secondly, the server compares the arrival time of each piece of data with the current time and divides the data into different categories of newness according to the preset time difference range, such as "newest", "available", "nearly outdated", and "risk of becoming outdated". Finally, the server writes a corresponding newness category tag for each piece of data, forming offset annotation content, and outputs this annotation content along with the data to the subsequent timeout data removal processing step.

[0043] In a preferred embodiment of the present invention, time-slice reorganization is performed on the data set. By reordering the remaining data within the same time slice, valid data to be processed after time-slice correction is generated. Specifically, this includes: First, the server receives the data set output from the previous elimination step, which contains only the remaining data that meets the timeliness requirements; Secondly, the server sorts the data in the dataset according to their timestamps, so that the data that arrived earlier and is still valid is given priority in the processing sequence. Furthermore, when both personnel location data and environmental sensor node data exist in the dataset, the server performs a secondary adjustment based on the sorted data according to the priority scheduling rules, so that high-priority data remains in the front position within that time slice. Finally, the server rewrites the adjusted sequence into the processing buffer corresponding to the time slice, forming valid processing data corrected by the time slice, and outputs it to the real-time pass-through processing step.

[0044] In a preferred embodiment of the present invention, real-time transparent transmission processing is performed based on valid data to be processed. This involves bypassing cross-type sorting or merging processes for high-priority data and generating synchronization packets based on their independent timestamp records, thereby generating real-time transparent transmission synchronization data, including: Based on the valid data to be processed, a pass-through trigger determination is performed. By identifying whether high-priority data meets the pass-through conditions in the current processing cycle, a trigger result for initiating pass-through is generated. Based on the pass-through trigger result, pass-through candidate filtering is performed. Target data that meets the synchronization conditions is selected from high-priority data to generate a pass-through candidate set. Synchronization order construction is performed based on the transparent candidate set. The transmission order is arranged according to the independent timestamp record table of the target data to generate the sequence content for synchronous packet construction. Synchronization packages are assembled according to their sequential content. By filling the data into the synchronization package fields in sequence, a synchronization package structure with cross-device synchronization requirements is generated. Based on the structure of the synchronization packet, transparent outgoing processing is performed. By directly outputting the synchronization packet to the cross-end push path and bypassing the cross-type sorting or merging process, real-time transparent synchronization data is generated.

[0045] In this embodiment of the invention, by performing pass-through trigger determination, candidate filtering, synchronization order construction, and synchronization package assembly on high-priority data, data requiring rapid synchronization can skip the cross-type sorting process and complete cross-device updates via a shorter path. Pass-through trigger determination can identify data meeting timeliness requirements within the processing cycle, enabling the pass-through process to be initiated while ensuring reliability. Candidate filtering avoids irrelevant or low-relevance data occupying the pass-through path, thus maintaining the relevance of the synchronized content. The synchronization order construction process, by combining timestamp information, ensures that the updated content transmitted through the pass-through maintains a reasonable order when presented on the terminal. The final synchronization package can be directly pushed without waiting for other data processing to complete, allowing the terminal to perceive key state changes more quickly.

[0046] In a preferred embodiment of the present invention, the method for setting the transparent transmission conditions specifically includes: First, during the deployment phase, a set of high-priority types that can trigger pass-through is pre-established, including data types that must be high-priority types and data arrival times that must be within the valid time window of the current processing cycle; Secondly, in conjunction with the time slice scheduling mechanism, set the timing conditions for triggering transparent transmission. For example, transparent transmission is only allowed when the transmission arrives in the early or middle part of a processing time slice and the length of the transmission channel's queue to be sent is lower than the preset upper limit. Secondly, set necessary conditions for data transmission at the operational level of the exercise. For example, data transmission should only be allowed when the exercise object corresponding to the data is in a critical area, critical stage, or critical role, so as to avoid non-critical updates occupying the fast link. Finally, the type conditions, timing conditions, and business necessity conditions are combined into a pass-through condition set, and each condition is verified at runtime. Only when all conditions are met will a pass-through trigger result be generated.

[0047] In a preferred embodiment of the present invention, the method for setting synchronization conditions specifically includes: First, referring to the existing window matching method for cross-terminal status synchronization, the topic scope of the current synchronization window is maintained on the server side, such as the current exercise stage, the area of ​​interest, or the set of objects of interest; Secondly, define synchronization relevance conditions for candidate data, such as the data source terminal belonging to the synchronization window coverage object, the data's geographical location falling into the synchronization window's attention area, and the data's timestamp being within the time range of the synchronization window's planned update. Secondly, set synchronization integrity conditions to ensure the continuity of cross-platform presentation. For example, when the same terminal generates multiple updates in a short period of time, only the latest update or the update with a change of a preset level is retained to avoid redundancy in synchronization packets. Finally, the relevance condition and the integrity condition are written together as synchronization conditions into the synchronization filtering rule table. At runtime, the high-priority data is matched one by one, and those that are successfully matched are entered into the pass-through candidate set.

[0048] In a preferred embodiment of the present invention, a synchronization order construction process is performed based on the transparent transmission candidate set. This process generates sequence content for synchronization packet construction by arranging the transmission order according to an independent timestamp record table of the target data. Specifically, this includes: First, the server extracts an independent timestamp record table of the target data from the pass-through candidate set. The record table contains the arrival order and time stamp of the target data in the local cache. Secondly, the server side initially sorts the candidate data according to the timestamp order to ensure that the state changes presented during cross-device synchronization conform to the actual order of occurrence. Furthermore, when multiple data entries have the same or very close arrival times, the server establishes a fixed order rule based on the data source terminal identifier or data type identifier. For example, personnel location data is prioritized, followed by environmental sensor node data, to avoid confusion in the update order at the terminal. Finally, the server side solidifies the sorting and order rules into sequential content output for subsequent synchronization package assembly steps.

[0049] In a preferred embodiment of the present invention, synchronization packet assembly is performed according to the sequential content. By filling the synchronization packet fields with data in sequence, a synchronization packet structure with cross-end synchronization requirements is generated, specifically including: First, the server determines the data arrangement list within the synchronization packet based on the sequential content, and defines unified fields for the synchronization packet, including but not limited to data type markers, timestamp markers, data payload area, and verification fields. The design of these fields is compatible with existing cross-end message packet formats. Secondly, the server writes data one by one according to the sorted list, writes the type mark and timestamp mark of each data item into the corresponding field, and then writes the data payload into the data payload area, thereby ensuring that the terminal can recognize the content without ambiguity. Furthermore, when the amount of data in the synchronization packet reaches the preset upper limit or the time slice ends, the server side completes the packet wrapping process, writes the length information and necessary verification information to prevent truncation or misreading during terminal parsing. Finally, the resulting synchronization packet structure is output to the transparent outgoing processing step.

[0050] In a preferred embodiment of the present invention, transparent outgoing processing is performed according to the synchronization packet structure. By directly outputting the synchronization packet to the cross-end push path and bypassing the cross-type sorting or merging process, real-time transparent synchronization data is generated, specifically including: First, the server marks the synchronization packet structure as a transparent type, so that it no longer enters the cross-type sorting or merging queue in the internal processing link, but directly enters the transparent channel. Secondly, the synchronization packet is sent to the cross-end push processing, which completes the sending and performs low-latency sending processing through the push link to send the synchronization packet to each exercise terminal. Secondly, after the server transmits the data, it records the timestamp of the transmission and the list of target terminals for subsequent necessary retransmissions or consistency checks to maintain stable cross-terminal synchronization. Finally, the transparent transmission outgoing process generates real-time transparent transmission synchronization data and sends it to the cross-end push processing, which then completes the actual transmission to the terminal.

[0051] In a preferred embodiment of the present invention, arrival interval fluctuation analysis is performed based on the growth analysis results. By identifying whether the arrival interval shows a concentrated shortening within a short period, a surge identification result is generated to indicate a surge trend, including: Based on the growth analysis results, short-cycle segment determination is performed by dividing the continuous arrival records into multiple time segments and identifying the segments with the most intensive changes, generating short-cycle segment content for subsequent fluctuation identification. Based on the short-period segment content, arrival interval difference analysis is performed. By comparing the interval change trend between adjacent records within the segment, difference analysis content reflecting whether the arrival interval is continuously shortening is generated. Based on the difference analysis, the interval concentration determination process is performed. By identifying whether the shortening of the interval shows a clustering phenomenon in multiple consecutive records, a concentration determination result is generated to indicate the degree of sudden increase. Based on the concentration determination results, a surge trend confirmation process is performed. By comparing the concentration range with the preset surge standard, a surge identification result is generated to indicate the surge trend.

[0052] In this embodiment of the invention, by identifying short-period segments, analyzing interval differences, determining interval concentration, and confirming sudden surges, the arrival interval fluctuation analysis process can achieve higher identification accuracy. When data reaches short periods with dense changes, this method can more quickly locate key intervals, avoid redundant analysis of the overall data, and improve identification efficiency. By determining whether the interval is continuously shortening through difference analysis, sudden increases in data reporting frequency can be detected in a timely manner. The determination of interval concentration further filters out sporadic changes, ensuring that surge identification is based on continuous data performance. The final surge identification result more accurately reflects the current load status, making subsequent sensitivity calculations and scheduling strategies more reasonable.

[0053] In a preferred embodiment of the present invention, the method for setting the concentration range and the preset surge standard specifically includes: First, during deployment, select an observation window length to characterize the degree of clustering at arrival intervals, which is consistent with the division length of short-cycle segments; Secondly, statistical analysis of the distribution characteristics of interval shortening under normal conditions is conducted in historical exercise data or trial operation data. For example, normal fluctuations are usually manifested as scattered shortening or discontinuous short-term shortening. Furthermore, features such as "the number of records with consecutive shortened intervals", "the proportion of short-cycle segments covered by shortened intervals", and "the number of time segments in which the clustering state lasts" are used as candidate elements for the surge criteria, and corresponding judgment boundaries are set for each candidate element so that it can distinguish between normal fluctuations and sudden situations. Finally, by comparing the differences in candidate elements between normal and sudden data, the boundary that can stably distinguish between the two states is selected and solidified into a concentration range and a preset sudden increase standard, which can be called in the sudden increase situation confirmation step during runtime.

[0054] In a preferred embodiment of the present invention, short-period segment determination processing is performed based on the growth analysis results. This involves dividing consecutive arrival records into multiple time segments and identifying the segments with the most dense changes to generate short-period segment content for subsequent fluctuation identification. Specifically, this includes: First, the server receives the growth analysis results and the corresponding timestamp data, and arranges the consecutive arrival records in the same data input channel in chronological order. Secondly, the continuous arrival records are divided into multiple continuous time segments according to a preset short duration. The short duration can be set to milliseconds or seconds according to the needs of the exercise. Next, the number of arrivals and the density of arrival intervals within each time segment are statistically analyzed and compared with adjacent segments to identify segments with a sudden increase in the number of arrivals or a significant contraction in the intervals. Finally, the segments with the highest density or the most concentrated variation are marked as short-period segments, and the content of the short-period segments is output to provide a defined range for subsequent arrival interval difference analysis.

[0055] In a preferred embodiment of the present invention, arrival interval difference analysis is performed based on the content of short-period segments. By comparing the interval change trends between adjacent records within the segment, difference analysis content reflecting whether the arrival interval is continuously shortening is generated, specifically including: First, the server extracts all arriving records within the short-period segment based on the content of the short-period segment, and reads the timestamps of adjacent records; Secondly, the "adjacent interval sequence" construction method commonly used in existing real-time monitoring is adopted to form an interval sequence by sequentially dividing the time intervals between adjacent records; Next, the interval sequence is compared segment by segment to determine whether the interval shows three trends: continuous shortening, phased shortening, or no obvious change. "Continuous shortening" means that multiple consecutive interval values ​​gradually decrease, while "phased shortening" means that the interval decreases significantly within a certain sub-interval. Finally, the trend determination results are compiled into a difference analysis output for the next step of interval concentration determination.

[0056] In a preferred embodiment of the present invention, interval concentration determination processing is performed based on the difference analysis content. By identifying whether the interval shortening exhibits a clustering phenomenon in multiple consecutive records, a concentration determination result representing the degree of sudden increase is generated, specifically including: First, the server reads the segments or sets of records marked as "shortened interval" from the difference analysis content; Secondly, check whether the shortening interval phenomenon forms a continuous distribution within the short period segment, for example, whether the shortening trend repeats at multiple adjacent interval positions; Furthermore, when the interval shortening occurs only sporadically, it is determined to be a non-clustered state; when the interval shortening occurs in multiple consecutive adjacent positions and covers the main part of the short-period segment, it is determined to be a clustered state. Finally, the determination of whether an area is clustered or not is recorded as the concentration determination result and output to the sudden surge situation confirmation and processing step.

[0057] In a preferred embodiment of the present invention, a surge trend confirmation process is performed based on the concentration determination result. By comparing the concentration range with a preset surge standard, a surge identification result for indicating the surge trend is generated, specifically including: First, the server reads the concentration determination result and obtains the preset surge standard corresponding to the data channel. The surge standard can be set during deployment. Secondly, the concentration determination results are compared with the sudden increase criteria item by item. When the aggregation state meets the sudden increase criteria, it is confirmed that the short-cycle segment is in a sudden increase state; when it does not meet the criteria, it is confirmed that it is a normal fluctuation. Next, the confirmation result is recorded as a sudden increase identification result and appended to the timestamp data of the data channel as the basis for subsequent sensitivity level classification and priority scheduling; Finally, the output burst recognition results can be used to promptly enter a high-sensitivity scheduling state when data bursts are reported.

[0058] In a preferred embodiment of the present invention, sensitivity level classification is performed based on the surge identification results. Multiple sensitivity levels are defined according to the strength of the growth trend and the degree of the surge, generating sensitivity level content to characterize the processing urgency of each data channel, including: Based on the burst identification results, the growth rate is evaluated and processed. By identifying the range of change of the queue growth in multiple intervals, growth rate content is generated to display the growth intensity. The severity of the surge is analyzed based on the growth rate. By determining whether the surge is continuous and whether it is accompanied by an expansion of the queue backlog, severity analysis content is generated to characterize the scope of the surge's impact. Based on the severity analysis content, a multi-level grade interval division process is performed. Multiple sensitivity intervals are divided according to the combination relationship between the growth rate and severity, and grade interval content is generated to support priority mapping. Sensitivity level confirmation is performed based on the content of the level range. By selecting the level range that best represents the urgency of the current channel, sensitivity level content is generated to characterize the processing urgency of each data channel.

[0059] In this embodiment of the invention, by assessing the magnitude of growth, analyzing the severity of sudden increases, dividing the data into multi-level intervals, and confirming the sensitivity level, the urgency classification process can be made more refined and stable. The magnitude of growth assessment quantifies the strength of queue changes, allowing sensitivity to be judged beyond a single indicator. Severity analysis further enhances the reliability of the judgment by determining whether the sudden increase is sustained or accompanied by an expanding backlog. The division into multi-level intervals allows for richer sensitivity expressions based on different combinations of growth, going beyond simple high, medium, and low categories. Final confirmation of these intervals ensures that the sensitivity results are consistent with the actual network state, reducing the possibility of deviations in scheduling.

[0060] In a preferred embodiment of the present invention, growth rate assessment is performed based on the surge identification result. By identifying the range of change in queue growth within multiple intervals, growth rate content for displaying growth intensity is generated, specifically including: First, the server receives the burst identification results and their corresponding growth analysis results, and locates the continuous time intervals related to the burst identification results; Secondly, the continuous time interval is divided into several adjacent evaluation intervals. The length of each evaluation interval can be consistent with the time period used in the aforementioned queue growth identification to ensure consistency in criteria. Next, within each evaluation interval, the changes in the number of arrivals or queue length of the channel are statistically analyzed, and the changes between adjacent evaluation intervals are compared to determine whether the growth shows a state of "continuous expansion", "phased expansion" or "tending to stabilize". Finally, the growth status and its corresponding range of change are compiled into a growth rate output for subsequent surge severity analysis, enabling those skilled in the art to reproduce the assessment process of growth strength without introducing formulas.

[0061] In a preferred embodiment of the present invention, a severity analysis of the surge is performed based on the growth rate. By determining whether the surge is continuous and whether it is accompanied by an expansion of the queue backlog, severity analysis content is generated to characterize the scope of the surge's impact. Specifically, this includes: First, the server reads the growth rate and the surge identification results to obtain the start and end intervals of the surge and the corresponding growth intensity. Secondly, determine whether the surge trend continues to occur across multiple adjacent assessment intervals, such as maintaining high growth or not declining in multiple consecutive intervals. Secondly, determine whether the surge is accompanied by an increase in queue backlog, that is, check whether the data to be processed in the queue continues to accumulate during the surge and is not fully processed and digested by the time slice. Finally, the combination of "persistent characteristics" and "backlog expansion characteristics" is summarized into the severity analysis output. For example, cases that simultaneously meet the criteria of persistence and backlog expansion are marked as higher severity, while cases that only occur briefly or do not expand backlog are marked as lower severity, so that they can be directly referenced in the subsequent grade interval division process.

[0062] In a preferred embodiment of the present invention, multi-level grade interval division is performed based on the severity analysis content. Multiple sensitivity intervals are divided according to the combination relationship between the growth rate and severity, generating grade interval content to support priority mapping. Specifically, this includes: First, the server reads the growth rate and severity analysis data, and extracts the state categories that express the strength of the growth and the severity categories that express the scope of the sudden increase. Secondly, set up a combination of "growth intensity category and severity category" for comparison. For example, high growth intensity and high severity correspond to the highest sensitivity range, and medium growth intensity and low severity correspond to the medium sensitivity range. Next, based on the comparison relationship, the combination category of the current channel is mapped to the corresponding sensitivity interval, forming multiple ordered interval sets to ensure that different channels can fall into different sensitivity intervals; Finally, the output of the level range content serves as the boundary basis for subsequent sensitivity level confirmation, so that sensitivity classification does not depend on a single factor, but forms a clear hierarchy based on the combination of relationships.

[0063] In a preferred embodiment of the present invention, sensitivity level confirmation processing is performed based on the content of the sensitivity interval. By selecting the sensitivity interval that best represents the urgency of the current channel, sensitivity level content for characterizing the processing urgency of each data channel is generated, specifically including: First, the server reads the content of the grade range and determines the sensitivity range corresponding to the current data input channel within this evaluation period; Secondly, the sensitivity intervals of the channel in multiple adjacent evaluation periods are compared to determine whether frequent cross-interval jumps occur. Furthermore, when a jump occurs, the existing control methods should be used to "maintain the previous interval" or "confirm according to the most recent stable interval" to avoid repeated rises and falls in sensitivity level within a short period. Finally, the final sensitivity level is confirmed and the sensitivity level content is output for subsequent priority mapping and data scheduling, so that the scheduling results can reflect sudden changes while maintaining the stability of the level.

[0064] In a preferred embodiment of the present invention, time slice refinement is performed based on the initial time slice ranking results. By further dividing the initial ranking interval into multiple continuous sub-intervals, a time slice refinement structure adapted to dynamic traffic is generated, including: Based on the initial time slice results, interval boundary identification is performed. By identifying the start and end positions and internal structure of the initial segment, interval boundary content for detailed division is generated. The sub-intervals are divided according to the content of the interval boundaries. The initial intervals are divided into multiple continuous, equal or unequal sub-intervals according to the processing requirements, generating a sub-interval structure for fine-grained time management. Traffic adaptation analysis is performed based on the sub-interval structure. By comparing the processing capacity of each sub-interval with the traffic changes of the data channel, sub-interval load content is generated to adapt to dynamic traffic. Time-slice adjustment is performed based on the load content of the sub-intervals. By expanding or compressing sub-intervals with insufficient or excessive processing capacity, a refined time-slice structure is generated to adapt to dynamic traffic.

[0065] In this embodiment of the invention, by identifying interval boundaries, dividing sub-intervals, performing traffic adaptation analysis, and adjusting time slices, the time slice structure can more flexibly respond to changes in data traffic at different stages. Interval boundary identification ensures that refined operations are processed on the correct segments, avoiding impact on non-critical segments. By further dividing segments into sub-intervals, time slice resource allocation can be more refined, improving the accuracy of data scheduling for different channels. Traffic adaptation analysis, by comparing the processing capacity of sub-intervals with actual traffic, can maintain a stable processing rhythm regardless of load increases or decreases. The time slice expansion or compression operations can be dynamically adjusted according to traffic changes, ensuring that the final refined time slice structure maintains reasonable processing performance under both high and low load scenarios.

[0066] In a preferred embodiment of the present invention, interval boundary identification processing is performed based on the initial time slice sorting results. By identifying the start and end positions and internal structure of the initial sorting segments, interval boundary content for refined division is generated, specifically including: First, the server receives the initial time-slice scheduling results and obtains the initial scheduling segment information of each data channel within the current processing cycle, including the start and end times of the segment in the processing cycle and the corresponding channel identifier. Secondly, the processing cycle is regarded as a continuous time axis, and each initial sorting segment is mapped onto this time axis to clarify the time position of each segment; Next, identify the boundary points of each initial row segment, namely the starting boundary point corresponding to the start time of the segment and the ending boundary point corresponding to the end time of the segment, and record whether there are any reserved empty intervals or intersection points of adjacent segments within the segment. Finally, the identified starting boundary points, ending boundary points, and internal structure information are organized into interval boundary content for output, which is used in subsequent sub-interval division steps to ensure that the refined division can be carried out within an accurate segment range.

[0067] In a preferred embodiment of the present invention, sub-interval division is performed based on the content of the interval boundaries. By dividing the initial interval into multiple continuous, equal-length, or unequal-length sub-intervals according to processing requirements, a sub-interval structure for refined time management is generated, specifically including: First, the server reads the interval boundary content to determine the effective time range that each initial sorted segment can be used for partitioning; Secondly, select the granularity of the segmentation according to the processing requirements. For example, use a smaller granularity when refining high-priority channels and a larger granularity when refining low-priority channels. Next, the initial section is continuously divided according to the selected granularity to form multiple contiguous sub-intervals. The start and end times and corresponding channels of each sub-interval are recorded to ensure that there is no overlap or gap between the sub-intervals. Finally, all the segmented sub-intervals and their attributes are organized into a sub-interval structure output, providing a foundation for subsequent traffic adaptation analysis.

[0068] In a preferred embodiment of the present invention, traffic adaptation analysis is performed based on the sub-interval structure. By comparing the processing capacity of each sub-interval with the traffic changes of the data channel, sub-interval load content for adapting to dynamic traffic is generated, specifically including: First, the server reads the sub-interval structure to obtain the duration of each sub-interval and its corresponding data channel, thereby determining the time resources available for processing by that channel in the current period for that sub-interval. Secondly, read the traffic performance of the channel in the adjacent period, such as changes in arrivals, queue backlog, and surge marking results, to reflect the current traffic trend of the channel; Next, the processing time available in the sub-interval is matched and compared with the traffic trend of the channel to determine whether the processing capacity of the sub-interval is sufficient to cover the expected traffic, or whether the processing capacity exceeds the current load demand. Finally, the load status of each sub-interval, such as "insufficient processing capacity", "matched processing capacity", and "abundant processing capacity", is recorded as the sub-interval load content output for the next time slice adjustment process.

[0069] In a preferred embodiment of the present invention, time-slice adjustment processing is performed based on the load content of the sub-intervals. By expanding or compressing sub-intervals with insufficient or excessive processing capacity, a refined time-slice structure adapted to dynamic traffic is generated, specifically including: First, the server reads the load content of the sub-intervals and filters out the sub-intervals marked as having insufficient processing capacity and the sub-intervals marked as having sufficient processing capacity. Secondly, extended processing is performed on sub-intervals with insufficient processing capacity. The extension method can be to extend the duration of the sub-interval or borrow some time resources from adjacent surplus sub-intervals. When borrowing, the duration of the sub-interval of the borrowed channel should not be lower than its minimum processing requirement. If it is lower, the borrowing will not be performed. Secondly, compression processing is performed on sub-intervals with surplus processing capacity. The compression method can be to shorten the duration of the sub-interval or release it as a shared time resource that can be used by other channels, thereby avoiding idle time resources. Finally, the expanded and compressed sub-intervals are recombined to form an updated time-slice refinement structure and output, so that the time-slice resource allocation can be dynamically adjusted to maintain a reasonable level as traffic changes.

[0070] In a preferred embodiment of the present invention, timeout data is removed based on the offset annotation content. By deleting data exceeding a preset offset threshold, a data set retaining only those meeting timeliness requirements is generated, including: Based on the offset annotation content, the data is processed in layers according to its age. By dividing the offset into multiple layers representing different time difference ranges, the data age content is generated for subsequent filtering. The threshold correspondence is determined based on the age of the data. By mapping each level to different processing and retention standards, the threshold relationship content used for elimination judgment is generated. Timeout judgment processing is performed based on the threshold relationship content. By comparing the level of each data with the corresponding retention standard one by one, a timeout judgment result is generated to indicate whether the data exceeds the offset limit. Based on the timeout judgment result, data removal is performed. By deleting data marked as timeout and retaining data that meets the timeliness requirements, a data set that only meets the timeliness requirements is generated.

[0071] In this embodiment of the invention, by stratifying data by age, determining threshold correspondences, implementing timeout checks, and removing data, the offset verification process can more accurately distinguish whether data still has processing value. By dividing the offset into multiple levels, the timeliness differences of different data can be effectively expressed, making the judgment no longer overly coarse due to reliance on a single threshold. Determining the threshold correspondences further maps the levels to clear retention standards, providing an executable basis for subsequent judgments. Timeout checks, through a line-by-line comparison, enable the timely identification of lagging data, thereby reducing invalid data from entering subsequent processing flows. The final data set retains only data that meets the current time-series requirements, helping to maintain the effectiveness of the processing chain and preventing backlogged data from interfering with overall scheduling.

[0072] In a preferred embodiment of the present invention, the data is processed by layering the newness and oldness based on the offset annotation content. By dividing the offset into multiple levels representing different time difference ranges, data newness and oldness content for subsequent filtering is generated, specifically including: First, the server receives the offset annotation content, which has been written with the time difference category between the arrival time and the current processing cycle for each data in the previous offset determination process. Secondly, multiple time difference range levels are preset, for example, from shorter time difference to longer time difference, several levels are set up in sequence, and each level represents a degree of data freshness; Next, the time difference category of each data point is matched with the preset level, the data is classified into the corresponding newness level, and a level identifier is written for the data. Finally, the system outputs data with hierarchical identifiers to indicate the age of the data, enabling subsequent processing to perform differentiated removal based on the hierarchy.

[0073] In a preferred embodiment of the present invention, a threshold correspondence determination process is performed based on the data's age. By mapping each level to different processing retention criteria, threshold relationship content for elimination judgment is generated, specifically including: First, the server reads the data to determine its age and identifies the number and meaning of the currently used levels. Secondly, pre-set corresponding retention criteria for each level. For example, newer levels adopt a direct retention strategy, near-obsolescence levels adopt a conditional retention strategy, and obsolescence risk levels adopt a priority removal strategy. Next, establish a rule table that corresponds one-to-one with the hierarchy and retention criteria, and write the rule table into the removal decision context of this processing cycle; Finally, the threshold relationship content is output so that subsequent timeout judgments can be made to make consistent decisions on whether to retain or remove data based on this relationship table.

[0074] In a preferred embodiment of the present invention, timeout judgment processing is performed based on threshold relationship content. This involves comparing the level of each data entry with the corresponding retention standard one by one to generate a timeout judgment result indicating whether the data exceeds the offset limit. Specifically, this includes: First, the server reads the threshold relationship content and the corresponding data age content to obtain the hierarchical identifier of each data and the retention standard corresponding to that level. Secondly, perform a retention judgment on each piece of data, that is, check whether the data level falls within the standard range of "needs to be removed", "can be retained" or "needs to be reconfirmed"; Secondly, for data that needs further confirmation, the channel priority status or data urgency of the current time slice should be considered for auxiliary judgment to avoid critical data being mistakenly deleted due to slight timeout. Finally, the judgment result is marked as "timeout" or "not timeout", forming a timeout judgment result output, which is used for the next step of elimination execution.

[0075] In a preferred embodiment of the present invention, data removal processing is performed based on the timeout judgment result. This involves deleting data marked as timeout and retaining data that meets the timeliness requirements, generating a data set that only retains data that meets the timeliness requirements. Specifically, this includes: First, the server receives the timeout result and locates the target data record marked as timeout in the corresponding data cache unit; Secondly, delete the timed-out data. The deletion method can be to directly dequeue and discard it or transfer it to a low-priority history area for auditing or backtracking. Secondly, retain data that has not been marked as timed out and maintain its original queue sequence to ensure that the subsequent processing order is not disrupted. Finally, the retained data is organized into a set that only retains data that meets the timeliness requirements and outputs it into the time-slice processing or real-time transparent transmission process, thereby ensuring that subsequent processing only deals with data that still has practical significance.

[0076] In a preferred embodiment of the present invention, a pass-through trigger determination process is performed based on valid data to be processed. By identifying whether high-priority data meets the pass-through conditions in the current processing cycle, a trigger result for initiating pass-through is generated, including: Based on the valid data to be processed, priority validity confirmation processing is performed. By identifying whether the data belongs to a preset high priority type, priority confirmation content is generated to ensure the premise of transparent transmission. Based on the priority of the confirmed content, timeliness verification is performed. By judging how close the arrival time of high-priority data is to the current processing cycle, a timeliness verification result is generated to indicate whether the data still has synchronization significance. Based on the timeliness verification results, pass-through adaptability analysis is performed. By comparing the matching degree between the current data processing window and the current synchronization window, an adaptation analysis result reflecting the synchronization adaptability is generated. Based on the adaptation analysis results, the pass-through start determination process is carried out. Through comprehensive priority confirmation, timeliness verification and adaptation comparison, a trigger result for starting pass-through is generated.

[0077] In this embodiment of the invention, priority validity confirmation, timeliness verification, adaptability analysis, and pass-through initiation determination ensure that the pass-through mechanism is triggered at the appropriate time. Priority validity confirmation ensures that the triggering process only targets data types that truly need to be transmitted quickly, reducing unnecessary pass-through overhead. Timeliness verification determines whether the data is still within the time window suitable for the current processing cycle, making the pass-through behavior meaningful. Adaptability analysis further compares the degree of fit between the processing window and the synchronization window, making the pass-through behavior more in line with cross-end synchronization requirements. The pass-through initiation determination, through a comprehensive evaluation of multiple judgments, makes the fulfillment of pass-through conditions more objective, maintaining stable decision-making logic even under load changes.

[0078] In a preferred embodiment of the present invention, the method for setting a preset high-priority type specifically includes: First, during the exercise business definition phase, all data categories were sorted out and ranked according to their impact on command decisions. For example, personnel location, vital signs, and key instructions belong to the high-impact category, while routine monitoring data from environmental sensor nodes belong to the medium-low impact category. Secondly, high-impact categories are defined as a set of high-priority types, and this set is limited to a preset set of high-priority types, and a type mapping relationship is established; Secondly, verify the reasonableness of the whitelist coverage during multiple rounds of drills or trial runs. For example, confirm whether there is any critical event data that is not included in the high-priority set, or whether some non-critical data is mistakenly included, leading to an increase in the pass-through load. Finally, based on the verification results, whitelist entries are added or removed to form the final preset high-priority type set, which serves as the basis for priority validity confirmation at runtime.

[0079] In a preferred embodiment of the present invention, priority validity confirmation processing is performed on valid data to be processed. By identifying whether the data belongs to a preset high-priority type, priority confirmation content is generated to ensure the premise of transparent transmission. Specifically, this includes: First, the server extracts the type tag and channel priority tag of each piece of data from the valid data to be processed. These tags are derived from the preceding priority scheduling and time slice correction process. Secondly, a set of high-priority types that can be triggered for transparent transmission is preset, such as setting personnel location data and key command instruction data as transparent transmission types; Next, the type tag of each data item is compared with the set of high-priority types one by one. If they match, the data is confirmed to belong to the high-priority valid range. If they do not match, they are marked as non-transparent targets. Finally, the comparison results are written into the priority confirmation content and output to the subsequent timeliness verification steps, so that the pass-through judgment is only performed on the truly critical data.

[0080] In a preferred embodiment of the present invention, timeliness verification processing is performed based on priority confirmation content. By judging the proximity of the arrival time of high-priority data to the current processing cycle, a timeliness verification result is generated to indicate whether the data still has synchronization significance. Specifically, this includes: First, the server reads the high-priority data that has been confirmed in the priority confirmation content and obtains its corresponding independent timestamp record; Secondly, a synchronization effective time range is set within the current processing cycle to represent the time interval that can be synchronized immediately; Next, the arrival time of high-priority data is compared with the effective synchronization time range. If the arrival time is within the effective range, the data is determined to still have real-time synchronization significance. If it is outside the effective range, its timeliness is determined to be insufficient. Finally, the judgment results are compiled into a timeliness verification result output for subsequent pass-through adaptability analysis, which can avoid directly passing through high-priority data that has expired.

[0081] In a preferred embodiment of the present invention, a pass-through adaptability analysis is performed based on the timeliness verification result. By comparing the matching degree between the current data processing window and the current synchronization window, an adaptation analysis result reflecting the synchronization adaptability is generated, specifically including: First, the server reads the data that is deemed valid from the timeliness verification results and determines its current processing window position, such as whether it is in the beginning or middle of the current time slice. Secondly, maintain a synchronization window on the server side to represent the rhythm interval in which the terminal side plans to receive updates; Next, the processing window containing the data is matched and checked with the synchronization window to determine whether the data would conflict with the current synchronization rhythm if it were immediately transmitted, such as whether it would cause the terminal to receive the data too densely or in reverse order. Finally, the matching check results are recorded as the adaptation analysis results output, providing a clear basis for the pass-through start determination, so that the pass-through behavior is consistent with the overall synchronization rhythm.

[0082] In a preferred embodiment of the present invention, a pass-through start determination process is performed based on the adaptation analysis results. This process, through comprehensive priority confirmation, timeliness verification, and adaptation comparison, generates a trigger result for initiating the pass-through, specifically including: First, the server side summarizes the priority confirmation content, timeliness verification results, and adaptation analysis results to form three judgment information for the same data; Secondly, the three judgment information are set as necessary conditions for starting transparent transmission, that is, the data can only enter the transparent transmission path when it simultaneously meets the requirements of high priority validity, timeliness validity and compatibility with the synchronization window. Secondly, data that does not meet the conditions simultaneously is rejected, so that it continues to wait for cross-type sorting or subsequent time slice scheduling according to the normal processing link, thus avoiding error pass-through; Finally, trigger results are generated for data that meets the conditions and output to the transparent candidate screening step to start the subsequent transparent process.

[0083] In a preferred embodiment of the present invention, a pass-through candidate filtering process is performed based on the pass-through triggering result. This involves filtering target data that meets the synchronization conditions from high-priority data to generate a pass-through candidate set, including: Based on the pass-through trigger result, the candidate range is determined by identifying the range of data that can participate in the filtering from the high-priority data in the current processing cycle, and generating the data range content for filtering. The necessity of synchronization is determined based on the data range and content. By identifying the degree of correlation between each data and the current synchronization window, a judgment result is generated to explain the necessity of synchronization. Based on the judgment results, conditional filtering is performed to remove data that is not related to or has insufficient correlation with the synchronization window, thereby generating a preliminary set of candidate data. Priority confirmation is performed on the candidate data set. Priority is determined based on the timestamp, source terminal, or type characteristics of each candidate data, and a transparent candidate set is generated.

[0084] In this embodiment of the invention, by determining the candidate range, judging the necessity of synchronization, conditional filtering, and confirming the priority order, the transparent transmission candidate screening process can focus on the data most suitable for entering the synchronization package. Determining the candidate range limits the screening target, keeping the processed objects within a reasonable range and preventing the screening efficiency from being affected by excessive data volume. Judging the necessity of synchronization identifies the true relationship between candidate data and the current synchronization window, avoiding the transmission of meaningless data. Conditional filtering eliminates data with insufficient relevance, giving the transparent transmission content more explicit value. Finally, confirming the priority order forms a reasonable arrangement based on different characteristics, making the data structure within the synchronization package clearer, thereby improving the accuracy of cross-end updates.

[0085] In a preferred embodiment of the present invention, a candidate range determination process is performed based on the pass-through trigger result. This process involves determining the range of data eligible for filtering from high-priority data within the current processing cycle and generating data range content for filtering. Specifically, this includes: First, the server receives the pass-through trigger result, which clearly indicates which data meets the pass-through start conditions. Secondly, extract all data marked as high priority and triggered pass-through from the valid unprocessed data corresponding to the current processing cycle, and remove data that has not triggered pass-through, so that the candidate objects are concentrated on the pass-through target. Secondly, the extracted data is grouped into multiple candidate subsets according to the data input channel or data type, so that the synchronization correlation can be judged by group in the future; Finally, the candidate subsets are summarized into a data range output, so that subsequent filtering has clear object boundaries and avoids ineffective filtering of the entire data.

[0086] In a preferred embodiment of the present invention, a synchronization necessity judgment is performed based on the data range content. By identifying the degree of correlation between each data and the current synchronization window, a judgment result illustrating the necessity of synchronization is generated, specifically including: First, the server reads the data range content and obtains the type marker, source terminal marker, and independent timestamp record of each candidate data item; Secondly, maintain the status topics that the current synchronization window is concerned with on the server side, such as the exercise phase, regional situation or personnel cluster range corresponding to the current window; Next, the attributes of the candidate data are matched with the status topics that the synchronization window is interested in. For example, it is determined whether the personnel location data belongs to the personnel set covered by the synchronization window, and whether the environmental sensor node data is located in the geographical area or time period that the synchronization window is interested in. Finally, the correlation between candidate data and the synchronization window is marked as "needs synchronization" or "can be delayed for synchronization", and the result of the synchronization necessity judgment is output to provide a basis for subsequent filtering.

[0087] In a preferred embodiment of the present invention, conditional filtering is performed based on the judgment result. By eliminating data that is not related to or has insufficient correlation with the synchronization window, a preliminary candidate data set is generated, specifically including: First, the server reads the synchronization necessity judgment result and locates the candidate data marked as "synchronization can be delayed" or "no synchronization is necessary"; Secondly, remove the aforementioned data that does not require synchronization from the candidate range to avoid them occupying the transparent transmission link; Furthermore, when there is an explicit correlation between candidate data, such as multiple location data reported by the same terminal in a very short period of time, the server can use existing redundancy removal strategies to retain the latest one or retain several data with larger changes to reduce duplicate synchronization. Finally, the filtered and deduplicated data is organized into a preliminary candidate data set for output, ensuring that the transparent content remains compact and consistent with current synchronization requirements.

[0088] In a preferred embodiment of the present invention, a priority order confirmation process is performed based on the candidate data set. This process determines the priority order based on the timestamp, source terminal, or type characteristics of each candidate data set, generating a transparent candidate set. Specifically, this includes: First, the server reads the preliminary filtered candidate data set and extracts the independent timestamp record, type tag, and source terminal tag for each data item. Secondly, candidate data are sorted first by timestamp to ensure that the synchronization order matches the actual order of occurrence. Furthermore, when there are multiple data entries with similar or identical timestamps, the server performs a secondary sorting based on preset order rules. For example, it prioritizes personnel location data and then environmental sensor node data, or prioritizes data reported by the command terminal, in order to avoid inconsistencies between the cross-terminal presentation order and the actual occurrence order. Finally, the sorted candidate data set is determined as the pass-through candidate set output, which is used for subsequent synchronization sequence construction and synchronization packet assembly, thereby ensuring the stability of the pass-through link and the clarity of the update order.

[0089] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for constructing a low-latency cross-terminal interactive network for disaster prevention and mitigation emergency drills, characterized in that, The method includes: Data type identification processing is performed on the raw exercise data from each exercise terminal. By identifying whether it belongs to personnel location data or environmental sensor node data, a classified data stream with type labels is generated. Independent caching and timestamp recording are performed based on the classified data stream. By establishing independent queues in their respective data cache units and recording their arrival time sequence, timestamp-marked data with queue sequence information is generated. Delay sensitivity calculation is performed based on timestamp-marked data. Priority scheduling data for data processing order control is generated by analyzing queue length changes and arrival time distribution. Based on priority scheduling data, time slice allocation and delay offset verification are performed. By allocating processing time slices to the corresponding data channels and deleting data that exceeds the preset offset threshold, valid data to be processed after time slice correction is generated. Real-time transparent transmission processing is performed on the valid unprocessed data. By bypassing the cross-type sorting or merging process for high-priority data and generating a synchronization package based on its independent timestamp record table, real-time transparent transmission synchronization data is generated. Cross-terminal push processing is performed based on real-time transparent transmission and synchronization data. By pushing the latest interaction status to each exercise terminal, a low-latency cross-terminal status update result is formed.

2. The method for constructing a low-latency cross-terminal interactive network for disaster prevention and mitigation emergency drills according to claim 1, characterized in that, Based on the timestamp-marked data, latency sensitivity calculations are performed. By analyzing queue length changes and arrival time distributions, priority scheduling data for data processing order control is generated, including: The queue growth identification process is performed on the timestamp-marked data. By comparing the changes in the number of arrivals over multiple consecutive time periods, growth analysis results representing the queue growth trend are generated. Based on the growth analysis results, arrival interval fluctuation analysis is performed. By identifying whether the arrival interval shows a concentrated shortening phenomenon in a short period, a surge identification result is generated to indicate the surge trend. Based on the results of the surge identification, sensitivity levels are classified. Multiple sensitivity levels are divided according to the strength of the growth trend and the degree of the surge, generating sensitivity level content to characterize the urgency of processing each data channel. Priority mapping is performed based on the sensitivity level content. By mapping higher sensitivity levels to higher processing priorities, priority scheduling data for controlling the order of data processing is generated.

3. The method for constructing a low-latency cross-terminal interactive network for disaster prevention and mitigation emergency drills according to claim 1, characterized in that, Based on priority scheduling data, time slice allocation and delay offset verification are performed. This involves allocating processing time slices to corresponding data channels and deleting data exceeding a preset offset threshold to generate time-sliced, valid data to be processed, including: Priority scheduling data is pre-arranged in time slices by arranging high-priority data channels in earlier processing intervals to generate initial time slice arrangement results for preliminary determination of processing time periods. Based on the initial time slice ranking results, the time slice is refined by further dividing the initial ranking interval into multiple continuous sub-intervals to generate a refined time slice structure that adapts to dynamic traffic. Offset determination is performed based on timestamp-marked data. By comparing the interval between the data arrival time and the current time, offset annotation content is generated to indicate the newness or oldness of the data. Data that has exceeded the time limit is removed based on the offset annotation. By deleting data that exceeds the preset offset threshold, a data set that only meets the timeliness requirements is generated. The dataset is processed within a time slice. By reordering the remaining data within the same time slice, valid data to be processed after time slice correction is generated.

4. The method for constructing a low-latency cross-terminal interactive network for disaster prevention and mitigation emergency drills according to claim 1, characterized in that, Real-time transparent transmission processing is performed on valid pending data. This involves bypassing cross-type sorting or merging processes for high-priority data and generating synchronization packets based on their independent timestamp records. This generates real-time transparent transmission synchronization data, including: Based on the valid data to be processed, a pass-through trigger determination is performed. By identifying whether high-priority data meets the pass-through conditions in the current processing cycle, a trigger result for initiating pass-through is generated. Based on the pass-through trigger result, pass-through candidate filtering is performed. Target data that meets the synchronization conditions is selected from high-priority data to generate a pass-through candidate set. Synchronization order construction is performed based on the transparent candidate set. The transmission order is arranged according to the independent timestamp record table of the target data to generate the sequence content for synchronous packet construction. Synchronization packages are assembled according to their sequential content. By filling the data into the synchronization package fields in sequence, a synchronization package structure with cross-device synchronization requirements is generated. Based on the structure of the synchronization packet, transparent outgoing processing is performed. By directly outputting the synchronization packet to the cross-end push path and bypassing the cross-type sorting or merging process, real-time transparent synchronization data is generated.

5. The method for constructing a low-latency cross-terminal interactive network for disaster prevention and mitigation emergency drills according to claim 2, characterized in that, Based on the growth analysis results, arrival interval fluctuation analysis is performed. By identifying whether the arrival interval shows a concentrated shortening within a short period, a surge identification result is generated to indicate a surge trend, including: Based on the growth analysis results, short-cycle segment determination is performed by dividing the continuous arrival records into multiple time segments and identifying the segments with the most intensive changes, generating short-cycle segment content for subsequent fluctuation identification. Based on the short-period segment content, arrival interval difference analysis is performed. By comparing the interval change trend between adjacent records within the segment, difference analysis content reflecting whether the arrival interval is continuously shortening is generated. Based on the difference analysis, the interval concentration determination process is performed. By identifying whether the shortening of the interval shows a clustering phenomenon in multiple consecutive records, a concentration determination result is generated to indicate the degree of sudden increase. Based on the concentration determination results, a surge trend confirmation process is performed. By comparing the concentration range with the preset surge standard, a surge identification result is generated to indicate the surge trend.

6. The method for constructing a low-latency cross-terminal interactive network for disaster prevention and mitigation emergency drills according to claim 2, characterized in that, Based on the surge identification results, sensitivity levels are categorized. Multiple sensitivity levels are assigned according to the strength of the growth trend and the degree of the surge, generating sensitivity level content to characterize the processing urgency of each data channel, including: Based on the burst identification results, the growth rate is evaluated and processed. By identifying the range of change of the queue growth in multiple intervals, growth rate content is generated to display the growth intensity. The severity of the surge is analyzed based on the growth rate. By determining whether the surge is continuous and whether it is accompanied by an expansion of the queue backlog, severity analysis content is generated to characterize the scope of the surge's impact. Based on the severity analysis, a multi-level grade interval division is performed. Multiple sensitivity intervals are divided according to the combination relationship between the growth rate and severity, and grade interval content is generated to support priority mapping. Sensitivity level confirmation is performed based on the content of the level range. By selecting the level range that best represents the urgency of the current channel, sensitivity level content is generated to characterize the processing urgency of each data channel.

7. The method for constructing a low-latency cross-terminal interactive network for disaster prevention and mitigation emergency drills according to claim 3, characterized in that, Based on the initial time-slot scheduling results, the time-slots are refined by further dividing the initial scheduling interval into multiple continuous sub-intervals, generating a refined time-slot structure to adapt to dynamic traffic, including: Based on the initial time slice results, interval boundary identification is performed. By identifying the start and end positions and internal structure of the initial segment, interval boundary content for detailed division is generated. The sub-intervals are divided according to the content of the interval boundaries. The initial intervals are divided into multiple continuous, equal or unequal sub-intervals according to the processing requirements, generating a sub-interval structure for fine-grained time management. Traffic adaptation analysis is performed based on the sub-interval structure. By comparing the processing capacity of each sub-interval with the traffic changes of the data channel, sub-interval load content is generated to adapt to dynamic traffic. Time-slice adjustment is performed based on the load content of the sub-intervals. By expanding or compressing sub-intervals with insufficient or excessive processing capacity, a refined time-slice structure is generated to adapt to dynamic traffic.

8. The method for constructing a low-latency cross-terminal interactive network for disaster prevention and mitigation emergency drills according to claim 3, characterized in that, Data exceeding the timeout threshold is removed based on the offset annotations. This process generates a dataset that retains only data meeting timeliness requirements by deleting data exceeding a preset offset threshold. This dataset includes: Based on the offset annotation content, the data is processed in layers according to its age. By dividing the offset into multiple layers representing different time difference ranges, the data age content is generated for subsequent filtering. The threshold correspondence is determined based on the age of the data. By mapping each level to different processing and retention standards, the threshold relationship content used for elimination judgment is generated. Timeout judgment processing is performed based on the threshold relationship content. By comparing the level of each data with the corresponding retention standard one by one, a timeout judgment result is generated to indicate whether the data exceeds the offset limit. Based on the timeout judgment result, data removal is performed. By deleting data marked as timeout and retaining data that meets the timeliness requirements, a data set that only meets the timeliness requirements is generated.

9. The method for constructing a low-latency cross-terminal interactive network for disaster prevention and mitigation emergency drills according to claim 4, characterized in that, Based on the valid data to be processed, a pass-through trigger determination process is performed. By identifying whether high-priority data meets the pass-through conditions in the current processing cycle, a trigger result for initiating pass-through is generated, including: Based on the valid data to be processed, priority validity confirmation processing is performed. By identifying whether the data belongs to a preset high priority type, priority confirmation content is generated to ensure the premise of transparent transmission. Based on the priority of the confirmed content, timeliness verification is performed. By judging how close the arrival time of high-priority data is to the current processing cycle, a timeliness verification result is generated to indicate whether the data still has synchronization significance. Based on the timeliness verification results, pass-through adaptability analysis is performed. By comparing the matching degree between the current data processing window and the current synchronization window, an adaptation analysis result reflecting the synchronization adaptability is generated. Based on the adaptation analysis results, the pass-through start determination process is carried out. Through comprehensive priority confirmation, timeliness verification and adaptation comparison, a trigger result for starting pass-through is generated.

10. The method for constructing a low-latency cross-terminal interactive network for disaster prevention and mitigation emergency drills according to claim 4, characterized in that, Based on the pass-through trigger result, pass-through candidate filtering is performed. Target data that meets the synchronization conditions is selected from high-priority data to generate a pass-through candidate set, including: Based on the pass-through trigger result, the candidate range is determined by identifying the range of data that can participate in the filtering from the high-priority data in the current processing cycle, and generating the data range content for filtering. The necessity of synchronization is determined based on the data range and content. By identifying the degree of correlation between each data and the current synchronization window, a judgment result is generated to explain the necessity of synchronization. Based on the judgment results, conditional filtering is performed to remove data that is not related to or has insufficient correlation with the synchronization window, thereby generating a preliminary set of candidate data. Priority confirmation is performed on the candidate data set. Priority is determined based on the timestamp, source terminal, or type characteristics of each candidate data, and a transparent candidate set is generated.

Images