A method for updating medical waste data across different stages

Abstract

This invention relates to the field of data update technology, specifically a method for cross-stage data update of medical waste. This method determines that the vehicle-mounted terminal has entered a signal blind zone when the network signal strength is detected to be below a preset blind zone threshold, and records the collected medical waste collection data as data to be uploaded. When the vehicle-mounted terminal enters a signal recovery zone, a random backoff delay is generated. A local timer is started and after waiting for the random backoff delay, the medical waste collection data to be uploaded is divided into M data fragments according to timestamp order, each fragment having a fragment number. An upload request is initiated to the server based on the fragment number. Upon receiving a response status code from the server, the current data fragment is considered successfully uploaded and updated, and the next data fragment upload request is initiated, until the Mth data fragment is successfully uploaded and updated. This ensures the stability of the medical waste collection data transmission and update process.

Description

Technical Field

[0001] This invention relates to the field of data update technology, specifically a method for updating medical waste data across different stages. Background Technology

[0002] The collection and management of medical waste is a crucial link in the urban medical waste disposal system. Collection vehicles need to continuously collect and upload waste collection data during their journeys to enable real-time information flow and traceability between different disposal stages. As the scale of the urban medical waste collection network expands, the reliability of data synchronization between onboard terminals and backend servers directly impacts the data integrity and compliance of the entire waste disposal chain.

[0003] However, urban waste collection routes inevitably pass through signal blind spots such as tunnels and underpasses. During these crossings, vehicle-mounted terminals cannot maintain communication with the server and must cache the continuously collected collection data locally. Due to the concentrated departure of multiple collection vehicles during morning and evening rush hours, the interval between adjacent vehicles is often only tens of seconds to several minutes. When the convoy exits a tunnel, a large number of vehicle-mounted terminals detect signal recovery almost simultaneously within a very short time window and simultaneously send upload requests to the server, each carrying a large amount of cached data accumulated in the blind spot. This sudden concurrent upload behavior creates instantaneous traffic spikes on the server side, causing server processing queue backlogs, leading to response tail latency jitter, and in severe cases, even causing partial data loss or upload failures, affecting the integrity and traceability of the entire medical waste data chain. Summary of the Invention

[0004] (1) Technical problems to be solved The purpose of this invention is to provide a method for updating medical waste data across different stages, which solves the problem of server congestion caused by sudden batch uploads of cached data by multiple medical waste collection vehicles at the boundary between signal blind spots and recovery zones, leading to latency jitter and the risk of data loss.

[0005] (2) Technical solution To achieve the above objectives, in one aspect, the present invention provides a method for updating medical waste data across different stages, the method comprising: S1. During operation, the vehicle-mounted terminal continuously collects medical waste collection and transportation data and monitors the current network signal strength in real time. When the network signal strength is detected to be lower than the preset blind zone judgment threshold, the vehicle-mounted terminal is determined to have entered the signal blind zone. The medical waste collection and transportation data collected thereafter is written into the local cache queue in order with the timestamp as the index and recorded as medical waste collection and transportation data to be uploaded.

[0006] S2. The vehicle terminal continuously polls the network signal strength in the signal blind zone; when the network signal strength is detected to be higher than the preset signal recovery judgment threshold, the vehicle terminal determines that it has entered the signal recovery zone and generates a random backoff delay.

[0007] S3. After generating a random backoff delay, the vehicle terminal starts a local timer and waits for the random backoff delay. Then, according to the preset data fragment size threshold, it divides the medical waste collection data to be uploaded into M data fragments in the order of timestamps. Each data fragment is accompanied by a fragment number and the corresponding timestamp range.

[0008] S4. The vehicle terminal initiates upload requests to the server piece by piece according to the fragment sequence number. If the vehicle terminal receives a response status code from the server carrying the corresponding fragment sequence number, it determines that the current data fragment upload and update is successful, and initiates the upload request for the next data fragment, until the Mth data fragment upload and update is successful.

[0009] Furthermore, the method for generating random backoff delay includes: Obtain the total amount of medical waste collection data to be uploaded in the local cache queue. Divide the total amount of data by the preset data volume normalization benchmark value to obtain the normalized data load ratio. Compare the normalized data load ratio with the preset upper and lower bound values ​​of the coefficient. When the normalized data load ratio is lower than the lower bound value, take the lower bound value. When the normalized data load ratio is higher than the upper bound value, take the upper bound value. When the normalized data load ratio falls between the preset lower and upper bound values ​​of the coefficient, directly take the normalized data load ratio. Use the truncated result as the dynamic backoff coefficient.

[0010] The number of other vehicle terminals currently performing upload operations is counted by monitoring the received signal strength indicator value, and the number of other vehicle terminals is calculated according to a preset linear growth rule to obtain the contention backoff increment; the random backoff delay is calculated based on the dynamic backoff coefficient and the contention backoff increment.

[0011] Furthermore, the method for counting the number of other vehicle terminals currently performing upload operations by monitoring the received signal strength indicator value includes: Within the channel sampling window, the received signal strength indicator value is continuously sampled at equal time intervals. Each sample value is compared with a preset occupancy decision threshold. When the sample value is lower than the occupancy decision threshold, the sampling point is marked as an idle point; when the sample value is not lower than the occupancy decision threshold, the sampling point is marked as an occupied point.

[0012] A sampling segment with at least three consecutive occupancy points is identified as a candidate transmission burst, and the start and end times of the candidate transmission burst are recorded as the burst duration. For each candidate transmission burst, the vehicle terminal attempts to demodulate and decode the radio frequency signal within the burst duration. If a complete data frame is successfully extracted and the unique identifier of the source vehicle terminal is parsed from the data frame header, the unique identifier of the source vehicle terminal is added to the set of identified terminals.

[0013] If demodulation and decoding fail but the duration of the candidate transmission burst falls within the preset normal medical waste collection data frame duration range, the candidate transmission burst will be counted in the unknown terminal burst count; if demodulation and decoding fail and the duration exceeds the preset normal medical waste collection data frame duration range, the candidate transmission burst will be regarded as a channel interference event and discarded.

[0014] After the channel sampling window ends, the unique identifiers of the source vehicle terminals in the identified terminal set are deduplicated to obtain the number of other identified vehicle terminals. The unknown terminal burst count is multiplied by a dynamic reduction factor to obtain the estimated number of unknown terminals. The number of other identified vehicle terminals is added to the estimated number of unknown terminals to obtain the number of other vehicle terminals currently performing the upload operation.

[0015] Furthermore, the method for determining the dynamic reduction factor includes: After the channel sampling window ends, the total number of candidate transmission bursts identified within the channel sampling window is counted. The number of candidate transmission bursts that successfully completed demodulation and decoding and extracted the unique identifier of the source vehicle terminal is taken as the number of successful decodings. The demodulation and decoding success rate of the current channel sampling window is obtained by dividing the number of successful decodings by the total number of candidate transmission bursts.

[0016] The burst duration of all candidate transmission bursts within the current channel sampling window is obtained. The arithmetic mean of the burst durations of all candidate transmission bursts is compared with the center value of the preset normal medical waste collection data frame duration interval to obtain the burst duration deviation ratio. When the burst duration deviation ratio exceeds the preset deviation tolerance range, the demodulation decoding success rate and the burst duration deviation ratio are weighted and fused to obtain a dynamic reduction coefficient. When the burst duration deviation ratio does not exceed the preset deviation tolerance range, the demodulation decoding success rate is used as the dynamic reduction coefficient.

[0017] Furthermore, the method for calculating the random backoff delay based on the dynamic backoff coefficient and the contention backoff increment includes: The dynamic backoff coefficient is multiplied by the preset upper limit of the basic backoff window to obtain the upper limit of the dynamic backoff window. The lower limit of the preset lower limit of the basic backoff window is used as the lower bound, and the upper limit of the dynamic backoff window is used as the upper bound to construct the dynamic backoff time window of the current vehicle terminal.

[0018] Within the dynamic backoff time window, random sample values ​​are generated according to a uniform random distribution, and these random sample values ​​are used as the initial backoff delay. The initial backoff delay is added to the contention backoff increment to obtain the random backoff delay.

[0019] Furthermore, the method for generating random sample values ​​according to a uniform random distribution within the dynamic backoff time window includes: Obtain the unique identifier of the vehicle terminal, perform a bitwise XOR operation between the unique identifier and the microsecond-level timestamp of the local clock of the vehicle terminal at the time of signal recovery, and use the result of the bitwise XOR operation as a random number seed; substitute the random number seed into the recursive formula of the linear congruence generator for iterative calculation to obtain the current item in the pseudo-random integer sequence.

[0020] The interval length of the dynamic backoff time window is converted into an integer interval length in microseconds. The current item is modulo the integer interval length to obtain a normalized integer offset. The normalized integer offset is added to the preset lower limit of the basic backoff window to obtain a random sampled value.

[0021] Furthermore, the method by which the vehicle-mounted terminal initiates upload requests to the server piece by piece according to the fragment sequence number includes: When the vehicle terminal initiates an upload request for the current data segment, it records the time when the upload request message is sent. When it receives a response status code from the server carrying the corresponding segment sequence number, it records the time when the response status code arrives. Based on the time when the upload request message is sent and the time when the response status code arrives, the round-trip time of the current data segment is calculated.

[0022] The round-trip time obtained after the first data shard is completed is used as the initial smoothed round-trip time. After each subsequent data shard is uploaded, the current round-trip time is weighted and merged with the previous round of smoothed round-trip time using an exponentially weighted moving average method to obtain a continuously updated smoothed round-trip time. A sliding observation window is established for the round-by-round update results of the smoothed round-trip time, and the dynamic time delay baseline is calculated within the sliding observation window.

[0023] Whenever the latest round-trip delay exceeds the dynamic delay baseline, it is determined that the concurrent uploading of multiple collection vehicles in the signal recovery area has caused congestion pressure on the server, triggering the adaptive transmission adjustment process to initiate an upload request.

[0024] Furthermore, the method for establishing a sliding observation window based on the round-by-round update results of the smooth round-trip delay, and calculating the dynamic delay baseline within the sliding observation window, includes: Before uploading the first data fragment, a fixed-length first-in-first-out queue is initialized as a sliding observation window. The smooth round-trip time obtained after each data fragment upload is completed is pushed into the tail of the sliding observation window. When the number of smooth round-trip times stored in the sliding observation window reaches the fixed length, the earliest smooth round-trip time at the head of the queue is popped out each time a new smooth round-trip time is pushed in.

[0025] After the sliding observation window has been filled with a fixed length of smooth round-trip delays, whenever it is necessary to calculate the dynamic delay baseline, the arithmetic mean of all smooth round-trip delays within the sliding observation window is calculated as the current window mean, and the arithmetic mean of the absolute deviations of all smooth round-trip delays from the current window mean is calculated as the current window average deviation. The sum of the current window mean and the current window average deviation is used as the dynamic delay baseline.

[0026] Furthermore, the method for triggering the adaptive sending adjustment process to initiate an upload request includes: The delay deviation ratio is obtained by dividing the difference between the smooth round-trip delay of the current round and the dynamic delay baseline by the dynamic delay baseline. The delay trend coefficient is obtained by dividing the smooth round-trip delay of the current round by the smooth round-trip delay of the previous round. A delay trend coefficient greater than one indicates that the congestion of concurrent uploads in the signal recovery area at the tunnel exit continues to worsen, while a delay trend coefficient less than one indicates that the congestion tends to ease.

[0027] Multiply the delay deviation ratio by the delay trend coefficient to obtain the comprehensive congestion adjustment factor; multiply the comprehensive congestion adjustment factor by the preset transmission interval benchmark value to obtain the transmission interval duration for the current round; after completing the upload request for the current data segment, the vehicle terminal waits for the transmission interval duration before initiating the upload request for the next data segment.

[0028] (3) Beneficial effects Compared with the prior art, the beneficial effects of the present invention are: 1. By generating a random backoff delay immediately after the vehicle terminal detects signal recovery, the cached data upload behavior triggered almost simultaneously by multiple collection vehicles in the signal recovery zone at the tunnel exit is actively dispersed on the time axis. This ensures that the backoff delay of each vehicle is statistically independent and evenly distributed. The batch upload behavior that was originally concentrated at the same time is naturally unfolded into a dispersed upload sequence with sufficient time staggering. This effectively eliminates the instantaneous traffic peak of the server caused by sudden concurrent uploads when entering the signal recovery zone and reduces the probability of server congestion.

[0029] 2. Furthermore, by continuously monitoring the round-trip latency during the upload process and comparing it with a dynamic baseline, the vehicle terminal can perceive the current congestion status of the server in real time and adaptively adjust the data fragmentation sending interval according to the degree of latency deviation. This achieves dynamic convergence control of the upload rate and a two-dimensional response to the current congestion severity and the direction of congestion change. Thus, even when the concurrent upload pressure still exists, the jitter of the server tail latency is further suppressed, ensuring the stability of the data transmission and update process. Attached Figure Description

[0030] Figure 1 This is a flowchart of a method for updating medical waste data across different stages according to the present invention. Detailed Implementation

[0031] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0032] Before providing examples, it's necessary to describe the application scenarios of this invention. This invention is a cross-stage data update method for medical waste, applicable to the entire process of medical waste collection and transportation from medical institutions to centralized disposal units, especially suitable for areas with unstable wireless network coverage, such as tunnels and underground parking lots where collection vehicles frequently enter and exit. In this scenario, medical waste collection vehicles are uniformly equipped with vehicle-mounted terminals with cellular mobile communication and local storage functions. The vehicle-mounted terminals and the backend server interact via a mobile communication network. The vehicle-mounted terminals need to upload medical waste collection data to the server in real-time or near real-time. This data typically includes fields such as information on the medical waste generating unit, medical waste category, medical waste weight, loading time, collection time, vehicle location, license plate number, driver information, and driving route. During vehicle operation, once the vehicle enters tunnels, underground parking garages, etc., the wireless signal strength received by the vehicle-mounted terminal will significantly weaken or even be completely lost, resulting in the medical waste collection data not being uploaded in a timely manner, causing data to accumulate locally on the vehicle-mounted terminal. When multiple collection vehicles regain signal at the same tunnel exit, underground parking garage exit, or mountain signal recovery area almost simultaneously and immediately initiate data upload, the server side and wireless access network will encounter a large number of concurrent requests in a short period of time, which can easily lead to instantaneous congestion. This may result in problems such as upload failure, timeout retry, and duplicate uploads, thereby affecting the integrity and timeliness of cross-institutional and cross-stage data updates for medical waste.

[0033] Example 1: As Figure 1As shown in the figure, this embodiment provides a method for updating medical waste data across different stages, the method including: S1. During operation, the vehicle-mounted terminal continuously collects medical waste collection and transportation data and monitors the current network signal strength in real time. When the network signal strength is detected to be lower than the preset blind zone judgment threshold, the vehicle-mounted terminal is determined to have entered the signal blind zone. The medical waste collection and transportation data collected thereafter is written into the local cache queue in order with the timestamp as the index and recorded as medical waste collection and transportation data to be uploaded.

[0034] For example, the vehicle-mounted terminal continuously collects medical waste collection data throughout the entire journey and continuously monitors the current network signal strength. The current network signal strength can be measured in dBm, directly output by the vehicle-mounted terminal's radio frequency module. During system deployment, maintenance personnel, combining operator wireless coverage test data and the vehicle-mounted terminal's receiving sensitivity, conduct on-site road tests and refer to network planning reports to select a blind zone determination threshold for identifying signal blind spots. The blind zone determination threshold is typically set as the lower limit of signal strength at which the vehicle-mounted terminal can still maintain data services but the bit error rate is significantly increased, minus a safety margin. For example, using the operator's given -108dBm as the edge coverage threshold, and adding an additional safety margin of approximately 2dB, the blind zone determination threshold is set to -110dBm.

[0035] The vehicle-mounted terminal reads the current network signal strength according to a predetermined detection cycle (e.g., every 100 milliseconds or every 200 milliseconds) and compares the detection results with a blind zone determination threshold. If multiple consecutive detection results are below the blind zone determination threshold, the vehicle-mounted terminal determines that the current geographical location is in a signal blind zone. After determining that it has entered a signal blind zone, the vehicle-mounted terminal no longer attempts to send newly generated medical waste collection data to the server in real time. Instead, it appends a timestamp generated by a local clock (the timestamp can be in millisecond or microsecond precision) to each newly generated medical waste collection data record and pushes these data records into the vehicle-mounted terminal's local cache queue in chronological order of the timestamps.

[0036] The local cache queue can adopt a first-in, first-out (FIFO) structure. Each queue element must include at least the medical waste collection data content, a record generation timestamp, and an upload status flag. As the vehicle travels through long tunnels or multiple consecutive blind spots, the onboard terminal continuously writes newly generated medical waste collection data into the local cache queue, ensuring that medical waste collection data is not lost during periods when wireless signals are unavailable. All cached data that has not yet been uploaded to the server is collectively referred to as medical waste collection data to be uploaded in this description.

[0037] S2. The vehicle terminal continuously polls the network signal strength in the signal blind zone; when the network signal strength is detected to be higher than the preset signal recovery judgment threshold, the vehicle terminal determines that it has entered the signal recovery zone and generates a random backoff delay. The method for generating random backoff delay includes: Obtain the total amount of medical waste collection data to be uploaded in the local cache queue. Divide the total amount of data by the preset data volume normalization benchmark value to obtain the normalized data load ratio. Compare the normalized data load ratio with the preset upper and lower bound values ​​of the coefficient. When the normalized data load ratio is lower than the lower bound value of the coefficient, take the lower bound value of the coefficient. When the normalized data load ratio is higher than the upper bound value of the coefficient, take the upper bound value of the coefficient. When the normalized data load ratio falls between the preset lower and upper bound values ​​of the coefficient, directly take the normalized data load ratio. Use the truncated result as the dynamic backoff coefficient. For example, to determine when the vehicle-mounted terminal enters the signal recovery state, the system needs to pre-set a signal recovery judgment threshold during the deployment phase. The signal recovery judgment threshold is determined through simulation and actual testing: on the one hand, it needs to be higher than the blind zone judgment threshold to ensure that the wireless link quality is sufficient to support data services when uploading resumes; on the other hand, it cannot be set too high to avoid situations where the signal has actually recovered and is usable but is mistakenly considered unsuitable for uploading. For example, in empirical testing, if the vehicle-mounted terminal can stably maintain data services near -95 dBm, the signal recovery judgment threshold can be set to -95 dBm.

[0038] When the vehicle-mounted terminal continuously polls the network signal strength within a signal blind zone, once it detects that the network signal strength has repeatedly exceeded the signal recovery threshold, the vehicle-mounted terminal determines that the current area is a signal recovery zone and meets the basic conditions for initiating data upload. To reduce the instantaneous concurrency impact caused by multiple vehicle-mounted terminals simultaneously initiating uploads at the same exit or in the same recovery zone, after the vehicle-mounted terminal enters the signal recovery zone, a random backoff delay is first calculated based on the amount of medical waste collection data to be uploaded and the number of other vehicle-mounted terminals performing upload operations in the wireless channel. Upload only begins after the timer has finished counting down.

[0039] Before generating random backoff delay, a dynamic backoff coefficient needs to be calculated. The vehicle terminal reads the total amount of medical waste collection data to be uploaded in the local cache queue. The total amount of data can be measured in various ways, such as: (1) by the number of data records; (2) by the total number of bytes; (3) by the weighted equivalent business volume (e.g., assigning different weights according to the importance of different types of records). In a simplified implementation, the number of data records can be used as the measure of total data volume. In order to facilitate the unified comparison of data volume under different vehicles and different backlog levels, the system determines a data volume normalization benchmark value during the design phase. The selection of the data volume normalization benchmark value can be done by statistical analysis: the operator counts the average number of backlog data records generated by each vehicle before and after entering the tunnel on a typical operating route within a certain period of time, and combines the concurrent processing capacity that the server can withstand to select a value that can represent the "medium backlog level" as the data volume normalization benchmark value, such as selecting 1000 records as the benchmark value.

[0040] The vehicle terminal divides the actual total data volume by the normalized baseline value to obtain the normalized data load ratio. For example, if the total data volume is 2500 records and the normalized baseline value is 1000 records, the normalized data load ratio is 2.5. To avoid the dynamic backoff coefficient being too low or too high in extreme cases, the system also presets a lower bound and an upper bound value during the design phase. The determination of the lower bound and upper bound value also relies on test data: maintenance personnel can measure the upload peak shifting effect corresponding to different backoff coefficients in a test environment, select the minimum backoff strength that ensures stable system operation as the lower bound value, such as 0.5; and select the maximum backoff strength that can still guarantee business timeliness within the acceptable maximum initial waiting time as the upper bound value, such as 3. The vehicle terminal compares the normalized data load ratio with the lower and upper bounds of the coefficient to ensure that the dynamic backoff coefficient reflects the amount of data to be uploaded and is kept within a reasonable range by the upper and lower bounds. This helps to automatically expand the backoff time window when the amount of data is large and shorten the backoff time window when the amount of data is small.

[0041] The number of other vehicle terminals currently performing upload operations is counted by monitoring the received signal strength indicator value. The method for counting the number of other vehicle terminals currently performing upload operations by monitoring the received signal strength indicator value includes: Within the channel sampling window, the received signal strength indicator value is continuously sampled at equal time intervals. Each sample value is compared with a preset occupancy decision threshold. When the sample value is lower than the occupancy decision threshold, the sampling point is marked as an idle point; when the sample value is not lower than the occupancy decision threshold, the sampling point is marked as an occupied point. A sampling segment with at least three consecutive occupancy points is identified as a candidate transmission burst, and the start and end times of the candidate transmission burst are recorded as the burst duration. For each candidate transmission burst, the vehicle terminal attempts to demodulate and decode the radio frequency signal within the burst duration. If a complete data frame is successfully extracted and the unique identifier of the source vehicle terminal is parsed from the data frame header, the unique identifier of the source vehicle terminal is added to the set of identified terminals. If demodulation and decoding fail but the duration of the candidate transmission burst falls within the preset normal medical waste collection data frame duration range, the candidate transmission burst will be counted in the unknown terminal burst count; if demodulation and decoding fail and the duration exceeds the preset normal medical waste collection data frame duration range, the candidate transmission burst will be regarded as a channel interference event and discarded. After the channel sampling window ends, the unique identifiers of the source vehicle terminals in the identified terminal set are deduplicated to obtain the number of other identified vehicle terminals. The unknown terminal burst count is then multiplied by the dynamic reduction factor to obtain the estimated number of unknown terminals. The method for determining the dynamic reduction factor includes: After the channel sampling window ends, the total number of candidate transmission bursts identified within the channel sampling window is counted. The number of candidate transmission bursts that successfully completed demodulation and decoding and extracted the unique identifier of the source vehicle terminal is taken as the number of successful decodings. The demodulation and decoding success rate of the current channel sampling window is obtained by dividing the number of successful decodings by the total number of candidate transmission bursts. The burst duration of all candidate transmission bursts within the current channel sampling window is obtained. The arithmetic mean of the burst durations of all candidate transmission bursts is compared with the center value of the preset normal medical waste collection data frame duration interval to obtain the burst duration deviation ratio. When the burst duration deviation ratio exceeds the preset deviation tolerance range, the demodulation decoding success rate and the burst duration deviation ratio are weighted and fused to obtain a dynamic reduction coefficient. When the burst duration deviation ratio does not exceed the preset deviation tolerance range, the demodulation decoding success rate is used as the dynamic reduction coefficient.

[0042] The number of other identified vehicle terminals is added to the estimated number of unknown terminals to obtain the number of other vehicle terminals currently performing the upload operation.

[0043] For example, after obtaining the dynamic backoff coefficient, the vehicle-mounted terminal also needs to estimate the number of other vehicle-mounted terminals simultaneously performing upload operations in the wireless channel. To this end, after entering the signal recovery zone, the vehicle-mounted terminal samples and monitors the current wireless channel for a period of time. Within a channel sampling window (the length of the channel sampling window is selected based on the communication frame structure and latency sensitivity, for example, from several hundred milliseconds to several seconds), the vehicle-mounted terminal continuously samples the received signal strength indication value at fixed time intervals. The sampling interval can be determined by comprehensively considering the communication frame length and processing capability, for example, 1 millisecond or 2 milliseconds, to ensure that multiple sampling points can be obtained within the transmission time of a typical medical waste collection data frame.

[0044] Within the channel sampling window, the received signal strength indication (RSI) value obtained from each sample needs to be compared with an occupancy decision threshold. The occupancy decision threshold distinguishes between an idle channel and a channel with transmission or interference. The value of the occupancy decision threshold can be determined through experimental calibration: record the distribution of RSI values ​​under no-uplink background noise, and record the corresponding RSI distribution when a single vehicle terminal is uploading data. Select a suitable boundary line between the two sets of data as the occupancy decision threshold, for example, -100dBm. The vehicle terminal further searches for consecutive occupancy points in the sampling point sequence. When the number of consecutive occupancy points is not less than a certain threshold, the corresponding time period is identified as a candidate transmission burst. The threshold for the number of consecutive occupancy points is set according to the sampling interval and the minimum data frame length. For example, if the shortest transmission time per frame is 2 milliseconds and the sampling interval is 0.5 milliseconds, then at least 4 consecutive occupancy points are needed to cover a complete minimum frame; therefore, the threshold can be set to 4 or greater. The vehicle terminal records the start and end times for each candidate transmission burst and calculates the burst duration.

[0045] During the duration of each candidate transmission burst, the vehicle terminal attempts to demodulate and decode the received radio frequency signal. If decoding is successful and the unique identifier of the source vehicle terminal can be parsed from the data frame header field (e.g., using the vehicle terminal device serial number, International Mobile Equipment Identity, or a virtual terminal identifier obtained through dedicated identifier mapping), the vehicle terminal adds the unique identifier to the "identified terminal set".

[0046] When demodulation and decoding of candidate transmission bursts fail, the vehicle-mounted terminal further filters the candidate transmission bursts using the "normal medical waste collection and transportation data frame duration range." The normal medical waste collection and transportation data frame duration range is determined jointly by the communication protocol design and link rate. During the system design phase, the theoretical frame duration is calculated based on the actual deployed uplink physical layer parameters and application layer message length, and combined with the actual frame duration distribution tested, an upper and lower limit range is determined. For example, if the theoretical frame duration is 3 milliseconds, and test results show that most frame durations are between 2 and 5 milliseconds, then the 2 to 5 millisecond range can be used as the normal medical waste collection and transportation data frame duration range.

[0047] If demodulation and decoding of a candidate transmission burst fails, but the burst duration falls within the normal medical waste collection data frame duration range, the vehicle terminal infers that the burst is highly likely to be medical waste collection data sent by a certain vehicle terminal, and therefore includes the candidate transmission burst in the "Unknown Terminal Burst Count". If the duration of the candidate transmission burst is much less than the lower limit of the interval or much greater than the upper limit of the interval, it is more likely to be other types of services or interference signals, and the vehicle terminal considers it a channel interference event and does not include it in the statistics. After a channel sampling window ends, the unique identifiers in the identified terminal set are deduplicated to obtain the number of other vehicle terminals clearly identified through demodulation and decoding. At the same time, the vehicle terminal also counts the total number of candidate transmission bursts identified within the channel sampling window, as well as the number of candidate transmission bursts successfully decoded.

[0048] To further improve the reasonableness of estimating the number of unknown terminals, the vehicle-mounted terminal also acquires the duration of all candidate transmission bursts within the channel sampling window, calculates the arithmetic mean of these durations, and compares it with the center value of the duration interval of normal medical waste collection data frames to obtain the burst duration deviation ratio. The center value of the duration interval of normal medical waste collection data frames is equal to the average of the upper and lower limits of the interval. The burst duration deviation ratio can be defined as the absolute value of the difference between the average burst duration and the center value of the interval divided by the center value of the interval. If the burst duration deviation ratio is greater than the preset deviation tolerance range, it indicates that the current candidate transmission burst differs significantly from the normal medical waste collection data frame in terms of average duration. The deviation tolerance range needs to be set in conjunction with test data, and is generally between 0.1 and 0.3. For example, 0.2 means that the average duration is allowed to fluctuate within ±20% of the center value and is still considered normal. The deviation tolerance range is also determined through experimental data analysis and business experience.

[0049] When the burst duration deviation rate exceeds the tolerance range, the vehicle terminal weights and fuses the demodulation / decoding success rate with the degree of burst duration deviation to obtain a dynamic reduction coefficient. This weighting formula can be used as follows: Dynamic Reduction Coefficient = ×Demodulation / decoding success rate + β × f (burst duration deviation ratio); where α and β are weighting coefficients, satisfying α + β = 1; f (burst duration deviation ratio) is a function that standardizes the burst duration deviation ratio, and can be designed as a function that monotonically decreases as the burst duration deviation ratio increases, for example f( ) = max(0,1- / ), α represents the maximum acceptable deviation ratio in engineering. The values ​​of α and β can be determined through simulation and experimental calibration. For example, setting α = 0.7 and β = 0.3 gives the decoding success rate a higher weight in the dynamic reduction coefficient, while the burst duration deviation only plays an auxiliary correction role.

[0050] When the deviation rate of the burst duration does not exceed the deviation tolerance range, it indicates that the duration of the candidate transmission burst is consistent with the overall duration of the normal medical waste collection and transportation data frame. In this case, the demodulation and decoding success rate can be directly used as the dynamic reduction coefficient.

[0051] The number of other vehicle terminals is calculated according to a preset linear growth rule to obtain the competition backoff increment; the random backoff delay is calculated based on the dynamic backoff coefficient and the competition backoff increment.

[0052] The method for calculating the random backoff delay based on the dynamic backoff coefficient and the competitive backoff increment includes: Multiply the dynamic backoff coefficient by the preset basic backoff window upper limit to obtain the dynamic backoff window upper limit. Use the preset basic backoff window lower limit as the lower bound and the dynamic backoff window upper limit as the upper bound to construct the dynamic backoff time window of the current vehicle terminal. Random sample values ​​are generated according to a uniform random distribution within the dynamic backoff time window; The method for generating random sample values ​​according to a uniform random distribution within the dynamic backoff time window includes: Obtain the unique identifier of the vehicle terminal, perform a bitwise XOR operation between the unique identifier and the microsecond-level timestamp of the local clock of the vehicle terminal at the signal recovery time, and use the bitwise XOR operation result as a random number seed; substitute the random number seed into the recursive formula of the linear congruence generator for iterative calculation to obtain the current item in the pseudo-random integer sequence; The interval length of the dynamic backoff time window is converted into an integer interval length in microseconds. The current item is modulo the integer interval length to obtain a normalized integer offset. The normalized integer offset is added to the preset lower limit of the basic backoff window to obtain a random sampled value.

[0053] Use the random sampled value as the initial backoff delay; add the initial backoff delay to the contention backoff increment to obtain the random backoff delay.

[0054] For example, the contention backoff increment can be obtained through the following linear relationship: Contention Backoff Increment = K × N; where N is the estimated number of other vehicle terminals concurrently uploading, and K is the time growth factor, in units of "time / terminal," used to represent the amount of time increase in random backoff latency when each additional concurrent vehicle terminal is added. The value of the time growth factor K needs to be determined through system-level simulation and actual road testing. During simulation, different values ​​of K can be simulated to observe the changing trends of server queuing length, response latency, and upload success rate. A K value that can effectively smooth out concurrency peaks without introducing excessive latency can be selected, for example, K can be set to 100 milliseconds / terminal or 200 milliseconds / terminal.

[0055] After obtaining the dynamic backoff coefficient and competitive backoff increment, the vehicle terminal begins to calculate the random backoff latency. First, during the system design phase, a basic backoff window upper limit and a basic backoff window lower limit are uniformly set for all vehicles. The basic backoff window upper limit can be determined based on the maximum acceptable initial waiting time for the business and the server's ability to handle instantaneous concurrency. For example, if testing shows that a wait of less than 5 seconds does not affect the management efficiency of medical waste collection, the basic backoff window upper limit can be set to 5 seconds. The basic backoff window lower limit can be set to avoid multiple vehicles simultaneously initiating uploads within a very small time difference due to excessively short wait times; for example, it can be set to 0.5 seconds or 1 second. The basic backoff window upper and lower limits are also calibrated through simulation or large-scale trial operation.

[0056] The lower bound of the dynamic backoff time window is taken from the lower limit of the basic backoff window, and the upper bound is taken from the upper limit of the dynamic backoff window, thus obtaining a backoff time range associated with the amount of data to be uploaded. For example, if the upper limit of the basic backoff window is 3 seconds, the lower limit of the basic backoff window is 1 second, and the dynamic backoff coefficient is 2, the dynamic backoff time window is [1 second, 6 seconds].

[0057] To avoid multiple vehicle terminals using highly correlated random number seeds at the same time, the vehicle terminal incorporates a unique identifier and a microsecond-level timestamp from the local clock at the signal recovery point into the random number seed construction. The vehicle terminal first reads its unique identifier. This unique identifier can be written by the manufacturer or system integrator during vehicle terminal production or installation; for example, it can be a globally unique serial number or a hashed combination of the license plate number and the device serial number. Bitwise XOR operations effectively mix vehicle-specific information with temporal randomness, ensuring good distinguishability of seeds generated by different vehicles at different times. The vehicle terminal uses a linear congruential generator to generate pseudo-random integers. The recursive formula for the linear congruential generator is: =(a× +c) mod m; where, Let be the pseudo-random integer value for the k-th iteration, where 'a' is the multiplier, 'c' is the increment constant, and 'm' is the modulus. The specific values ​​of 'a', 'c', and 'm' can be determined using classic parameter combinations or through randomness testing. For example, 'm' can be chosen as 2³² or 2³¹⁻¹, and 'a' and 'c' can be chosen as integers that satisfy long periods and good distribution characteristics. The random number seed is used as... Substituting into the above formula, we can calculate the result. and will As the current pseudo-random integer.

[0058] The vehicle-mounted terminal converts the interval length (upper bound minus lower bound) of the dynamic backoff time window into an integer length L in microseconds. For example, if the dynamic backoff time window is [1 second, 6 seconds], then the interval length is 5 seconds, corresponding to L = 5,000,000 microseconds. The vehicle-mounted terminal converts the pseudo-random integer... Modulo L, we obtain a normalized integer offset within the range [0, L-1]. The vehicle terminal then converts the lower limit of the basic backoff window into microseconds and adds it to the normalized integer offset to obtain a random sampled value within the dynamic backoff time window. This random sampled value, after unit conversion, is used as the initial backoff delay.

[0059] The vehicle-mounted terminal adds the initial backoff delay to the contention backoff increment to obtain the final random backoff delay. When there are many other vehicle-mounted terminals, the contention backoff increment is larger, and the final random backoff delay also increases accordingly, thus dispersing the upload start time of concurrent vehicle-mounted terminals over a longer period of time. S3. After generating a random backoff delay, the vehicle terminal starts a local timer and waits for the random backoff delay. Then, according to the preset data fragment size threshold, it divides the medical waste collection data to be uploaded into M data fragments in the order of timestamps. Each data fragment is accompanied by a fragment number and the corresponding timestamp range. For example, after acquiring the random backoff delay, the vehicle-mounted terminal starts a local timer, and begins data uploading when the timer expires. Before entering the upload phase, the medical waste collection data to be uploaded is divided into multiple data fragments according to a preset data fragmentation strategy. The data fragmentation strategy is determined by the operations and maintenance team based on network bandwidth, server processing capacity, and the maximum allowed message length for a single HTTP or TCP request. A data fragment size threshold can be set in the system configuration, for example, by configuring the maximum number of records per fragment (e.g., 500 records) or the maximum number of bytes (e.g., 256KB).

[0060] Starting from the head of the local cache queue, the vehicle-mounted terminal sequentially adds the medical waste collection and transportation data to be uploaded to the current data segment according to timestamp order. This continues until the number of records or the data size of the current segment reaches the data segment size threshold. At this point, the current segment is encapsulated into an independent data segment and assigned a segment number. The segment number increments from 1. The vehicle-mounted terminal also calculates the minimum and maximum timestamp values ​​recorded in the segment and appends this timestamp range to the segment's control information. After all the medical waste collection and transportation data to be uploaded has been assigned to various data segments, a total of M data segments are obtained.

[0061] S4. The vehicle terminal sends upload requests to the server piece by piece according to the fragment sequence number; The method by which the vehicle-mounted terminal initiates upload requests to the server piece by piece according to the fragment sequence number includes: When the vehicle terminal initiates an upload request for the current data segment, it records the time when the upload request message is sent. When it receives a response status code from the server carrying the corresponding segment sequence number, it records the time when the response status code arrives. The round-trip time of the current data segment is calculated based on the time when the upload request message is sent and the time when the response status code arrives. For example, when initiating each data fragment upload request, the vehicle-mounted terminal constructs an upload request message. This message includes fields such as fragment sequence number, timestamp range, unique identifier of the vehicle-mounted terminal, and medical waste collection data. The vehicle-mounted terminal records its local clock time as the "upload request sent moment" the instant it sends the upload request message. After successfully receiving and processing a data fragment, the server returns a response status code containing the corresponding fragment sequence number to the vehicle-mounted terminal. The vehicle-mounted terminal records the "response status code arrival moment" the instant it receives the response status code, using the difference between the two as the round-trip time for that data fragment upload.

[0062] The round-trip time obtained after the first data slice is completed is used as the initial smoothed round-trip time. After each subsequent data slice is uploaded, the current round-trip time is weighted and merged with the previous round of smoothed round-trip time using an exponentially weighted moving average method to obtain a continuously updated smoothed round-trip time. A sliding observation window is established for the round-by-round update results of the smoothed round-trip time, and a dynamic time delay baseline is calculated within the sliding observation window. The method for establishing a sliding observation window based on the round-by-round update results of the smooth round-trip delay, and calculating the dynamic delay baseline within the sliding observation window, includes: Before uploading the first data fragment, a fixed-length first-in-first-out queue is initialized as a sliding observation window. The smooth round-trip time obtained after each data fragment upload is completed is pushed into the tail of the sliding observation window. When the number of smooth round-trip times stored in the sliding observation window reaches a fixed length, the earliest smooth round-trip time at the head of the queue is popped out each time a new smooth round-trip time is pushed in. After the sliding observation window has been filled with a fixed length of smooth round-trip delays, whenever it is necessary to calculate the dynamic delay baseline, the arithmetic mean of all smooth round-trip delays within the sliding observation window is calculated as the current window mean, and the arithmetic mean of the absolute deviations of all smooth round-trip delays from the current window mean is calculated as the current window average deviation. The sum of the current window mean and the current window average deviation is used as the dynamic delay baseline.

[0063] For example, for the first data fragment that has been uploaded, the vehicle terminal directly uses the corresponding round-trip time (RTT) as the initial smoothed RTT. When uploading the second and subsequent data fragments, the vehicle terminal calculates the actual RTT for that fragment and updates the smoothed RTT using an exponentially weighted moving average method. The formula for updating the smoothed RTT can be expressed as: ;in, The smooth round-trip latency obtained after uploading the k-th data fragment; Smooth round-trip latency after the previous data fragment is uploaded; The actual round-trip time for the k-th data segment; For smoothing coefficients, The value range is (0,1). The settings can be adjusted by analyzing network latency fluctuations and convergence speed requirements. For example, when network fluctuations are relatively smooth, you can select... =0.1, which gives a relatively large weight to historical data, preventing short-term jitter from having an excessive impact on smoothing round-trip latency; when network fluctuations are severe, it can be appropriately increased. To improve the speed of response to new changes. The optimal value can be determined through multiple rounds of network testing and simulation.

[0064] To monitor the trend of smoothed round-trip latency (RTL) over a period of time, the vehicle-mounted terminal establishes a fixed-length sliding observation window for the RTL sequence. Before uploading the first fragment, the vehicle-mounted terminal initializes a first-in-first-out (FIFO) queue as the sliding observation window, with the queue length set to a fixed integer, such as 10 or 20. The queue length is determined by the operations team based on the average number of fragments in each upload task and the historical latency range to be observed. Whenever a data fragment is uploaded and a new RTL is calculated, the vehicle-mounted terminal pushes that RTL to the tail of the sliding observation window queue. If the queue length has not yet reached the set value, it continues to accumulate; when the number of RTLs in the queue reaches the fixed length, each time a new RTL is pushed in, the earliest RTL needs to be popped from the head of the queue to ensure that the window length remains unchanged. The dynamic latency baseline reflects both the general level of network latency during the observation period and the average amplitude of latency fluctuations, thus it can be used to identify abnormally high latency.

[0065] Whenever the latest round-trip delay exceeds the dynamic delay baseline, it is determined that the concurrent uploading of multiple collection vehicles in the signal recovery area has caused congestion pressure on the server, triggering the adaptive transmission adjustment process to initiate an upload request.

[0066] The method for triggering the adaptive sending adjustment process to initiate an upload request includes: The time delay deviation ratio is obtained by dividing the difference between the smooth round-trip time delay of the current round and the dynamic time delay baseline by the dynamic time delay baseline; the time delay trend coefficient is obtained by dividing the smooth round-trip time delay of the current round by the smooth round-trip time delay of the previous round. A time delay trend coefficient greater than one indicates that the congestion of concurrent upload in the signal recovery area at the tunnel exit continues to worsen, while a time delay trend coefficient less than one indicates that the congestion tends to ease. Multiply the delay deviation ratio by the delay trend coefficient to obtain the comprehensive congestion adjustment factor; multiply the comprehensive congestion adjustment factor by the preset transmission interval benchmark value to obtain the transmission interval duration for the current round; after completing the upload request for the current data segment, the vehicle terminal waits for the transmission interval duration before initiating the upload request for the next data segment.

[0067] If the vehicle terminal receives a response status code from the server carrying the corresponding fragment sequence number, it determines that the current data fragment upload and update has been successful, and initiates the upload request for the next data fragment, until the Mth data fragment is successfully uploaded and updated.

[0068] For example, after each subsequent fragment upload and smoothed round-trip time update, the vehicle terminal compares the latest smoothed round-trip time with the current dynamic latency baseline. If the smoothed round-trip time exceeds the dynamic latency baseline, it indicates that the current network latency is significantly higher than the recent average level, which is highly likely caused by multiple vehicle terminals concurrently uploading in the signal recovery area. At this time, the vehicle terminal initiates an adaptive transmission adjustment process to adjust the transmission interval of subsequent fragment uploads, thereby smoothing the server load.

[0069] In the adaptive transmission adjustment process, the vehicle-mounted terminal first calculates the latency deviation ratio. If the latency deviation ratio is positive and large, it indicates that the current latency is significantly higher than the baseline in absolute terms. If the latency trend coefficient is greater than 1, it indicates that the latency has been increasing between the last two uploads, and congestion is worsening; if the latency trend coefficient is less than 1, it indicates that the congestion has eased. The comprehensive congestion adjustment factor characterizes the absolute degree of latency deviation and the impact of latency change trends on congestion. During the system design phase, a transmission interval baseline value is set for all vehicle-mounted terminals. The transmission interval baseline value can be understood as the default time interval between two adjacent data fragment upload requests when there is no significant congestion. The transmission interval baseline value can be determined by evaluating server processing capacity and business real-time requirements, for example, 1 second or 500 milliseconds.

[0070] When the overall congestion adjustment factor is greater than 1, the transmission interval will be greater than the transmission interval baseline value, thereby reducing the subsequent fragment upload rate and helping to alleviate congestion; when the overall congestion adjustment factor is close to 1 or slightly less than 1, the transmission interval will be close to or slightly less than the transmission interval baseline value, in order to maintain or slightly increase the upload rate.

[0071] After successfully uploading each data fragment and receiving a response status code from the server, the vehicle-mounted terminal waits according to the current round's transmission interval. It only initiates the next fragment upload request after the interval has elapsed. If the vehicle-mounted terminal successfully receives a response status code from the server carrying the corresponding fragment sequence number, it considers the current data fragment upload successful and marks the corresponding data as uploaded or deletes it from its local cache queue. Then, it continues uploading the next fragment in sequence. The vehicle-mounted terminal repeats this process until the server confirms the successful upload of all M data fragments.

[0072] Finally, it should be noted that although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for updating medical waste data across different stages, characterized in that, The method includes: During operation, the vehicle-mounted terminal continuously collects medical waste collection data and monitors the current network signal strength in real time. When the network signal strength is detected to be lower than the preset blind zone determination threshold, the vehicle-mounted terminal is determined to have entered the signal blind zone. The medical waste collection data collected thereafter is written into the local cache queue in order with the timestamp as the index and recorded as medical waste collection data to be uploaded. The vehicle-mounted terminal continuously polls the network signal strength in the signal blind zone; when the network signal strength is detected to be higher than the preset signal recovery judgment threshold, the vehicle-mounted terminal determines that it has entered the signal recovery zone and generates a random backoff delay. After generating a random backoff delay, the vehicle terminal starts a local timer and waits for the random backoff delay. Then, according to the preset data fragment size threshold, it divides the medical waste collection data to be uploaded into M data fragments in the order of timestamps. Each data fragment is accompanied by a fragment number and the corresponding timestamp range. The vehicle terminal initiates upload requests to the server piece by piece according to the fragment sequence number; if the vehicle terminal receives a response status code from the server carrying the corresponding fragment sequence number, it determines that the current data fragment upload and update is successful, and initiates the upload request for the next data fragment, until the Mth data fragment upload and update is successful.

2. The method for updating medical waste data across different stages according to claim 1, characterized in that, The method for generating random backoff delay includes: Obtain the total amount of medical waste collection data to be uploaded in the local cache queue. Divide the total amount of data by the preset data volume normalization benchmark value to obtain the normalized data load ratio. Compare the normalized data load ratio with the preset upper and lower bound values ​​of the coefficient. When the normalized data load ratio is lower than the lower bound value of the coefficient, take the lower bound value of the coefficient. When the normalized data load ratio is higher than the upper bound value of the coefficient, take the upper bound value of the coefficient. When the normalized data load ratio falls between the preset lower and upper bound values ​​of the coefficient, directly take the normalized data load ratio. Use the truncated result as the dynamic backoff coefficient. The number of other vehicle terminals currently performing upload operations is counted by monitoring the received signal strength indicator value, and the number of other vehicle terminals is calculated according to a preset linear growth rule to obtain the contention backoff increment; the random backoff delay is calculated based on the dynamic backoff coefficient and the contention backoff increment.

3. The method for updating medical waste data across different stages according to claim 2, characterized in that, The method for counting the number of other vehicle terminals currently performing upload operations by monitoring the received signal strength indicator value includes: Within the channel sampling window, the received signal strength indicator value is continuously sampled at equal time intervals. Each sample value is compared with a preset occupancy decision threshold. When the sample value is lower than the occupancy decision threshold, the sampling point is marked as an idle point; when the sample value is not lower than the occupancy decision threshold, the sampling point is marked as an occupied point. A sampling segment with at least three consecutive occupancy points is identified as a candidate transmission burst, and the start and end times of the candidate transmission burst are recorded as the burst duration. For each candidate transmission burst, the vehicle terminal attempts to demodulate and decode the radio frequency signal within the burst duration. If a complete data frame is successfully extracted and the unique identifier of the source vehicle terminal is parsed from the data frame header, the unique identifier of the source vehicle terminal is added to the set of identified terminals. If demodulation and decoding fail but the duration of the candidate transmission burst falls within the preset normal medical waste collection data frame duration range, the candidate transmission burst will be counted in the unknown terminal burst count; if demodulation and decoding fail and the duration exceeds the preset normal medical waste collection data frame duration range, the candidate transmission burst will be regarded as a channel interference event and discarded. After the channel sampling window ends, the unique identifiers of the source vehicle terminals in the identified terminal set are deduplicated to obtain the number of other identified vehicle terminals. The unknown terminal burst count is multiplied by a dynamic reduction factor to obtain the estimated number of unknown terminals. The number of other identified vehicle terminals is added to the estimated number of unknown terminals to obtain the number of other vehicle terminals currently performing the upload operation.

4. The method for updating medical waste data across different stages according to claim 3, characterized in that, The method for determining the dynamic reduction factor includes: After the channel sampling window ends, the total number of candidate transmission bursts identified within the channel sampling window is counted. The number of candidate transmission bursts that successfully completed demodulation and decoding and extracted the unique identifier of the source vehicle terminal is taken as the number of successful decodings. The demodulation and decoding success rate of the current channel sampling window is obtained by dividing the number of successful decodings by the total number of candidate transmission bursts. The burst duration of all candidate transmission bursts within the current channel sampling window is obtained. The arithmetic mean of the burst durations of all candidate transmission bursts is compared with the center value of the preset normal medical waste collection data frame duration interval to obtain the burst duration deviation ratio. When the burst duration deviation ratio exceeds the preset deviation tolerance range, the demodulation decoding success rate and the burst duration deviation ratio are weighted and fused to obtain a dynamic reduction coefficient. When the burst duration deviation ratio does not exceed the preset deviation tolerance range, the demodulation decoding success rate is used as the dynamic reduction coefficient.

5. A method for updating medical waste data across different stages according to claim 2, characterized in that, The method for calculating the random backoff delay based on the dynamic backoff coefficient and the competitive backoff increment includes: Multiply the dynamic backoff coefficient by the preset basic backoff window upper limit to obtain the dynamic backoff window upper limit. Use the preset basic backoff window lower limit as the lower bound and the dynamic backoff window upper limit as the upper bound to construct the dynamic backoff time window of the current vehicle terminal. Within the dynamic backoff time window, random sample values ​​are generated according to a uniform random distribution, and these random sample values ​​are used as the initial backoff delay. The initial backoff delay is added to the contention backoff increment to obtain the random backoff delay.

6. A method for updating medical waste data across different stages according to claim 5, characterized in that, The method for generating random sample values ​​according to a uniform random distribution within the dynamic backoff time window includes: Obtain the unique identifier of the vehicle terminal, perform a bitwise XOR operation between the unique identifier and the microsecond-level timestamp of the local clock of the vehicle terminal at the signal recovery time, and use the bitwise XOR operation result as a random number seed; substitute the random number seed into the recursive formula of the linear congruence generator for iterative calculation to obtain the current item in the pseudo-random integer sequence; The interval length of the dynamic backoff time window is converted into an integer interval length in microseconds. The current item is modulo the integer interval length to obtain a normalized integer offset. The normalized integer offset is added to the preset lower limit of the basic backoff window to obtain a random sampled value.

7. A method for updating medical waste data across different stages according to claim 1, characterized in that, The method by which the vehicle-mounted terminal initiates upload requests to the server piece by piece according to the fragment sequence number includes: When the vehicle terminal initiates an upload request for the current data segment, it records the time when the upload request message is sent. When it receives a response status code from the server carrying the corresponding segment sequence number, it records the time when the response status code arrives. The round-trip time of the current data segment is calculated based on the time when the upload request message is sent and the time when the response status code arrives. The round-trip time obtained after the first data slice is completed is used as the initial smoothed round-trip time. After each subsequent data slice is uploaded, the current round-trip time is weighted and merged with the previous round of smoothed round-trip time using an exponentially weighted moving average method to obtain a continuously updated smoothed round-trip time. A sliding observation window is established for the round-by-round update results of the smoothed round-trip time, and a dynamic time delay baseline is calculated within the sliding observation window. Whenever the latest round-trip delay exceeds the dynamic delay baseline, it is determined that the concurrent uploading of multiple collection vehicles in the signal recovery area has caused congestion pressure on the server, triggering the adaptive transmission adjustment process to initiate an upload request.

8. A method for updating medical waste data across different stages according to claim 7, characterized in that, The method for establishing a sliding observation window based on the round-by-round update results of the smooth round-trip delay, and calculating the dynamic delay baseline within the sliding observation window, includes: Before uploading the first data fragment, a fixed-length first-in-first-out queue is initialized as a sliding observation window. The smooth round-trip time obtained after each data fragment upload is completed is pushed into the tail of the sliding observation window. When the number of smooth round-trip times stored in the sliding observation window reaches a fixed length, the earliest smooth round-trip time at the head of the queue is popped out each time a new smooth round-trip time is pushed in. After the sliding observation window has been filled with a fixed length of smooth round-trip delays, whenever it is necessary to calculate the dynamic delay baseline, the arithmetic mean of all smooth round-trip delays within the sliding observation window is calculated as the current window mean, and the arithmetic mean of the absolute deviations of all smooth round-trip delays from the current window mean is calculated as the current window average deviation. The sum of the current window mean and the current window average deviation is used as the dynamic delay baseline.

9. A method for updating medical waste data across different stages according to claim 7, characterized in that, The method for triggering the adaptive sending adjustment process to initiate an upload request includes: The time delay deviation ratio is obtained by dividing the difference between the smooth round-trip time delay of the current round and the dynamic time delay baseline by the dynamic time delay baseline; the time delay trend coefficient is obtained by dividing the smooth round-trip time delay of the current round by the smooth round-trip time delay of the previous round. A time delay trend coefficient greater than one indicates that the congestion of concurrent upload in the signal recovery area at the tunnel exit continues to worsen, while a time delay trend coefficient less than one indicates that the congestion tends to ease. Multiply the delay deviation ratio by the delay trend coefficient to obtain the comprehensive congestion adjustment factor; multiply the comprehensive congestion adjustment factor by the preset transmission interval benchmark value to obtain the transmission interval duration for the current round; after completing the upload request for the current data segment, the vehicle terminal waits for the transmission interval duration before initiating the upload request for the next data segment.

Images