Vehicle-mounted buried point data weak network adaptive transmission method, device, equipment and system

Abstract

The application relates to a vehicle-mounted embedded point data weak network adaptive transmission method, device, equipment and system, and belongs to the technical field of vehicle-mounted data transmission. The vehicle-mounted embedded point data weak network adaptive transmission method comprises the following steps: performing multi-dimensional value scoring on accessed vehicle-mounted embedded point data to obtain data priority; storing the vehicle-mounted embedded point data into cache partitions according to the data priority; performing hierarchical compression coding on the cached vehicle-mounted embedded point data to generate three structured data with different granularities, namely a basic layer, an abstract layer and a complete layer; detecting a network state in real time and determining a current network quality level, selecting a corresponding transmission level according to the network quality level to perform data transmission; and recording local time stamps and time service time stamps during data collection, and correcting the time sequence of data to be transmitted according to the time service time stamps when detecting that the local clock has time drift. The application can guarantee reliable transmission of vehicle-mounted core embedded point data in a weak network environment.

Description

Technical Field

[0001] This invention relates to the field of vehicle data transmission technology, and in particular to a method, apparatus, device and system for adaptive transmission of vehicle embedded data in weak network conditions. Background Technology

[0002] Intelligent connected vehicles continuously generate a large amount of onboard data during operation, including core data such as vehicle status, autonomous driving logs, driving behavior, and fault events. This data is crucial for vehicle fault analysis and user habit research. Vehicle driving scenarios are complex, and areas such as tunnels, mountainous regions, and underground parking garages often experience weak networks, network outages, and network fluctuations, which significantly impact the reliable transmission of onboard data.

[0003] Existing in-vehicle data transmission solutions have rigid strategies and significant shortcomings: traditional solutions use a unified FIFO caching mechanism, which cannot distinguish the value of data services, and high-value security data is easily squeezed out and lost by low-value logs; fixed transmission strategies cannot adapt to weak or extremely weak network scenarios, resulting in low success rates for uploading core data; long-term vehicle offline operation causes local clock drift, resulting in disordered cloud data timing, which cannot support accident backtracking and user habit research; and bursts of uploads after network recovery easily cause repeated network congestion, resulting in poor transmission stability. Existing technologies cannot meet the high-reliability and high-precision data transmission requirements of embedded data points in complex driving scenarios. Summary of the Invention

[0004] In view of this, it is necessary to provide a method, device, equipment and system for adaptive transmission of vehicle-mounted embedded data in weak network environments, so as to solve the problems of easy loss of core embedded data in vehicles, low transmission success rate and timing disorder caused by offline clock drift in the existing technology.

[0005] To address the aforementioned problems, in a first aspect, the present invention provides a method for adaptive transmission of vehicle-mounted embedded data in weak network conditions, comprising: The accessed vehicle data points are evaluated for value across multiple dimensions to determine data priority. The vehicle-mounted data points are stored in the cache partitions according to the data priority. The cached vehicle tracking data is hierarchically compressed and encoded to generate structured data of three different granularities: basic layer, summary layer, and complete layer. The system detects the network status in real time and determines the current network quality level. Based on the network quality level, it selects the corresponding transmission layer from the base layer, the digest layer, and the full layer, and dynamically adjusts the corresponding number of concurrent paths and the transmission rate to adapt to the current network conditions for data transmission. Record the local timestamp and time synchronization timestamp when data is collected. When time drift is detected in the local clock, perform timing correction on the data to be transmitted according to the time synchronization timestamp. The timing correction is used to adapt the offline retransmission data to the standard time base after network failure recovery, and complete the weak network adaptive transmission of vehicle embedded data.

[0006] In one possible implementation, the step of performing multi-dimensional value scoring on the vehicle-mounted data to obtain data priority includes: Calculate the business weight score, timeliness decay score, and scenario association score of the vehicle-mounted embedded data respectively; A comprehensive score is calculated based on the business weight score, the timeliness decay score, and the scenario association score. Data priority is determined based on the comprehensive score.

[0007] In one possible implementation, the cache partition includes a first cache partition, a second cache partition, and a third cache partition, and the cache capacity of the three is allocated in a fixed ratio. The data priority includes a first data priority, a second data priority, a third data priority, a fourth data priority, and a fifth data priority, with the scores decreasing sequentially. The step of storing the vehicle-mounted data points into cache partitions according to the data priority includes: The vehicle-mounted data points corresponding to the first data priority and the second data priority are stored in the first cache partition, and no elimination trigger condition is set for the first cache partition. The vehicle-mounted data points corresponding to the third data priority and the fourth data priority are stored in the second cache partition. When the real-time storage capacity of the second cache partition reaches the first preset ratio of the partition quota, the data is eliminated based on the data idle time and data priority. The vehicle-mounted data points corresponding to the fifth data priority are stored in the third cache partition. When the real-time storage capacity of the third cache partition reaches the second preset ratio of the partition quota, an elimination is triggered, and a first-in-first-out elimination strategy is executed.

[0008] In one possible implementation, the base layer includes at least a core identifier field for the data, and the base layer is encoded using field filtering and Huffman compression. The summary layer adds core business fields on top of the base layer. Based on the encoding of the base layer, the summary layer uses differential compression for incremental encoding of continuously changing time-series values. The complete layer contains all the original fields collected, and the complete layer is deduplicated and compressed while retaining all the original fields.

[0009] In one possible implementation, the network quality level includes a first network quality level, a second network quality level, a third network quality level, and a fourth network quality level; The step of selecting the corresponding transport layer from the base layer, the digest layer, and the full layer based on the network quality level, and dynamically adjusting the corresponding concurrent paths and transmission rate to adapt to the current network conditions for data transmission, includes: When the network quality level is the first network quality level, the complete layer is selected for transmission, and multiple concurrent transmissions are enabled at the first transmission rate. When the network quality level is the second network quality level or the third network quality level, the base layer and / or the digest layer are selected for transmission to reduce the number of concurrent transmission paths and transmit at the second transmission rate or the third transmission rate, wherein the first transmission rate, the second transmission rate, and the third transmission rate are decreased sequentially. When the network quality level is the fourth network quality level, the transmission task is terminated and the network is awaited to recover.

[0010] In one possible implementation, the method further includes: Detect packet loss rate during data transmission; When the packet loss rate is less than the first packet loss rate threshold, maintain the current sending rate; When the packet loss rate is greater than or equal to the first packet loss rate threshold and less than the second packet loss rate threshold, the sending rate is reduced. When the packet loss rate is greater than or equal to the second packet loss rate threshold, batch transmission is paused and a backup transmission is performed.

[0011] In one possible implementation, the timing correction adopts a piecewise linear mapping method. For each offline interval, the start and end times of the offline interval are used as the reference anchor points, and the corresponding local clock time is used as the reference point. The local acquisition time of the data to be corrected is linearly mapped to calculate the corrected standard time.

[0012] Secondly, the present invention also provides an adaptive transmission device for vehicle-mounted embedded data in weak network conditions, comprising: The value scoring module is used to perform multi-dimensional value scoring on the accessed vehicle tracking data to obtain data priority; The cache management module is used to store the vehicle-mounted embedded data into cache partitions according to the data priority. The hierarchical compression and encoding module is used to perform hierarchical compression and encoding on the cached vehicle tracking data to generate structured data with three different granularities: the base layer, the summary layer, and the complete layer. The network adaptive transmission module is used to detect the network status in real time and determine the current network quality level. Based on the network quality level, it selects the corresponding transmission layer from the base layer, the digest layer, and the full layer, and dynamically adjusts the corresponding number of concurrent paths and the transmission rate to adapt to the current network conditions for data transmission. The timing correction module is used to record the local timestamp and the time synchronization timestamp during data acquisition. When a time drift is detected in the local clock, the timing correction is performed on the data to be transmitted according to the time synchronization timestamp. The timing correction is used to adapt the offline retransmission data to the standard time base after the network is restored, and to complete the weak network adaptive transmission of vehicle embedded data.

[0013] Thirdly, the present invention also provides an electronic device, including a data collector, a memory, and a processor, wherein the data collector is used to access vehicle-mounted embedded data. The memory is used to store programs; The processor, coupled to the memory, is used to execute the program stored in the memory to implement the steps in the vehicle-mounted embedded data weak network adaptive transmission method described in any of the above implementations.

[0014] Fourthly, the present invention also provides a vehicle-mounted embedded data weak network adaptive transmission system, the vehicle-mounted embedded data weak network adaptive transmission system including the electronic device, data access device and data output device as described above, the data access device and the data output device are respectively connected to the electronic device, the data access device is used to access vehicle-mounted embedded data, and the data output device is used to perform data transmission.

[0015] The beneficial effects of this invention are as follows: The vehicle-mounted embedded data weak network adaptive transmission method provided by this invention prioritizes vehicle-mounted embedded data through multi-dimensional value scoring, replacing the traditional indiscriminate caching method, and ensuring the priority retention of high-value security data from the source; by storing data of different priorities separately in corresponding cache partitions, combined with differentiated eviction strategies, it effectively solves the problem of core data being squeezed out and lost by low-value logs; by performing hierarchical compression encoding on cached data to generate structured data of three different granularities—basic layer, summary layer, and complete layer—it achieves gradient simplification of data volume, providing a precise basis for data granularity selection for network adaptive transmission; by detecting network status in real time and dynamically matching transmission level, concurrent paths, and sending rate, it achieves adaptive transmission capability in all scenarios from strong networks to extremely weak networks, significantly improving the transmission success rate of core data in weak network environments; through a dual recording mechanism of local timestamp and time-synchronized timestamp, combined with time drift detection and timing correction, it completely solves the clock drift problem in long-term offline scenarios, ensuring the accuracy of cloud data timing. This successfully solves the technical problems of easy loss of vehicle core embedded data in weak network environments, low transmission success rate, and timing disorder caused by offline clock drift in existing technologies, and realizes high reliability and high precision transmission of vehicle embedded data in all scenarios of weak network, offline, and network fluctuation. Attached Figure Description

[0016] Figure 1 A flowchart illustrating an embodiment of the vehicle-mounted embedded data weak network adaptive transmission method provided by the present invention; Figure 2 This is a schematic diagram of the overall process of an embodiment of the vehicle-mounted embedded data weak network adaptive transmission method provided by the present invention; Figure 3 This is a system architecture diagram of an embodiment of the vehicle-mounted embedded data weak network adaptive transmission method provided by the present invention; Figure 4 For the present invention Figure 1 A schematic diagram of an embodiment of S102; Figure 5 For the present invention Figure 1 A schematic diagram of an embodiment of S104; Figure 6 This is a functional block diagram of a vehicle-mounted embedded data weak network adaptive transmission device provided in an embodiment of this application; Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0017] 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 a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0018] In the description of the embodiments of the present invention, unless otherwise stated, "multiple" means two or more. "And / or" describes the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone.

[0019] The terms "first," "second," etc., used in the embodiments of this invention are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a technical feature defined with "first" or "second" may explicitly or implicitly include at least one of that feature.

[0020] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0021] This invention provides a method, apparatus, device, and system for adaptive transmission of vehicle-mounted embedded data in weak network conditions, which will be described below.

[0022] Figure 1 This is a flowchart illustrating an embodiment of the vehicle-mounted embedded data weak network adaptive transmission method provided by the present invention. The execution subject of the vehicle-mounted embedded data weak network adaptive transmission method can be a computer, a vehicle terminal, a portable smart device, or a cloud server, etc. The specific execution subject is the storage medium in the computer, and this embodiment does not limit it.

[0023] Furthermore, a schematic diagram of the overall process in practical applications can be referenced. Figure 2As shown. The process starts with data access and executes in the following order: "Value Classification → Cache Management → Encoding and Compression → Adaptive Transmission → Time-Sequence Repair": 1. Data Access and Value Classification: After the vehicle-mounted data is accessed, the data type is first identified, and then a multi-dimensional value score is performed based on business weight, timeliness decay, and scenario relevance to complete the P0-P4 five-level priority labeling. 2. Hot / Warm / Cold Zone Cache Management: Data is entered into the corresponding cache zone according to priority; when the cache is full, low-value data is intelligently eliminated to ensure that high-priority data is retained first. 3. Three-Layer Progressive Encoding: Data is encoded into L1 (basic layer), L2 (summary layer), and L3 (complete layer) to adapt to different transmission scenarios. 4. Network Status Detection and Tiered Transmission: After network status assessment, strong networks transmit the complete L3 layer; weak networks transmit L1+L2 layers (2 concurrent); extremely weak networks transmit only L1 layer data (1 concurrent) for P0 / P1 data; offline data is buffered and transmitted only after network recovery. 5. GPS Time Anchor and Timing Correction: T_local and T_gps are recorded synchronously during data transmission. The cloud detects clock drift; if the error exceeds 5 seconds, linear interpolation is performed for correction, ultimately outputting standardized data with accurate timing for cloud storage and analysis. The corresponding system architecture based on the above process can be referenced. Figure 3 As shown, the system adopts a five-layer decoupled architecture, with each layer having independent responsibilities and coordinating at each level to achieve a closed-loop end-to-end process from data acquisition to time-series repair: 1. Data Generation Layer: The bottom-level data source, responsible for collecting full-dimensional vehicle-mounted data from CAN bus, body sensors, autonomous driving logs, user behavior, fault monitoring, and other sources. 2. Value Assessment Layer: The core identification layer, which completes P0-P4 classification through data type identification, a multi-dimensional value scoring engine, and a five-level priority labeler, providing a basis for subsequent differentiated processing. 3. Cache Management Layer: The data storage and scheduling layer, which achieves high-value data retention and low-value data eviction through hot / warm / cold three-zone cache pools, a dynamic space allocator, and an intelligent eviction controller, solving the problem of limited cache resources. 4. Transmission Layer: The adaptive communication layer, which achieves transmission strategy adaptation under different network conditions through a real-time network detector, a three-layer hierarchical encoder, a dynamic concurrency controller, and an intelligent congestion controller. 5. Timing Correction Layer: This layer refines data processing by using GPS time anchor recording, local clock monitoring, cloud NTP synchronization, and timing interpolation to resolve clock drift in offline scenarios and ensure that data timing errors are within ±5 seconds.

[0024] Figure 1 This is a schematic flowchart of an embodiment of the vehicle-mounted embedded data weak network adaptive transmission method provided by the present invention, as shown below. Figure 1 As shown, the adaptive transmission method for vehicle-mounted embedded data in weak network conditions includes: S101. Perform multi-dimensional value scoring on the accessed vehicle-mounted data points to obtain data priority.

[0025] In this embodiment, after the vehicle-mounted data is accessed, the traditional method of indiscriminate reception and storage is abandoned. All newly added data is evaluated for value in real time. By integrating a three-dimensional scoring model that considers business weight, time decay, and scenario relevance, individual data is quantitatively scored, and data priorities are accurately assigned. This addresses the root cause of the traditional process's lack of value differentiation and the mixing of high- and low-value data.

[0026] In some embodiments, step S101 includes: calculating the business weight score, timeliness decay score, and scenario association score of the vehicle-mounted data respectively; calculating a comprehensive score based on the business weight score, the timeliness decay score, and the scenario association score; and determining the data priority based on the comprehensive score.

[0027] Specifically, overall score S The calculation formula is:

[0028] in, W Score based on business weight. T Score for time-related decay, C Score based on scene relevance; α, β, γ For the corresponding dimension, a fixed weight coefficient is set, and the following conditions are met: After extensive calibration in real-world vehicle scenarios, the optimal calibration coefficients were selected: 、 、 Prioritize the security attributes of data services while taking into account data timeliness and scenario adaptability.

[0029] Furthermore, the specific assignment criteria for the business weight score W are as follows: High-risk safety data (vehicle collisions, fault reports, autonomous driving failures, sudden braking loss of control): W=90-100; Core driving status data (vehicle speed, heading, vehicle attitude, braking status): W=70-89; Regular driving logs (gear changes, light status, air conditioning adjustments): W=40-69; Low-value auxiliary data (system background logs, minor parameter fine-tuning records): W=0-39.

[0030] Time-related decay score T (range 0-100): Employs an exponential decay algorithm to ensure higher priority for recent data. The formula is as follows:

[0031] in, λ The attenuation coefficient is calibrated to 0.005. Δt Data generation time (in seconds), the instant data is generated. Δt=0, T=100The longer the data generation time, the lower the timeliness score; after 300 seconds, the timeliness score tends to stabilize at a low value.

[0032] The assignment rules for the scene association score C (range 0-100) are as follows: the score is dynamically assigned based on the real-time driving scene of the vehicle. Scene recognition is determined by multiple parameters such as vehicle speed, road conditions, and driving status. The specific rules are as follows: High-risk scenes (high-speed driving, curves, slopes, congested road sections, autonomous driving): associated with safety and driving status data, C=80-100; Normal driving scenes (smooth urban road conditions, low-speed driving): associated with normal driving log data, C=40-79; Static scenes (vehicle parked, engine off and in standby): only updated by system background data, C=0-39.

[0033] S102. Store the vehicle-mounted embedded data into the cache partition according to the data priority.

[0034] It should be noted that after data priority classification is completed, the system automatically distributes data of different levels to the corresponding cache partitions, replacing the traditional single FIFO cache mode. At the same time, it clarifies the capacity ratio of each partition, data migration, and cache eviction mechanism, accurately solving the problem of core data loss.

[0035] It should be noted that the cache partitions include: a first cache partition, a second cache partition, and a third cache partition, and the cache capacity of the three is allocated in a fixed ratio. The data priorities include the first data priority, the second data priority, the third data priority, the fourth data priority, and the fifth data priority, which are ranked in descending order of score.

[0036] Specifically, the five-level priority judgment thresholds, data types, and processing rules are as follows: Five priority levels are defined based on a comprehensive score S (0-100), with clearly defined thresholds, corresponding data types, and subsequent processing standards for each level, leaving no ambiguity. The specific definitions are as follows: P0 Top-Tier Core Data (S≥90): Data types include high-risk safety events such as vehicle collisions, serious malfunctions, autonomous driving failures, and emergency braking; Processing rules: Highest priority, permanently retained, prioritized transmission in extremely weak networks, and not involved in any cache eviction. P1 Advanced Core Data (80≤S<90): Data types include real-time vehicle speed, vehicle posture, steering angle, braking pressure, and autonomous driving status logs; Processing rules: Second highest priority, permanently retained, prioritized transmission in weak and extremely weak networks. P2 Intermediate Regular Data (60≤S<80): Data types include gear shifting, turn signal operation, and constant speed driving status records; Processing rules: Regularly retained, using an optimized LRU eviction mechanism, selectively transmitted in weak network conditions. P3 Low-level routine data (40≤S<60): Data types include air conditioning temperature adjustment, volume adjustment, and window opening / closing records; Processing rules: Limited retention, priority elimination when cache is tight, transmission only on strong networks and normal weak networks. P4 Extremely low-value data (S<40): Data types include system background startup logs, records of minor parameter drifts, and redundant duplicate data; Processing rules: Temporary retention, immediate elimination when cache is saturated, batch transmission only in strong network environments.

[0037] Furthermore, after completing the data priority classification, the system automatically distributes data of different levels to corresponding hot, warm, and cold cache partitions, replacing the traditional single FIFO cache mode. It also clearly defines the capacity allocation, data migration, and cache eviction mechanisms for the three zones, precisely addressing the issue of core data loss. The first, second, and third cache partitions are the hot data area, warm data area, and cold data area, respectively. The cache capacity of the three is allocated according to a fixed ratio, specifically as follows: Hot data area (dedicated to P0 and P1 data): 30% of the total cache (153.6MB), dedicated quota, independent partition, prohibited from expansion or encroachment, storing only high-value, secure data, and not sharing space with other partitions. Warm data area (dedicated to P2 and P3 data): 50% of the total cache (256MB), the core buffer partition, supporting small-scale dynamic adjustments, and the main storage area for regular data. Cold data area (dedicated to P4 data): 20% of the total cache (102.4MB), a dynamically floating partition, automatically cleaned up after capacity saturation, with no retention priority.

[0038] In some embodiments of the present invention, such as Figure 4 As shown, step S102 includes: S401. Store the vehicle-mounted data points corresponding to the first data priority and the second data priority in the first cache partition. The first cache partition is not set with an elimination trigger condition.

[0039] S402. Store the vehicle-mounted data points corresponding to the third data priority and the fourth data priority in the second cache partition. When the real-time storage capacity of the second cache partition reaches the first preset ratio of the partition quota, trigger the elimination. Elimination is performed based on the data idle time and data priority.

[0040] S403. Store the vehicle-mounted data points corresponding to the fifth data priority in the third cache partition. When the real-time storage capacity of the third cache partition reaches the second preset ratio of the partition quota, trigger the elimination and execute the first-in-first-out elimination strategy.

[0041] Specifically, the first cache partition has a dedicated quota, is an independent partition, and is prohibited from being expanded or encroached upon. It stores only high-value, secure data and does not share space with other partitions. For the second cache partition, eviction is triggered when the real-time storage capacity reaches 90% of the partition's quota. A value-weighted LRU algorithm is used, combining data idle time and overall data value to determine eviction priority. The formula for calculating the eviction priority value using the value-weighted LRU algorithm is:

[0042] in, ΔT_idle The duration of data that has been idle and not accessed. S To score the overall value of the data, k 1. k 2 is the weighting coefficient ( k 1 = 0.8 k 2=0.2). The larger the W_LRU value, the higher the data eviction priority.

[0043] The algorithm execution logic is as follows: when the warm data area cache triggers eviction, it calculates the W_LRU value of all existing data in the area in batches and prioritizes eviction of the data with the largest value; for data with the same idle time, it prioritizes eviction of data with a lower value score; for data with the same value score, it prioritizes eviction of data with a longer idle time. This achieves the optimization goal of prioritizing the eviction of low-value and long-idle data, avoiding the shortcomings of traditional LRU in eviction of high-value, unpopular core data.

[0044] For the third cache partition, when the real-time storage capacity reaches 85% of the partition quota, an eviction process is triggered, and a full first-in-first-out eviction policy is implemented. The earliest stored P4 data is evicted first to quickly release storage space, and the process stops when the capacity falls back below 60% of the quota.

[0045] The system also supports dynamic migration across partitions: if the comprehensive score of warm and cold partition data is improved to P0 / P1 level due to changes in scenario or data timeliness updates, it will be immediately migrated to the hot data area and permanently retained; if the data score of warm data area drops to P4 level, it will be automatically migrated to the cold data area.

[0046] Furthermore, the cache eviction triggering conditions and complete mechanism are as follows: Hot data area: No eviction trigger condition. Regardless of whether the cache capacity is saturated or the data has been idle for a long time, P0 and P1 data are permanently retained. Active eviction is prohibited and data cannot be squeezed out by other data.

[0047] Warm Data Zone: The trigger condition is that the real-time storage capacity is ≥90% of the partition quota; after triggering, the optimized value LRU algorithm is started to batch evict the bottom 20% of low-value data until the capacity drops below 70% of the quota, at which point the evictment stops.

[0048] Cold data zone: The trigger condition is that the real-time storage capacity is ≥85% of the partition quota; after triggering, a full first-in-first-out (FIFO) eviction process is executed, prioritizing the eviction of the earliest stored P4 data to quickly release storage space, and stopping when the capacity falls back to below 60% of the quota.

[0049] S103. Perform hierarchical compression encoding on the cached vehicle tracking data to generate structured data of three different granularities: basic layer, summary layer, and complete layer.

[0050] It should be noted that after the cache storage is completed, the system performs gradient layered encoding on the original data points, and uses a custom lightweight compression encoding algorithm to generate structured data with three different granularities: L1 base layer, L2 summary layer, and L3 complete layer, providing accurate data support for network adaptive transmission.

[0051] In this embodiment, the base layer includes at least the core identifier field of the data, and the base layer is encoded using field filtering and Huffman compression; the summary layer adds core business fields on the basis of the base layer, and the summary layer uses differential compression for incremental encoding of continuously changing time-series values ​​on the basis of the encoding of the base layer; the complete layer includes all fields of the original collection, and the complete layer is deduplicated and compressed while retaining all original fields.

[0052] Specifically, the L1 base layer (minimalist base layer / core backup data) retains only key traceability fields, including dual timestamps (local time and GPS absolute time), event ID, event type code, data priority level, vehicle / equipment ID, and fault status identifier, with no redundant fields. The L1 base layer employs precise field filtering to eliminate all dynamically redundant fields; text-based event codes and equipment IDs undergo Huffman compression; and timestamps and status identifiers are stored using fixed-point quantization, maximizing data volume reduction. The compressed data size is 8%-12% of the original data.

[0053] The L2 summary layer (regular analysis data) adds core business fields to all fields in the L1 base layer, including real-time vehicle speed, vehicle attitude parameters (pitch angle, roll angle, yaw angle), braking status, steering status, throttle opening, autonomous driving mode status, and road segment type. Building upon the L1 base layer encoding, the L2 summary layer uses differential compression for continuously changing time-series values ​​such as vehicle speed and attitude, storing only the changes between adjacent data points instead of storing all values, reducing data redundancy, and reducing the compressed data size to 30%-40% of the original.

[0054] The L3 complete layer (deep retrospective data) includes all fields from the L1 base layer and the L2 summary layer, while retaining all original acquired fields, including raw sensor sampling data, CAN bus messages, device operating parameters, environmental data (temperature, humidity, light intensity), full user operation logs, data checksums, and acquired device status data. The L3 complete layer, while retaining all original fields, performs global deduplication and compression on duplicate messages and redundant logs, and uses adaptive precision compression on the raw sampling data. This slightly reduces the original data volume without losing valid information, resulting in a compressed volume of 85%-90% of the original data.

[0055] S104. Real-time detection of network status and determination of the current network quality level; selection of the corresponding transmission layer from the base layer, the digest layer and the complete layer based on the network quality level; and dynamic adjustment of the corresponding concurrent paths and transmission rate to adapt to the current network conditions for data transmission.

[0056] In this embodiment, after the data encoding is completed, the system detects multi-dimensional parameters of the vehicle network in real time, accurately determines the network level, dynamically matches the transmission level, concurrency, and transmission rate, and, together with congestion control and breakpoint resume mechanisms, achieves adaptive transmission across the entire network scenario.

[0057] Specifically, network quality levels include a first network quality level, a second network quality level, a third network quality level, and a fourth network quality level. The network quality determination logic and threshold standards are as follows: the system collects four core parameters in real time: network bandwidth, packet loss rate, latency, and signal strength. In this embodiment, a weighted scoring method is used to determine the network quality level. The network scoring formula is as follows:

[0058] in, B For bandwidth normalization score, L For packet loss rate, D For the time delay normalized value, S The signal strength is scored and ultimately divided into four levels: strong network, normal weak network, extremely weak network, and offline.

[0059] Specifically, the threshold classification rules are as follows: Strong network: N≥85, bandwidth≥2Mbps, packet loss rate≤3%, latency≤100ms, excellent signal strength; Regular weak network: 60≤N<85, bandwidth500Kbps-2Mbps, packet loss rate3%-10%, latency100-300ms; Extremely weak network: 30≤N<60, bandwidth<500Kbps, packet loss rate10%-30%, latency>300ms; Offline: N<30, no network connection, signal interruption.

[0060] In some embodiments of the present invention, such as Figure 5 As shown, step S104 includes: S501. When the network quality level is the first network quality level, select the complete layer for transmission and enable multi-channel concurrent transmission at the first transmission rate.

[0061] S502. When the network quality level is the second network quality level or the third network quality level, select the base layer and / or the summary layer for transmission, reduce the number of concurrent transmission paths, and transmit at the second transmission rate or the third transmission rate.

[0062] S503. When the network quality level is the fourth network quality level, terminate the transmission task and wait for the network to recover.

[0063] Specifically, the dedicated transmission strategy for each network level is as follows: When the network quality level is the first network quality level, the full layer (L3 layer) is selected for transmission, with 4 concurrent transmissions at the first transmission rate (512Kbps), no data priority restrictions, and full synchronization of all cached data; when the network quality level is the second network quality level (normal weak network), the base layer and the digest layer (L1+L2 layers) are selected for combined transmission, with 2 concurrent transmissions at the second transmission rate (256Kbps), prioritizing the transmission of P0-P2 high-resolution data. For medium-priority data, P3 and P4 low-value data are temporarily suspended; when the network quality level is the third network quality level (extremely weak network), the basic layer (L1 layer) is selected for transmission, and only the L1 basic layer that transmits P0 and P1 core data is transmitted. One single concurrent transmission is started at the third transmission rate (64Kbps) to ensure the upload of critical security data, and all low-value data transmission is suspended; when the network quality level is the fourth network quality level (offline), the transmission task is terminated, all newly added data is stored in the corresponding cache partition, and the interruption is resumed after the network is restored.

[0064] In some embodiments of the present invention, the method further includes a congestion control step: real-time monitoring of the packet loss rate; when the packet loss rate is less than a first packet loss rate threshold (5%), maintaining the current sending rate and steadily advancing data upload; when the packet loss rate is greater than or equal to the first packet loss rate threshold and less than a second packet loss rate threshold (15%), linearly reducing the sending rate by 20% every 10 seconds until the packet loss rate falls back to the normal range; when the packet loss rate is greater than or equal to the second packet loss rate threshold (15%), pausing batch transmission and performing a safety net transmission, retaining only P0 data for safety net upload to avoid continuous congestion and retransmission failure; after the network condition improves, adopting a step-by-step speed-up strategy to avoid instantaneous traffic bursts causing congestion again.

[0065] In some embodiments of the present invention, the method further includes a fragmented breakpoint resumption step: the data to be transmitted is fragmented into fixed sizes, with each fragment size uniformly set to 4KB, and each fragment carries a unique fragment ID, total number of fragments, data checksum, and timestamp information; during transmission, uploaded fragment records are synchronized in real time to the local cache and the cloud; when the network is interrupted or transmission fails, the current breakpoint position and the list of incomplete fragments are accurately recorded; after the network is restored, the system reads the breakpoint record, without re-uploading completed fragments, and only re-transmits fragments that have not been uploaded or whose transmission failed, and automatically splices the complete data after the re-transmission is completed; after each data fragment is uploaded, a CRC check is performed, and if the check fails, the current fragment is automatically retransmitted, with a maximum of 3 retransmissions per fragment; if the maximum is exceeded, an error is marked and a retry is postponed.

[0066] S105. Record the local timestamp and time synchronization timestamp during data acquisition. When time drift is detected in the local clock, perform timing correction on the data to be transmitted based on the time synchronization timestamp.

[0067] It should be noted that throughout the entire process of data acquisition, caching, and transmission, the system synchronously executes timing calibration logic. Through a full-process mechanism of dual timestamp recording, clock drift detection, linear interpolation correction, multi-interval splicing, and post-verification, it completely solves the defects of offline clock drift and timing disorder. Through timing calibration, offline retransmitted data can be adapted to the standard time base after network outage recovery.

[0068] In some embodiments of the present invention, the timing correction adopts a piecewise linear mapping method. For each offline interval, the timing of the start and end times of the offline interval is used as the reference anchor point, and the corresponding local clock time is used as the reference point. The local acquisition time of the data to be corrected is linearly mapped to calculate the corrected standard time.

[0069] Specifically, the system automatically triggers clock drift detection each time the network is restored. It connects to the cloud's NTP standard time to obtain the current accurate cloud time T_ntp, and simultaneously compares it with the local current time T_local and the latest GPS time T_gps, calculating the drift error ΔT = |T_local - T_ntp| between the local clock and the standard time. The preset error threshold is 5 seconds. If ΔT ≤ 5 seconds, it is considered that there is no significant drift and no correction is needed; if ΔT > 5 seconds, it is considered that clock drift exists, and the timing correction process is immediately initiated. The correction formula is:

[0070] in: T_gps_start , T_gps_end The absolute GPS time (reference anchor point) at the start and end times of the offline interval. T_local_start , T_local_end This refers to the start and end times of the local clock within the offline interval. T_local_curr This refers to the local collection time of a single data entry during offline periods; T_correct This is the final, accurate standard time after calibration.

[0071] For scenarios involving multiple intermittent offline events for vehicles, the system automatically splits the network into multiple independent offline intervals based on the network disconnection and recovery time, marks the GPS time anchor point and local time interval for each interval, performs linear interpolation correction on each offline interval separately, sorts all corrected segmented data globally in time series using GPS absolute time as a unified benchmark, and eliminates time series gaps and overlaps by aligning time anchor points to form a continuous and complete global timeline.

[0072] After calibration, a two-layer verification is performed: single-point precision verification ensures that all data errors are ≤ ±5 seconds; timing logic verification ensures that the data generation order and event association logic are reasonable, with no time reversal or sequence disorder issues. Verified data is uploaded to the cloud, while unqualified data is recalibrated.

[0073] This embodiment performs multi-dimensional value scoring on the accessed vehicle-mounted data to obtain data priority; based on the data priority, the vehicle-mounted data is stored in a cache partition; the cached vehicle-mounted data is subjected to hierarchical compression encoding to generate structured data of three different granularities: basic layer, summary layer, and complete layer; the network status is detected in real time and the current network quality level is determined, and the corresponding transmission layer is selected for data transmission based on the network quality level; the local timestamp and time synchronization timestamp at the time of data collection are recorded, and when time drift of the local clock is detected, the timing of the data to be transmitted is corrected according to the time synchronization timestamp. Therefore, it can ensure that the core vehicle-mounted data is not lost, transmitted stably, and time-accurate in weak network environments, achieving highly reliable and high-precision transmission of vehicle-mounted data in all scenarios of weak networks, offline, and network fluctuations.

[0074] Based on the same idea as the vehicle-mounted embedded data weak network adaptive transmission method in the above embodiments, this application also provides a vehicle-mounted embedded data weak network adaptive transmission device 600, which can be used to execute the above-described vehicle-mounted embedded data weak network adaptive transmission method. Figure 6 As shown, the vehicle-mounted embedded data weak network adaptive transmission device 600 specifically includes a value scoring module 601, a cache management module 602, a hierarchical compression encoding module 603, a network adaptive transmission module 604, and a timing correction module 605. In some embodiments, the above modules can be programmable software instructions stored in memory and executable by a processor. It is understood that in other embodiments, the above modules can also be program instructions or firmware embedded in the processor.

[0075] The value scoring module 601 is used to perform multi-dimensional value scoring on the accessed vehicle-mounted data to obtain data priority. Cache management module 602 is used to store the vehicle embedded data into cache partitions according to the data priority; The hierarchical compression coding module 603 is used to perform hierarchical compression coding on the cached vehicle tracking data to generate structured data with three different granularities: basic layer, summary layer and complete layer. The network adaptive transmission module 604 is used to detect the network status in real time and determine the current network quality level. Based on the network quality level, it selects the corresponding transmission layer from the base layer, the digest layer and the full layer, and dynamically adjusts the corresponding number of concurrent paths and the transmission rate to adapt to the current network conditions for data transmission. The timing correction module 605 is used to record the local timestamp and the time synchronization timestamp during data acquisition. When a time drift is detected in the local clock, the timing correction is performed on the data to be transmitted according to the time synchronization timestamp. The timing correction is used to adapt the offline retransmission data to the standard time base after the network is restored, and to complete the weak network adaptive transmission of vehicle embedded data.

[0076] Please refer to Figure 7 , Figure 7 This is a schematic diagram of an embodiment of the electronic device of this application. In this embodiment of the invention, the electronic device 700 includes a processor 701, a memory 702, a display 703, and a data acquisition device 704. Figure 7 Only some components of the electronic device 700 are shown, but it should be understood that it is not required to implement all the components shown, and more or fewer components may be implemented instead.

[0077] In some embodiments, processor 701 may be a central processing unit (CPU), microprocessor, or other data processing chip, used to run program code stored in memory 702 or process data, such as the vehicle-mounted embedded data weak network adaptive transmission method of the present invention.

[0078] In some embodiments, display 703 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen. Display 703 is used to display information from electronic device 700 and to display visual user applications. Components 701-704 of electronic device 700 communicate with each other via a system bus.

[0079] In some embodiments, the data acquisition unit 704 can be one or more combinations of a CAN bus interface, an in-vehicle Ethernet interface, an OBD (On-Board Diagnostics) interface, a serial communication interface (such as RS232, RS485), a USB interface, an AHD (Analog High Definition) video acquisition interface, a GMSL (Gigabit Multimedia Serial Link) interface, an FPD-Link (Flat Panel Display Link) interface, and a GPIO (General-Purpose Input / Output) interface. The data acquisition unit 704 is used to access in-vehicle embedded data, which includes, but is not limited to, CAN bus messages, raw sensor sampling data, vehicle status data, autonomous driving log data, GPS positioning data, and user operation behavior data.

[0080] In one embodiment, when the processor 701 executes the vehicle-mounted embedded data weak network adaptive transmission program in the memory 702, the following steps can be implemented: The accessed vehicle data points are evaluated for value across multiple dimensions to determine data priority. The vehicle-mounted data points are stored in the cache partitions according to the data priority. The cached vehicle tracking data is hierarchically compressed and encoded to generate structured data of three different granularities: basic layer, summary layer, and complete layer. The system detects the network status in real time and determines the current network quality level. Based on the network quality level, it selects the corresponding transmission layer from the base layer, the digest layer, and the full layer, and dynamically adjusts the corresponding number of concurrent paths and the transmission rate to adapt to the current network conditions for data transmission. Record the local timestamp and time synchronization timestamp when data is collected. When time drift is detected in the local clock, perform timing correction on the data to be transmitted according to the time synchronization timestamp. The timing correction is used to adapt the offline retransmission data to the standard time base after network failure recovery, and complete the weak network adaptive transmission of vehicle embedded data.

[0081] It should be understood that when the processor 701 executes the vehicle embedded data weak network adaptive transmission program in the memory 702, in addition to the functions mentioned above, it can also perform other functions, as can be found in the description of the corresponding method embodiments above.

[0082] Furthermore, the embodiments of the present invention do not specifically limit the type of the electronic device 700 mentioned. The electronic device 700 can be a vehicle terminal, a T-Box (Telematics Control Unit), a domain controller, a mobile phone, a tablet computer, or other portable electronic devices. Exemplary embodiments of portable electronic devices include, but are not limited to, portable electronic devices running iOS, Android, Microsoft, or other operating systems. The aforementioned portable electronic device can also be other portable electronic devices, such as a laptop computer with a touch-sensitive surface (e.g., a touch panel). It should also be understood that in some other embodiments of the present invention, the electronic device 700 may not be a portable electronic device, but rather a desktop computer with a touch-sensitive surface (e.g., a touch panel).

[0083] Those skilled in the art will understand that all or part of the processes of the methods described in the above embodiments can be implemented by a computer program instructing related hardware (such as a processor, controller, etc.), and the computer program can be stored in a computer-readable storage medium. The computer-readable storage medium may be a disk, optical disk, read-only memory, or random access memory, etc.

[0084] The above provides a detailed description of the vehicle-mounted embedded data weak network adaptive transmission method, apparatus, device, and system provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for adaptive transmission of vehicle-mounted embedded data in weak network conditions, characterized in that, include: The accessed vehicle data points are evaluated for value across multiple dimensions to determine data priority. The vehicle-mounted data points are stored in the cache partitions according to the data priority. The cached vehicle tracking data is hierarchically compressed and encoded to generate structured data of three different granularities: basic layer, summary layer, and complete layer. The system detects the network status in real time and determines the current network quality level. Based on the network quality level, it selects the corresponding transport layer from the base layer, the digest layer, and the full layer, and dynamically adjusts the corresponding number of concurrent paths and the transmission rate to adapt to the current network conditions for data transmission. Record the local timestamp and time synchronization timestamp when data is collected. When time drift is detected in the local clock, perform timing correction on the data to be transmitted according to the time synchronization timestamp. The timing correction is used to adapt the offline retransmission data to the standard time base after network failure recovery, and complete the weak network adaptive transmission of vehicle embedded data.

2. The vehicle-mounted embedded data weak network adaptive transmission method according to claim 1, characterized in that, The process of performing multi-dimensional value scoring on the vehicle-mounted data to obtain data priority includes: Calculate the business weight score, timeliness decay score, and scenario association score of the vehicle-mounted embedded data respectively; A comprehensive score is calculated based on the business weight score, the timeliness decay score, and the scenario association score. Data priority is determined based on the comprehensive score.

3. The vehicle-mounted embedded data weak network adaptive transmission method according to claim 1, characterized in that, The cache partitions include: a first cache partition, a second cache partition, and a third cache partition, and the cache capacity of the three is allocated in a fixed ratio. The data priorities include a first data priority, a second data priority, a third data priority, a fourth data priority, and a fifth data priority, with the scores decreasing sequentially. The step of storing the vehicle-mounted data points into cache partitions according to the data priority includes: The vehicle-mounted data points corresponding to the first data priority and the second data priority are stored in the first cache partition, and no elimination trigger condition is set for the first cache partition. The vehicle-mounted data points corresponding to the third data priority and the fourth data priority are stored in the second cache partition. When the real-time storage capacity of the second cache partition reaches the first preset ratio of the partition quota, the data is eliminated based on the data idle time and data priority. The vehicle-mounted data points corresponding to the fifth data priority are stored in the third cache partition. When the real-time storage capacity of the third cache partition reaches the second preset ratio of the partition quota, an elimination is triggered, and a first-in-first-out elimination strategy is executed.

4. The vehicle-mounted embedded data weak network adaptive transmission method according to claim 1, characterized in that, The base layer contains at least a core identifier field for the data, and the base layer is encoded using field filtering and Huffman compression. The summary layer adds core business fields on top of the base layer. Based on the encoding of the base layer, the summary layer uses differential compression for incremental encoding of continuously changing time-series values. The complete layer contains all the original fields collected, and the complete layer is deduplicated and compressed while retaining all the original fields.

5. The vehicle-mounted embedded data weak network adaptive transmission method according to claim 1, characterized in that, The network quality levels include a first network quality level, a second network quality level, a third network quality level, and a fourth network quality level; The step of selecting the corresponding transport layer from the base layer, the digest layer, and the full layer based on the network quality level, and dynamically adjusting the corresponding concurrent paths and transmission rate to adapt to the current network conditions for data transmission, includes: When the network quality level is the first network quality level, the complete layer is selected for transmission, and multi-path concurrent transmission is enabled at the first transmission rate. When the network quality level is the second network quality level or the third network quality level, the base layer and / or the digest layer are selected for transmission to reduce the number of concurrent transmission paths and transmit at the second transmission rate or the third transmission rate, wherein the first transmission rate, the second transmission rate, and the third transmission rate are decreased sequentially. When the network quality level is the fourth network quality level, the transmission task is terminated and the network is awaited to recover.

6. The vehicle-mounted embedded data weak network adaptive transmission method according to claim 1, characterized in that, The method further includes: Detect packet loss rate during data transmission; When the packet loss rate is less than the first packet loss rate threshold, maintain the current sending rate; When the packet loss rate is greater than or equal to the first packet loss rate threshold and less than the second packet loss rate threshold, the sending rate is reduced. When the packet loss rate is greater than or equal to the second packet loss rate threshold, batch transmission is paused and a backup transmission is performed.

7. The vehicle-mounted embedded data weak network adaptive transmission method according to claim 1, characterized in that, The timing correction adopts a piecewise linear mapping method. For each offline interval, the start and end times of the offline interval are used as the reference anchor points, and the corresponding local clock time is used as the reference point. The local acquisition time of the data to be corrected is linearly mapped to calculate the corrected standard time.

8. A vehicle-mounted embedded data weak network adaptive transmission device, characterized in that, include: The value scoring module is used to perform multi-dimensional value scoring on the accessed vehicle tracking data to obtain data priority; The cache management module is used to store the vehicle-mounted embedded data into cache partitions according to the data priority. The hierarchical compression and encoding module is used to perform hierarchical compression and encoding on the cached vehicle tracking data to generate structured data with three different granularities: the base layer, the summary layer, and the complete layer. The network adaptive transmission module is used to detect the network status in real time and determine the current network quality level. Based on the network quality level, it selects the corresponding transmission layer from the base layer, the digest layer, and the full layer, and dynamically adjusts the corresponding number of concurrent paths and the transmission rate to adapt to the current network conditions for data transmission. The timing correction module is used to record the local timestamp and the time synchronization timestamp during data acquisition. When a time drift is detected in the local clock, the timing correction is performed on the data to be transmitted according to the time synchronization timestamp. The timing correction is used to adapt the offline retransmission data to the standard time base after the network is restored, and to complete the weak network adaptive transmission of vehicle embedded data.

9. An electronic device, characterized in that, It includes a data collector, a memory, and a processor, wherein the data collector is used to access vehicle-mounted embedded data; The memory is used to store programs; The processor, coupled to the memory, is used to execute the program stored in the memory to implement the steps in the vehicle-mounted embedded data weak network adaptive transmission method according to any one of claims 1 to 7.

10. A vehicle-mounted embedded data weak network adaptive transmission system, characterized in that, The vehicle-mounted embedded data weak network adaptive transmission system includes an electronic device, a data access device, and a data output device as described in claim 9. The data access device and the data output device are respectively connected to the electronic device. The data acquisition device is used to access vehicle-mounted embedded data, and the data output device is used to perform data transmission.

Images