Vehicle-mounted network recovery method and device for pre-searching network signals, vehicle and medium

Abstract

The application mainly relates to the technical field of automobile engineering. The application discloses a vehicle-mounted network recovery method for pre-searching network signals, a device, a vehicle and a medium. The method comprises the following steps: based on space-time driving data and a network signal space-time mapping library, the position relationship between a vehicle driving track and a signal blind area is determined; before entering the blind area, a surrounding service area in the exit direction of the blind area is searched, candidate service areas are determined, and network connection parameters of the candidate service areas are obtained; when being in the blind area, searching is stopped, and the parameters are saved to the local; when driving out of the blind area, target parameters are read from the local, and network communication is recovered by connecting with the corresponding service area. The application realizes rapid network reconnection by completing target cell connection preparation in advance, and effectively guarantees the continuity and reliability of vehicle-mounted safety services.

Description

Technical Field

[0001] This invention relates to the field of automotive engineering technology, specifically to a method, apparatus, vehicle, and medium for recovering in-vehicle networks by pre-searching for network signals. Background Technology

[0002] With the rapid development of vehicle-to-everything (V2X) and intelligent driving technologies, in-vehicle communication terminals have become a critical infrastructure supporting core services such as remote monitoring, online navigation, software upgrades, and emergency rescue. Currently, mainstream in-vehicle terminals widely adopt Long Term Evolution (LTE) and New Radio (NR) cellular communication modules, relying on ground base stations to achieve wide-area coverage. In actual road driving, vehicles inevitably enter typical signal blind spots such as tunnels, underground parking garages, mountain valleys, and densely populated urban areas. Due to physical obstruction and multipath fading, the wireless link between the terminal and the current serving base station will be interrupted, resulting in network service unavailability.

[0003] To restore communication connections, existing vehicle-mounted terminals generally employ standard cellular network reselection and reconnection mechanisms. This mechanism relies on the terminal's physical layer to perform real-time scanning and synchronization detection of candidate cell broadcast signals. When the terminal detects that the current signal quality is below a preset threshold, it triggers a cell search process across the entire or a portion of the frequency band, sequentially completing synchronization signal detection, primary and secondary synchronization signal parsing, system information block reading, and random access. However, the above process reveals significant shortcomings at the boundary of signal blind zones: on the one hand, the high-speed movement of vehicles results in extremely short dwell times for the terminal within the blind zone, making it difficult for the traditional passive response mode of "loss first, search later" to complete the full reconnection process within a limited time window; on the other hand, the standard search algorithm, in order to maintain versatility, does not pre-determine and optimize for the geographical distribution characteristics and signal recovery patterns at the blind zone boundary, causing the terminal to often consume a large amount of synchronization detection time on invalid frequency points where recovery conditions are not available.

[0004] More importantly, existing solutions lack proactive sensing mechanisms for the location of blind zone exits and the timing of signal recovery, making it impossible to initiate frequency and time slot synchronization preparations for the target base station before leaving the blind zone. Since the synchronization acquisition process itself requires milliseconds to hundreds of milliseconds of time, this passive waiting delay directly accumulates to the total service interruption duration at higher vehicle speeds. Real-world test data shows that in a typical underground parking garage exit scenario, the latency from when the terminal leaves the blind zone to when network reconnection is completed is generally between 3 and 8 seconds, far exceeding the second-level or even sub-second recovery requirements for safety-related services such as emergency alarms and remote driving in intelligent connected vehicles.

[0005] In summary, the lack of a proactive, low-power pre-synchronization mechanism based on precise spatiotemporal location triggering for typical discontinuous signal blind zone scenarios in vehicles leads to excessively high network reconnection latency, failing to meet the second-level recovery requirements of secure services. This issue has become a significant technical bottleneck restricting the continuity of highly reliable in-vehicle communication services. Summary of the Invention

[0006] This invention provides a method, apparatus, vehicle, and medium for restoring in-vehicle networks by pre-searching for network signals. It can complete the network connection preparation of the target cell before the vehicle leaves the signal blind zone, thereby achieving rapid network reconnection and effectively ensuring the continuity and reliability of in-vehicle security services.

[0007] This invention provides a method for restoring vehicular network signals after pre-search, the method comprising: Based on the spatiotemporal driving data of the target vehicle and the spatiotemporal mapping library data of the network signal, the positional relationship between the driving trajectory of the target vehicle and the target signal blind zone is determined. When the target vehicle is on its driving trajectory before entering the target signal blind zone, the surrounding service areas in the direction of the exit of the target signal blind zone are searched, and the network connection parameters of the candidate service area are obtained after determining at least one candidate service area. When the target vehicle is in the target signal blind zone, the search for the surrounding service area is stopped, and each network connection parameter is saved to the local storage space. When the target vehicle is on a driving trajectory that leaves the target signal blind zone, the target network connection parameters are read from the local storage space, and a network connection is established with the candidate service area corresponding to the target network connection parameters to restore the network communication of the target vehicle.

[0008] Optionally, the specific methods for determining the positional relationship between the target vehicle's trajectory and the target signal blind spot include: The predicted location of the target vehicle after a preset time period is calculated based on the spatiotemporal driving data of the target vehicle. The predicted location is matched with the spatiotemporal mapping relationship in the preset spatiotemporal mapping library to obtain the signal receiving power corresponding to the predicted location; When the signal reception power at the predicted location is lower than a preset signal strength threshold and the target vehicle is traveling in the direction of the target signal blind zone, it is determined that the target vehicle will enter the target signal blind zone after the preset future time period.

[0009] Optionally, the specific methods for determining that the target vehicle has entered the target signal blind zone include: Obtain the vehicle-mounted map of the target vehicle, wherein the vehicle-mounted map contains pre-marked static blind spot labels for signal blocking structures, and the static blind spot labels include at least the blind spot entrance; The predicted location of the target vehicle is matched with the static blind spot labels in the vehicle map; When the predicted position is within a preset forward distance from the blind spot entrance defined by the static blind spot label and the target vehicle's driving direction is pointing towards the blind spot entrance, it is determined that the target vehicle has entered the target signal blind spot.

[0010] Optionally, the vehicular network recovery method for the pre-searched network signal further includes: The future preset time period and the preset signal strength threshold are determined based on the current speed of the target vehicle, wherein the current speed is positively correlated with the future preset time period and the preset signal strength threshold.

[0011] Optionally, the specific methods for constructing the spatiotemporal mapping library include: Collect historical spatiotemporal driving data of vehicles, historical network dwell results, and historical neighboring cell measurement data; The collected data is divided into multiple grids according to geographical location and time period, and the network environment characteristics of each grid within the target time period are statistically analyzed. The network environment characteristics include the reference signal received power distribution of the service area and neighboring cells based on the historical neighboring cell measurement data, and the blind zone label generated based on the comparison of the historical network dwell results and the signal strength threshold. The historical cell identifiers of the target grid that have the highest frequency of occurrence or the most stable residence within the corresponding time period are used as the historical network residence results, and are associated and stored with the geographical location, time period and network environment characteristics corresponding to the target grid to form a spatiotemporal mapping relationship corresponding to the target grid. The spatiotemporal mapping library is generated based on the spatiotemporal mapping relationship corresponding to each of the grids.

[0012] Optionally, the vehicular network recovery method for the pre-searched network signal further includes: When the target vehicle is on a driving trajectory leaving the target signal blind zone, feedback data of crossing the target signal blind zone is collected. The feedback data includes at least a list of candidate service areas and the network signal quality of each candidate service area, the location of the target signal blind zone and the entry time. The feedback data is compared with the data recorded in the target raster in the spatiotemporal mapping library to generate an optimization decision; Based on the optimization decision, the network environment characteristics and historical network dwell results of the target grid are updated.

[0013] Optionally, the vehicular network recovery method for the pre-searched network signal further includes: Generate a list of candidate service areas related to the target vehicle's frequently driven area and current travel route, and save the list of candidate service areas to the target vehicle's local storage space. The list of candidate service areas is used to provide the target vehicle with selectable candidate service areas and network connection parameters for the candidate service areas. When the target vehicle is on its driving trajectory before entering the target signal blind zone, a communication channel is established between the target vehicle and the roadside edge node in the current area, wherein the current area is a geographical area centered on the location of the target vehicle and with a preset distance as the radius; The latest information of candidate service areas in the exit direction of the target signal blind zone is obtained from the roadside edge node through the communication channel. The latest information includes multiple candidate service area identifiers uploaded by each vehicle within the preset range of the roadside edge node and the network connection parameters corresponding to each candidate service area identifier. The candidate service area list is updated using the latest information.

[0014] The present invention also provides a vehicle network recovery device for pre-searching network signals, the device comprising: The prediction module is used to determine the positional relationship between the target vehicle's driving trajectory and the target signal blind zone based on the target vehicle's spatiotemporal driving data and the spatiotemporal mapping library data of network signals. The search module is used to search for surrounding service areas in the direction of the exit of the target signal blind zone when the target vehicle is on the driving trajectory before entering the target signal blind zone, and to obtain the network connection parameters of the candidate service area after determining at least one candidate service area. The storage module is used to stop searching for the surrounding service area when the target vehicle is in the target signal blind zone, and to save each network connection parameter to the local storage space. The communication module is used to read target network connection parameters from the local storage space and establish a network connection with the candidate service area corresponding to the target network connection parameters when the target vehicle is on a driving trajectory that leaves the target signal blind zone, so as to restore the network communication of the target vehicle.

[0015] The present invention also provides a vehicle, the vehicle including a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the vehicle network recovery method for pre-searching network signals as described in any of the preceding claims.

[0016] The present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the vehicular network recovery method for pre-searching network signals as described in any of the preceding claims.

[0017] The present invention has at least the following beneficial effects: First, this solution relies on spatiotemporal driving data and a network signal mapping library to accurately determine the relative position of the vehicle and the blind spot. This determination provides a reliable triggering time for subsequent actions, ensuring that all operations are initiated at the correct geographical location node. Second, before the vehicle enters the blind spot, the solution performs a pre-search on surrounding service areas in the direction of the blind spot exit and obtains the network connection parameters of candidate service areas in advance. This step moves the most time-consuming search and parameter parsing work forward, allowing the vehicle to complete the identification and configuration preparation of the target cell while there is still signal coverage. Subsequently, when the vehicle is inside the blind spot, the solution actively stops unnecessary surrounding searches and saves the acquired parameters locally. This avoids the waste of power and latency caused by invalid searches within the blind spot and ensures that the parameters can be retrieved instantly when the vehicle leaves the blind spot through local storage. Finally, the moment the vehicle leaves the blind spot, the solution directly reads the target parameters from the local storage and initiates a connection, eliminating the need for real-time search, synchronization, and parsing. This continuous spatiotemporal triggering process fundamentally reduces the processing time at the exit point, enabling network reconnection to be completed within seconds, thereby meeting the stringent requirements of security services for the timeliness of communication recovery. Attached Figure Description

[0018] The accompanying drawings are provided to further understand the technical solutions of the present invention and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the technical solutions of the present invention, and do not constitute a limitation on the technical solutions of the present invention.

[0019] Figure 1 This is a flowchart illustrating the steps of a method for restoring an in-vehicle network by pre-searching for network signals. Figure 2 This is a flowchart illustrating the steps involved in iteratively optimizing a spatiotemporal mapping library in a pre-search network signal recovery method for vehicular networks. Figure 3 This is a flowchart illustrating the steps involved in obtaining candidate service areas through road test edge nodes in a method for recovering vehicular networks by pre-searching network signals. Figure 4 This is a schematic diagram of the structure of an in-vehicle network recovery device that pre-searches for network signals; Figure 5 This is a schematic diagram of the architecture of an in-vehicle network recovery system that pre-searches for network signals. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0021] The researchers who developed this technical solution discovered that the existing publicly available technologies closest to this solution fall into two categories: (1) Standardized fast recovery mechanism: 3GPP has defined basic mechanisms such as RRC inactive state fast recovery and fast carrier aggregation in the LTE / 5G NR standard. These mechanisms can only optimize the handover latency in short-term signal attenuation scenarios and cannot be adapted to long-term deep signal blind zone scenarios.

[0022] (2) Optimization solutions from academia and industry: Li et al.'s "5G NR Fast Handover Algorithm for High-Speed ​​Rail" published at IEEE TVT 2021 optimized handover latency for high-speed rail scenarios with continuous signal attenuation; Chen et al.'s "Location-Aware Predictive Handover Scheme for Vehicle-to-Everything (V2X)" published at IEEE TVT 2022 optimized the target cell selection efficiency for continuous handover through location prediction. Existing solutions all focus on high-speed handover scenarios with continuous signal attenuation and have not optimized for long-term deep signal blind zone scenarios on vehicles.

[0023] The core problems with existing technologies include four aspects. First, the network recovery latency is too long, failing to meet the compliance requirements for security services. Existing mass-produced T-Boxes use native closed-source algorithms, requiring 15 to 23 seconds to reconnect to the network and restore data transmission after a vehicle leaves signal blind spots such as underground parking garages or highway tunnels. This falls far short of the second-level recovery requirement, directly causing the failure of functions such as emergency calls, remote control, and real-time warnings for new energy vehicles, resulting in serious security risks. Second, the passive response logic fails to overcome the inherent limitations of full-band blind scanning after network outages. Existing solutions all adopt a passive response mode that initiates cell search only after network outages, failing to complete pre-search and pre-synchronization of target cells before the vehicle enters a signal blind spot. This cannot fundamentally eliminate the latency redundancy caused by full-band blind scanning, leading to severe performance degradation in long-term offline scenarios. Third, indiscriminate pre-search results in excessive power consumption, making it unsuitable for mass-produced automotive scenarios. Some existing pre-search solutions use indiscriminate full-band neighbor cell measurement without precise triggering based on the vehicle's spatiotemporal location, resulting in too many invalid measurements and a significant increase in power consumption of the vehicle terminal, which does not meet the requirements for automotive-grade low-power design. Fourth, it is not adapted to typical blind spot scenarios in vehicles, resulting in poor scenario versatility. The existing pre-search solution is only designed for continuous switching scenarios on fixed routes such as high-speed rail and highways, and is not adapted to blind spot scenarios with high-frequency discontinuous signals in vehicles such as underground parking garages, urban tunnels, and elevated road obstructions, which cannot meet the full scenario adaptation requirements of mass-produced models.

[0024] To address the aforementioned technical problems, this technical solution provides a method, apparatus, vehicle, and medium for restoring vehicular network signals before the vehicle leaves a signal blind spot. This allows for pre-emptive preparation of the target cell network connection before the vehicle leaves the signal blind spot, thereby achieving rapid network reconnection and effectively ensuring the continuity and reliability of vehicular security services. The following are various embodiments of this technical solution.

[0025] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating the steps of a method for restoring an in-vehicle network by pre-searching for network signals.

[0026] This embodiment provides a method for restoring an in-vehicle network by pre-searching for network signals, including: S101. Based on the spatiotemporal driving data of the target vehicle and the spatiotemporal mapping library data of the network signal, determine the positional relationship between the driving trajectory of the target vehicle and the target signal blind zone.

[0027] S102. When the target vehicle is on its driving trajectory before entering the target signal blind zone, search for surrounding service areas in the direction of the exit of the target signal blind zone, and obtain the network connection parameters of the candidate service area after determining at least one candidate service area.

[0028] S103. When the target vehicle is in the target signal blind zone, stop searching for surrounding service areas and save each network connection parameter to local storage.

[0029] S104. When the target vehicle is on a driving trajectory that leaves the target signal blind zone, read the target network connection parameters from the local storage space and establish a network connection with the candidate service area corresponding to the target network connection parameters in order to restore the network communication of the target vehicle.

[0030] In some embodiments, target service area information can be obtained in advance through a vehicle-cloud collaborative platform, which can obtain information about all cells around the blind spot exit in advance, replacing the end-side baseband pre-search; or the target cell access information can be obtained through cell information broadcast by the roadside RSU unit, further reducing end-side measurement power consumption.

[0031] In this embodiment, firstly, the solution relies on spatiotemporal driving data and a network signal mapping library to accurately determine the relative position of the vehicle and the blind spot. This determination provides a reliable triggering opportunity for subsequent actions, ensuring that all operations are initiated at the correct geographical location node. Secondly, before the vehicle enters the blind spot, the solution performs a pre-search on surrounding service areas in the direction of the blind spot exit and obtains the network connection parameters of candidate service areas in advance. This step moves the most time-consuming search and parameter parsing work forward, allowing the vehicle to complete the identification and configuration preparation of the target cell while there is still signal coverage. Subsequently, when the vehicle is inside the blind spot, the solution actively stops unnecessary surrounding searches and saves the acquired parameters locally. This avoids the waste of power consumption and latency caused by invalid searches within the blind spot and ensures that the parameters can be instantly retrieved when the vehicle leaves the blind spot through local storage. Finally, at the moment the vehicle leaves the blind spot, the solution directly reads the target parameters from the local storage and initiates a connection, eliminating the need for real-time search, synchronization, and parsing. This continuous spatiotemporal triggering process fundamentally reduces the processing time at the exit point, enabling network reconnection to be completed within seconds, thereby meeting the stringent requirements of security services for the timeliness of communication recovery.

[0032] In some embodiments, the specific methods for determining the positional relationship between the target vehicle's trajectory and the target signal blind spot include: The predicted position of the target vehicle is calculated based on the spatiotemporal driving data of the target vehicle after a preset time period in the future; the predicted position is matched with the spatiotemporal mapping relationship in the preset spatiotemporal mapping library to obtain the signal receiving power corresponding to the predicted position; when the signal receiving power of the predicted position is lower than the preset signal strength threshold and the driving direction of the target vehicle is pointing towards the target signal blind zone, it is determined that the target vehicle will enter the target signal blind zone after a preset time period in the future.

[0033] Understandably, this technical solution achieves accurate prediction of blind spot entry points. By calculating the predicted location and matching signal power, the system can identify blind spot boundaries before the vehicle actually arrives, ensuring that pre-search actions are triggered only when it is confirmed that the vehicle will enter the blind spot, thus avoiding blind operations. This enhancement further improves the accuracy of spatiotemporal triggering and the effectiveness of action execution, thereby more reliably compressing network reconnection latency to the second level, effectively guaranteeing the continuity and reliability of in-vehicle safety services.

[0034] In some embodiments, the specific methods for determining that a target vehicle has entered a target signal blind zone include: Obtain the vehicle-mounted map of the target vehicle, wherein the vehicle-mounted map pre-marks static blind spot labels for signal-blocking structures, and the static blind spot labels include at least the blind spot entrance; match the predicted position of the target vehicle with the static blind spot labels in the vehicle-mounted map; when the predicted position is within a preset forward distance range of the blind spot entrance defined by the static blind spot label and the target vehicle's driving direction is pointing towards the blind spot entrance, it is determined that the target vehicle has entered the target signal blind spot.

[0035] Specifically, the vehicle-mounted map is pre-labeled with static blind spot tags for various signal-obstructing structures at the factory. Taking a tunnel on the vehicle's current route as an example, the vehicle-mounted map stores the static blind spot tags for that tunnel. The tags include the precise latitude and longitude coordinates of the tunnel entrance and exit, as well as the blind spot type being tunnel type. These static tags are derived from surveying data and have the highest confidence level.

[0036] As the vehicle approaches the tunnel, it collects its own spatiotemporal driving data in real time, including its current geographical location, speed, and heading. Based on the current speed, the system calculates the predicted location two seconds from now. The system then spatially matches the predicted location with all static blind spot labels stored in the vehicle's onboard map.

[0037] When the predicted location enters within 50 meters of the tunnel entrance, the system further detects the vehicle's heading. The system determines that the vehicle's heading is consistently pointing towards the tunnel entrance for more than one second. At this point, both conditions are met, and the system determines that the vehicle is about to enter the tunnel blind zone, generating a pre-search trigger command.

[0038] When determining the target cells for pre-search, the system employs a collaborative approach of static labels and a dynamic spatiotemporal mapping database. The static labels provide the tunnel's exit location and direction, which the system uses to query the historical network dwell results of the tunnel exit raster in the dynamic spatiotemporal mapping database. The database shows two cells with the highest historical dwell success rates at the tunnel exit, identified as cells A and B. The system selects these two cells as candidate cells for this pre-search.

[0039] When static blind spot labels exist in the vehicle map, the system always prioritizes the geofence defined by the static labels as the spatial basis for predictive triggering. This is because the mapping accuracy of vehicle maps is generally better than the boundaries fitted by the dynamic mapping library through crowdsourced data. Meanwhile, exit neighboring cell information that static labels cannot provide is supplemented by the dynamic mapping library. Together, they complete the entire predictive process from determining the trigger timing to selecting candidate cells.

[0040] Understandably, this technical solution achieves independent prediction of blind spot entrances through pre-marked static blind spot labels in the vehicle map. Matching the predicted location with the labels allows for early identification of blind spot boundaries even without real-time signal measurement, reducing reliance on signal strength thresholds and improving trigger robustness in obscured scenarios such as tunnels. This approach ensures that the pre-search action reliably initiates before approaching the entrance, further guaranteeing reconnection latency compressed to the second level, enhancing the continuity and reliability of vehicle safety services.

[0041] In some embodiments, the above-described method for restoring the vehicular network signal before pre-searching further includes: The preset time period and preset signal strength threshold are determined based on the current speed of the target vehicle. The current speed is positively correlated with the preset time period and the preset signal strength threshold.

[0042] Understandably, this technical solution enables adaptive adjustment of predicted parameters to vehicle speed. At high speeds, the preset time period is extended and the signal strength threshold is increased, allowing the system to initiate searches from a greater distance, reserving sufficient processing time for rapid reconnection. At low speeds, the time period is shortened and the threshold is lowered to avoid premature triggering and reduced parameter timeliness. This dynamic adaptation mechanism ensures that the pre-search action is executed accurately and reliably at different vehicle speeds, further guaranteeing that network reconnection latency is stably compressed to the second level, meeting the continuity and reliability requirements of in-vehicle safety services.

[0043] In some embodiments, the specific methods for constructing a spatiotemporal mapping library include: Historical spatiotemporal driving data, historical network dwell results, and historical neighbor cell measurement data of vehicles are collected. The collected data is divided into multiple grids according to geographical location and time period, and the network environment characteristics of each grid within the target time period are statistically analyzed. The network environment characteristics include the reference signal received power distribution of the service area and neighbor cells based on historical neighbor cell measurement data, and blind zone labels generated by comparing historical network dwell results with signal strength thresholds. The historical dwell cell identifiers with the highest frequency or most stable dwell within the corresponding time period of the target grid are used as historical network dwell results and are associated and stored with geographical location, time period, and network environment characteristics corresponding to the target grid to form a spatiotemporal mapping relationship for the target grid. A spatiotemporal mapping library is generated based on the spatiotemporal mapping relationship corresponding to each grid.

[0044] Specifically, during daily driving, the vehicle continuously collects three types of driving data. The first type is spatiotemporal driving data, which includes the vehicle's geographical coordinates, speed, heading, and corresponding timestamps recorded at regular intervals. The second type is network dwell results, which includes the serving cell identifier, frequency band, and reference signal received power for each network connection. The third type is neighbor cell measurement data, which includes periodic measurement reports of the signal strength of surrounding neighboring cells while the vehicle is connected.

[0045] The system divides the collected data into geographic location grids and time periods. Each geographic location grid has a side length of fifty meters, and each grid corresponds to a square geographic area. The time period is in hours, dividing the day into twenty-four time slots. For the data of a specific grid within a specific time slot, the system statistically analyzes two dimensions of network environment characteristics.

[0046] The first dimension of the network environment feature is the distribution of reference signal received power (RSP) between the serving cell and neighboring cells. The system averages the RSP values ​​of all serving cells and neighboring cells collected for that grid during that time period to form the signal strength distribution map of that grid for that time period. The second dimension of the network environment feature is the dead zone label. When the network dwell time for all vehicles in a grid is no service, or the average RSP value is consistently below -110 milliwatt decibels, the system marks that grid as a signal dead zone during that time period.

[0047] For each raster, the system counts the number of times and duration of historical dwellings for each cell within each time period. The system selects the cell with the highest number of dwellings and the longest dwelling time as the historical network dwelling result for that raster within that time period. The system then associates and stores this historical network dwelling result with the corresponding geographic location raster number, time period number, and the aforementioned network environment characteristics, forming a spatiotemporal mapping record. The records for all rasteres across all time periods collectively constitute the spatiotemporal mapping relationship.

[0048] After the mapping library is built, the system updates it daily or when the vehicle network status changes. When a vehicle passes a grid and collects new network data, the system compares this data with existing records for the corresponding grid and time period. If the new reference signal received power value or the number of cells in use differs from existing records, the system updates the signal strength distribution using a weighted average and includes the new cells in the statistics. In the updated mapping library, the location matching error of each spatiotemporal mapping record does not exceed fifty meters, and the update cycle does not exceed twenty-four hours.

[0049] Understandably, this technical solution constructs a high-precision spatiotemporal mapping library through rasterized statistical analysis of historical data. This library associates and stores geographical location, time period, network environment characteristics, and stable cell identifiers, providing historical pattern support for the prediction process and improving the accuracy of identifying candidate cells at blind zone boundaries and outgoing directions, avoiding reliance on single measurements. This enhancement provides a reliable data foundation for pre-search actions, ensuring that reconnection latency is consistently compressed to the second level, meeting the requirements for secure service continuity and reliability.

[0050] Please refer to Figure 2 , Figure 2 This is a flowchart illustrating the steps involved in iteratively optimizing a spatiotemporal mapping library in a pre-search network signal vehicular network recovery method.

[0051] In some embodiments, the above-described method for restoring the vehicular network signal before pre-searching further includes: S201. When the target vehicle is on a driving trajectory leaving the target signal blind zone, collect feedback data on crossing the target signal blind zone. The feedback data includes at least a list of candidate service areas, the network signal quality of each candidate service area, the location of the target signal blind zone, and the entry time.

[0052] S202. Compare the feedback data with the data recorded in the target raster in the spatiotemporal mapping library to generate an optimization decision.

[0053] S203. Based on optimization decisions, update the network environment characteristics and historical network dwell results of the target grid.

[0054] Specifically, a vehicle successfully traversed a signal blind spot in an underground parking garage. After exiting the garage, the vehicle directly established a connection with the target cell using pre-stored candidate cell access parameters, and the network was successfully restored. At this point, the system automatically collected all feedback data from this blind spot traversal.

[0055] The first part of the feedback data is the predicted trigger data. The system records the vehicle's actual geographical location when the prediction module issues the trigger command; this location is 80 meters away from the garage entrance. The system also records the moment of the predicted trigger and the time it takes for the vehicle to travel from that location to the actual location of the network outage, which is 4 seconds.

[0056] The second part of the feedback data is the pre-search execution data. The system records the identifiers of the three candidate cells locked during the pre-search in the exit direction before entering the blind zone. For each candidate cell, the system records the reference signal received power value measured during the pre-search phase. The reference signal received power of the first candidate cell is -95 mW dB, the second is -100 mW dB, and the third is -105 mW dB.

[0057] The third part of the feedback data is the network recovery result data. The target cell the vehicle actually stayed in after leaving the garage was the second candidate cell. The total latency from detecting the signal in this cell to completing the connection was 0.8 seconds. After recovery, network service remained stable without any disconnections.

[0058] The system compared the feedback data with existing records of the garage exit grid in the spatiotemporal mapping library. The comparison revealed two discrepancies that required optimization.

[0059] The first discrepancy involves the blind zone geofence boundary. The blind zone end position labeled in the mapping library for this garage exit grid deviates from the actual vehicle network recovery position. The actual vehicle network recovery position is 15 meters ahead of the labeled position. The system determines that the blind zone exit boundary needs to be corrected 15 meters inwards towards the garage interior.

[0060] The second difference involves the priority of candidate cells in the outgoing direction. In this pre-search, the three candidate cells were prioritized as follows: Cell 1 highest, Cell 2 second highest, and Cell 3 lowest. However, Cell 2 was actually the one camped, and its reference signal received power was not the highest. The system determined that Cell 2 had a higher actual connectivity success rate at the outgoing direction, therefore prioritizing Cell 2 to first place, relegating Cell 1 to second place, and keeping Cell 3 in third place.

[0061] The system also evaluated the lead time parameter for the predicted trigger. The actual time from triggering to network disconnection was four seconds, while the preset lead time was two seconds. The system determined that the lead time was too short and adjusted the predicted lead time parameter for the blind spot of the garage from two seconds to three seconds, giving subsequent vehicles more time to pre-search.

[0062] Based on the above optimization decisions, the system performs three update operations on the garage exit grid. The first is to update the blind zone label, moving the coordinates of the blind zone exit boundary fifteen meters inwards into the garage. The second is to update the historical network dwell results, adjusting the priority ranking of candidate cells to Cell 2, Cell 1, and Cell 3. The third is to update the prediction trigger parameters, modifying the pre-emptive time parameter corresponding to the blind zone to three seconds.

[0063] After the update, the record of the garage exit grid in the spatiotemporal mapping library is closer to the actual network state. When the next car passes through the same garage, the predicted trigger time will be more accurate, the priority of candidate cells searched will be more accurate, and network recovery will be faster and more reliable.

[0064] Understandably, this technical solution introduces a closed-loop update mechanism based on traversal feedback data. By collecting the candidate service area list, signal quality, and location / time information at the blind zone exit, and comparing it with historical records in the spatiotemporal mapping library, optimization decisions are generated to dynamically update the network environment characteristics and dwell results of the corresponding grid. This mechanism enables the mapping library to continuously learn and improve, effectively enhancing its adaptability to dynamic network environment changes and predictive accuracy, thereby ensuring that reconnection latency is stably compressed to the second level, providing reliable assurance for vehicle-mounted safety services.

[0065] Please refer to Figure 3 , Figure 3 This is a flowchart illustrating the steps involved in obtaining candidate service areas through road test edge nodes in a method for restoring vehicular networks by pre-searching network signals.

[0066] In some embodiments, the above-described method for restoring the vehicular network signal before pre-searching further includes: S301. Generate a list of candidate service areas related to the target vehicle's frequently driven areas and current travel route, and save the list of candidate service areas to the target vehicle's local storage space. The list of candidate service areas is used to provide the target vehicle with optional candidate service areas and network connection parameters for the candidate service areas.

[0067] S302. When the target vehicle is on its driving trajectory before entering the target signal blind zone, a communication channel is established between the target vehicle and the roadside edge node in the current area. The current area is a geographical area centered on the location of the target vehicle and with a preset distance as the radius.

[0068] S303. Obtain the latest information of candidate service areas in the direction of the exit of the target signal blind zone from the roadside edge node through the communication channel. The latest information includes multiple candidate service area identifiers uploaded by each vehicle within the preset range where the roadside edge node is located, as well as the network connection parameters corresponding to each candidate service area identifier.

[0069] S304. Update the list of candidate service areas using the latest information.

[0070] Specifically, during daily driving, a subset of the spatiotemporal mapping library is stored locally on the vehicle. This subset covers the areas the vehicle frequently travels, including the owner's residence, workplace, and signal blind spots around shopping malls and hospitals that are frequently passed. For each blind spot, the vehicle stores the geographic coordinates of the blind spot's entrance and exit, the blind spot type label, and a list of candidate cells with the highest historical success rate of re-entry in the exit direction. When the vehicle is in an area with no network coverage or where edge nodes are unreachable, prediction and pre-search are performed independently entirely based on this locally stored data.

[0071] Within the same city, multiple vehicles with similar capabilities are operating on the roads. Each vehicle, after traversing a signal blind spot and successfully restoring network access, reports its traversal data to the roadside edge nodes covering that area. This data includes the actual location of the network outage, the actual location of network restoration, the identifier of the successfully camped cell at the exit, and the restoration latency. The edge nodes aggregate the data reported by all vehicles within their service area, continuously updating the list and priority ranking of candidate cells for each blind spot in the area through aggregation analysis. For example, if an operator adds a new base station near a tunnel exit, the edge nodes can discover and verify the availability of the new cell within minutes by using the new cell camping records reported by the first few vehicles crossing the tunnel.

[0072] The vehicle is currently traveling along a highway and is about to enter a tunnel. Based on high-precision maps or historical data, the vehicle anticipates entering the tunnel's blind spot in thirty seconds. At this point, the vehicle is still in a signal-enabled area outside the tunnel entrance, and it sends a query request to the edge nodes covering this section of the road, carrying the tunnel's blind spot identifier.

[0073] Upon receiving the request, the edge node retrieves the latest candidate cell list maintained for the tunnel blind zone exit direction. Due to the presence of the aforementioned newly added base station, the list returned by the edge node includes one more new cell than the list stored locally on the vehicle. The edge node also returns pre-synchronization auxiliary information such as the frequency point and physical cell identifier corresponding to each of these three candidate cells.

[0074] After receiving the response message from the edge node, the vehicle replaces the locally stored list of candidate cells for the tunnel exit with the latest list returned by the edge node. Originally, the list contained only two old cells; now it is updated to three, with the newly added cells ranked first, followed by the two existing cells. Subsequently, before entering the tunnel blind zone, the vehicle performs a pre-synchronization operation on these three candidate cells according to the updated list.

[0075] When the vehicle exits the tunnel, the strongest signal is found in the newly added cell at the exit. Since the access parameters for this cell were obtained in advance during the pre-synchronization phase, the vehicle directly initiates random access to that cell, and the network recovers in less than a second. Without obtaining the latest information from the edge nodes, the vehicle would rely on the locally stored old list for pre-synchronization, potentially missing this newly added optimal cell, leading to increased recovery latency or access to a suboptimal cell.

[0076] Understandably, this technical solution achieves multi-source dynamic maintenance of candidate service area information by combining a locally pre-stored candidate list with real-time updates from edge nodes. The local list ensures basic availability, while edge nodes provide the latest network status reported by surrounding vehicles, making candidate information for blind spot exit directions more accurate and timely. This mechanism effectively responds to changes in the network environment, ensuring that the parameters obtained through pre-search match the actual signal conditions, further improving the reconnection success rate at the exit, and providing a reliable guarantee for compressing latency to the second level.

[0077] This technical solution also provides an embodiment of a method for restoring an in-vehicle network after pre-searching for network signals. The specific implementation process is as follows: After the vehicle is powered on, the GNSS data acquisition module collects the vehicle's spatiotemporal driving data in real time at a frequency of no less than 1Hz, including the current GPS geographical location, driving direction unit vector, real-time driving speed and positioning accuracy correction coefficient, and transmits the data synchronously to the blind spot prediction module and the spatiotemporal mapping library module.

[0078] The spatiotemporal mapping library module dynamically constructs and updates the location-network spatiotemporal mapping library based on real-time vehicle driving data, network dwell results, and neighboring cell measurement data. The communication mapping relationship of this mapping library is defined as the mapping from the set of geographic locations and time dimensions to the set of network environment features and historical dwell results. The update cycle does not exceed 24 hours, and the location matching error is controlled within 50 meters.

[0079] The blind spot prediction module calculates the vehicle's predicted position after a preset time period based on current spatiotemporal driving data. This preset time period is 2 seconds by default and can be adaptively adjusted based on vehicle speed. The calculation formula is the current geographical location plus the product of speed, time period, and direction vector, plus a positioning accuracy correction coefficient. Then, the predicted position is queried from a spatiotemporal mapping library to obtain signal strength and blind spot label information. When the reference signal received power at the predicted position is lower than a preset threshold and the driving direction continuously points towards the signal blind spot for more than 1 second, a pre-search trigger command is output; otherwise, continuous acquisition is resumed to avoid invalid measurements.

[0080] After receiving the trigger command, the pre-search execution module searches for 4G LTE and 5G NR neighboring cells around the blind zone exit. The baseband module completes the measurement, search, and pre-synchronization of the target neighboring cells, acquiring full access information including operator, standard, frequency band, frequency point, physical cell identifier, synchronization signal block information, and timing advance, and stores this information in the spatiotemporal mapping library. If the network is interrupted when the vehicle enters the blind zone, the system locks this information, prohibiting unnecessary updates and modifications.

[0081] After the vehicle leaves the blind spot, the GNSS module pushes a departure trigger signal, and the baseband communication module directly calls the pre-stored target cell access information from the mapping library, skipping the full-band blind scan process, and immediately initiates network registration and data link establishment, achieving network recovery in seconds.

[0082] After the network is successfully restored, the system will synchronize all the data from this pre-search and recovery to the spatiotemporal mapping library, iteratively optimize blind zone labels, neighboring zone information and prediction parameters, and continuously improve the prediction accuracy and pre-search effectiveness.

[0083] Please refer to Figure 4 , Figure 4 This is a schematic diagram of the structure of an in-vehicle network recovery device that pre-searches for network signals.

[0084] This embodiment also provides an in-vehicle network recovery device for pre-searching network signals, comprising: The prediction module 401 is used to determine the positional relationship between the target vehicle's driving trajectory and the target signal blind zone based on the target vehicle's spatiotemporal driving data and the spatiotemporal mapping library data of the network signal.

[0085] The search module 402 is used to search for surrounding service areas in the direction of the exit of the target signal blind zone when the target vehicle is on the driving trajectory before entering the target signal blind zone, and to obtain the network connection parameters of the candidate service area after determining at least one candidate service area.

[0086] The storage module 403 is used to stop searching for surrounding service areas and save each network connection parameter to local storage space when the target vehicle is in the target signal blind zone.

[0087] The communication module 404 is used to read the target network connection parameters from the local storage space and establish a network connection with the candidate service area corresponding to the target network connection parameters when the target vehicle is on a driving trajectory that leaves the target signal blind zone, so as to restore the network communication of the target vehicle.

[0088] The specific implementation methods and technical effects of the embodiments of the present invention can be referred to the specific implementation methods and technical effects of the vehicle network recovery method for pre-searched network signals in any of the above embodiments.

[0089] Please refer to Figure 5 , Figure 5 This is a schematic diagram of the architecture of an in-vehicle network recovery system that pre-searches for network signals.

[0090] This embodiment provides an in-vehicle network recovery system that pre-searches for network signals. It adopts a modular design, with each module working together to achieve rapid network reconnection in blind spot scenarios.

[0091] The GNSS data acquisition module is bidirectionally connected to the blind spot prediction module and the spatiotemporal mapping library module. It is responsible for collecting the vehicle's spatiotemporal driving data in real time at a sampling frequency of no less than 1Hz, including the current GPS geographical location, driving direction unit vector, real-time driving speed and positioning accuracy information, and synchronously transmitting the data to the relevant modules to provide basic data support for blind spot prediction and mapping library updates.

[0092] The spatiotemporal mapping library module is bidirectionally connected to the GNSS data acquisition module, blind zone prediction module, and pre-search execution module. It is used for local storage and real-time dynamic updating of the location-network spatiotemporal mapping library. This mapping library contains four core dimensions of data: geographic location information, including GPS coordinates, road segment identification, blind zone labels, and scene type; surrounding communication cell information, including operator, standard, frequency band, frequency point, physical cell identifier, and synchronization signal block information; historical dwell information, recording dwell time, service quality, and handover success rate; and time dimension information, including busy / idle time labels and network quality time variation patterns.

[0093] The blind spot prediction module's input is connected to the GNSS data acquisition module and the spatiotemporal mapping library module, while its output is connected to the pre-search execution module. This module calculates the vehicle's predicted position based on the current spatiotemporal driving data and, in conjunction with the spatiotemporal mapping library, determines whether the vehicle is about to enter a signal blind spot. When the triggering conditions are met, it outputs a pre-search trigger command, making it the core decision-making unit of the entire system.

[0094] The input of the pre-search execution module is connected to the blind zone prediction module, and the output is connected to the baseband communication module and the spatiotemporal mapping library module. After receiving the pre-search trigger command, this module controls the baseband module to complete the pre-search and pre-synchronization of neighboring cells around the blind zone exit, obtain the full access information of the target cell, and store the information in the spatiotemporal mapping library for subsequent use.

[0095] The baseband communication module is bidirectionally connected to the pre-search execution module and the protocol stack module. It is responsible for measuring, searching and pre-synchronizing 4G LTE and 5G NR full-mode neighbor cells according to the instructions of the pre-search execution module, and outputting full access information of the cell. After the vehicle leaves the blind zone, the baseband communication module directly initiates network registration based on the pre-stored cell information, skipping the full-band blind scan process.

[0096] This system, implemented using domestically produced 4G communication chips, is planned for mass production application in the Hongqi E009 model. The software has passed full-function testing, and the T-Box has obtained OTS approval. Verification results in a simulated underground parking garage environment show that, with LTE neighboring cell coverage at the garage exit, the 95% confidence intervals for network recovery times at offline durations of 5 seconds, 20 seconds, 10 minutes, and 20 minutes are 2 to 4 seconds, 2.8 to 3.2 seconds, 5 to 8 seconds, and 3 to 4 seconds, respectively, with an average recovery latency of 4.47 seconds. Compared to the traditional baseline solution's recovery time of 15 to 23 seconds, this solution reduces the average latency by 70.2%, fully validating the system's stable recovery capability in long-term offline scenarios.

[0097] The beneficial effects of the above embodiments are as follows: Network recovery latency is significantly reduced, meeting compliance requirements. By predicting signal blind spots in advance and completing pre-search and pre-synchronization of surrounding cells before entering the blind spot, vehicles can directly initiate registration based on pre-stored information after exiting the vehicle. This reduces recovery latency from 15-23 seconds in traditional solutions to less than 10 seconds. The average recovery latency in short-term offline scenarios in underground parking garages is as low as 3.1 seconds, fully meeting the second-level recovery requirements of safety services such as emergency calls, eliminating safety hazards. Precise pre-search triggering ensures power consumption control meets automotive-grade requirements. Based on the vehicle's spatiotemporal location and blind spot labels in the mapping library, precise triggering only initiates measurement when the vehicle is predicted to enter a blind spot, avoiding indiscriminate full-band scanning. The increase in terminal static power consumption does not exceed 5%, meeting automotive-grade low-power design requirements. Full-scenario and full-system compatibility with strong mass production adaptability. It is compatible with all typical vehicle signal blind spot scenarios such as underground parking garages, highway tunnels, and urban elevated road obstructions. It also supports 4G LTE and 5G NR full-system network adaptation without customized modifications, possessing strong industrial promotion value. Improved communication reliability and reduced user complaints resulted in a 90% decrease in network-related user experience issues in mass-produced vehicles after implementation. The number of vehicle networking service work orders was significantly reduced, greatly improving the continuity of in-vehicle communication services and user experience.

[0098] This invention also provides a vehicle control device, including a memory, a processor, and a program stored in the memory and executable on the processor. When the program is executed by the processor, it implements the vehicle network recovery method for pre-searching network signals described in the above embodiments.

[0099] Taking the example of a processor and memory in a vehicle controller being connected via a bus, the memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, the memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, the memory may optionally include memory remotely located relative to the control processor, and these remote memories can be connected to the control device via a network. The non-transitory software programs and instructions required to implement the control methods of the above embodiments are stored in the memory, and when executed by the processor, the control methods of the above embodiments are performed.

[0100] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0101] This invention also provides a vehicle, including the vehicle control device described in the above embodiments.

[0102] The vehicle can be a private car, such as a sedan, SUV, MPV, or pickup truck. It can also be a commercial vehicle, such as a van, bus, small truck, or large semi-trailer. The vehicle must have an electric motor capable of outputting power or acting as a generator to store mechanical energy. When the vehicle is a new energy vehicle, it can be a hybrid or a pure electric vehicle.

[0103] Since the vehicle applies all the technical solutions of the above-mentioned control device or vehicle controller, it has at least all the beneficial effects brought about by the technical solutions of the above embodiments, which will not be repeated here.

[0104] Furthermore, one embodiment of the present invention provides a computer-readable storage medium storing computer-executable instructions for performing the above-described method for restoring vehicular networks by pre-searching for network signals.

[0105] It is worth noting that, since the computer-readable storage medium of the present invention is capable of executing the vehicle network recovery method for pre-searching network signals of any of the above embodiments, the specific implementation and technical effects of the computer-readable storage medium of the present invention can be referred to the specific implementation and technical effects of the vehicle network recovery method for pre-searching network signals of any of the above embodiments.

[0106] Furthermore, one embodiment of the present invention also provides a computer program product, including a computer program or computer instructions, which are stored in a computer-readable storage medium. A processor of a computer device reads the computer program or computer instructions from the computer-readable storage medium and executes the computer program or computer instructions, causing the computer device to perform the above-described method for restoring vehicular networks by pre-searching for network signals.

[0107] It is worth noting that, since the computer program product of this embodiment can execute the vehicle network recovery method for pre-searching network signals of any of the above embodiments, the specific implementation method and technical effects of the computer program product of this embodiment can be referred to the specific implementation method and technical effects of the vehicle network recovery method for pre-searching network signals of any of the above embodiments.

[0108] It will be understood by those skilled in the art that all or some of the steps and systems in the methods disclosed above can be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components can be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, which can include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. Furthermore, as is known to those skilled in the art, communication media typically include computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0109] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

Claims

1. A method for restoring a vehicular network by pre-searching for network signals, characterized in that, The method includes: Based on the spatiotemporal driving data of the target vehicle and the spatiotemporal mapping library data of the network signal, the positional relationship between the driving trajectory of the target vehicle and the target signal blind zone is determined. When the target vehicle is on its driving trajectory before entering the target signal blind zone, the surrounding service areas in the direction of the exit of the target signal blind zone are searched, and the network connection parameters of the candidate service area are obtained after determining at least one candidate service area. When the target vehicle is in the target signal blind zone, the search for the surrounding service area is stopped, and each network connection parameter is saved to the local storage space. When the target vehicle is on a driving trajectory that leaves the target signal blind zone, the target network connection parameters are read from the local storage space, and a network connection is established with the candidate service area corresponding to the target network connection parameters to restore the network communication of the target vehicle.

2. The method according to claim 1, characterized in that, The specific methods for determining the positional relationship between the target vehicle's trajectory and the target signal blind spot include: The predicted location of the target vehicle after a preset time period is calculated based on the spatiotemporal driving data of the target vehicle. The predicted location is matched with the spatiotemporal mapping relationship in the preset spatiotemporal mapping library to obtain the signal receiving power corresponding to the predicted location; When the signal reception power at the predicted location is lower than a preset signal strength threshold and the target vehicle is traveling in the direction of the target signal blind zone, it is determined that the target vehicle will enter the target signal blind zone after the preset future time period.

3. The method according to claim 2, characterized in that, The specific methods for determining that the target vehicle has entered the target signal blind zone include: Obtain the vehicle-mounted map of the target vehicle, wherein the vehicle-mounted map contains pre-marked static blind spot labels for signal blocking structures, and the static blind spot labels include at least the blind spot entrance; The predicted location of the target vehicle is matched with the static blind spot labels in the vehicle map; When the predicted position is within a preset forward distance from the blind spot entrance defined by the static blind spot label and the target vehicle is traveling in the direction of the blind spot entrance, it is determined that the target vehicle has entered the target signal blind spot.

4. The method according to claim 2, characterized in that, The method further includes: The future preset time period and the preset signal strength threshold are determined based on the current speed of the target vehicle, wherein the current speed is positively correlated with the future preset time period and the preset signal strength threshold.

5. The method according to claim 2, characterized in that, The specific methods for constructing the spatiotemporal mapping library include: Collect historical spatiotemporal driving data of vehicles, historical network dwell results, and historical neighboring cell measurement data; The collected data is divided into multiple grids according to geographical location and time period, and the network environment characteristics of each grid within the target time period are statistically analyzed. The network environment characteristics include the reference signal received power distribution of the service area and neighboring cells based on the historical neighboring cell measurement data, and the blind zone label generated based on the comparison of the historical network dwell results and the signal strength threshold. The historical cell identifiers of the target grid that have the highest frequency of occurrence or the most stable residence within the corresponding time period are used as the historical network residence results, and are associated and stored with the geographical location, time period and network environment characteristics corresponding to the target grid to form a spatiotemporal mapping relationship corresponding to the target grid. The spatiotemporal mapping library is generated based on the spatiotemporal mapping relationship corresponding to each of the grids.

6. The method according to claim 5, characterized in that, The method further includes: When the target vehicle is on a driving trajectory leaving the target signal blind zone, feedback data of crossing the target signal blind zone is collected. The feedback data includes at least a list of candidate service areas and the network signal quality of each candidate service area, the location of the target signal blind zone and the entry time. The feedback data is compared with the data recorded in the target raster in the spatiotemporal mapping library to generate an optimization decision; Based on the optimization decision, the network environment characteristics and historical network dwell results of the target grid are updated.

7. The method according to claim 1, characterized in that, The method further includes: Generate a list of candidate service areas related to the target vehicle's frequently driven area and current travel route, and save the list of candidate service areas to the target vehicle's local storage space. The list of candidate service areas is used to provide the target vehicle with selectable candidate service areas and network connection parameters for the candidate service areas. When the target vehicle is on its driving trajectory before entering the target signal blind zone, a communication channel is established between the target vehicle and the roadside edge node in the current area, wherein the current area is a geographical area centered on the location of the target vehicle and with a preset distance as the radius; The latest information of candidate service areas in the exit direction of the target signal blind zone is obtained from the roadside edge node through the communication channel. The latest information includes multiple candidate service area identifiers uploaded by each vehicle within the preset range of the roadside edge node and the network connection parameters corresponding to each candidate service area identifier. The candidate service area list is updated using the latest information.

8. A vehicle network recovery device for pre-searching network signals, characterized in that, The device includes: The prediction module is used to determine the positional relationship between the target vehicle's driving trajectory and the target signal blind zone based on the target vehicle's spatiotemporal driving data and the spatiotemporal mapping library data of network signals. The search module is used to search for surrounding service areas in the direction of the exit of the target signal blind zone when the target vehicle is on the driving trajectory before entering the target signal blind zone, and to obtain the network connection parameters of the candidate service area after determining at least one candidate service area. The storage module is used to stop searching for the surrounding service area when the target vehicle is in the target signal blind zone, and to save each network connection parameter to the local storage space. The communication module is used to read target network connection parameters from the local storage space and establish a network connection with the candidate service area corresponding to the target network connection parameters when the target vehicle is on a driving trajectory that leaves the target signal blind zone, so as to restore the network communication of the target vehicle.

9. A vehicle, characterized in that, The vehicle includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the vehicle network recovery method for pre-searching network signals as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the vehicle network recovery method for pre-searched network signals as described in any one of claims 1 to 7.

Images