Method for dynamically setting APN of 4G camera device and cloud server system thereof

Abstract

The invention discloses a method for dynamically setting an APN of 4G camera equipment and a cloud server system thereof, and the method comprises the following steps: the 4G camera equipment collects and generates network access environment feature information which at least comprises at least two of SIM card identification information, resident network identification information and wireless network quality information; the 4G camera device sends the network access environment feature information to a cloud server to request APN configuration; the cloud server calls a pre-trained APN prediction model based on the network access environment feature information to generate and issue an APN configuration packet to the 4G camera device, the APN configuration packet at least comprises a candidate APN list, a confidence coefficient parameter corresponding to each candidate APN, candidate APN sorting information, a hierarchical verification strategy, a budget parameter, a rollback strategy, a blacklist strategy, an anti-rollback counter and signature information.

Description

Technical Field

[0001] This invention relates to the field of camera technology, and in particular to a method for dynamically setting the APN of a 4G camera device and its cloud server system. Background Technology

[0002] 4G camera devices (such as 4G low-power cameras, mobile security cameras, vehicle / field monitoring terminals, etc.) typically use cellular communication networks for data reporting, remote control, and video / image transmission. To enable a terminal to establish a data connection within a cellular network, the Access Point Name (APN) parameter usually needs to be configured in the communication module. APN parameters may differ between different operators, regions, and SIM card types. Furthermore, the available APN and network quality for the same terminal may vary at different times or locations due to roaming, changes in network coverage, and changes in base station load.

[0003] In existing technologies, the main methods for configuring APNs in 4G camera devices include: pre-setting a fixed APN at the factory, manually configuring the APN by the user in an app or backend platform, or polling multiple candidate APNs in a fixed order on the device side. However, these methods still have the following shortcomings in practical applications: 1) Insufficient Adaptability: Pre-configured fixed APNs are often only applicable to specific operators or regions. Changing SIM card operators, using the device across regions, or roaming can easily lead to network access failures or low success rates. 2) High Dependence on Manual Configuration: When a terminal cannot access the network, users need to manually configure the APN or receive remote guidance, resulting in high maintenance costs and unsuitability for unattended scenarios. For a large number of distributed camera devices, manual configuration is inefficient and prone to errors. 3) High Cost of Blind Trial and Error: Device-side polling and attempting multiple APNs often lacks effective resource constraints and failure convergence mechanisms, potentially leading to prolonged network access failures, increased traffic consumption, increased power consumption, and excessively long service recovery times. 4) Uncontrollable Network Quality: Even if an APN can establish a connection, issues such as high latency, large packet loss, significant jitter, or service detection failures may still exist, resulting in unstable video transmission, increased remote control latency, or frequent disconnections and reconnections. 5) Frequent Switching and Oscillations: In scenarios with short-term network fluctuations, the lack of a reasonable anti-jitter strategy may cause the terminal to frequently switch APNs, leading to repeated dialing by the communication module, service reconnection, and abnormal power consumption, affecting the stable operation of the terminal. 6) Security risks and configuration degradation issues: Without a trusted verification and anti-rollback mechanism for the distributed configuration, there is a risk that the configuration package may be tampered with or the old configuration package may be replayed, leading to policy degradation and affecting the stability and security of terminal network access. 7) Lack of closed-loop optimization mechanism: Existing solutions usually cannot form an effective feedback loop and cannot use the verification results and network quality data on the device side to adaptively correct subsequent policies, resulting in long-term unstable performance in complex network environments.

[0004] Therefore, how to achieve automatic, fast, and stable APN setting for 4G camera devices in complex network environments with multiple operators, multiple regions, and multiple APN differences, and how to continuously optimize network access strategies under the premise of resource constraints and security and reliability, remains a technical problem that needs to be solved. Summary of the Invention

[0005] To address the problems existing in the prior art, this invention provides a method for dynamically setting the APN of a 4G camera device and its cloud server system.

[0006] To achieve the above objectives, the present invention provides a method for dynamically setting the APN of a 4G camera device, comprising the following steps: Step S1: The 4G camera device collects and generates network access environment feature information, which includes at least two of the following: SIM card identification information, network identification information, and wireless network quality information. Step S2: The 4G camera device sends the network access environment feature information to the cloud server to request APN configuration; Step S3: The cloud server generates and sends an APN configuration package to the 4G camera device based on the network access environment feature information by calling the pre-trained APN prediction model. The APN configuration package includes at least: a candidate APN list, confidence parameters for each candidate APN, candidate APN sorting information, hierarchical verification strategy, budget parameters, rollback strategy, blacklist strategy, anti-rollback counter, and signature information. Step S4: The 4G camera device performs signature verification and anti-rollback verification on the APN configuration package. After the verification is successful, it performs a confidence gating strategy based on the confidence parameter to determine a subset of candidate APNs to be tried from the candidate APN list. Step S5: For each candidate APN in the subset of candidate APNs to be tried, the 4G camera device sequentially performs connection-level verification, resolution-level verification, link-level verification, and service-level verification according to the hierarchical verification strategy. If any verification stage fails, an early stop mechanism is triggered and subsequent verification of the current candidate APN is stopped. At the same time, the current candidate APN is added to the blacklist according to the failure stage or error code. Furthermore, the early stop mechanism is triggered when the cumulative time consumed in any verification stage reaches the duration budget parameter or the cumulative traffic reaches the traffic budget parameter. Step S6: The 4G camera device performs multiple sampling evaluations on candidate APNs that have passed at least a preset grade verification to obtain the average value and stability range of the network connection quality evaluation index, and calculates the candidate APN quality score accordingly. Step S7: The 4G camera device determines the target APN from the verified candidate APNs according to the selection rules with hysteresis and writes it into the 4G communication module. The selection rules with hysteresis include at least the following: switching is allowed only when the quality score of the candidate APN improves by a preset number of consecutive times relative to the quality score of the currently used APN, and a preset cooling time is entered after switching to prevent switching again. Step S8: After the 4G camera device applies the target APN, it monitors the network status. When the target APN fails continuously for a preset number of times within a preset time window, it rolls back to the last known available APN according to the rollback strategy and restricts switching to the target APN where the rollback occurred. Step S9: The 4G camera device uploads structured backhaul data to the cloud server. The structured backhaul data includes at least candidate APN identifiers, success / failure stages of hierarchical verification, corresponding error codes, time consumption of each stage, network connection quality assessment indicators, and traffic consumption information during the verification process. The cloud server updates the candidate APN generation strategy based on the structured backhaul data and generates subsequent APN configuration packages, which are then sent to the 4G camera device to correct and adjust the candidate APN list, hierarchical verification strategy, or budget parameters.

[0007] Preferably, in step S1, the network access environment characteristic information includes at least one or more of the following: IMSI prefix, ICCID field, current MCC / MNC, roaming flag, LTE frequency band information, TAC or cell identifier, RSRP or RSRQ, signal strength information; Furthermore, the network access environment characteristic information further includes at least one piece of historical statistical information, which includes at least: historical successful APN identifier, historical failed APN identifier, historical failed error code distribution, historical failed stage distribution, and the network identifier information at the time of the most recent successful network connection; The historical statistical information is stored in a structured manner by the 4G camera device as statistical records "by operator, by region, or by cell" and uploaded along with this request to improve the accuracy of candidate APN ranking information and confidence parameters.

[0008] Preferably, before step S2, the method further includes: performing two-way authentication when the mobile terminal APP and the 4G camera device are paired via Bluetooth, wherein the two-way authentication includes the exchange of identity credentials between the two parties and digital signature verification, and generating a session key after successful authentication; The network access environment feature information is encrypted and its integrity is verified using the session key during the upload process, and it also carries device identification information, random number and timestamp information, so that the cloud server can verify the legality of the request and suppress replay requests; The device identification information includes at least one or more of the following: device serial number, module identifier, and firmware version number, to support the cloud server in adopting a differentiated candidate APN generation strategy for different hardware / firmware versions.

[0009] Preferably, in step S3, the APN prediction model is trained based on historical network access data, which includes at least the correlation data between network access environment characteristic information and the availability, network latency, packet loss rate or connectivity stability of candidate APNs. The cloud server outputs candidate APN confidence parameters, candidate APN ranking information, and parameter adjustment information for failure correction based on the historical network access data. The parameter adjustment information includes at least one or more of the following: The phase enable / skip flags of the hierarchical verification strategy; The number of retries and timeout thresholds for each verification stage; Duration budget parameters, traffic budget parameters, or power consumption budget parameters; The blacklist strategy specifies the blacklist duration for different error code categories. The preset threshold Δ, the number of consecutive satisfactions N, and the cooling time T are in the hysteresis selection rule; This enables subsequent APN configuration packages to perform targeted convergence and correction for different failure types.

[0010] Preferably, in step S4, the confidence gating strategy includes: When the highest confidence parameter of a candidate APN is not less than a preset confidence threshold, only candidate APNs with confidence parameters not less than the preset confidence threshold are included as the subset of candidate APNs to be tried. When the highest confidence parameter of a candidate APN is less than the preset confidence threshold, the top M candidate APNs in the candidate APN ranking information are taken as the subset of candidate APNs to be tried, where M is a preset positive integer. Furthermore, when determining the subset of candidate APNs to be tried, the 4G camera device performs blacklist filtering and deduplication processing, prioritizing the exclusion of blacklisted APNs within the effective blacklist duration and duplicates with parameters consistent with already tried APNs, in order to reduce traffic consumption, power consumption and network waiting latency caused by invalid dialing attempts.

[0011] Preferably, the hierarchical verification strategy includes at least: Connection-level verification: Perform PDP activation and obtain IP address, and record the PDP activation result and the type of IP assigned; DNS resolution verification: Perform DNS resolution and determine the validity of the resolution result, and record the DNS resolution time and the return status of the resolution server; Link-level verification: Perform a TLS handshake on the specified server and record the handshake time, the reason for the handshake failure, and the certificate verification result; Service-level verification: Perform minimum service detection on the camera service endpoint and determine whether the detection is successful. The minimum service detection includes at least sending a detection request of a preset length to the service endpoint and receiving a response. The early shutdown mechanism includes at least the following: If any tiered verification stage fails, the subsequent verification of the current candidate APN is immediately terminated; The subsequent verification of the current candidate APN is terminated when the cumulative time spent in any tiered verification stage reaches the time budget parameter. The subsequent verification of the current candidate APN is terminated when the cumulative traffic in any tiered verification stage reaches the traffic budget parameter. Furthermore, the 4G camera device associates and maps "failure stage" with "error code" to a blacklist error category, so as to support different blacklist duration strategies for different types of failures.

[0012] Preferably, in step S6, the multiple sampling evaluation includes performing K network probe samplings on each candidate APN that has passed at least a preset grade verification, where K is a preset integer greater than or equal to 2; The network connection quality assessment metrics include at least one or more of the following: round-trip time (RTT), packet loss rate, jitter, number of disconnections, TLS handshake time, and service probe success rate; and the network connection quality assessment metrics are normalized. The stability interval is used to characterize at least the fluctuation range of the round-trip time (RTT) or the packet loss rate in the K samples. Furthermore, the candidate APN quality score is obtained by normalizing and weighting the network connection quality assessment index, and the stability interval is used to penalize and correct the candidate APN quality score to reduce the probability of selecting a candidate APN with large fluctuations, thereby reducing network jitter and service interruption rates.

[0013] Preferably, in step S7, the selection rule with hysteresis includes: Switching to the candidate APN is only permitted when the quality score of the candidate APN improves by more than a preset threshold Δ relative to the quality score of the currently used APN, and this improvement is satisfied for a preset number of times N. Furthermore, after switching, a preset cooling time T is entered to prevent APN switching from being triggered again within the preset cooling time T; The preset threshold Δ is used to avoid erroneous switching due to short-term fluctuations, the preset number of times N is used to constrain the sustainability of candidate APN quality improvement, and the preset cooling time T is used to prevent frequent switching from causing communication module reset, service reconnection, and abnormal power consumption.

[0014] Preferably, the blacklist strategy includes: Different blackout durations are set according to the error code category. The error code categories include at least PDP activation failure, DNS resolution failure, TLS handshake failure, and service probe failure, and different blackout durations are set for different categories. The rollback strategy includes: when the target APN fails consecutively a preset number of times within a preset time window, it will roll back to the last known available APN, and will prohibit switching to the target APN that has rolled back within a preset restriction period; The rollback prevention verification includes: the 4G camera device stores the rollback prevention counter value of the most recently verified rollback prevention. The current APN configuration package is only allowed to be applied when the rollback prevention counter value of the current APN configuration package is greater than the most recently verified rollback prevention counter value, so as to prevent the old configuration package from being replayed and causing the candidate APN list, budget parameters or hierarchical verification strategy to be downgraded. Furthermore, the structured backhaul data further includes: stage time statistics, budget consumption statistics, and joint distribution of failure stages and error codes for each candidate APN, so as to be used by the cloud server to generate a correction strategy that is more in line with the actual network conditions.

[0015] The present invention also provides a cloud server system, including a feature parsing module, a model prediction module, a candidate list generation module, a configuration package encapsulation module, a signature and anti-rollback module, and a feedback correction module; The feature parsing module is used to parse network access environment feature information from 4G camera devices; The model prediction module is used to call the APN prediction model to generate a candidate APN list, candidate APN ranking information, and candidate APN confidence parameters. The candidate list generation module is used to trim or expand the candidate APNs based on the comparison results between the confidence parameter and the preset confidence threshold, and generate a candidate APN list corresponding to the exploratory APN configuration package or the exploitation APN configuration package. The configuration package encapsulation module is used to encapsulate the candidate APN list, confidence parameters, sorting information, hierarchical verification strategy, budget parameters, rollback strategy, blacklist strategy, and parameter adjustment information for failure correction to form an APN configuration package; The signature and anti-rollback module is used to add signature information to the APN configuration package and generate an anti-rollback counter; The feedback correction module is used to receive and process the structured backhaul data transmitted by the 4G camera device, and update the candidate APN generation strategy, hierarchical verification strategy, budget parameters and blacklist strategy based on the structured backhaul data, thereby generating a subsequent APN configuration package and sending it to the 4G camera device.

[0016] The technical solution of this invention has the following beneficial effects: This invention, through a closed-loop mechanism of "network access environment feature collection—cloud-based configuration package generation—device-side hierarchical verification—quality scoring and optimization—rollback protection—structured feedback correction," can achieve at least the following beneficial effects: 1) Significantly improved network access success rate: By collecting SIM card identifier, residency network identifier, wireless network quality and historical statistical information, and generating a candidate APN list and sorting / confidence parameters by the cloud based on the features, the device can quickly obtain a more suitable APN configuration for different operators, regions and roaming scenarios, reducing the probability of failure to access the network and improving the success rate and stability of the first network access.

[0017] 2) Reduce reliance on manual configuration and improve the efficiency of large-scale deployment. By distributing configuration packages from the cloud and automatically performing verification, optimization and writing on the device, the reliance on users to manually set the APN is avoided. This is especially suitable for unattended, distributed batch deployment scenarios, which can reduce operation and maintenance costs and reduce human configuration errors.

[0018] 3) The trial and error cost is controllable, reducing traffic and power consumption. The duration budget parameter and traffic budget parameter (optional power consumption budget parameter) are introduced, and an early stop mechanism is triggered during the graded verification process. This allows the device to terminate invalid attempts in a timely manner when verification fails or resource consumption reaches the threshold, reducing traffic loss, network waiting latency and power consumption caused by blind dialing.

[0019] 4) Tiered verification improves the accuracy of judgment and avoids "seemingly connected but unavailable services". It adopts a tiered verification strategy of connection-level verification, resolution-level verification, link-level verification and service-level verification, and records the failure stage and error code. It can more accurately distinguish abnormal situations such as "IP has been obtained but DNS / TLS / service is unavailable", improve the reliability of APN availability judgment, and thus improve service availability and stability.

[0020] 5) Quality scoring ensures network quality and improves service continuity. Verified candidate APNs are sampled and evaluated multiple times. The quality score is calculated by combining indicators such as RTT, packet loss rate, jitter, number of disconnections, TLS handshake time, and service probe success rate. Stability interval penalty correction is introduced to reduce the probability of links with large fluctuations being selected, thereby improving the continuity and stability of video / image uploads and remote control.

[0021] 6) Hysteresis optimization and cooling-off period anti-jitter: avoid frequent switching oscillations. By limiting the switching frequency through hysteresis rules (threshold Δ, number of consecutive satisfactions N) and cooling time T, the erroneous switching caused by short-term network fluctuations can be effectively suppressed, reducing repeated dialing of communication modules and repeated reconnection of services, improving system stability and reducing abnormal power consumption.

[0022] 7) The rollback strategy improves fault tolerance and shortens the recovery time. When the target APN fails continuously within a preset time window and reaches a threshold, it automatically rolls back to the last known available APN and sets switching restrictions on the failed APN to avoid getting caught in a cycle of "repeated failures - repeated switching", thereby shortening the recovery time after network outage and improving overall reliability.

[0023] 8) The blacklist strategy accelerates convergence and improves verification efficiency. By mapping the failure stage and error code to error category and setting differentiated blacklist duration, it can avoid repeated attempts of APN with a high probability of failure in a short period of time, improve the convergence speed of the verification process, and reduce the impact of invalid attempts on latency, traffic and power consumption.

[0024] 9) Signature verification and anti-rollback mechanisms enhance security and trustworthiness. Signature verification is performed on configuration packages distributed from the cloud, and an anti-rollback counter is introduced to prevent configuration packages from being tampered with or old configuration packages from being replayed, which could lead to policy degradation and ensure the trustworthiness and stability of configuration policies.

[0025] 10) Structured backhaul and cloud-based correction form a closed-loop adaptive optimization. The device backhauls candidate APN verification results, failure stages / error codes, time consumption, network quality and budget consumption to the cloud in a structured manner. The cloud updates the candidate APN generation strategy, verification strategy and budget parameters accordingly and sends out subsequent configuration packages to achieve continuous adaptive correction, adapt to different regions, different operators and dynamic network conditions, and improve long-term operating performance. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0027] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0028] Example 1: Reference Figure 1 This embodiment provides a method for dynamically setting the APN of a 4G camera device. This method is applicable to network environments with multiple operators, multiple regions, and multiple APN parameter differences. It can quickly and stably select the optimal APN for network access while ensuring that resource constraints such as traffic and duration are controllable, and achieve continuous adaptive optimization through structured backhaul and cloud correction.

[0029] The method for dynamically setting the APN of a 4G camera device includes the following steps: Step S1: Network access environment feature information collection and generation. The 4G camera device collects and generates network access environment feature information, which includes at least two of the following: SIM card identification information (e.g., IMSI prefix, ICCID field, etc.); registered network identification information (e.g., MCC / MNC, roaming flag, TAC or cell identifier, etc.); wireless network quality information (e.g., RSRP / RSRQ, signal strength, current frequency band information, etc.). The network access environment characteristic information also includes any one or more of the following: LTE frequency band information, signal strength information, TAC or cell identifier, IP type capability (IPv4 / IPv6 / dual-stack support), etc. 4G camera devices can use local historical statistical information as part of the network access environment characteristic information. This historical statistical information includes, for example, historical successful APN identifiers, historical failed APN identifiers; historical failure error code distribution, historical failure stage distribution; and registered network identifier information at the time of the most recent successful network connection; to improve the accuracy of cloud-based candidate APN ranking and confidence assessment.

[0030] Furthermore, the historical statistical information can be stored in a structured statistical record format, and statistical records can be established according to the following dimensions: by operator dimension (MCC / MNC); by region dimension (country / province / city / geofencing); by cell dimension (TAC / Cell ID).

[0031] Each statistical record can include at least: number of successes, number of failures, distribution of failure stages, distribution of failure error codes, the most recent successful APN, and the identifier of the most recent successful network. By uploading the above statistical records, the cloud can more accurately sort and assess the confidence level of APN performance under the same operator / same region / same cell, thereby reducing invalid dialing attempts.

[0032] Step S2: Feature Reporting and APN Configuration Request. The 4G camera device sends its network access environment feature information to the cloud server to request APN configuration. This network access environment feature information is encapsulated in a structured data format, such as key-value pairs / field tables carrying SIM-related fields, network residency fields, wireless quality fields, historical statistics fields, etc. Further, prior to step S2, when the mobile terminal APP and the 4G camera device perform Bluetooth network pairing, two-way authentication can be performed. This two-way authentication includes both parties exchanging identity credentials and performing digital signature verification. After successful authentication, a session key is generated. During the upload of the network access environment feature information, the session key can be used for encryption and integrity verification, and it carries device identification information, a random number, and a timestamp, enabling the cloud server to verify the legitimacy of the request and suppress replay requests. The device identification information preferably includes one or more of the following: device serial number, module identifier, and firmware version number, to support the cloud server in using a differentiated candidate APN generation strategy for different hardware / firmware versions.

[0033] Step S3: Generate and distribute APN configuration package in the cloud. Based on the network access environment feature information, the cloud server calls the pre-trained APN prediction model to generate the APN configuration package and distributes it to the 4G camera device.

[0034] The APN configuration package includes at least: a candidate APN list, confidence parameters for each candidate APN, candidate APN ranking information, tiered verification strategy, budget parameters, rollback strategy, blacklist strategy, anti-rollback counter, and signature information. The candidate APN list may include multiple candidate APN configuration parameters, each of which includes at least: APN name, username / password (if any), authentication method, and IP type (IPv4 / IPv6 / dual-stack). Furthermore, the APN prediction model can be trained based on historical network access data, which includes at least the correlation data between network access environment characteristics and candidate APN availability, network latency, packet loss rate, or connectivity stability. The cloud server outputs candidate APN confidence parameters and candidate APN ranking information based on the historical network access data, and can further generate parameter adjustment information for failure correction.

[0035] The parameter adjustment information may include at least one or more of the following: the phase enable / skip flag of the hierarchical verification strategy; the number of retries and timeout thresholds for each verification phase; the duration budget parameter, traffic budget parameter, or power consumption budget parameter; the blacklist duration corresponding to different error code categories in the blacklist strategy; and the preset threshold Δ, the number of consecutive satisfactions N, and the cooling time T in the hysteresis optimization rule.

[0036] Based on the failure types and resource consumption in the structured backhaul data, the cloud performs targeted convergence and correction on the aforementioned parameters. For example, when the DNS resolution failure rate increases, the resolution-level verification timeout threshold is raised or an alternative resolution strategy is issued; when the TLS handshake failure rate increases, the server endpoint or handshake parameters for link-level verification are adjusted; when budget consumption is high, the number of candidate APNs is reduced or the verification phase budget is reduced. The preset confidence threshold, budget parameters, sampling count K, number of candidates M, hysteresis parameters Δ / N / T, preset time window, and preset limitation period can be issued by the cloud server in the APN configuration package, or pre-stored as default values ​​locally on the 4G camera device. These default values ​​are used when not issued to ensure the method's feasibility.

[0037] Step S4: Device-side signature verification and rollback prevention verification. After receiving the APN configuration package, the 4G camera device performs signature verification and rollback prevention verification: Signature verification: Verify the signature information of the configuration package sent from the cloud. Only after the signature verification is passed can the configuration package be parsed and applied; Rollback prevention verification: Compare the rollback prevention counter. Preferably, the current configuration package is only allowed to be applied when the rollback prevention counter of the current configuration package is greater than the counter value of the most recent verification passed on the device side, so as to avoid the old configuration package being replayed, which would cause the candidate APN list, budget parameters or hierarchical verification strategy to be downgraded.

[0038] Step S5: Confidence gating and hierarchical verification (including early budget stop and blacklist). After the signature verification and anti-rollback verification are passed, the 4G camera device executes the confidence gating strategy according to the confidence parameters of each candidate APN, determines the subset of candidate APNs to be tried from the candidate APN list, and performs verification on each candidate APN in the subset of candidate APNs to be tried according to the hierarchical verification strategy.

[0039] (1) Confidence gating: The confidence gating strategy may include the following rules: when the highest confidence of a candidate APN is not less than a preset confidence threshold, only candidate APNs with a confidence of not less than the threshold are selected to enter the subset to be tried; when the highest confidence of a candidate APN is less than the preset confidence threshold, the top M candidate APNs are selected to enter the subset to be tried, where M is a preset positive integer (e.g., M=2~5). Furthermore, blacklisted APNs are filtered when forming the subset to be tried, and duplicate parameter items are deduplicated to reduce traffic consumption, power consumption, and waiting latency caused by invalid dialing. For example, blacklisted APNs within the valid blacklist duration can be excluded first, and duplicate processing is performed on candidate items with the same parameters such as APN name, authentication method, username / password, etc.

[0040] (2) Hierarchical verification: For each candidate APN in the subset to be tested, the hierarchical verification strategy is executed sequentially: connection-level verification, resolution-level verification, link-level verification, and service-level verification. If any stage fails, an early stop mechanism is triggered, stopping the subsequent verification of the current candidate APN, and adding the candidate APN to the blacklist according to the failure stage or error code. For ease of implementation, this embodiment provides an optional implementation method for each stage: Connection-level verification: Write the candidate APN parameters to the 4G communication module and initiate PDP activation to obtain the IP address; record the PDP activation result, failure reason, and assigned IP type, etc. Resolution-level verification: Perform DNS resolution after the connection is established to determine whether the resolution result is valid; record the resolution time, the resolution server return status, or the error code. Link-level verification: Perform a TLS handshake on the preset cloud service address and record the handshake time, handshake failure reason, and certificate verification result. Service-level verification: Perform minimum service probe on the preset service interface address, such as sending a preset length probe request and receiving a response, to determine whether the service link is available.

[0041] (3) Budget parameters and early stop mechanism: To control trial and error costs, a budget parameter is introduced into the tiered verification process. The budget parameter includes at least a duration budget parameter or a traffic budget parameter, and preferably also a power consumption budget parameter. When the cumulative time consumed in any verification stage reaches the duration budget parameter or the cumulative traffic reaches the traffic budget parameter, the early stop mechanism is triggered to terminate the verification of the current candidate APN, thereby avoiding excessive traffic consumption and reducing network access latency. The cumulative time consumed can be obtained based on the start and end timestamps of each verification stage, and the cumulative traffic can be obtained based on the number of bytes sent / received reported by the communication module or the cumulative network interface counter.

[0042] (4) Blacklist strategy: The device can map "failure stage + error code" to blacklist error categories, such as PDP activation failure, DNS resolution failure, TLS handshake failure, and service probe failure, and set different blacklist durations for different categories. During the blacklist duration, the device will prioritize excluding the blacklisted APN to avoid repeated failures. Preferably, the blacklist duration can be set according to the error category, for example: PDP activation failure: 1-10 minutes; DNS resolution failure: 1-5 minutes; TLS handshake failure: 2-10 minutes; service probe failure: 1-5 minutes.

[0043] Step S6: Multiple Sampling Evaluation and Stability Range. For candidate APNs that have passed at least the preset graded verification, perform multiple sampling evaluations to obtain the average value and stability range of network connection quality evaluation indicators, and calculate the candidate APN quality score. Perform K network probe samplings for each candidate APN, where K is a preset integer ≥ 2 (e.g., K = 3~10). Network connection quality evaluation indicators include at least one or more of the following: Round-Trip Time (RTT), packet loss rate, jitter, number of disconnections, TLS handshake time, and service probe success rate. The stability range can be used to characterize the fluctuation range of the K samplings, for example: RTT stability range = maximum RTT value. Minimum RTT; Packet loss rate stability range = Maximum packet loss rate Minimum packet loss rate.

[0044] Furthermore, a stability interval is introduced into the scoring penalty term to apply a penalty correction to candidate APNs with large fluctuations, thereby reducing the probability of selecting candidate APNs with high jitter and thus reducing the service interruption rate. For example, the candidate APN quality score can be expressed as: Score = Σ(w i ·Norm(x i )) λ·Penalty (stability interval), where x i For metrics such as RTT and packet loss rate, w i λ is the weighting coefficient, Norm(·) is the normalization function, λ is the penalty coefficient, and Penalty(·) is used to penalize and correct candidate APNs with large fluctuations.

[0045] Step S7: Hysteresis Selection and Cooling-Off Anti-jitter Switching. The device determines the target APN from the verified candidate APNs and writes it into the 4G communication module based on hysteresis selection rules. The hysteresis selection rules include at least the following: switching is allowed only when the quality score of the candidate APN improves by more than a preset threshold Δ relative to the quality score of the currently used APN, and this improvement is satisfied consecutively for a preset number of times N; after switching, a preset cooling-off time T is entered, during which switching is prohibited to avoid frequent switching leading to communication module reset, service reconnection, and abnormal power consumption. Here, Δ is used to suppress false switching caused by short-term fluctuations, N is used to constrain the continuity of quality improvement, and T is used to limit the switching frequency. Preferably, the hysteresis parameters can take the following example ranges: Δ can be a score improvement of 3% to 20%; N can be satisfied consecutively 2 to 5 times; T can be 30 seconds to 10 minutes.

[0046] Step S8: Network Access Monitoring and Rollback Restriction. After the target APN is applied, the device continuously monitors the network status. When the target APN fails a preset number of times consecutively within a preset time window, it rolls back to the last known available APN according to the rollback policy, and restricts switching to the target APN where the rollback occurred. This restriction can be set to a preset restriction period or can be lifted after a new configuration package is issued from the cloud for correction, to avoid prolonged offline time or excessive power consumption caused by repeated failures.

[0047] Step S9: Structured Backhaul and Cloud Correction Closed Loop. The device uploads structured backhaul data to the cloud server. This structured backhaul data includes at least: candidate APN identifiers, success / failure stages of hierarchical verification, corresponding error codes, time consumption for each stage, network connection quality assessment indicators, and traffic consumption information during the verification process. The structured backhaul data also includes candidate APN stage time statistics, budget consumption statistics, and the joint distribution of failure stages and error codes, so that the cloud can generate a correction strategy that better reflects the actual network conditions. The cloud server updates the candidate APN generation strategy based on the structured backhaul data and generates subsequent APN configuration packages, which are then sent to the device. This corrects and adjusts the candidate APN list, hierarchical verification strategy, or budget parameters, forming a dynamic adaptive closed loop. Furthermore, the cloud update can be an online correction of candidate APN ranking, confidence parameters, budget parameters, and verification thresholds; or it can be an offline incremental training update of the APN prediction model.

[0048] Example 2: This second embodiment provides a 4G camera device that can execute the method for dynamically setting the APN described in the first embodiment.

[0049] The 4G camera device may include: a processor, a memory, a 4G communication module, a communication interface (which may include at least one of a Bluetooth module, a Wi-Fi module, or an Ethernet interface), and an antenna unit connected to the 4G communication module.

[0050] The memory stores a computer program. When the processor executes the computer program, it causes the 4G camera device to perform the following functional modules or steps: 1) Feature Acquisition Module: Collects and generates network access environment feature information, including at least two of the following: SIM card identification information, residing network identification information, and wireless network quality information; and optionally generates historical statistical information. 2) Feature Reporting Module: Reports network access environment feature information to the cloud server to request APN configuration. 3) Configuration Packet Reception and Verification Module: Receives APN configuration packets from the cloud and performs signature verification and anti-rollback verification. 4) Gating and Filtering Module: Determines a subset of candidate APNs to be tried from the candidate APN list based on a confidence level gating strategy, and performs blacklist filtering and deduplication. 5) Hierarchical Verification and Budget Control Module: Verifies candidate APNs sequentially at the connection level, resolution level, link level, and service level, and triggers an early stop mechanism based on budget parameters. 6) Scoring and Optimization Module: Performs K sampling evaluations on verified candidate APNs to obtain quality scores, and determines the target APN based on hysteresis optimization rules. 7) Write and Rollback Module: Used to write the target APN to the 4G communication module and monitor network status. When consecutive failures reach a threshold, it rolls back to the last known available APN and restricts further switching. 8) Backhaul and Feedback Module: Used to upload structured backhaul data to the cloud server to support cloud-based error correction and optimization.

[0051] Furthermore, the 4G camera device may include a secure storage area or a secure chip for storing the public key or certificate chain required for signature verification and the anti-rollback counter value to improve the security of configuration packet verification.

[0052] Example 3: This embodiment three provides a cloud server system, which corresponds to the method for dynamically setting APN described in embodiment one. It is used to parse the network access environment feature information uploaded by 4G camera devices, generate candidate APN configurations and distribute them, and provide feedback correction on the structured backhaul data returned by the devices, so as to achieve continuous optimization of candidate APN generation strategy, hierarchical verification strategy and budget parameters.

[0053] The cloud server system includes: a feature parsing module, a model prediction module, a candidate list generation module, a configuration package encapsulation module, a signature and anti-rollback module, and a feedback correction module.

[0054] The system includes the following modules: A Feature Parsing Module: This module parses network access environment feature information from 4G camera devices. This information includes at least two of the following: SIM card identification information, network identification information, and wireless network quality information. Preferably, it also parses historical statistical information uploaded by the device to improve the accuracy of candidate APN ranking and confidence parameter output. A Model Prediction Module: This module calls an APN prediction model and outputs, based on the network access environment feature information, a candidate APN list, confidence parameters for each candidate APN, and candidate APN ranking information. Preferably, it can also output parameter adjustment information for failure correction, such as enabling / skipping tiered verification strategies, retry count and timeout threshold, budget parameters, blacklist duration, and hysteresis parameters. A Candidate List Generation Module (Exploratory / Exploitative Configuration Package): This module trims or expands candidate APNs based on the comparison between the confidence parameters and preset thresholds, and generates configuration package types matching different confidence scenarios. Exploratory APN configuration package: Generated when the highest confidence level is less than a preset threshold, with a larger number of candidates and a stricter budget, used for exploration in low-confidence scenarios. Exploitative APN configuration package: Generated when the highest confidence level is not less than a preset threshold, with a smaller number of candidates and a more lenient budget, used for rapid network access in high-confidence scenarios. Configuration package encapsulation module: Used to encapsulate the candidate APN list, confidence parameters, sorting information, tiered verification strategy, budget parameters, rollback strategy, blacklist strategy, and parameter adjustment information to form an APN configuration package. Signature and anti-rollback module: Used to add signature information to the APN configuration package and generate an anti-rollback counter to ensure the trustworthiness and replay resistance of the configuration package.

[0055] The feedback correction module receives and processes structured feedback data from the device, and updates the candidate APN generation strategy, hierarchical verification strategy, budget parameters, and blacklist strategy based on the structured feedback data, thereby generating subsequent configuration packages to be sent to the device. Preferably, the structured feedback data includes candidate APN identifiers, verification success / failure stages, error codes, stage duration, network quality indicators, and budget consumption information; and may further include stage duration statistics, budget consumption statistics, and the joint distribution of failure stages and error codes to enhance the accuracy of the correction strategy.

[0056] Example 4: This embodiment four provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the method steps for dynamically setting the APN of a 4G camera device as described in embodiment one, including: collecting and reporting network access environment feature information, receiving and verifying cloud configuration packets, confidence gating, hierarchical verification and budget early stop, sampling scoring and hysteresis optimization, rollback limitation, structured backhaul and cloud correction closed loop.

[0057] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural transformations made using the contents of the present invention's specification and drawings under the inventive concept of the present invention, or direct / indirect applications in other related technical fields, are included within the patent protection scope of the present invention.

Claims

1. A method for dynamically setting the APN of a 4G camera device, characterized in that, Includes the following steps: Step S1: The 4G camera device collects and generates network access environment feature information, which includes at least two of the following: SIM card identification information, network identification information, and wireless network quality information. Step S2: The 4G camera device sends the network access environment feature information to the cloud server to request APN configuration; Step S3: The cloud server generates and sends an APN configuration package to the 4G camera device based on the network access environment feature information by calling the pre-trained APN prediction model. The APN configuration package includes at least: a candidate APN list, confidence parameters for each candidate APN, candidate APN sorting information, hierarchical verification strategy, budget parameters, rollback strategy, blacklist strategy, anti-rollback counter, and signature information. Step S4: The 4G camera device performs signature verification and anti-rollback verification on the APN configuration package. After the verification is successful, it performs a confidence gating strategy based on the confidence parameter to determine a subset of candidate APNs to be tried from the candidate APN list. Step S5: For each candidate APN in the subset of candidate APNs to be tried, the 4G camera device sequentially performs connection-level verification, resolution-level verification, link-level verification, and service-level verification according to the hierarchical verification strategy. If any verification stage fails, an early stop mechanism is triggered and subsequent verification of the current candidate APN is stopped. At the same time, the current candidate APN is added to the blacklist according to the failure stage or error code. Furthermore, the early stop mechanism is triggered when the cumulative time consumed in any verification stage reaches the duration budget parameter or the cumulative traffic reaches the traffic budget parameter. Step S6: The 4G camera device performs multiple sampling evaluations on candidate APNs that have passed at least a preset grade verification to obtain the average value and stability range of the network connection quality evaluation index, and calculates the candidate APN quality score accordingly. Step S7: The 4G camera device determines the target APN from the verified candidate APNs according to the selection rules with hysteresis and writes it into the 4G communication module. The selection rules with hysteresis include at least the following: switching is allowed only when the quality score of the candidate APN improves by a preset number of consecutive times relative to the quality score of the currently used APN, and a preset cooling time is entered after switching to prevent switching again. Step S8: After the 4G camera device applies the target APN, it monitors the network status. When the target APN fails continuously for a preset number of times within a preset time window, it rolls back to the last known available APN according to the rollback strategy and restricts switching to the target APN where the rollback occurred. Step S9: The 4G camera device uploads structured backhaul data to the cloud server. The structured backhaul data includes at least candidate APN identifiers, success / failure stages of hierarchical verification, corresponding error codes, time consumption of each stage, network connection quality assessment indicators, and traffic consumption information during the verification process. The cloud server updates the candidate APN generation strategy based on the structured backhaul data and generates subsequent APN configuration packages, which are then sent to the 4G camera device to correct and adjust the candidate APN list, hierarchical verification strategy, or budget parameters.

2. The method for dynamically setting the APN of a 4G camera device according to claim 1, characterized in that, In step S1, the network access environment characteristic information includes at least one or more of the following: IMSI prefix, ICCID field, current MCC / MNC, roaming flag, LTE frequency band information, TAC or cell identifier, RSRP or RSRQ, and signal strength information; Furthermore, the network access environment characteristic information further includes at least one piece of historical statistical information, which includes at least: historical successful APN identifier, historical failed APN identifier, historical failed error code distribution, historical failed stage distribution, and the network identifier information of the last successful network connection.

3. The method for dynamically setting the APN of a 4G camera device according to claim 1, characterized in that, The steps preceding step S2 also include: When the mobile terminal APP and the 4G camera device perform Bluetooth pairing, two-way authentication is performed. The two-way authentication includes the exchange of identity credentials between the two parties and digital signature verification. After successful authentication, a session key is generated. The network access environment feature information is encrypted and its integrity is verified using the session key during the upload process. It also carries device identification information, random numbers, and timestamp information so that the cloud server can verify the legitimacy of the request and suppress replay requests.

4. The method for dynamically setting the APN of a 4G camera device according to claim 1, characterized in that, In step S3, the APN prediction model is trained based on historical network access data, which includes at least the correlation data between network access environment characteristic information and the availability, network latency, packet loss rate or connectivity stability of candidate APNs. The cloud server outputs candidate APN confidence parameters, candidate APN ranking information, and parameter adjustment information for failure correction based on the historical network access data; wherein, the parameter adjustment information includes at least one or more of the following: phase enable / skip flags for the hierarchical verification strategy; retry count and timeout threshold for each verification phase; Duration budget parameters, traffic budget parameters, or power consumption budget parameters; The blacklist strategy includes the blacklist duration corresponding to different error code categories; the preset threshold Δ, the number of consecutive satisfactions N, and the cooldown time T in the hysteresis optimization rule; so that the subsequent APN configuration package can perform targeted convergence and correction for different failure types.

5. The method for dynamically setting the APN of a 4G camera device according to claim 1, characterized in that, In step S4, the confidence gating strategy includes: When the highest confidence parameter of a candidate APN is not less than a preset confidence threshold, only candidate APNs with confidence parameters not less than the preset confidence threshold are included as the subset of candidate APNs to be tried. When the highest confidence parameter of a candidate APN is less than the preset confidence threshold, the top M candidate APNs in the candidate APN ranking information are taken as the subset of candidate APNs to be tried, where M is a preset positive integer.

6. The method for dynamically setting the APN of a 4G camera device according to claim 1, characterized in that, The hierarchical verification strategy includes at least the following: Connection-level verification: Perform PDP activation and obtain IP address, and record the PDP activation result and the type of IP assigned; DNS resolution verification: Perform DNS resolution and determine the validity of the resolution result, and record the DNS resolution time and the return status of the resolution server; Link-level verification: Perform a TLS handshake on the specified server and record the handshake time, the reason for the handshake failure, and the certificate verification result; Service-level verification: Perform minimum service detection on the camera service endpoint and determine whether the detection is successful. The minimum service detection includes at least sending a detection request of a preset length to the service endpoint and receiving a response. The early shutdown mechanism includes at least the following: If any tiered verification stage fails, the subsequent verification of the current candidate APN is immediately terminated; The subsequent verification of the current candidate APN is terminated when the cumulative time spent in any tiered verification stage reaches the time budget parameter. The subsequent verification of the current candidate APN is terminated when the cumulative traffic in any tiered verification stage reaches the traffic budget parameter.

7. The method for dynamically setting the APN of a 4G camera device according to claim 1, characterized in that, In step S6, the multiple sampling evaluation includes performing K network probe samplings on each candidate APN that has passed at least a preset grade verification, where K is a preset integer greater than or equal to 2; The network connection quality assessment metrics include one or more of the following: round-trip time (RTT), packet loss rate, jitter, number of disconnections, TLS handshake time, and service probe success rate, and the network connection quality assessment metrics are normalized.

8. The method for dynamically setting the APN of a 4G camera device according to claim 1, characterized in that, In step S7, the selection rule with hysteresis includes: Switching to the candidate APN is only permitted when the quality score of the candidate APN improves by more than a preset threshold Δ relative to the quality score of the currently used APN, and this improvement is satisfied for a preset number of times N. Furthermore, after switching, a preset cooling time T is entered to prevent APN switching from being triggered again within the preset cooling time T.

9. The method for dynamically setting the APN of a 4G camera device according to claim 1, characterized in that, The blacklist strategy includes: Different blackout durations are set according to the error code category. The error code categories include at least PDP activation failure, DNS resolution failure, TLS handshake failure, and service probe failure, and different blackout durations are set for different categories. The rollback strategy includes: when the target APN fails consecutively a preset number of times within a preset time window, it will roll back to the last known available APN, and will prohibit switching to the target APN that has rolled back within a preset restriction period; The rollback prevention verification includes: the 4G camera device stores the rollback prevention counter value of the most recently verified rollback prevention. The current APN configuration package is only allowed to be applied when the rollback prevention counter value of the current APN configuration package is greater than the most recently verified rollback prevention counter value, so as to prevent the old configuration package from being replayed and causing the candidate APN list, budget parameters or hierarchical verification strategy to be downgraded. Furthermore, the structured backhaul data further includes: stage time statistics, budget consumption statistics, and joint distribution of failure stages and error codes for each candidate APN, so as to be used by the cloud server to generate a correction strategy that is more in line with the actual network conditions.

10. A cloud server system, characterized in that, It includes a feature parsing module, a model prediction module, a candidate list generation module, a configuration package encapsulation module, a signature and anti-rollback module, and a feedback correction module; The feature parsing module is used to parse network access environment feature information from 4G camera devices; The model prediction module is used to call the APN prediction model to generate a candidate APN list, candidate APN ranking information, and candidate APN confidence parameters; The candidate list generation module is used to trim or expand the candidate APNs based on the comparison results between the confidence parameter and the preset confidence threshold, and generate a candidate APN list corresponding to the exploratory APN configuration package or the exploitation APN configuration package. The configuration package encapsulation module is used to encapsulate the candidate APN list, confidence parameters, sorting information, hierarchical verification strategy, budget parameters, rollback strategy, blacklist strategy, and parameter adjustment information for failure correction to form an APN configuration package; The signature and anti-rollback module is used to add signature information to the APN configuration package and generate an anti-rollback counter; The feedback correction module is used to receive and process the structured backhaul data transmitted by the 4G camera device, and update the candidate APN generation strategy, hierarchical verification strategy, budget parameters and blacklist strategy based on the structured backhaul data, thereby generating a subsequent APN configuration package and sending it to the 4G camera device.

Images