Automatic test system and method for multi-device call test

Abstract

The invention discloses an automatic test system and method for a multi-device call test. The system comprises a device state control module, an interface recognition and answering control module, a cross-device call time sequence control module, a double-side call event acquisition module and a unified time axis log fusion module. The system establishes connection between a calling device and a called device, executes dialing, ringing confirmation and interface acquisition, and sequentially performs interface structure identification, text identification and image matching to determine an answering area, thereby automatically completing answering operation. Through setting time sequence check points of dialing completion, ringing confirmation, answering confirmation, call stability, hang-up completion and the like, sequence consistency of cross-device call actions is realized; corresponding relations are established between key events of a calling side and key events of a called side according to timestamps, and double-side logs and interface screenshots are fused and stored according to a unified time axis, so that event alignment and centralized tracing are realized.

Description

Technical Field

[0001] This invention relates to the field of mobile communication testing technology, and in particular to an automated testing system and method for multi-device call testing. Background Technology

[0002] Current mobile terminal call testing mostly relies on manual operation or single-device scripting, lacking unified control over the call process between multiple devices. In existing testing procedures, the calling and called devices are usually operated by different personnel, making it difficult to maintain consistency in the timing of actions. This results in inaccurate alignment of key data such as call setup time, connection sequence, and hang-up sequence, thus affecting fault location and performance analysis.

[0003] Furthermore, in the ringing interface of the called device, due to significant differences in interfaces across different manufacturers, system versions, and language environments, recognition failures are prone to occur when relying on a single identification method (such as relying solely on text or a fixed layout), leading to unstable answering actions and unreproducible test results. Existing technologies also generally lack the ability to collect call events from both the calling and called ends, making it difficult to establish correlations between key events such as outgoing calls, answering, connection, call holding, and hanging up on the two devices, resulting in a lack of a unified timeline for the entire call chain. Simultaneously, logs from the calling and called devices are generally stored in a scattered manner, lacking a unified timeline fusion mechanism. Cross-device problem analysis is complex and inefficient, and most tools cannot automatically restore to a consistent state and re-execute when anomalies occur, nor do they possess a rollback call checkpoint mechanism.

[0004] Therefore, existing multi-device call testing technologies have significant shortcomings in terms of automation, interface recognition reliability, cross-device timing consistency, bilateral event correlation, and unified log traceability. There is an urgent need for an automated call testing system and method that can coordinate cross-device control, has multi-dimensional interface recognition capabilities, supports timing checkpoints and automatic recovery mechanisms, and can complete bilateral event alignment and provide unified timeline log fusion, in order to solve the problems of low efficiency, poor stability, and difficulty in problem localization of existing technologies.

[0005] Therefore, existing technologies still need to be improved. Summary of the Invention

[0006] Given the significant deficiencies in the aforementioned multi-device call tests regarding automated control, interface recognition stability, cross-device action timing consistency, bilateral event correlation, and unified log traceability, this invention proposes an automated call testing system and method that enables collaborative control between the calling and called devices, possesses multi-dimensional interface recognition capabilities, supports key timing checkpoints and anomaly recovery mechanisms in the call process, and can integrate bilateral call events and logs along a unified timeline. This effectively improves the repeatability, timing accuracy, and fault location efficiency of the tests.

[0007] The technical solution of the present invention is as follows: This invention provides an automated testing system for multi-device call testing, comprising: The device status control module is used to establish connections with the calling device and the called device respectively, obtain the device status, and perform dialing, hanging up and status query operations on the calling device, and perform status detection and interface acquisition on the called device. The interface recognition and answering control module is used to recognize the interface image when the called device enters the ringing state, determine the answering area through interface structure recognition, text recognition and image template matching, and automatically execute the answering operation. The cross-device call timing control module is used to set timing checkpoints during dialing, ring confirmation, answer confirmation, call hold and hang-up to ensure the consistency of the order of cross-device call actions. The dual-sided call event acquisition module is used to record key events such as call time, answer recognition completion time, connection time, call duration and hang-up time for both the calling and called sides. The unified timeline log fusion module is used to merge and store the event records, logs and screenshots of the calling and called devices according to timestamps to form an indexable unified call record.

[0008] In one embodiment, the interface recognition and answering control module sequentially performs text matching, optical character recognition, and interface screenshot template matching to determine the answering area in a multi-level recognition manner.

[0009] In one embodiment, the cross-device call timing control module executes test tasks in a multi-threaded manner, sets interruptible checkpoints at each timing node, and executes a preset recovery strategy when an anomaly is detected.

[0010] In one embodiment, the dual-sided call event acquisition module performs failure type classification statistics on failed events and generates call connection rate trend records.

[0011] In one embodiment, the unified timeline log fusion module names log files using device identifiers and time tags, and stores interface screenshots and call records in chronological order of events.

[0012] In one embodiment, the system supports the deployment of multiple test sites and synchronizes test results and logs to the scheduling center through secure communication to achieve cross-regional parallel testing.

[0013] In one embodiment, the system includes a parameter configuration module for setting the test device combination, number of calls, call duration, task interval and exception handling method, and archiving the configuration corresponding to the test session.

[0014] In another aspect, the present invention provides an automated call testing method based on multi-device collaboration, comprising: S1. Detect the connection status of the calling and called devices and establish a test session; S2. Generate test tasks based on test parameters and add them to the scheduling queue; S3. The calling device performs the dialing operation and records the call time; S4. Detect the ringing status on the called device, identify the answer interface and perform the answer operation, and record the connection time; S5. Maintain the call for the set duration and monitor the device status and abnormal events; S6. After the hang-up conditions are met, perform the hang-up operation and record the hang-up time and call result; S7. Save test logs according to device identification and time sequence, and generate test result analysis records.

[0015] In one embodiment, the execution order of answering recognition in step S4 is text matching, optical character recognition, and image template matching, and an exception branch is triggered if recognition fails within a timeout period.

[0016] In one embodiment, in step S5, when a device is detected to be offline, low on power, or experiencing continuous interface recognition failures, abnormal recovery is achieved by isolating the abnormal device, recording the abnormal log, and scheduling a backup device to continue executing the unfinished task.

[0017] In summary, this invention achieves full-process automation and precise timing alignment from dialing and answering to hanging up by constructing a collaborative control mechanism between the calling and called devices, combined with key technologies such as multi-dimensional interface recognition, cross-device call timing control, dual-side event acquisition, and unified timeline log fusion. Furthermore, timing checkpoints and anomaly recovery strategies ensure the stability and repeatability of the testing process. This technical solution effectively overcomes the shortcomings of existing multi-device call testing methods, such as reliance on manual intervention, unstable recognition, asynchronous actions, scattered logs, and difficulty in problem localization. It provides a highly reliable, highly consistent, and reproducible automated testing capability for multi-device communication scenarios.

[0018] Compared with existing multi-device call testing technologies that rely on manual operation, single identification methods, and distributed recording, the system and method proposed in this invention produce technical effects that cannot be expected by existing technologies in the following aspects: First, this invention enables consistent synchronization of call actions between the calling and called devices at the millisecond-level timestamps through the collaboration of five modules: device status control, interface recognition, call timing control, dual-side event collection, and unified timeline log fusion. This overcomes the inherent limitation of existing technologies where cross-device actions cannot be precisely aligned, and significantly improves the relevance and reproducibility of call link events.

[0019] Secondly, this invention adopts a combination of multi-dimensional interface recognition strategies such as structure recognition, text recognition and template matching, which enables the answering interface recognition to maintain stability and high success rate in different brands, system versions and language environments. This is significantly better than the high failure rate caused by relying on a single OCR or fixed layout in the existing technology, and greatly improves cross-device adaptability.

[0020] Furthermore, by setting multiple interruptible checkpoints such as dial completion, ring confirmation, answer completion, call hold and hang-up completion, the present invention enables multi-device call actions to have a controllable sequential structure, and can continue to execute unfinished tasks through recovery strategies when an anomaly occurs, realizing an automated self-recovery capability that is not available in the prior art, and enabling high reliability for long-term, large-scale continuous testing.

[0021] Furthermore, this invention is the first to achieve the fusion and storage of key events and interface screenshots on both the calling and called sides along a unified timeline. This eliminates the need for manual log comparison in problem localization, allowing for direct viewing of the complete call path in an indexable manner. This enables a dual-side event tracing capability that is unattainable by existing technologies.

[0022] Finally, this invention supports cross-site collaborative testing, which can synchronize data from multiple test sites to the scheduling center in a unified format, forming a time consistency analysis structure in a global test environment. This provides a new technical capability for large-scale cross-regional call behavior evaluation, a system-level scalability capability that is completely absent in existing technologies.

[0023] In summary, this invention has achieved technological advancements in several key aspects, such as call action consistency, interface recognition stability, cross-device event alignment, self-recovery capability, and unified log fusion, which are beyond the expectations of existing technologies. It has significant technical effects and inventiveness. Attached Figure Description

[0024] The present invention will be further described below with reference to the accompanying drawings and embodiments. In the accompanying drawings: Figure 1 The present invention provides a structural block diagram of an automated testing system for multi-device call testing; Figure 2 A flowchart of the method steps for an automated testing method for multi-device call testing provided by the present invention; Figure 3 An example diagram of a tool for an automated testing system for multi-device call testing provided by the present invention; Figure 4 Example diagram showing detailed recording of test results for an automated testing method for multi-device call testing provided by this invention; Figure 5 This is an example diagram showing the statistical information of test results for an automated testing method for multi-device call testing provided by the present invention. Detailed Implementation

[0025] To make the objectives, technical solutions, and effects of this invention clearer and more explicit, the invention is further described in detail below. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. The embodiments of the invention are described below in conjunction with the accompanying drawings.

[0026] One embodiment of the present invention provides an automated testing system for multi-device call testing; please refer to [link to relevant documentation]. Figure 1 It includes five core functional components: device status control module 1, interface recognition and answering control module 2, cross-device call timing control module 3, dual-sided call event acquisition module 4, and unified timeline log fusion module 5. Each module works collaboratively to automate the execution of multi-device call testing during system operation.

[0027] The device status control module 1 establishes independent communication connections with both the calling and called devices via a USB debugging interface, ADB channel, or wireless debugging link. It periodically acquires the device's IMEI, network registration status, SIM card status, signal strength, battery level, and whether it is in a callable state. During task execution, this module performs operations such as dialing, hanging up, call holding, and interface refresh on the calling device according to scheduling instructions, and performs device status detection and interface screenshot capture on the called device to ensure that the operating conditions of both devices can be monitored in real time.

[0028] The interface recognition and answering control module 2 initiates an interface acquisition loop during the ringing phase of the called device. It acquires an interface image at fixed intervals and performs a multi-level recognition process. This module prioritizes extracting layout features from the interface structure, such as button positions, text bar areas, and background block shapes. If structural feature judgment is unstable, it performs text recognition, identifying language features such as "answer" and "answer a call." When text recognition still cannot confirm the answering area, the module performs image feature matching between the screenshot and a preset answering button template, selecting the area with the highest similarity from multiple candidate areas as the answering target. Upon successful recognition, the module injects a simulated click event into the called device to achieve automatic answering. This multi-level recognition mechanism ensures a high recognition success rate across different Android UIs and different manufacturer interfaces.

[0029] The cross-device call timing control module 3 sets multiple timing checkpoints during the call process according to the test task workflow, including dial completion confirmation, called party ring confirmation, answer interface recognition completion, call stability maintenance, and hang-up completion checks. The module continuously monitors the status changes of both devices and ensures that each action is completed in the correct order through preset time thresholds and condition judgments. When a timing condition is not met, the module immediately triggers an exception branch and executes a recovery strategy, such as reconnecting the device, restarting the task, or isolating the abnormal device, to ensure the consistency and reproducibility of cross-device actions during the test.

[0030] The dual-sided call event collection module 4 records key call events on both the calling and called devices, including call time, ringing detection time, answer recognition completion time, actual connection time, call duration, and hang-up time. This module assigns a precise timestamp to each event for subsequent timing alignment; it also records the failure type (e.g., interface not recognized, call timeout, call not established) when a test fails, for further analysis and statistical optimization. After independent collection of dual-sided events, they will be correlated during the log fusion phase.

[0031] The unified timeline log fusion module 5 receives event logs, system logs, and screenshots from both the calling and called devices after event collection, and constructs a unified timeline based on timestamps. By sorting and matching the time sequences, the module maps events from both devices one-to-one, achieving a serialized presentation of the entire call process. The fused log files are saved by test task and named according to device identifiers and timestamps, enabling users to quickly locate the specific actions on both sides of the call, significantly improving the efficiency of problem review and fault location.

[0032] Through the collaboration of the above five modules, the system can achieve automated, reproducible, timing-accurate and stable testing capabilities in multi-device call scenarios, fully covering key actions and abnormal situation handling in the call process.

[0033] In a further embodiment, the interface recognition and answering control module 2 sequentially performs text matching, optical character recognition, and interface screenshot template matching to determine the answering area in a multi-level recognition manner.

[0034] Specifically, in this embodiment, the interface recognition and answering control module 2 initiates a recognition loop after the called device enters the ringing state. To improve the recognition success rate across different brands, system versions, and language environments, this module sequentially executes a three-layer recognition strategy on the acquired interface images. The first layer is interface structure judgment, which uses common layout features in the Android UI ringing interface, such as the answer button usually being located in the bottom middle area and the background area having a semi-transparent overlay, to perform initial screening based on layout position and geometry. If the structure judgment fails to effectively locate the answering area, the second layer, text recognition, is entered, which uses a language-adapted text dictionary to recognize keywords such as "answer," "Answer," and "receive a phone call," thereby improving multilingual compatibility. If the device manufacturer uses non-standard text or dynamic icons on the interface, making text recognition unstable, the third layer, image template matching, is executed. This involves using pre-collected multi-version answer button templates to perform feature matching on the interface, selecting the target area based on feature similarity, and generating a click event after confirmation to inject into the called device, thus enabling the answering operation. The superposition of the three-layer recognition mechanism ensures the system's high adaptability and high reliability in diverse interface environments.

[0035] In a further embodiment, the cross-device call timing control module 3 uses a multi-threaded approach to execute test tasks, sets interruptible checkpoints at each timing node, and executes a preset recovery strategy when an anomaly is detected.

[0036] Specifically, in another embodiment, the cross-device call timing control module 3 employs a multi-threaded scheduling structure to manage the entire call process in stages. This module creates an independent task execution thread for each test task, and within this thread, it sequentially sets interruptible timing checkpoints according to five stages: dialing, ring confirmation, answer confirmation, call hold, and hang-up completion. Each checkpoint is accompanied by explicit completion conditions, timeout thresholds, and exception judgment rules, enabling the system to impose fine-grained constraints on the execution order of cross-device actions.

[0037] During the dialing phase, checkpoints are used to confirm whether the calling device has successfully initiated the call; the system comprehensively judges whether the dialing has truly been completed by considering factors such as the dialing interface status, the call status string, and the call progress status. Once the dialing is successful, the timing control module immediately triggers the next stage's ringing detection thread.

[0038] During the ringing confirmation phase, the module periodically checks the interface status of the called device, including ringing prompts, ringing progress, and incoming call notifications in the notification bar. When the called device is detected to have entered the ringing interface, the module records the ringing confirmation time and proceeds to the answering identification phase.

[0039] During the call confirmation phase, the timing control module and the interface recognition and call answering control module 2 work together. After the automatic call answering operation is initiated, the module will monitor whether the call answering is successfully executed within the set time, such as successfully transitioning from the "incoming call interface" state to the "call interface" state, or detecting an audio channel establishment event in the system log. If the call is not connected within the time limit, the module immediately determines that the call answering has failed and enters the abnormal recovery process.

[0040] During the call hold phase, the module starts a continuous monitoring thread to poll the operating status of the calling and called devices in real time, including signal strength, battery level, CPU usage, call process status, and whether the interface has abnormally redirected. If any abnormality occurs on either side, the module will trigger recovery strategies such as pausing the current test, isolating the abnormal device, and switching to a backup device, and re-establish the call process to ensure that the task can continue to be executed.

[0041] During the hang-up phase, the module monitors whether the hang-up signal and hang-up action of the calling device are executed correctly. If the hang-up fails or the call interface cannot be exited, the module will execute a recovery strategy that forcibly terminates the call process and record the exception type.

[0042] The combination of the multi-threaded structure and interruptible checkpoints ensures that the entire call process has strict action sequence constraints and automatic recovery capabilities in abnormal scenarios. This ensures that cross-device testing can maintain stable operation under long-term, multi-round, and high-pressure conditions, and that the entire task will not be interrupted due to the failure of a single stage.

[0043] In a further embodiment, the dual-sided call event acquisition module 4 performs failure type classification statistics on failed events and generates call connection rate trend records.

[0044] Specifically, in this embodiment, the dual-sided call event acquisition module 4 not only collects data such as call out time, answer recognition time, connection time, call duration, and hang-up time, but also classifies and records all failure events. When a failure occurs, the module automatically determines the failure type based on device logs, interface screenshots, and timing information, such as: failure to recognize the answer interface, failure to execute the answer action, network signal failure causing the call not to be established, and no ringing interface on the called device. All failure types are stored in a unified format, and statistical tables are generated according to device, time, and failure reason dimensions. Furthermore, after multiple rounds of testing, the module automatically generates a call connection rate trend chart (not involving graphics, only table data logic) to evaluate call stability under different devices, network environments, and system versions. Through this structured approach, the system can support quality statistics and anomaly localization in large-scale testing scenarios.

[0045] In a further embodiment, the unified timeline log fusion module 5 names the log files using device identifiers and time tags, and stores interface screenshots and call records in chronological order of events.

[0046] More specifically, in one embodiment of the present invention, the unified timeline log fusion module 5 establishes directory structures for the calling device log, the called device log, and the test process log, and names each log file using the format of "device identifier + timestamp" to ensure uniqueness and indexability. The module not only saves system logs, event records, and timestamp data, but also saves screenshots of each interface, call process screenshots, and auxiliary images during the recognition phase (such as cropped candidate answering area images). After testing, the module packages and archives these logs and image files, enabling users to directly locate key behaviors based on the event timeline and quickly trace problems. For example, if a call recognition fails, users can immediately see the interface screenshot and recognition feature comparison at that time through the fused logs, thus having a clear path to problem localization. The unified timeline log fusion module 5 also supports exporting logs for the entire call task set by test round, providing support for large-scale regression testing.

[0047] In a further embodiment, the system supports the deployment of multiple test sites and synchronizes test results and logs to the scheduling center through secure communication to achieve cross-regional parallel testing.

[0048] Specifically, in another embodiment of the present invention, the system can be deployed at multiple test sites, each containing several calling and called devices. The sites interact with the dispatch center via encrypted communication channels. Each site executes test tasks locally, periodically synchronizing event data, screenshots, log files, etc., to the dispatch center incrementally during execution. The dispatch center intelligently distributes tasks based on the number of devices, real-time load, and network conditions at each site, enabling parallel testing across regions and sites. During the fault diagnosis phase, the dispatch center can retrieve logs in a unified format from different sites, achieving global summary and analysis of remote test results. This architecture improves testing efficiency, enabling the system to support large-scale call performance testing, regression testing, and multi-environment device difference analysis.

[0049] In a further embodiment, the system is provided with a parameter configuration module for setting the test device combination, number of calls, call duration, task interval and exception handling method, and archiving the configuration and test session accordingly.

[0050] Specifically, in another embodiment of the present invention, the parameter configuration module in the system is used to centrally manage test-related parameters, including device combination methods, role allocation of calling and called devices, number of single-round calls, task interval, call duration, number of retry attempts, and exception handling strategies. Before the test task starts, the user can set the above parameters through the interface or configuration file. The system automatically generates a parameter snapshot corresponding to this test and writes it to the test session record. Throughout the test, all modules execute according to this parameter snapshot to ensure consistency of test conditions. After the test is completed, the system archives the parameter snapshot and log file as combined information, so that subsequent problem reproduction or test regression can completely reuse the same parameters, improving repeatability and comparative analysis value.

[0051] Another aspect of this invention provides an automated call testing method based on multi-device collaboration. Please refer to [link to relevant documentation]. Figure 2 ,include: Step S1: Detect the connection status of the calling device and the called device and establish a test session; After system startup, the connection status of the calling and called devices is sequentially detected via USB, ADB, or wireless debugging channel. This includes checking if the devices are online, if debugging permissions are available, SIM card status, network registration status, and whether they have dialing capabilities. If any device is abnormal, the system executes an automatic reconnection and permission restoration process, marking the device as unavailable if restoration fails. Once the connection is stable, the system generates a unique test session identifier for this test and writes the device identifier, configuration parameters, and initial status into the test session record for subsequent timing association.

[0052] In a further embodiment, the method further includes: Step S2: Generate test tasks based on test parameters and add them to the scheduling queue; The configuration module reads the user-defined number of calls, call hold duration, task interval, failure retry policy, and exception handling method. Based on these parameters, it generates a complete test task and pushes it to the scheduling queue. The scheduling module arranges the task execution time according to the current system load and generates corresponding subtasks for actions such as dialing, interface detection, answering, call hold, and hanging up, giving the entire task a clear execution path and a controllable timing structure.

[0053] In a further embodiment, the method further includes: Step S3: The calling device performs the dialing operation and records the call time; When the scheduling module triggers task execution, device status control module 1 sends a dialing command to the calling device. The calling device then initiates the dialing interface and performs number input and call execution. The system records the call timestamp simultaneously with the command; this timestamp serves as one of the reference points for subsequent timing alignment. If the dialing interface malfunctions or the number is not entered correctly, the system will record the error and automatically retry or switch calling devices according to the established policy.

[0054] In a further embodiment, the method further includes: Step S4: Detect the ringing status on the called device, identify the answer interface and perform the answer operation, and record the connection time; Once the called device enters the ringing state, the interface recognition and answering control module 2 continuously captures screenshots of the interface and sequentially determines the answering area in the order of structure recognition, text recognition, and image template matching. Upon successful recognition, the system automatically injects a click event into the called device to complete the answer. Upon successful recognition, the system immediately records the answer recognition completion time and the final successful connection time for time-series comparison with the calling end's event. If the answering fails to be completed within the preset time, the system handles it as an exception and records the failure event type.

[0055] In a further embodiment, the method further includes: Step S5: Maintain the call for the set duration and monitor the device status and any abnormal events; After a call is established, the system starts a call hold timer and continuously monitors risk factors such as network signal strength, battery level, CPU usage, abnormal interface transitions, and whether the call process is reclaimed by the system. When network interruption, device offline, or unrecognizable interface is detected, the system immediately executes exception handling strategies, such as reconnecting to the device, pausing the task, or switching to a backup device, to ensure the repeatability and consistency of test results.

[0056] In a further embodiment, the method further includes: Step S6: After the hang-up conditions are met, perform the hang-up operation and record the hang-up time and call result; When the call hold timer reaches the set duration, or the system receives a test task end instruction, device status control module 1 sends a hang-up instruction to the calling device. After the hang-up is completed, the hang-up timestamp is recorded, and the effective call duration is calculated. The system determines whether the call was successfully completed based on the events collected throughout the process and generates test result labels, such as "Call Successful," "Call Failed to Answer," "Connection Timeout," and "Network Interruption."

[0057] In a further embodiment, the method further includes: Step S7: Save the test logs according to the device identifier and time sequence, and generate test result analysis records.

[0058] After the task is completed, the unified timeline log fusion module 5 sorts and merges the event information, system logs, interface screenshots, and identification auxiliary images of the calling and called devices according to timestamps, forming a complete call timeline chain. The merged logs are named with the test session identifier and device identifier and stored in the results directory. The system automatically generates an analysis record for the call task, including an event alignment table, failure reason classification, call connection result statistics, etc., providing a basis for subsequent quality analysis and problem localization.

[0059] In a further embodiment, the execution order of answering recognition in step S4 is text matching, optical character recognition, and image template matching, and an abnormal branch is triggered when recognition fails to occur within a timeout period.

[0060] In another embodiment of the present invention, to ensure that the recognition of the called device's answering interface remains stable under different manufacturers' UIs, different system versions, and different screen resolutions, the interface recognition and answering control module 2 executes the recognition process in the order of text matching, optical character recognition (OCR), and interface screenshot template matching. In the text matching stage, the system quickly scans the text area in the screenshot, recognizing words such as "answer," "Answer," and "incoming call answering" through a language adaptation dictionary. If the interface uses icons or has no text, the system enters the OCR stage, performing multi-directional character recognition on the complete screenshot to improve the recognition rate under conditions of mixed languages ​​or font deformation. If the device uses a button without text or an icon-based design, causing text recognition failure, the system enters the template matching stage, comparing the screenshot with multiple preset Android system answering button templates, selecting the area with the highest similarity from the candidate areas, and automatically clicking it. If the above three recognition methods still fail to locate the answering area within the set timeout period, a recognition failure event is recorded, and an abnormal branch is triggered for the scheduling module to perform task recovery or device switching.

[0061] In a further embodiment, in step S5, when a device is detected to be offline, low on power, or experiencing continuous interface recognition failures, the abnormal recovery is accomplished by isolating the abnormal device, recording the abnormal log, and scheduling a backup device to continue executing the unfinished task.

[0062] In another embodiment of the invention, the system continuously monitors the operating status of the calling and called devices in real time during the call hold phase, including signal strength, network disconnection, battery threshold, whether the interface jumps abnormally, and whether background processes are reclaimed by the system. When the system detects abnormal situations such as device offline, battery below the threshold, continuous interface recognition failure, or device process being killed, it immediately enters the abnormal recovery process. The recovery process includes: marking the currently abnormal device as unavailable, saving the abnormal log, stopping the execution of the relevant threads of the device, and transferring the original task to a backup device. The selection of the backup device is based on a preset priority strategy to ensure that the task can be seamlessly resumed and to avoid interruption of the entire test process. The abnormal recovery mechanism ensures the continuity and robustness of the system under large-scale testing conditions, so that a single point of failure will not affect the entire test task.

[0063] In summary, this invention establishes a collaborative control mechanism between the calling and called devices, and combines key technologies such as multi-level interface recognition, interruptible call timing control, precise dual-side event acquisition, and unified timeline log fusion to achieve automated execution and high-precision synchronous control of multi-device call processes. By enabling the functional modules to operate collaboratively in a layered manner, this invention not only improves the reproducibility and consistency of actions in the call process but also provides automatic recovery capabilities in abnormal scenarios, ensuring stability and engineering usability for large-scale, long-term call testing. The above embodiments are all used to illustrate the technical solutions of this invention. Those skilled in the art can adjust and extend these embodiments without departing from the core ideas of this invention, and these equivalent solutions should all fall within the protection scope of this invention.

Claims

1. An automated testing system for multi-device call testing, characterized in that, include: The device status control module is used to establish connections with the calling device and the called device respectively, obtain the device status, and perform dialing, hanging up and status query operations on the calling device, and perform status detection and interface acquisition on the called device. The interface recognition and answering control module is used to recognize the interface image when the called device enters the ringing state, determine the answering area through interface structure recognition, text recognition and image template matching, and automatically execute the answering operation. The cross-device call timing control module is used to set timing checkpoints during dialing, ring confirmation, answer confirmation, call hold and hang-up to ensure the consistency of the order of cross-device call actions. The dual-sided call event acquisition module is used to record key events such as call time, answer recognition completion time, connection time, call duration and hang-up time for both the calling and called sides. The unified timeline log fusion module is used to merge and store the event records, logs, and screenshots of the calling and called devices according to timestamps, so as to form an indexable unified call record.

2. The automated testing system according to claim 1, characterized in that, The interface recognition and answering control module sequentially performs text matching, optical character recognition, and interface screenshot template matching to determine the answering area in a multi-level recognition manner.

3. The automated testing system according to claim 1 or 2, characterized in that, The cross-device call timing control module uses a multi-threaded approach to execute test tasks, sets interruptible checkpoints at each timing node, and executes a preset recovery strategy when an anomaly is detected.

4. The automated testing system according to any one of claims 1 to 3, characterized in that, The dual-sided call event acquisition module performs failure type classification statistics on failed events and generates call connection rate trend records.

5. The automated testing system according to any one of claims 1 to 4, characterized in that, The unified timeline log fusion module names log files using device identifiers and timestamps, and stores screenshots and call records in chronological order of events.

6. The automated testing system according to any one of claims 1 to 5, characterized in that, The system supports deployment at multiple test sites and synchronizes test results and logs to the scheduling center via secure communication to achieve cross-regional parallel testing.

7. The automated testing system according to any one of claims 1 to 6, characterized in that, The system is equipped with a parameter configuration module, which is used to set the test equipment combination, number of calls, call duration, task interval and exception handling method, and archive the configuration and test session accordingly.

8. A method for automated call testing based on multi-device collaboration, characterized in that, include: S1. Detect the connection status of the calling and called devices and establish a test session; S2. Generate test tasks based on test parameters and add them to the scheduling queue; S3. The calling device performs the dialing operation and records the call time; S4. Detect the ringing status on the called device, identify the answer interface and perform the answer operation, and record the connection time; S5. Maintain the call for the set duration and monitor the device status and abnormal events; S6. After the hang-up conditions are met, perform the hang-up operation and record the hang-up time and call result; S7. Save test logs according to device identification and time sequence, and generate test result analysis records.

9. The automated testing method according to claim 8, characterized in that, In step S4, the execution order of call recognition is text matching, optical character recognition, and image template matching, and an exception branch is triggered if recognition fails within the timeout period.

10. The automated testing method according to claim 8 or 9, characterized in that, In step S5, when a device is detected to be offline, low on power, or experiencing continuous interface recognition failures, the abnormal device is isolated, an error log is recorded, and a backup device is scheduled to continue executing the unfinished task to complete the abnormal recovery.

Images