Background process priority management system and methods

Abstract

This invention discloses a background process management system and method based on user-defined priority sorting, belonging to the field of mobile terminal operating system technology. The system includes a priority classification module, a resource scheduler, a status monitoring engine, and a scene configuration module. It achieves fine-grained resource management through a five-level priority classification system (real-time guarantee level, task continuation level, resource retention level, intelligent release level, and forced reclamation level). The resource scheduler adopts a kernel-level policy execution unit and a preemptive hierarchical scheduling model, while the status monitoring engine saves process states using copy-on-write technology. The scene configuration module supports policy switching based on time period and geographical location, and employs a double-buffering mechanism to ensure uninterrupted switching. This invention solves the problems of lack of user customization capabilities in background process resource allocation and high risk of task interruption, achieving technical effects such as response latency ≤1 millisecond, memory fragmentation rate ≤5%, and scene switching time ≤200 milliseconds, making it suitable for multi-tasking operating systems such as HarmonyOS NEXT.

Description

Technical Field

[0001] This invention belongs to the field of mobile terminal operating system technology, specifically relating to a background process management system and method based on user-defined priority sorting, applicable to mobile operating systems with multi-tasking capabilities such as HarmonyOS NEXT. Background Technology

[0002] Current mobile terminals generally suffer from the following technical defects:

[0003] 1. Lack of background process management mechanism: The application runs in the background without an explicit way to close it, resulting in continuous consumption of memory and computing resources.

[0004] 2. Insufficient user strategy customization capabilities: The existing system only provides fixed priority templates and cannot customize resource allocation rules according to usage scenarios.

[0005] 3. Significant risk of task interruption: When an application is switched to the background, the system directly terminates any unfinished task processes.

[0006] 4. Inefficient implementation of management solutions: Existing backend optimization solutions require setting whitelists one by one after application startup, which is a complex process and cannot achieve global configuration. Summary of the Invention

[0007] This invention constructs a system-level background process management architecture, achieving the following technical breakthroughs through a user-preset five-level priority classification mechanism and scenario-based configuration strategies:

[0008] 1. Complete global priority configuration before the application starts.

[0009] 2. Establish a resource scheduling strategy that is deeply coupled with the operating system kernel.

[0010] 3. Ensure the integrity of critical mission execution.

[0011] 4. Enable dynamic switching of resource allocation strategies across multiple scenarios.

[0012] Technical solution

[0013] This system consists of a priority classification module, a resource scheduler, a status monitoring engine, and a scenario configuration module. Specific implementation features include:

[0014] I. Five-level priority classification system

[0015] 1. Real-time Assurance Level: Core communication service applications employ hard real-time scheduling strategies to ensure 100% resource availability.

[0016] 2. Task Continuation Level: For tasks that need to be executed continuously, full resource supply is maintained in the background until the task is completed.

[0017] • Allows overlaying with resource retention or smart release categories.

[0018] • When stacking resources, the allocation right is determined in the following order: Real-time protection level > Task continuation level > Resource retention level > Smart release level.

[0019] 3. Resource Retention Level: Inactive applications maintain process survival when system resources are sufficient.

[0020] • When resources reach a threshold, they will be downgraded according to the user-preset order.

[0021] 4. Intelligent Release Stage: Gradually reclaims resources based on a dynamic load assessment model.

[0022] 5. Forced Reclamation Level: Immediately releases all resources upon exiting the foreground.

[0023] Each priority category has its own internal sorting rules.

[0024] • All four categories—Real-time Protection, Task Continuation, Resource Retention, and Smart Release—support user-defined application sorting.

[0025] The sorting rules are stored using a linear linked list structure, ensuring that the time complexity of the operation during order adjustment is O(1).

[0026] • When allocating resources, priority should be given to meeting the needs of applications that are ranked higher.

[0027] II. Kernel-level scheduling mechanism

[0028] 1. Embed a priority policy execution unit in the operating system's process management module.

[0029] 2. Construct a memory allocation strategy matrix, mapping the five priority levels to physical memory page management algorithms.

[0030] 3. CPU time slice allocation adopts a preemptive hierarchical scheduling model.

[0031] III. State Persistence Guarantee

[0032] 1. Perform a full state snapshot for task-continuation level applications.

[0033] 2. Employ copy-on-write technology to reduce the performance overhead of state preservation.

[0034] IV. Scenario-based configuration strategy

[0035] 1. Supports setting scene strategies based on two dimensions: time period and geographical location.

[0036] • Time period dimension: differentiates between weekdays / rest days and time intervals (accurate to 30-minute granularity)

[0037] • Geographic location dimension: Identifies pre-defined areas such as companies and homes through base station positioning.

[0038] 2. Automatically switch to the corresponding priority configuration scheme when the scenario policy takes effect.

[0039] 3. Scene switching employs a double-buffering mechanism to ensure that ongoing tasks are not interrupted.

[0040] Beneficial effects

[0041] 1. Increased certainty in resource allocation

[0042] Real-time assurance-grade application response latency ≤1ms

[0043] • Background process memory usage fluctuations decreased by 80%

[0044] 2. Task integrity assurance

[0045] • Zero task progress loss during process resumption

[0046] • I / O channel reconstruction time ≤ 5ms

[0047] 3. System energy efficiency optimization

[0048] • Memory fragmentation rate drops below 5%

[0049] • CPU idle time percentage increased to over 30%

[0050] 4. Scene adaptation optimization

[0051] • Resource configuration strategy switching time ≤ 200ms

[0052] Scene recognition accuracy ≥ 95% Detailed Implementation

[0053] I. System Architecture Implementation

[0054] 1. Priority Configuration Module

[0055] • Add an "Application Priority Management" function interface to the system settings.

[0056] • Displays the list of installed applications using a tree-like topology.

[0057] • Configure five-level categories and internal sorting through drag-and-drop operations.

[0058] 2. Resource Scheduler

[0059] • Insert scheduling policy hook functions into the kernel process management subsystem

[0060] • Real-time protection-level processes are bound to independent CPU cores

[0061] • Enable memory locking protection for task continuation processes

[0062] 3. Scene Configuration Module

[0063] • Six preset basic scene templates (weekday office, home entertainment, etc.)

[0064] • Supports setting time-period coverage rules for scenario-based strategies

[0065] • Utilizing base station fingerprint recognition technology to achieve location determination with an accuracy of 500 meters.

[0066] II. Key Technology Implementation Details

[0067] 1. Dynamic priority stacking mechanism

[0068] • When an application is assigned to multiple priority categories simultaneously

[0069] • Calculate the final priority according to the preset weight formula:

[0070] Final priority = max(base category weight, superimposed category weight)

[0071] The weight values ​​are set as follows: Real-time protection level (100), Task continuation level (80), Resource retention level (60), and Smart release level (40).

[0072] 2. Process state snapshot technology

[0073] • Employ differential incremental saving mechanism to reduce storage overhead

[0074] • Configure a dedicated memory area to store process context information

[0075] 3. Scene strategy switching algorithm

[0076] • Establish scene feature vectors (time, location, network type)

[0077] • A sliding window algorithm is used to achieve smooth transitions between scene states.

[0078] Technical effect verification

[0079] The following proofs have been completed using formal verification tools:

[0080] 1. The scheduling strategy has no deadlock risk (verified by Petri net modeling).

[0081] 2. Resource allocation meets the monotonicity requirement (verified by Hall logic).

[0082] 3. Proof of process termination completeness (π calculus modeling analysis).

Claims

1. A background process management system based on user-defined prioritization, comprising: It includes a priority classification module, a resource scheduler, a status monitoring engine, and a scene configuration module, among which: The priority classification module is used to establish a five-level priority classification system, including real-time guarantee level, task continuation level, resource retention level, intelligent release level, and forced reclamation level. Among them, the real-time guarantee level application is bound to an independent CPU core and adopts a hard real-time scheduling strategy; the task continuation level application maintains a complete resource supply in the background until the task is completed; the resource retention level application keeps the process alive when system resources are sufficient; the intelligent release level application gradually reclaims resources based on a dynamic load assessment model; and the forced reclamation level application releases all resources immediately after exiting the foreground. The resource scheduler is built into the operating system kernel and includes a priority policy execution unit, a memory allocation policy matrix, and a preemptive hierarchical scheduling model, which is used to map five priority levels to physical memory page management algorithms and allocate CPU time slices. The state monitoring engine uses copy-on-write technology to perform a full state snapshot of the task-continuation level application and stores process context information in a dedicated memory area. The scene configuration module supports setting scene strategies by time period and geographical location, and adopts a double buffering mechanism to achieve uninterrupted switching of configuration schemes.

2. The system according to claim 1, characterized in that, The five-level priority classification system satisfies the following rules: (1) Real-time protection level, task continuation level, resource retention level and intelligent release level support user-defined application sorting, and the sorting rules are stored in a linear linked list structure. (2) The task continuation level can be superimposed with the resource retention level or the smart release level. When superimposed, the resource allocation right is determined in the order of real-time guarantee level > task continuation level > resource retention level > smart release level. (3) When the system resources reach the threshold, the resource-reserved application will be downgraded according to the user-preset order.

3. The system according to claim 1, characterized in that, The operation of the scene configuration module includes: (1) The time cycle dimension distinguishes between weekdays and rest days, and the time interval is set in 30-minute increments; (2) Geographic location dimension achieves area determination with an accuracy of 500 meters through base station fingerprint recognition technology; (3) When the scenario strategy takes effect, the corresponding priority configuration scheme will be automatically switched.

4. The system according to claim 1, characterized in that, The resource scheduler also includes a dynamic priority stacking mechanism, which calculates the final priority according to the following formula when an application is assigned to multiple priority categories: Final priority = MAX(base category weight, superimposed category weight), where the weight of real-time guarantee level is 100, task continuation level is 80, resource retention level is 60, and intelligent release level is 40.

5. The system according to claim 1, characterized in that, The state monitoring engine uses a differential incremental saving mechanism to reduce the storage overhead of process state snapshots and achieves input / output channel reconstruction time ≤ 5 milliseconds when the process is restored.

6. A background process management method based on user-defined priority sorting, characterized in that, Includes the following steps: Step 1: Before the application starts, complete the five-level priority global configuration through the tree topology interface, including category division and internal sorting; Step 2: Dynamically switch the corresponding priority configuration scheme according to the user's preset time period or geographical location scenario strategy; Step 3: Allocate independent CPU cores to real-time guarantee-level processes through kernel-level scheduling mechanisms, and enable memory locking protection for task continuity-level processes; Step 4: When an application switches to the background, resource allocation strategies are executed according to priority. For applications in the task continuation category, a full state snapshot is triggered, while for applications in the smart release category, resources are released gradually according to the dynamic load assessment model. Step 5: When system resources reach the threshold, downgrade applications that are reserved for resources according to the user's preset order.

7. The method according to claim 6, characterized in that, In step 2, the scene strategy switching adopts a sliding window algorithm to achieve a smooth transition, with a switching time of ≤200 milliseconds and a scene recognition accuracy of ≥95%.

8. The method according to claim 6, characterized in that, Step 4 further includes: immediately releasing all resources and terminating the relevant processes of the forced recycling application after it exits the foreground.