Capability abstraction method of industrial Internet of Things ubiquitous operating system

Abstract

The invention provides a capability abstraction method for an industrial Internet of Things ubiquitous operating system, and belongs to the field of operating systems. According to the method, industrial Internet of Things resources are abstracted into capabilities, abstraction of resource access and authority control of the industrial Internet of Things resources is provided for applications, ubiquitous resources are abstracted through the most basic resource virtualization technology, and core functions which must be supported by different industrial Internet of Things ubiquitous operating systems are constructed. The method can solve the management problem of massive, heterogeneous, heterogeneous, autonomous and other ubiquitous resources of the industrial internet of things, is adaptive to dynamic and changeable upper-layer application scenes, meets the frequently changing application requirements, can support a flexible and diversified resource virtualization mechanism and a heterogeneous resource management function, focuses on application development and management support in a novel application mode, and has a wide application prospect. The method has a wide application prospect.

Description

Technical Field

[0001] The present invention provides a capability abstraction method for an operating system, and specifically relates to a capability abstraction method for an industrial Internet of Things ubiquitous operating system, and belongs to the field of operating systems. Background Art

[0002] Industrial Internet of Things is the connection of sensors and instruments to industrial applications through the Internet. The network uses IoT sensing and communication technologies to collect industrial application data and analyze the data. The purpose is to optimize the production process, improve production efficiency, reduce manufacturing costs, improve product quality, and ultimately achieve the upgrading of traditional industries to a new stage of intelligence.

[0003] Ubiquitous computing believes that computing is widely present. In the future, computers can be embedded in various objects in human life, production and social activities. Through data collection and processing, information integration and analysis, and knowledge generation and representation, an "intelligent environment" can be generated in real time, expanding the perception, cognitive understanding, analytical decision-making capabilities in all aspects of human life.

[0004] The Industrial Internet of Things Ubiquitous Operating System is a system software platform that serves the development, operation and adaptive evolution of diverse ubiquitous computing applications in the Industrial Internet of Things environment. As an operating system, it is essentially still a software layer that provides resource virtualization and application programming interfaces. The core is how to provide appropriate abstractions. Therefore, to build an Industrial Internet of Things Ubiquitous Operating System, we need to study how to understand the computing environment of the Industrial Internet of Things and the various heterogeneous resources in the environment that support the realization of perception, computing, communication, execution, service and other capabilities, as well as the essential characteristics of ubiquitous applications in the Industrial Internet of Things scenario, and provide an effective abstract model for it to better condense the commonality of applications and manage resources more effectively. Summary of the invention

[0005] The purpose of the present invention is to provide a capability abstraction method for an Industrial Internet of Things ubiquitous operating system, which abstracts ubiquitous resources through the most basic resource virtualization technology, constructs core functions that must be supported by different Industrial Internet of Things ubiquitous operating systems, and is used to solve the management problems of massive, heterogeneous, heterogeneous, and autonomous ubiquitous resources in Industrial Internet of Things scenarios, while adapting to dynamic and changeable upper-level application scenarios to meet frequently changing application requirements.

[0006] The technical solution of the present invention is as follows:

[0007] A capability abstraction method for an industrial IoT ubiquitous operating system abstracts industrial IoT resources into capabilities and provides abstraction of resource access and permission control of ubiquitous resources for applications. The method comprises the following steps:

[0008] A. Processing of ubiquitous resources through the resource abstraction layer, including:

[0009] A1. Encapsulate all ubiquitous resources provided to upper-layer applications in a specific industrial IoT scenario. The ubiquitous resources include traditional computer hardware resources, physical hardware resources, and abstract resources. The abstract resources include databases in IoT systems, AI inference frameworks in industrial IoT operating systems, etc.

[0010] A2. Use drivers and program runtime libraries to transform different types of ubiquitous resources into a unified programming interface, so that the system does not need to perceive the specific usage of ubiquitous resources, the specific implementation of drivers, and other details involving the uniqueness of heterogeneous resources, thereby achieving unified management of various complex resources in industrial IoT scenarios and ensuring the ubiquity of the system;

[0011] A3. In response to changes in resource types and quantities under the Industrial Internet of Things, the resource management set is added and deleted through the new driver plug-in and unload method, and new resources are included in the scope of system management resources to ensure the flexibility of the ubiquitous operating system, reduce the cost of system regeneration introduced by resource changes, and manage resources through a unified programming interface;

[0012] B. Establish the capability abstraction of industrial IoT based on the ubiquitous resource abstraction, including:

[0013] B1. Based on the programming interface provided by the ubiquitous resource abstraction layer, the corresponding capability module will be generated based on the scope of resources managed, each resource will be registered, and the upper-layer application will use the resource through the capability;

[0014] B2. Provide complete management for all registered resources through the capability definition layer, including startup, shutdown, revocation and allocation capabilities based on permission control;

[0015] B3. When new ubiquitous resources are added or resources are uninstalled, the resource management list can be modified according to the programming interface provided by the resource abstraction layer to achieve flexible adjustment of capabilities;

[0016] C. Submit the capabilities to the capability management layer of the ubiquitous operating system for management, including:

[0017] C1. After the resource is started, the capability provided by the resource can be found in the capability list, and the capability can be invoked from the upper layer. At the same time, the capability model also embeds the capability transfer (delegate) mechanism and capability recovery (revoke) function, so that the capability can be effectively managed;

[0018] C2. Applications can obtain capabilities through capability transfer, use capabilities, and manage underlying resources through capabilities. Different applications obtain the right to use capabilities through capability transfer. When capabilities need to be reclaimed, the capability management layer controls the recovery of capabilities allocated to applications;

[0019] C3. Capability management uses communication restrictions to limit capability transfer. Only trusted objects can communicate directly with the capability definition layer to exchange capabilities, thereby ensuring the security of capability transfer. The capability management layer uses the capability transfer method of generating intermediate capability nodes to simplify capability recovery;

[0020] D. Monitor the usage and operation status of the capabilities in real time, and schedule the capabilities used by each task in the Industrial Internet of Things ubiquitous operating system, including:

[0021] D1. Capability scheduling is performed by the capability scheduling layer, which maps resources (R) to functions f:R->AM on automata (AM). Capability-based task management includes two planes: resource definition plane and control plane. The resource definition plane describes all resources contained in the system and the connection relationships between them. In addition to the automata containing capability abstractions, it also includes a set of connection relationships between different resources.

[0022] D2. The control plane is responsible for scheduling and managing computing tasks. By separating the resource definition plane from the control plane, the upper-layer tasks do not need to care about the resource scheduling of the lower layer, and are only responsible for writing the logic of the tasks, thus simplifying the complexity of application development;

[0023] D3. The capability scheduling layer defines a task as a function t:D╳AM->D, which converts data d∈D into new data d'∈D through the processing of a certain automaton (the automaton comes from the abstraction of resources). At the same time, the cost function C:T->REAL of the task operation is defined, which maps the task to the generalized real number field REAL. The cost function C represents the comprehensive cost of task operation, which includes all factors related to the quality of task execution, such as time, power consumption, and user experience. In particular, the cost of task execution failure is defined as +∞. Based on the above definition, the task scheduling that the control plane is responsible for can be expressed as an optimization problem, that is, decomposing a task t into several subtask sequences t1⊙t2⊙t3…⊙tn, so that ∑C(ti) is minimized.

[0024] Furthermore, the step B2 is specifically as follows: when registering a resource, the operation corresponding to the resource is passed through parameters, and after the resource is registered and started, the operation function corresponding to the resource is queried and called from the capability abstraction; heterogeneous resources include various operations and provide an extensible resource programming interface.

[0025] Beneficial effects of the present invention:

[0026] The present invention proposes a capability abstraction method for an Industrial Internet of Things ubiquitous operating system, which abstracts Industrial Internet of Things resources into capabilities and provides applications with abstraction of resource access and permission control of Industrial Internet of Things resources. It can solve the management problems of massive, heterogeneous, isomorphic, and autonomous ubiquitous resources of the Industrial Internet of Things, and at the same time adapt to dynamic and changeable upper-level application scenarios to meet frequently changing application requirements. It can support flexible and diverse resource virtualization mechanisms and heterogeneous resource management functions, and focus on application development and management support under new application models. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is the overall architecture diagram of the present invention;

[0028] Figure 2 It is a schematic diagram of the capability management of the Industrial Internet of Things ubiquitous operating system. DETAILED DESCRIPTION

[0029] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0030] The overall structure of the present invention is shown in FIG. Figure 1 As shown, it includes ubiquitous resources, resource abstraction layer, capability definition layer, capability management layer and capability scheduling layer. The capability abstraction method of the industrial Internet of Things ubiquitous operating system of the present invention is implemented on this architecture, and the industrial Internet of Things resources are abstracted into capabilities to provide applications with abstraction of resource access and permission control of ubiquitous resources. The method includes the following steps:

[0031] A. Processing of ubiquitous resources through the resource abstraction layer, including:

[0032] A1. Encapsulate all resources provided to upper-layer applications in specific industrial IoT scenarios, including but not limited to traditional computer hardware resources, physical hardware resources, and other abstract resources;

[0033] The resources are as follows Figure 1 As shown in the ubiquitous resources of Figure 1 As shown in the resource abstraction layer of the Industrial Internet of Things, the ubiquitous operating system abstracts and manages ubiquitous resources through the resource abstraction layer, and ubiquitous resources are abstracted into virtual resources;

[0034] A2. Use drivers and program runtime libraries to transform different types of ubiquitous resources into a unified programming interface, so that the system does not need to perceive the specific usage of ubiquitous resources, the specific implementation of drivers, and other details involving the uniqueness of heterogeneous resources, thereby achieving unified management of various complex resources in industrial IoT scenarios and ensuring the ubiquity of the system;

[0035] A3. In response to changes in resource types and quantities under the Industrial Internet of Things, the resource management set is added and deleted through the new driver plug-in and unload method, and new resources are included in the scope of system management resources to ensure the flexibility of the ubiquitous operating system, reduce the cost of system regeneration introduced by resource changes, and manage resources through a unified programming interface;

[0036] B. Establish the capability abstraction of industrial IoT based on the ubiquitous resource abstraction, including:

[0037] B1. Based on the programming interface provided by the ubiquitous resource abstraction layer, the corresponding capability module will be generated based on the scope of resources managed, and each resource will be registered. The upper-layer application uses the resource through the capability;

[0038] The generated capability modules are as follows: Figure 1 As shown in the capability definition layer in , the capability definition registers virtual resources as capabilities, and upper-layer applications use resources through capabilities;

[0039] B2. For all registered resources, the capability definition layer provides complete management, including startup, shutdown, revocation, and allocation capabilities based on permission control. Specifically, when registering a resource, the corresponding operation of the resource (such as the core functions of the file system such as open, read, write, and update) will be passed through parameters. After the resource is registered and started, the corresponding operation function of the resource can be queried and called from the capability abstraction. Heterogeneous resources include various operations and provide an extensible resource programming interface;

[0040] B3. When new ubiquitous resources are added or resources are uninstalled, the capability definition layer can also modify the resource management list according to the programming interface provided by the resource abstraction layer to achieve flexible adjustment of capabilities;

[0041] C. Submit the capabilities to the capability management layer of the ubiquitous operating system for management, including:

[0042] After the resource is started, C1 can find the capability provided by the resource in the capability list and call (Invoke) the capability from the upper layer. At the same time, the capability model also embeds the capability transfer (delegate) mechanism and capability recovery (revoke) function, so that the capability can be effectively managed;

[0043] The functions embedded in the capability model are Figure 1 The capability management layer in the system provides that virtual resources are abstracted into capabilities and then the capabilities are managed by the capability management layer.

[0044] C2. Applications can obtain capabilities through capability transfer, use capabilities, and manage underlying resources through capabilities. Different applications obtain the right to use capabilities through capability transfer. When capabilities need to be reclaimed, the capability management layer controls the recovery of capabilities allocated to applications;

[0045] C3. The capability management layer uses communication restrictions to limit the transfer of capabilities. Only trusted objects can communicate directly with the capability definition layer to exchange capabilities, thereby ensuring the security of capability transfer. Figure 2 As shown in the left picture;

[0046] If an object is not trusted, its access permissions are restricted and capabilities are denied to the object. Figure 2 In the left figure, Task 1 has the ability to access resources and passes the ability to Task 2, so Task 2 also has the ability to access. If the system decides not to trust Task 2, it will restrict Task 1's access to Task 2, that is, Task 1 cannot access Task 2 and cannot pass the ability to Task 2;

[0047] The capability definition layer uses the capability transfer method of generating intermediate capability nodes to simplify capability recovery, such as Figure 2 As shown in the right figure; in the figure, task 1 has access capability. If task 1 wants to pass the access capability to task 2, task 1 first generates a transfer station and a port. The capability can access the transfer station, and the transfer station can access resources through the port. Task 1 passes the capability to task 2, and task 2 has the capability to access the transfer station and can access resources through the transfer station. If task 1 wants to revoke task 2's access capability, it can just revoke the port. Task 2 can still access the transfer station, but cannot access resources.

[0048] D. Monitor the usage and operation status of the capabilities in real time, and schedule the capabilities used by each task in the Industrial Internet of Things ubiquitous operating system, including:

[0049] D1. Map the resource (R) to the function f:R->AM on the automaton (AM) and schedule the capacity therein, where the capacity scheduling is determined by Figure 1 The capability scheduling layer in the system schedules capabilities and effectively uses ubiquitous operating system resources, including:

[0050] Task management and scheduling are performed through the capability scheduling layer. Capability-based task management includes two planes: resource definition plane and control plane. The resource definition plane describes all resources contained in the system and the connection relationships between them. In addition to the automaton containing capability abstraction, it also includes the connection relationship set between different resources;

[0051] D2. The control plane is responsible for scheduling and managing computing tasks. By separating the resource definition plane from the control plane, the upper-layer tasks do not need to care about the resource scheduling of the lower layer, and are only responsible for writing the logic of the tasks, thus simplifying the complexity of application development;

[0052] D3. The capability scheduling layer defines a task as a function t:D╳AM->D, which converts the data d∈D into new data d'∈D through the processing of a certain automaton (the automaton comes from the abstraction of resources). At the same time, the cost function C:T->REAL of the task operation is defined, which maps the task to the generalized real number field REAL. The cost function C represents the comprehensive cost of task operation, which includes all factors related to the quality of task execution, such as time, power consumption, and user experience. In particular, the cost of task execution failure is defined as +∞. Based on the above definition, the task scheduling that the control plane is responsible for can be expressed as an optimization problem, that is, decomposing a task t into several subtask sequences t1⊙t2⊙t3…⊙tn, so that ∑C(ti) is minimized;

[0053] Based on capability abstraction, users can write applications to access the services provided by resources, thereby changing the behavior of resources to meet the diverse needs of applications for resources.

[0054] Finally, it should be noted that the purpose of publishing the embodiments is to help further understand the present invention, but those skilled in the art can understand that various substitutions and modifications are possible without departing from the spirit and scope of the present invention and the appended claims. Therefore, the present invention should not be limited to the contents disclosed in the embodiments, and the scope of protection claimed by the present invention shall be subject to the scope defined in the claims.

Claims

1. A capability abstraction method for industrial IoT ubiquitous operating system, It is characterized in that Industrial IoT resources are abstracted into capabilities to provide applications with abstraction of resource access and permission control of ubiquitous resources. The steps of this method include: A. Processing of ubiquitous resources through the resource abstraction layer, including: A1. Encapsulate all ubiquitous resources provided to upper-layer applications in a specific industrial IoT scenario. The ubiquitous resources include traditional computer hardware resources, physical hardware resources, and abstract resources. The abstract resources include databases in IoT systems and AI inference frameworks in industrial IoT operating systems. A2. Use drivers and runtime libraries to transform different types of ubiquitous resources into a unified programming interface, and achieve unified management of various complex resources in industrial IoT scenarios; A3. In response to changes in the types and quantities of resources under the Industrial Internet of Things, the resource management collection is added and deleted through the new driver plug-in and unload method, the new resources are included in the scope of system management resources, and the resources are managed through a unified programming interface; B. Establish the capability abstraction of industrial IoT based on the ubiquitous resource abstraction, including: B1. Based on the programming interface provided by the ubiquitous resource abstraction layer, the corresponding capability module will be generated based on the scope of resources managed, each resource will be registered, and the upper-layer application will use the resource through the capability; B2. Provide complete management for all registered resources through the capability definition layer, including startup, shutdown, revocation and allocation capabilities based on permission control; B3. When new ubiquitous resources are added or resources are uninstalled, the capability definition layer modifies the resource management list according to the programming interface provided by the resource abstraction layer to achieve flexible adjustment of capabilities; C. Submit the capabilities to the capability management layer of the ubiquitous operating system for management, including: C1. After the resource is started, the capability provided by the resource is found in the capability list and called from the upper layer. At the same time, the capability model also embeds the capability transfer mechanism and capability recovery functions, so as to effectively manage the capability; C2. Applications obtain capabilities through capability transfer, use capabilities, and manage underlying resources through capabilities. Different applications obtain the right to use capabilities through capability transfer. When capabilities need to be reclaimed, the capability management layer controls the reclaiming of capabilities allocated to applications. C3. Capability management uses communication restriction to limit capability transfer. Only trusted objects can communicate directly with the capability definition layer to exchange capabilities, thereby ensuring the security of capability transfer. The capability management layer uses the capability transfer method of generating intermediate capability nodes to simplify capability recovery. D. Monitor the usage and operation status of the capabilities in real time, and schedule the capabilities used by each task in the Industrial Internet of Things ubiquitous operating system, including: D1. Capability scheduling is performed by the capability scheduling layer, which maps resources (R) to functions f:R->AM on automata (AM); Capability-based task management includes two planes: resource definition plane and control plane. The resource definition plane describes all resources contained in the system and the connection relationships between them. In addition to the automata containing capability abstractions, it also includes the connection relationship set between different resources; D2. The control plane is responsible for scheduling and managing computing tasks, separating the resource definition plane from the control plane. The upper-layer tasks do not need to care about the resource scheduling of the lower layer, but are only responsible for writing the logic of the tasks, thus simplifying the complexity of application development. D3. The capability scheduling layer defines a task as a function t:D╳AM->D, which converts data d∈D into new data d'∈D through the processing of a certain automaton (the automaton comes from the abstraction of resources). At the same time, it defines the cost function C:T->REAL of the task operation, which maps the task to the generalized real number field REAL. The cost function C represents the comprehensive cost of task operation, which includes factors related to the quality of task execution: time, power consumption, and user experience. The cost of task execution failure is specifically defined as +∞. Based on the above definition, the task scheduling that the control plane is responsible for is expressed as an optimization problem, that is, decomposing a task t into several subtask sequences t1⊙t2⊙t3…⊙tn, so that ∑C(ti) is minimized.

2. The method according to claim 1, It is characterized in that The step B2 is specifically as follows: when registering a resource, the operation corresponding to the resource is passed through parameters, and after the resource is registered and started, the operation function corresponding to the resource is queried and called from the capability abstraction.

Images