Method and system for calling computing power and network capability for cloud VR service

Abstract

This disclosure relates to a method and system for invoking computing power and network capabilities for cloud VR services, belonging to the field of distributed computing technology. The method includes: an optimization processing module in the control layer, based on the user-focused area of ​​a cloud-based graphics application, dividing the screen area of ​​the application in the application layer into blocks to obtain a screen area block result, and generating a block rendering request based on the screen area block result; a distributed computing network resource scheduling module in the control layer responding to the block rendering request, parsing the block rendering request, obtaining the screen area block result and the corresponding graphics rendering computing power and network transmission capability requirements; and the distributed computing network resource scheduling module matching target computing power nodes in the resource layer for the block rendering task corresponding to the screen area block result based on the graphics rendering computing power and network transmission capability requirements. This disclosure improves the flexibility of rendering task implementation.

Description

Technical Field

[0001] This disclosure relates to the fields of network technology and security technology, and more specifically, to a method and system for invoking computing power and network capabilities for cloud VR services. Background Technology

[0002] In the relevant technical solutions, for graphics cloud applications, due to the variability of user scale, the uniqueness of the operating environment, and the extreme requirements of network transmission, if a general computing power network resource calling method is used for resource calling, the following defects will occur: on the one hand, it will increase the complexity of resource calling, thus making the resource calling efficiency lower; on the other hand, it will reduce the screen rendering efficiency of graphics cloud applications.

[0003] It should be noted that the information in the background section above is only used to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0004] The purpose of this disclosure is to provide a method and system for invoking computing and network capabilities for cloud VR services, thereby overcoming, to some extent, the problems of low resource mobilization efficiency and low screen rendering efficiency caused by the limitations and defects of related technologies.

[0005] According to one aspect of this disclosure, a method for invoking computing power and network capabilities for cloud VR services is provided, comprising:

[0006] The optimization processing module of the control layer divides the screen area of ​​the graphics cloud application in the application layer into blocks according to the user attention area of ​​the graphics cloud application running on the cloud, obtains the screen area block result, and obtains the block rendering request based on the screen area block result.

[0007] The distributed computing network resource scheduling module in the control layer responds to the segmented rendering request, parses the segmented rendering request, and obtains the screen area segmentation result and the graphics rendering computing power requirement and network transmission capability requirement corresponding to the screen area segmentation result.

[0008] The distributed computing network resource scheduling module matches the target computing power node in the resource layer with the block rendering task corresponding to the block rendering result of the screen area according to the graphics rendering computing power requirement and network transmission capability requirement.

[0009] In one exemplary embodiment of this disclosure, the screen region of the graphics cloud application in the application layer is segmented according to the user attention area of ​​the graphics cloud application running in the cloud, to obtain the screen region segmentation result, including:

[0010] The user attention area of ​​the graphical cloud application is determined, and based on the user attention area, the attention surrounding area, the peripheral area, and the temporarily unperceived area adjacent to the peripheral area are determined.

[0011] Determine the required regional rendering precision and regional refresh rate for the graphics cloud application in the user's attention area, the surrounding area, the outer area, and the area that cannot be perceived temporarily.

[0012] The overall screen size of the graphics cloud application's screen area is determined, and the screen area of ​​the graphics cloud application is divided into blocks according to the overall screen size and the area rendering precision and area refresh frequency required in different display areas, to obtain the screen area block result.

[0013] In one exemplary embodiment of this disclosure, determining the user-focused area of ​​the graphics cloud application includes:

[0014] Determine the field of view and / or the human eye gaze position of the cloud-based graphics application, and determine the user attention area based on the field of view and / or the human eye gaze position.

[0015] In one exemplary embodiment of this disclosure, obtaining a block rendering request based on the block segmentation result of the screen region includes:

[0016] Determine the graphics rendering computing power requirements required to execute the block rendering task corresponding to the block rendering result of the screen region, and the network transmission capability requirements required to transmit the block rendering result corresponding to the block rendering result of the screen region.

[0017] Based on the image region segmentation results, graphics rendering computing power requirements, and network transmission capability requirements, a segmented rendering request corresponding to the image region segmentation results is generated.

[0018] In one exemplary embodiment of this disclosure, matching a target computing power node in the resource layer for a segmented rendering task corresponding to the image region segmentation result, based on the graphics rendering computing power requirements and network transmission capability requirements, includes:

[0019] The computational and network requirements of the segmented rendering task are correlated with the results of the segmented image region, the graphics rendering computing power requirements required to execute the segmented rendering task corresponding to the segmented image region, and the network transmission capability requirements required to transmit the segmented rendering result corresponding to the segmented image region, to obtain the computational and network requirements fusion result of the segmented rendering task.

[0020] The alternative computing power capabilities and alternative network transmission capabilities of the alternative computing power nodes included in the resource layer are determined, and the alternative computing power nodes, alternative computing power capabilities, and alternative network transmission capabilities are fused to obtain the fusion result of the alternative computing power node capabilities.

[0021] Based on the fusion results of the computing network requirements and the fusion results of the candidate node capabilities, a target computing node is matched for the block rendering task from the candidate computing nodes included in the resource layer.

[0022] In one exemplary embodiment of this disclosure, the graphics rendering computing power requirements for executing the segmented rendering task corresponding to the segmented rendering result, and the network transmission capability requirements for transmitting the segmented rendering result corresponding to the segmented rendering result are correlated to obtain a computing and network requirement fusion result for the segmented rendering task, including:

[0023] Based on the position of the image region segmentation result in the image cloud application, the image region segmentation result is sorted to obtain the segmentation sorting result, and a segmentation identifier is configured for the image region segmentation result according to the segmentation sorting result.

[0024] The image region segmentation result, segmentation identifier, graphics rendering computing power requirements required to execute the segmented rendering task corresponding to the image region segmentation result, and network transmission capability requirements required to transmit the segmented rendering result corresponding to the image region segmentation result are correlated to obtain the computing and network requirements fusion result of the segmented rendering task.

[0025] In one exemplary embodiment of this disclosure, matching a target computing power node for the block rendering task from the candidate computing power nodes included in the resource layer, based on the computing network demand fusion result and the candidate node capability fusion result, includes:

[0026] Based on the network transmission capacity requirements in the network demand fusion result, the target node capability fusion result is matched with the candidate node capability fusion result;

[0027] Based on the graphics rendering computing power requirements in the network demand fusion result and the alternative computing power capabilities in the target node capability fusion result, a target computing power node is matched from the alternative computing power nodes for the block rendering task corresponding to the block division result of the screen area.

[0028] In one exemplary embodiment of this disclosure, determining the alternative computing power capabilities and alternative network transmission capabilities of the alternative computing power nodes included in the resource layer includes:

[0029] The system receives node attribute information and remaining node computing power of candidate computing power nodes reported by the resource layer, and determines the candidate computing power capability of the candidate computing power nodes based on the node attribute information and remaining node computing power; wherein, the node attribute information includes at least one of chip model, number of chips, memory size and computing power.

[0030] The system receives the current node IP address and original egress bandwidth resources of the candidate computing power nodes reported by the resource layer, and determines the candidate network transmission capabilities of the candidate computing power nodes based on the current node IP address and original egress bandwidth resources.

[0031] In one exemplary embodiment of this disclosure, determining the candidate network transmission capability of the candidate computing power node based on the current node IP address and the original egress bandwidth resources includes:

[0032] The computing power of the candidate computing power nodes is evaluated to obtain the computing power evaluation results, and it is determined whether the candidate computing power nodes can be used as rendering task execution nodes based on the computing power evaluation results.

[0033] If the candidate computing power node can serve as the rendering task execution node, then the current node IP address and original outbound bandwidth resources of the candidate computing power node are synchronized to the network layer, and the candidate network transmission capabilities corresponding to the candidate computing power node are received from the network layer.

[0034] In one exemplary embodiment of this disclosure, the alternative network transmission capability is obtained by the network layer in the following manner:

[0035] Based on the current node IP address and the original outbound bandwidth resources, the current network environment of the candidate computing power node is evaluated to obtain the current network environment evaluation result; wherein, the current network environment evaluation result includes at least one of the current link latency, the current available outbound bandwidth of the network, and the current network transmission stability;

[0036] Based on the current network environment assessment results, determine whether the current bandwidth resources and / or current transmission links of the candidate computing power nodes need to be enhanced;

[0037] If the current bandwidth resources and / or the current transmission links need to be enhanced, the current bandwidth resources are accelerated, and / or the current transmission links are optimized, and the alternative network transmission capabilities are determined based on the accelerated current bandwidth resources and / or the optimized current transmission links.

[0038] In one exemplary embodiment of this disclosure, the method for invoking computing power and network capabilities further includes:

[0039] The distributed computing network resource scheduling module sends the target node IP address of the target computing power node to the application layer;

[0040] The application layer performs node addressing on the target computing power node based on the target node's IP address, thereby establishing a first communication link between the application layer and the target computing power node.

[0041] In one exemplary embodiment of this disclosure, the method for invoking computing power and network capabilities further includes:

[0042] The distributed computing network resource scheduling module generates a segmented rendering task based on the screen region segmentation result, the segmentation identifier of the screen region segmentation result, the graphics rendering computing power requirement required to execute the segmented rendering task corresponding to the screen region segmentation result, the network transmission capability required to transmit the segmented rendering result corresponding to the screen region segmentation result, the target computing node, and the target computing power capability of the target computing node, and then distributes the segmented rendering task to the target computing node through the resource layer.

[0043] In one exemplary embodiment of this disclosure, the distributed computing network resource scheduling module distributes the block rendering task to the target computing power node via the resource layer, including:

[0044] The distributed computing network resource scheduling module synchronizes the target network transmission capability of the target computing power node to the network layer;

[0045] The network layer evaluates the target network environment of the target computing node based on the target network transmission capability and the network transmission capability requirements of the chunked rendering task, obtains the target network environment evaluation result, and determines whether the target network transmission capability needs to be enhanced based on the target network environment evaluation result.

[0046] If so, the network layer enhances the target network transmission capability and notifies the distributed computing network resource scheduling module to issue the segmented rendering task; if not, it directly notifies the distributed computing network resource scheduling module to issue the segmented rendering task.

[0047] Based on the received notification, the distributed computing network resource scheduling module distributes the block rendering task to the target computing power node via the resource layer.

[0048] In one exemplary embodiment of this disclosure, the method for invoking computing power and network capabilities further includes:

[0049] The target computing node executes the segmented rendering task, obtains the task execution result, and encapsulates the segment identifier and the task execution result in the segmented rendering task to obtain the segmented rendering result.

[0050] The target computing node feeds back the block rendering result to the application layer based on the first communication link;

[0051] The application layer sequentially splices the segmented rendering results according to the segmentation identifiers in the segmented rendering results to obtain the overall screen rendering result of the cloud-based graphics application, and then feeds the overall screen rendering result back to the terminal layer.

[0052] According to one aspect of this disclosure, a system for invoking computing power and network capabilities for cloud VR services is provided, comprising a control layer, an application layer, and a resource layer. The control layer includes an optimization processing module and a distributed computing network resource scheduling module; wherein:

[0053] The optimization processing module is used to divide the screen area of ​​the graphics cloud application in the application layer into blocks according to the user attention area of ​​the graphics cloud application running in the cloud, to obtain the screen area block result, and to obtain the block rendering request based on the screen area block result.

[0054] The distributed computing network resource scheduling module is used to respond to the segmented rendering request, parse the segmented rendering request, obtain the screen region segmentation result, and the corresponding graphics rendering computing power requirement and network transmission capacity requirement; and

[0055] Based on the graphics rendering computing power requirements and network transmission capability requirements, target computing power nodes in the resource layer are matched for the block rendering tasks corresponding to the block rendering results of the screen area.

[0056] In one exemplary embodiment of this disclosure, the system for invoking computing power and network capabilities further includes:

[0057] The network layer, which is communicatively connected to the control layer, is used to evaluate the target network environment of the target computing power node and determine whether the target network transmission capability of the target computing power node needs to be enhanced based on the evaluation results.

[0058] In one exemplary embodiment of this disclosure, the resource layer includes multiple computing power nodes, which are used to execute the block rendering task issued by the distributed computing network resource scheduling module, obtain the block rendering result, and feed the block rendering result back to the application layer.

[0059] In one exemplary embodiment of this disclosure, the system for invoking computing power and network capabilities further includes:

[0060] The terminal layer, communicating with the application layer, is used to request the display screen of a cloud-based graphical application from the application layer; and

[0061] Receive the overall screen display result corresponding to the display screen from the application layer, and display the overall screen display result.

[0062] According to one aspect of this disclosure, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the method for invoking computing power and network capabilities for cloud VR services as described in any of the preceding claims.

[0063] According to one aspect of this disclosure, an electronic device is provided, comprising:

[0064] Processor; and

[0065] Memory for storing the executable instructions of the processor;

[0066] The processor is configured to execute the above-described method for invoking computing and network capabilities for cloud VR services by executing the executable instructions.

[0067] This disclosure provides a method for invoking computing power and network capabilities for cloud VR services. On one hand, the optimization processing module in the control layer divides the screen area of ​​the graphics cloud application in the application layer into blocks based on the user's attention area, obtaining a screen area block result, and generating a block rendering request based on the screen area block result. Then, the distributed computing network resource scheduling module in the control layer responds to the block rendering request and parses it to obtain the screen area block result and the corresponding graphics rendering computing power and network transmission capability requirements. Finally, the distributed computing network resource scheduling module matches the target computing power node in the resource layer with the block rendering task corresponding to the screen area block result based on the graphics rendering computing power and network transmission capability requirements. This reduces the complexity of calling the target computing node, solves the problem of low resource calling efficiency caused by complex resource calling in related technical solutions, and improves the calling efficiency of the target computing node. On the other hand, it improves the screen rendering efficiency of the screen area block result.

[0068] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0069] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure. It is obvious that the drawings described below are merely some embodiments of this disclosure, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort.

[0070] Figure 1 The flowchart illustrates a method for invoking computing and network capabilities for cloud VR services according to an example embodiment of this disclosure.

[0071] Figure 2 This diagram schematically illustrates a hierarchical example of a system for invoking computing and network capabilities for cloud VR services according to an example embodiment of this disclosure.

[0072] Figure 3 The diagram illustrates an application scenario example of a method for invoking computing power and network capabilities for cloud VR services according to an exemplary embodiment of this disclosure.

[0073] Figure 4 This diagram schematically illustrates a scene example of a screen region segmentation result obtained according to an exemplary embodiment of the present disclosure.

[0074] Figure 5 The illustration shows a scene example diagram with a field of view according to an exemplary embodiment of the present disclosure.

[0075] Figure 6 The diagram illustrates an example of a process for determining a target computing node according to an exemplary embodiment of the present disclosure.

[0076] Figure 7 The illustration shows a scenario example of segmenting and marking the results of segmenting a screen region according to an example embodiment of the present disclosure.

[0077] Figure 8 The illustration shows an example scenario of allocating chunked rendering tasks to target computing nodes according to an example embodiment of the present disclosure.

[0078] Figure 9 The diagram schematically illustrates a block diagram of an apparatus for invoking computing and network capabilities for cloud VR services according to an exemplary embodiment of the present disclosure.

[0079] Figure 10 An electronic device is illustrated in accordance with an example embodiment of this disclosure for implementing a method of invoking computing and network capabilities for cloud VR services. Detailed Implementation

[0080] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided to make this disclosure more comprehensive and complete, and to fully convey the concept of the example embodiments to those skilled in the art. The described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a full understanding of embodiments of this disclosure. However, those skilled in the art will recognize that the technical solutions of this disclosure can be practiced with one or more of the specific details omitted, or other methods, components, apparatus, steps, etc., can be employed. In other instances, well-known technical solutions are not shown or described in detail to avoid obscuring various aspects of this disclosure.

[0081] Furthermore, the accompanying drawings are merely illustrative of this disclosure and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0082] For increasingly complex graphics rendering cloud applications, single cloud computing power is insufficient to fundamentally meet their comprehensive needs for computing, rendering, storage, network bandwidth, and latency. Therefore, it is necessary to adopt a synergistic approach that integrates computing power and network capabilities to achieve the rendering of graphics cloud applications, so as to provide comprehensive assurance of both comprehensive computing power and network transmission determinism during the rendering process of graphics cloud applications.

[0083] It is for the reasons mentioned above that the concept of computing power networks has emerged and gained popularity. However, the commonly used computing power network technology solutions on the market all require many complex steps such as computing power acquisition, computing power scheduling, computing power routing, and computing power matching to provide the final computing power service. Furthermore, the provision of computing power services also requires reliance on architectures such as SDN (Software-Defined Network) and NFV (Network Functions Virtualization). On this basis, existing network routing and other equipment also need to be upgraded and transformed. Therefore, the computing power network technology in the existing solutions seriously restricts its large-scale deployment.

[0084] To address these issues, the current common approach is to encapsulate and integrate computing power and network requirements into a computing power routing architecture based on Software-Defined Networking (SDN) and Network Functions Virtualization (NFV). This routing is then forwarded via intelligent network element routing, and finally, the distributed computing power allocation and addressing are obtained by parsing the computing power routes. However, this approach has several drawbacks: First, it requires rebuilding a new computing power routing protocol, thus reducing computing power allocation efficiency. Second, it necessitates upgrading network element devices or virtual network element devices to support the corresponding computing power routing protocol, further complicating computing power allocation. Third, ensuring that the routing protocol clearly represents the rendering computation requirements of the corresponding graphics cloud application and matches it with numerous computing power nodes remains a challenge.

[0085] Therefore, while the aforementioned computing power allocation methods offer advantages in matching capabilities to various business needs and scheduling computing power / network resources, they also have the following drawbacks for personalized cloud-based graphics cloud applications: Firstly, they lack sufficient flexibility, resulting in lower accuracy of the allocated computing power. Secondly, they require unified management by cloud resource providers and application service providers, which not only severely restricts the flexibility of business implementation but also raises the entry barrier for participants, limiting effective interconnection. Furthermore, for graphics cloud applications, the variability in user scale, the uniqueness of the operating environment, and the extreme requirements of network transmission mean that using general computing power and network resource scheduling methods would further increase the complexity of resource scheduling. For example, common technical solutions are difficult to meet the requirements of efficiency and convenience, while also increasing deployment costs and scaling pressures.

[0086] The method for invoking computing and network capabilities for cloud VR services provided in this exemplary embodiment uses a flexible distributed computing network resource scheduling mechanism similar to distributed storage and computing separation. It builds a computing network capability matching mechanism based on network capability and computing capability providers. By interfacing with open capability providers of network capabilities and computing capability providers, it achieves the matching and mapping of graphics rendering computing power by uniformly subpackaging and identifying graphics cloud application rendering tasks and integrating them with network capabilities according to the user experience requirements of graphics cloud applications.

[0087] The method for invoking computing and network capabilities for cloud VR services provided in this exemplary embodiment will be incorporated into the iTU-T-related cloud VR protocol international standard as a standard-related patent, serving as the cloud graphics application cloud VR (Virtual Reality) network capability and rendering computing power scheduling processing flow and interaction protocol. Simultaneously, at the standard level, this part, as a cloud VR service and a component of cloud resource computing power scheduling and network capability matching and integration, improves user experience by using a distributed computing network scheduling mechanism to achieve integrated scheduling of cloud VR network capabilities and rendering computing power, providing matching and mapping of graphics rendering computing power and network capabilities.

[0088] This exemplary embodiment first provides a method for invoking computing power and network capabilities for cloud VR services. This method can run on servers, server clusters, or cloud servers, etc. Of course, those skilled in the art can also run the method disclosed herein on other platforms as needed, and this exemplary embodiment does not impose any special limitations on this. Specifically, refer to... Figure 1 As shown, the method for invoking computing and network capabilities for cloud VR services may include the following steps:

[0089] Step S110. The optimization processing module of the control layer divides the screen area of ​​the graphics cloud application in the application layer into blocks according to the user attention area of ​​the graphics cloud application running on the cloud, obtains the screen area block result, and obtains the block rendering request based on the screen area block result.

[0090] Step S120. The distributed computing network resource scheduling module in the control layer responds to the block rendering request, parses the block rendering request, and obtains the screen area block result and the graphics rendering computing power requirement and network transmission capability requirement corresponding to the screen area block result.

[0091] Step S130. The distributed computing network resource scheduling module matches the target computing power node in the resource layer for the block rendering task corresponding to the block rendering result of the screen area according to the graphics rendering computing power requirement and network transmission capability requirement.

[0092] In the aforementioned method for invoking computing power and network capabilities for cloud VR services, on the one hand, the optimization processing module in the control layer divides the screen area of ​​the graphics cloud application in the application layer into blocks based on the user's attention area in the cloud-based graphics cloud application, obtaining the screen area block results, and generating block rendering requests based on the screen area block results; then, the distributed computing network resource scheduling module in the control layer responds to the block rendering requests, parses the block rendering requests, and obtains the screen area block results and the corresponding graphics rendering computing power requirements and network transmission capability requirements; finally, the distributed computing network resource scheduling module matches the target computing power nodes in the resource layer with the block rendering tasks corresponding to the screen area block results based on the graphics rendering computing power requirements and network transmission capability requirements, reducing the complexity of calling the target computing nodes, solving the problem of low resource calling efficiency caused by the complexity of resource calling in related technical solutions, and improving the calling efficiency of the target computing nodes; on the other hand, it improves the screen rendering efficiency of the screen area block results.

[0093] The following will provide a detailed explanation and description of the method for invoking computing power and network capabilities for cloud VR services as described in the exemplary embodiments of this disclosure, with reference to the accompanying drawings.

[0094] First, the terms used in the exemplary embodiments of this disclosure will be explained and described.

[0095] Graphics cloud applications refer to applications that complete online processing or operation based on cloud rendering, including cloud gaming, cloud VR (Virtual Reality), and cloud AR (Augmented Reality) services. The graphics cloud applications described in the exemplary embodiments of this disclosure specifically refer to applications requiring large-scale scene cloud rendering; these applications can include different scenarios such as film and television scene rendering and digital city rendering; in the specific rendering process, a distributed scheduling module is needed to allocate rendering tasks to distributed computing nodes, and the distributed computing nodes complete the corresponding rendering tasks in a parallel processing manner.

[0096] Computing-Network Convergence: In the cloud rendering process of graphics cloud applications, not only is sufficient computing power required as support, but also because film and television scene rendering and digital city rendering have high requirements for experience latency and transmission bandwidth, it is necessary to converge computing power and network capabilities. At the same time, the computing-network convergence described in the example embodiments of this disclosure refers to matching the network transmission capabilities and rendering computing power of computing power nodes through unified computing-network convergence scheduling, thereby achieving the dual goal of a good experience in rendering computing power and network transmission capabilities of graphics cloud applications.

[0097] Secondly, the technical implementation principles of the exemplary embodiments of this disclosure will be explained and described.

[0098] Specifically, the method for invoking computing and network capabilities for cloud VR services provided in this exemplary embodiment can, based on open network capabilities and computing power services, construct a distributed scheduling and matching mechanism for graphics rendering and network transmission for graphics cloud applications. It adopts a method that integrates network transmission capabilities with unified segmentation, identification conventions, and adaptive service experience quality, distributing graphics rendering tasks to different distributed rendering computing nodes to jointly complete the matching and mapping of network capabilities and graphics rendering capabilities. Furthermore, the method for invoking computing and network capabilities for cloud VR services provided in this exemplary embodiment can be developed around the following points: Firstly, it achieves flexible matching of computing and network capabilities based on demand, eliminating the need for computing and network routing processing, thereby avoiding the complexity brought about by network transformation. That is, in practical applications, this exemplary embodiment adopts a network data packet encapsulation mechanism based on graphics rendering requirements, and uses a centralized scheduling and addressing mode to find and match computing resources for computing and network capability matching. This allows for the matching and integration of computing and network resources on the existing IP network architecture without the need for computing power routing forwarding based on network devices. On the other hand, the unified segmentation / identification of graphics rendering tasks connects with the transmission requirements of the small-granularity network, thereby meeting the personalized rendering needs of different regions. That is, the exemplary embodiments of this disclosure are geared towards graphics cloud applications, especially large-scale metaverse processing scenarios. Due to the significant differences in user experience requirements caused by regional variations in rendering needs, as well as the existence of different resolution processing mechanisms and network bandwidth / latency requirements, unified segmentation / identification and finer-grained integration with network transmission requirements can flexibly achieve the connection between small-granular graphics computing power requirements and network transmission requirements. Furthermore, based on a distributed scheduling mechanism, it can more efficiently connect with distributed computing resources. That is, compared to a centralized scheduling mechanism, the distributed scheduling mechanism of the exemplary embodiments of this disclosure enables connection with different computing power providers and different computing power centers within the same provider, achieving more efficient fusion of heterogeneous graphics rendering computing power.

[0099] Meanwhile, the method for invoking computing power and network capabilities for cloud VR services described in the example embodiments of this disclosure can effectively solve the following problems: On the one hand, it solves the problem that the overall graphics cloud application cannot achieve computing power fusion between graphics block computing power and fine-grained network computing power. That is, the current methods of integrating computing power resources and network resources all treat the cloud graphics application as a whole and then use an integrated computing power routing mode to provide computing power and network fusion services for the cloud graphics application, which cannot achieve flexible matching between the screen area segmentation results and fine-grained computing network in a personalized way. The example embodiments of this disclosure adopt a unified method to divide the rendering task into graphics blocks, which can not only match different computing power graphics processors to the divided graphics blocks, but also match network transmission capabilities according to the network resources required to transmit the graphics blocks, thereby achieving fine-grained fusion of computing power and network capabilities. On the other hand, the exemplary embodiments of this disclosure can construct a computing network resource scheduling mechanism for graphics cloud applications at the application level, solving the problem that existing network devices do not support computing power routing. That is, the computing power routing schemes in related technical solutions are based on NFV (Network Functions Virtualization), SDN (Software Defined Network) architectures, and corresponding intelligent network element devices and customized network routing protocols. However, existing network devices cannot support these, resulting in significant limitations in application specifications and deployment scope. Based on the method provided by the exemplary embodiments of this disclosure, the above problems can be solved at the application layer, satisfying the issue of existing network element devices not supporting the solution, which is conducive to its large-scale deployment and promotion. Furthermore, it solves the problem of high-quality computing power and network resources being squeezed out due to the single processing mechanism of graphics cloud applications. That is, the exemplary embodiments of this disclosure adopt a block-based processing mechanism for cloud graphics applications, which can divide different graphics blocks into different graphics rendering nodes and corresponding network transmission configurations according to different user experience needs, thereby achieving a fine distinction between computing power and network capabilities. This effectively and flexibly adapts computing power and network transmission capabilities, avoiding resource contention.

[0100] Furthermore, the system for invoking computing power and network capabilities for cloud VR services involved in the exemplary embodiments of this disclosure will be explained and described. Specifically, refer to... Figure 2As shown, this system for invoking computing and network capabilities for cloud VR services can include a control layer 210, an application layer 220, and a resource layer 230. The control layer includes an optimization processing module and a distributed computing network resource scheduling module. In practical applications, the optimization processing module can segment the screen area of ​​the graphics cloud application in the application layer according to the user's attention area, obtaining a screen area segmentation result, and generating a segmented rendering request based on the segmentation result. The distributed computing network resource scheduling module can respond to the segmented rendering request, parse the request, obtain the screen area segmentation result and the corresponding graphics rendering computing power and network transmission capability requirements, and match the target computing power node in the resource layer with the segmented rendering task corresponding to the screen area segmentation result based on the graphics rendering computing power and network transmission capability requirements.

[0101] In one exemplary embodiment, the computing power and network capability adjustment system for cloud VR services described above further includes a network layer 240, which can communicate with the control layer to evaluate the target network environment of the target computing power node and determine whether the target network transmission capability of the target computing power node needs to be enhanced based on the evaluation results. In one exemplary embodiment, the resource layer may include multiple computing power nodes, which can be used to execute the block rendering tasks issued by the distributed computing network resource scheduling module, obtain the block rendering results, and feed the block rendering results back to the application layer.

[0102] In one example embodiment, reference is made to... Figure 3 As shown, the computing power and network capability call system for cloud VR services described above also includes a terminal layer 310, which can communicate with the application layer to request the display screen of the graphics cloud application running on the cloud from the application layer; and receive the overall screen display result corresponding to the display screen from the application layer, and display the overall screen display result.

[0103] The following will further explain and illustrate the computing power and network capability invocation system for cloud VR services described above. Specifically, the example embodiment of this disclosure adopts a distributed computing network resource scheduling capability matching, connects to the network capability open platform and the computing power scheduling platform, and uses a unified method to divide and identify graphics rendering tasks, and bundles different user experience rendering needs and network transmission needs; on this basis, suitable graphics rendering resource computing power is matched to different graphics rendering task blocks according to network transmission capabilities, thereby realizing fine-grained processing of network transmission capabilities and computing power capabilities. Furthermore, after the graphics rendering task is completed, the block rendering results can be integrated according to the block identifiers in the obtained graphics rendering task, ultimately realizing flexible processing of graphics cloud applications. Furthermore, in the actual application process, the specific functions to be executed by each functional layer can be as follows:

[0104] 1) Application Layer: As the initiator of the computing network requirements for the graphics cloud application, the application layer needs to perform the following functions: ① Encapsulate the computing network requirements according to the experience requirements of the graphics cloud application, and upload the encapsulation results to the distributed computing network resource scheduling module of the control layer; ② Perform corresponding addressing communication according to the allocated computing power nodes; ③ Receive the segmented rendering results of the screen region segmentation results completed by the computing power nodes, parse the segmentation identifiers of the screen region segmentation results, and perform image integration and output based on the parsed segmentation identifiers. Among these:

[0105] First, the specific implementation process of encapsulating and uploading computing network resources according to business needs is as follows: The graphics cloud application is divided into blocks according to two aspects: resource rendering needs factors such as user attention area, frequency of interaction and complexity of screen changes, and network transmission needs factors such as display requirements and latency. The computing network resource requests, including the screen region block results, resource rendering needs and network transmission needs, are then uploaded to the distributed computing network resource scheduling module in the control layer.

[0106] Secondly, the specific implementation process of computing node addressing is as follows: according to the computing resource allocation results fed back by the distributed computing network resource scheduling module, the addressing and docking of the corresponding allocated computing nodes are completed, and the interactive operation between the application layer and the computing nodes is realized.

[0107] Then, the specific implementation process of parsing the image area segmentation result identifier is as follows: receive the segmented rendering result processed and sent by the computing power node, and parse the segment identifier in the segmented rendering result to facilitate subsequent overall output.

[0108] Finally, the specific implementation process of the integrated output of the graphics application is as follows: the received block rendering results are integrated and output according to the parsed block identifiers.

[0109] 2) Distributed computing network resource scheduling module in the control layer: Specifically, to meet the service requirements of the graphics cloud application, the module parses the service requirements (i.e., computing network resource requests) of the graphics cloud application, then allocates block rendering tasks and merges corresponding fine-grained computing network requirements, ultimately matching network transmission capacity requirements with graphics rendering computing power requirements. Among these:

[0110] First, the specific implementation process of business requirement parsing is as follows: parse the business requirements sent by the application layer (i.e., the block rendering request or the computing network resource request), and obtain the block rendering result of the graphics cloud application, the graphics rendering computing power requirement required to render the block rendering result, and the network transmission capability requirement required to transmit the block rendering result.

[0111] Secondly, the specific implementation process of the block identification / network requirement fusion is as follows: based on the requirement analysis results, the block results of each screen area to be rendered are identified accordingly, and the block identification, graphics rendering computing power requirement and network transmission capability requirement corresponding to the block results of the screen area are mapped.

[0112] Then, the specific implementation process of network transmission capacity block matching is as follows: according to the fine-grained network transmission capacity requirements of each screen area block result, and combined with the evaluation of the network environment of the target computing power node or the processing result of the network enhancement service, the candidate computing power node area that meets the network transmission capacity requirements is determined.

[0113] Finally, graphics rendering computing power matching: within the region (i.e., the candidate computing power node region) that meets the network transmission requirements of different screen area segmentation results, target computing power nodes that can meet their graphics rendering capability requirements are selected, and the corresponding matching is completed.

[0114] 3) Network Layer: As the interface for network capabilities, it provides services to assess the current network environment capabilities of different computing nodes and to enhance the network capabilities of computing nodes that do not meet network transmission requirements. Among these:

[0115] First, the specific implementation process of network environment assessment for computing power nodes is as follows: For different computing power nodes, environmental assessments of transmission indicators such as network transmission latency and transmission bandwidth are conducted so that the service range of computing power nodes that meet the network transmission capacity requirements can be located by dividing the service range during resource scheduling.

[0116] Secondly, the specific implementation process of network enhancement services is as follows: for computing power nodes that need network enhancement, the corresponding network capabilities of the computing power nodes are enhanced; for example, the transmission capabilities are enhanced based on bandwidth increase or multi-path transmission.

[0117] 4) Resource Layer: As the executor of the block rendering task, it first reports the graphics computing capabilities of the computing nodes, then performs parallel processing of graphics block rendering; after the parallel processing of graphics rendering is completed, it encapsulates the block rendering result based on the block identifier task execution result to obtain the block rendering result, and then distributes the block rendering result in parallel, etc. Among these:

[0118] The specific implementation process of graphics computing capability reporting is as follows: the relevant computing power nodes report their currently used computing power and available remaining computing power to the distributed computing network resource scheduling module; at the same time, they synchronize their corresponding network location to the network layer so that the network layer can evaluate the current network connectivity of the computing power node.

[0119] Secondly, the specific implementation process of parallel processing of the segmented rendering task is as follows: the distributed computing network resource scheduling module matches the graphics rendering computing power of the computing power node and then distributes the allocated screen area segmented rendering task to the corresponding target computing power node, and performs parallel processing on the segmented rendering task corresponding to the segmented result of the screen area based on the target computing power node.

[0120] Then, the specific implementation process of the task execution result identification encapsulation is as follows: the target computing power node encapsulates the relevant results of the parallel processing of the completed graphics block rendering (i.e., the task execution result) according to the block identifier marked by the distributed computing network resource scheduling module, and encapsulates the task execution result and the block identifier corresponding to the block result of the screen area to obtain the block rendering result.

[0121] Finally, the specific implementation process of parallel distribution of the block rendering results is as follows: each target computing power node distributes the block rendering results it has processed in parallel to the application layer in parallel, so that the application layer can integrate and output the graphics application according to the corresponding block identifier.

[0122] The following will combine Figure 2 as well as Figure 3 right Figure 1 The method for invoking computing power and network capabilities for cloud VR services, as shown in the example embodiments of this disclosure, is explained and described in detail. Specifically, the implementation process of the method for invoking computing power and network capabilities for cloud VR services as described in the example embodiments of this disclosure can include two parts: a pre-configuration process and a business processing process. The pre-configuration process mainly completes pre-processing functions such as computing power reporting of relevant computing power nodes and network capability assessment, so as to improve the efficiency of computing power node matching and network capability docking during the processing of graphics cloud applications. Specifically, the pre-configuration process can be implemented in the following way:

[0123] Step S001: The candidate computing power nodes in the resource layer complete the collection of corresponding node computing power resources and network information. The process of collecting node computing power resources is as follows: the graphics processing capability of the corresponding candidate computing power node is collected. The graphics processing capability may include, but is not limited to, the chip model, number of chips, memory size, and computing power of the graphics processing chip, as well as the public computing power resources and remaining computing power node capabilities. At the same time, the process of collecting location information is as follows: the network-related information, including the current IP address and available outbound bandwidth of the candidate computing power node, is collected.

[0124] Step S002: The resource layer reports the graphics computing capabilities of the candidate computing power nodes; specifically, the computing power nodes report their node graphics computing capabilities and related network information collection results to the distributed computing network resource scheduling module of the control layer via the resource layer.

[0125] Step S003: The distributed computing network resource scheduling module of the control layer evaluates the capabilities of the computing power nodes and records the computing power status. Specifically, the distributed computing network resource scheduling module evaluates the graphics computing capabilities uploaded by the computing power nodes. When the graphics rendering computing capabilities that can be reallocated reach the corresponding graphics computing capability threshold (such as a certain number of graphics processing chips of a certain model), the candidate computing power node is used as the graphics rendering computing power resource to be allocated.

[0126] Step S000: The distributed computing network resource scheduling module of the control layer synchronizes the network information of the computing power nodes; specifically, for computing power nodes that have reached the graphics computing power threshold, the distributed computing network resource scheduling module of the control layer will synchronize their corresponding network information to the network layer.

[0127] Step S005: The network layer performs a corresponding network environment assessment for the synchronized computing nodes; the network environment assessment mainly includes several dimensions such as link latency, available network egress bandwidth, and network transmission quality.

[0128] Step S006: The network layer evaluates the relevant network capabilities enhancements, such as bandwidth acceleration and link priority transmission, for the computing power nodes that require network enhancement.

[0129] Step S007: The network layer feeds back the network environment of the computing power nodes to the distributed computing network resource scheduling module of the control layer; specifically, the network layer feeds back the current bandwidth, latency, stability and other status of the corresponding computing power node network to the distributed computing network resource scheduling module, and simultaneously shows the potential for network capability enhancement, etc.

[0130] Step S008: The distributed computing network resource scheduling module of the control layer records the network status of the computing power nodes and integrates it with the computing capabilities. Specifically, the distributed computing network resource scheduling module integrates and records the network capabilities corresponding to the computing power nodes with their graphics rendering capabilities so that the computing power and network environment capabilities can be matched during business execution.

[0131] It should be noted here that the following configuration rules must also be followed during the pre-configuration process:

[0132] On the one hand, regarding step S004 above, before the computing power nodes synchronize network information, it is necessary to first determine the graphics computing power threshold. The implementation process is as follows: First, determine whether the computing power nodes can meet the relevant requirements for providing graphics rendering capabilities, that is, whether they have graphics processing cards related to graphics rendering and the hardware and software environment to run these graphics processing cards; Second, determine whether the relevant service thresholds are met: In order to facilitate the subsequent division of graphics rendering tasks, the graphics rendering capabilities of the relevant computing power nodes need to reach a certain level or indicator (i.e., threshold). For example, determine the corresponding graphics card model and quantity according to the type of graphics cloud application. In order to achieve complementarity and redundancy of relevant computing power resources, each computing power node should have at least twice the basic graphics rendering capability. The basic graphics rendering capability refers to the rendering capability of the smallest rendering unit for graphics cloud applications, such as the rendering capability of the FOV area in cloud VR, etc.

[0133] On the other hand, regarding step S006 above, the network layer can enhance the network transmission capabilities of computing nodes in the following ways: ① Network speed-up: Utilize the network capabilities of communication operators to increase the outbound bandwidth of the corresponding computing node. Of course, the computing node needs to have the link capability for the bandwidth increase during the enhancement process; ② Prioritized transmission: Improve the QoS (Quality of Service) level of the outbound traffic of the corresponding computing node to enhance its ability to seize network resources; ③ Reduce link latency: Reduce network path forwarding and related transmission time by optimizing the transmission links of the relevant computing nodes, thereby reducing the overall link latency; ④ Network slicing capability: Construct a network transmission environment with isolation attributes for the corresponding rendering nodes that require an isolated network transmission environment based on hard slicing or soft slicing.

[0134] Furthermore, after completing the pre-configured processing procedures, the corresponding computing-network convergence scheduling processing procedures for graphics cloud applications can be implemented. Specifically, firstly, in Figure 1 The methods for invoking computing and network capabilities for cloud VR services are shown below:

[0135] In step S110, the optimization processing module of the control layer divides the screen area of ​​the graphics cloud application in the application layer into blocks according to the user attention area of ​​the graphics cloud application running on the cloud, obtains the screen area block result, and obtains the block rendering request based on the screen area block result.

[0136] In this example embodiment, firstly, the screen area segmentation result is determined. Specifically, the specific process for determining the screen area segmentation result can be implemented as follows: Determine the user-focused area of ​​the graphics cloud application, and based on the user-focused area, determine the surrounding area adjacent to the user-focused area, the outer area adjacent to the surrounding area, and the temporarily imperceptible area adjacent to the outer area; determine the area rendering precision and area refresh rate required by the graphics cloud application for the user-focused area, the surrounding area, the outer area, and the temporarily imperceptible area; determine the overall screen size of the screen area of ​​the graphics cloud application, and segment the screen area of ​​the graphics cloud application according to the overall screen size and the area rendering precision and area refresh rate required in different display areas to obtain the screen area segmentation result. The user-focused area mentioned here can be referenced... Figure 4 As shown in 401, the surrounding area adjacent to the user's area of ​​interest can be referenced in 402 of the figure, and the outer area adjacent to the surrounding area of ​​interest can be referenced in 402 of the figure. Figure 4 As shown in 403, the temporarily undetectable area adjacent to the outer area can be referenced. Figure 4 As shown in 404; it should also be noted that the region rendering precision and region refresh rate required for the user's focus area are the highest, and the others decrease in that order; in the specific block segmentation process, the higher the required region rendering precision and region refresh rate, the larger the area of ​​the corresponding block; the lower the required region rendering precision and region refresh rate, the smaller the area of ​​the corresponding block.

[0137] In one exemplary embodiment, determining the user attention area of ​​a graphics cloud application can be achieved by: determining the field of view and / or the user's eye gaze position of the cloud-based graphics cloud application, and determining the user attention area based on the field of view and / or the user's eye gaze position. The field of view described herein can also be referred to as the field of view angle; specific scene example diagrams can be found in the provided image. Figure 5 As shown; at the same time, the field of view described here can be used to determine the user's attention area in VR scenes; furthermore, the user's attention area in naked-eye 3D scenes can be determined based on the human eye's gaze position; at the same time, the human eye's gaze position can be determined based on the captured human eye images.

[0138] Secondly, a segmented rendering request is obtained based on the image region segmentation result. Specifically, this can be achieved as follows: determine the graphics rendering computing power requirements needed to execute the segmented rendering task corresponding to the image region segmentation result, and the network transmission capability requirements needed to transmit the segmented rendering result corresponding to the image region segmentation result; generate a segmented rendering request corresponding to the image region segmentation result based on the image region segmentation result, the graphics rendering computing power requirements, and the network transmission capability requirements. That is, the obtained segmented rendering request includes the image region segmentation result itself, the graphics rendering computing power requirements needed to execute the segmented rendering task corresponding to the image region segmentation result, and the network transmission capability requirements needed to transmit the segmented rendering result corresponding to the image region segmentation result.

[0139] In step S120, the distributed computing network resource scheduling module in the control layer responds to the segmented rendering request, parses the segmented rendering request, and obtains the screen area segmentation result and the graphics rendering computing power requirement and network transmission capability requirement corresponding to the screen area segmentation result.

[0140] In other words, in practical applications, when the optimization processing module receives a segmented rendering request, it can synchronize the segmented rendering request to the distributed computing network resource scheduling module. When the distributed computing network resource scheduling module senses the segmented rendering request, it can respond to the segmented rendering request and then parse it to obtain the segmented result of the screen area, as well as the corresponding graphics rendering computing power requirements and network transmission capability requirements.

[0141] In step S130, the distributed computing network resource scheduling module matches the target computing power node in the resource layer with the block rendering task corresponding to the block rendering result of the screen area according to the graphics rendering computing power requirements and network transmission capability requirements.

[0142] For details, please refer to Figure 6 As shown, to match the target computing power node in the resource layer with the block rendering task corresponding to the block rendering result of the screen area, based on the graphics rendering computing power requirements and network transmission capability requirements, the following steps may be included:

[0143] Step S610: The image region segmentation result, the graphics rendering computing power requirement required to execute the segmented rendering task corresponding to the image region segmentation result, and the network transmission capability requirement required to transmit the segmented rendering result corresponding to the image region segmentation result are correlated to obtain the computing and network requirement fusion result of the segmented rendering task.

[0144] Specifically, the process of determining the network-computer interface requirements fusion result can be implemented as follows: Based on the position of the image region segmentation result in the graphics cloud application, the image region segmentation result is sorted to obtain a segmentation sorting result, and a segmentation identifier is configured for the image region segmentation result according to the segmentation sorting result; the image region segmentation result, the segmentation identifier, the graphics rendering computing power requirement required to execute the segmented rendering task corresponding to the image region segmentation result, and the network transmission capability requirement required to transmit the segmented rendering result corresponding to the image region segmentation result are associated to obtain the network-computer interface requirements fusion result for the segmented rendering task. Specifically, in the segmentation sorting process, the sorting can be performed in the order from inside to outside and from left to right; at the same time, "inside" here refers to the user's attention area; "from inside to outside" means: from the user's attention area → the surrounding area → the outer area → the temporarily unperceived area; "from left to right" means from the left side of the image region to the right side of the image region; at the same time, the obtained segmentation identifier can refer to Figure 7 As shown.

[0145] Step S620: Determine the alternative computing power capabilities and alternative network transmission capabilities of the alternative computing power nodes included in the resource layer, and fuse the alternative computing power nodes, alternative computing power capabilities, and alternative network transmission capabilities to obtain the fusion result of the alternative computing power node capabilities.

[0146] Specifically, the process of determining the alternative computing power capabilities and alternative network transmission capabilities of the candidate computing power nodes can be implemented as follows: receiving the node attribute information and remaining node computing power of the candidate computing power nodes reported by the resource layer, and determining the alternative computing power capabilities of the candidate computing power nodes based on the node attribute information and remaining node computing power; wherein, the node attribute information includes at least one of chip model, number of chips, memory size, and computing power; receiving the current node IP address and original outbound bandwidth resources of the candidate computing power nodes reported by the resource layer, and determining the alternative network transmission capabilities of the candidate computing power nodes based on the current node IP address and original outbound bandwidth resources.

[0147] In one example embodiment, determining the alternative network transmission capability of the alternative computing power node based on the current node IP address and the original outgoing bandwidth resources can be achieved as follows: The computing power capability of the alternative computing power node is evaluated to obtain a computing power capability evaluation result, and based on the computing power capability evaluation result, it is determined whether the alternative computing power node can serve as a rendering task execution node; if the alternative computing power node can serve as a rendering task execution node, the current node IP address and the original outgoing bandwidth resources of the alternative computing power node are synchronized to the network layer, and the alternative network transmission capability corresponding to the alternative computing power node is received from the network layer. The alternative network transmission capabilities described herein are obtained by the network layer in the following manner: Based on the current node IP address and the original egress bandwidth resources, the current network environment of the alternative computing node is evaluated to obtain a current network environment evaluation result; wherein the current network environment evaluation result includes at least one of current link latency, current available egress bandwidth, and current network transmission stability; based on the current network environment evaluation result, it is determined whether the current bandwidth resources and / or current transmission links of the alternative computing node need to be enhanced; if the current bandwidth resources and / or current transmission links need to be enhanced, the current bandwidth resources are accelerated, and / or the current transmission links are optimized, and the alternative network transmission capabilities are determined based on the accelerated current bandwidth resources and / or optimized current transmission links.

[0148] Step S630: Based on the fusion result of the computing network requirements and the fusion result of the candidate node capabilities, a target computing node is matched for the block rendering task from the candidate computing nodes included in the resource layer.

[0149] Specifically, the matching process for target computing nodes can be implemented as follows: Based on the network transmission capacity requirements in the network demand fusion result, the target node capability fusion result is matched from the candidate node capability fusion result; based on the graphics rendering computing power requirements in the network demand fusion result and the candidate computing power capabilities in the target node capability fusion result, a target computing power node is matched from the candidate computing power nodes for the block rendering task corresponding to the image region block segmentation result. That is, in practical applications, matching is performed first based on the network transmission capacity requirements dimension, and then matching is performed based on the graphics rendering computing power requirements dimension. This approach can improve matching efficiency.

[0150] In one example embodiment, after the target computing power node is matched, it is also necessary to establish a first communication link between the application layer and the target computing power node. Specifically, this can be achieved as follows: the distributed computing network resource scheduling module sends the target node IP address of the target computing power node to the application layer; the application layer performs node addressing on the target computing power node according to the target node IP address to establish the first communication link between the application layer and the target computing power node.

[0151] In an exemplary embodiment, after the target computing power node is matched, it is also necessary to generate and distribute a segmented rendering task corresponding to the screen region segmentation result. Specifically, this can be achieved as follows: the distributed computing network resource scheduling module generates a segmented rendering task based on the screen region segmentation result, the segmentation identifier of the screen region segmentation result, the graphics rendering computing power requirement required to execute the segmented rendering task corresponding to the screen region segmentation result, the network transmission capability required to transmit the segmented rendering result corresponding to the screen region segmentation result, the target computing power node, and the target computing power capability of the target computing power node, and distributes the segmented rendering task to the target computing power node via the resource layer. The distributed computing network resource scheduling module can achieve the following when distributing the segmented rendering task to the target computing power node via the resource layer: The distributed computing network resource scheduling module synchronizes the target network transmission capability of the target computing power node to the network layer; the network layer evaluates the target network environment of the target computing power node based on the target network transmission capability and the network transmission capability requirements of the segmented rendering task, obtains the target network environment evaluation result, and determines whether the target network transmission capability needs to be enhanced based on the evaluation result; if so, the network layer enhances the target network transmission capability and notifies the distributed computing network resource scheduling module to distribute the segmented rendering task; if not, it directly notifies the distributed computing network resource scheduling module to distribute the segmented rendering task; the distributed computing network resource scheduling module, based on the received notification, distributes the segmented rendering task to the target computing power node via the resource layer; An example diagram of the scenario for the segmented rendering task to be executed by the target computing power node can be found in the following diagram. Figure 8 As shown.

[0152] In one exemplary embodiment, after the target computing power node receives the segmented rendering task, it needs to execute the rendering task and transmit the segmented rendering result to the application layer. Specifically, this can be achieved as follows: the target computing power node executes the segmented rendering task, obtains the task execution result, and encapsulates the segment identifier and the task execution result in the segmented rendering task to obtain the segmented rendering result; the target computing power node feeds back the segmented rendering result to the application layer based on the first communication link.

[0153] In one exemplary embodiment, after the application layer receives the segmented rendering result, it also needs to feed back the overall screen rendering result to the terminal layer. Specifically, this can be achieved as follows: the application layer sequentially splices the segmented rendering results according to the segmentation identifier in the segmented rendering result to obtain the overall screen rendering result of the cloud-based graphics application, and then feeds back the overall screen rendering result to the terminal layer.

[0154] Secondly, based on the above-mentioned content, the specific processing flow related to the convergence scheduling of computing and network can be implemented in the following way:

[0155] Step S001': The application layer encapsulates business requirements; specifically, firstly, it determines the computing power requirements; that is, in actual application, the application layer first considers the overall rendering graphics size, FOV (Field of View), etc. Based on factors such as the view (field of view) range and the rendering precision / refresh frequency of different regions, the image quality requirements of different graphics rendering regions are encapsulated according to the business experience, and this is used as the basis for evaluating rendering computing power; then, network requirements are determined; that is, in the actual application process, the application layer also needs to give corresponding network transmission requirements based on the experience quality, mainly for the differences in latency indicators of different regions, etc.

[0156] Step S002', the application layer reports the graphics rendering requirements: the application layer reports the corresponding graphics rendering requirements to the optimization processing module of the control layer, including the relevant computing power requirements and network requirements;

[0157] Step S003': The optimization processing module of the control layer conducts corresponding evaluations on the computing power and network requirements related to the business. That is, it analyzes the rendering requirements and network transmission requirements of different regions of the graphics rendering task from the perspective of experience, so as to lay the foundation for subsequent graphics rendering task segmentation.

[0158] Step S004': The optimization processing module of the control layer determines the screen area segmentation result and generates a segmented rendering request; that is, in the actual application process, the optimization processing module analyzes the rendering requirements of different areas of the graphics rendering task and the network transmission requirements, divides the corresponding graphics rendering task into blocks according to different graphics rendering requirements, obtains the screen area segmentation result, generates a segmented rendering request based on the screen area segmentation result, and then synchronizes the segmented rendering request to the distributed computing network resource scheduling module.

[0159] Step S005': The distributed computing network resource scheduling module of the control layer generates the computing network demand fusion result. Specifically, after the distributed computing network resource scheduling module senses the block rendering request, it divides the corresponding segments according to the position of the block result in the screen area, in the order from inside to outside and from left to right, to obtain the block identifier. It then associates the screen area block result with the corresponding computing power, network and other different requirements, so as to allocate relevant computing power nodes according to the corresponding computing power and network requirements in the future.

[0160] Step S006': The distributed computing network resource scheduling module of the control layer performs block matching of network transmission capabilities; specifically, in the actual application process, it is necessary to map the network requirements related to different block rendering tasks with the network transmission condition evaluation of each computing power node that has been pre-configured and completed the process, and match different graphics rendering blocks with computing power nodes that meet the network transmission conditions respectively.

[0161] Step S007': The distributed computing network resource scheduling module of the control layer performs graphics rendering computing power matching; specifically, in the process of matching computing power nodes with the network requirements of different block rendering tasks, the computing power nodes that meet both computing power and network requirements can be matched according to the graphics rendering capabilities of each computing power node and the rendering requirements of each rendering task block.

[0162] Step S008': The distributed computing network resource scheduling module of the control layer maps computing power nodes; specifically, in the actual application process, it is necessary to map different graphics rendering block tasks with candidate computing power nodes that meet the corresponding network requirements and graphics rendering computing power.

[0163] Step S009': The distributed computing network resource scheduling module of the control layer sends a network capability assessment request for the computing power node to the network layer. Specifically, in actual application, the distributed computing network resource scheduling module of the control layer needs to send the matched computing power node and its corresponding network requirements to the network layer, and send a current network capability assessment request for the corresponding computing power node.

[0164] Step S010': The network layer evaluates the network environment of the matched computing power nodes and provides enhanced network services. Specifically, in practical applications, the network layer evaluates the network environment of the matched computing power nodes (i.e., the target computing power nodes) and provides enhanced network services for nodes whose network environments do not meet the requirements.

[0165] Step S011': The target computing power node of the resource layer performs parallel processing on the block rendering task; specifically, in the actual application process, the computing power node that completes the corresponding network environment assessment or network service enhancement processing receives the block rendering task and performs the corresponding parallel rendering processing according to the requirements of the block rendering task.

[0166] In step S012', the target computing node of the resource layer encapsulates the task execution result and the block identifier to obtain the block rendering result; specifically, in actual application, the target computing node can encapsulate the corresponding completed task execution result and the corresponding block identifier according to the block identifier uniformly allocated by the distributed computing network resource scheduling module to obtain the corresponding block rendering result.

[0167] Step S013': The target computing node of the resource layer distributes the block rendering results in parallel; specifically, in actual application, the target computing node distributes the corresponding graphics rendering results (i.e., block rendering results) that have completed the task identifier encapsulation to the application layer.

[0168] Step S014': The application layer parses the segmentation identifiers in the segmented rendering result. Specifically, in actual application, the application layer receives the segmented rendering result data of the screen area (i.e., the segmented rendering result), parses it, and identifies the corresponding segmentation identifiers for subsequent integration and output.

[0169] Step S015': The application integrates and outputs the graphics application; specifically, in the actual application process, the application layer integrates the block rendering results according to the corresponding block identifier and completes the corresponding output in a unified manner.

[0170] It should be noted here that the following rules must also be followed in the computing-network convergence scheduling process:

[0171] On the one hand, regarding the above step S001', the specific implementation of encapsulating relevant business requirements can follow the following rules:

[0172] ① For the encapsulation of computing power requirements, the goal can be to improve the user experience. The computing power requirements can be encapsulated according to the corresponding rendering graphics resolution, frame rate and other indicators. At the same time, during the encapsulation process, the graphics rendering quality requirements can be mapped to different graphics rendering requirements based on the characteristics of different regions.

[0173] ② For areas of focus (i.e., user-focused areas), such as the FOV (Field of View) in VR, these areas, being the primary focus, require higher resolution and frame rate in graphics rendering, and their rendering power requirements are also higher than other areas. Meanwhile, for surrounding areas: as secondary areas adjacent to the user-focused area, their resolution and frame rate requirements are lower than the user-focused area, and their rendering power requirements are also correspondingly lower. Furthermore, for peripheral areas: as areas adjacent to the user-focused area, although they are perceptible, their image precision is lower than the user-focused area because they are on the periphery, and their latency is lower due to the reduced attention they receive, thus their rendering power requirements are also lower than the user-focused area. Finally, for areas that are temporarily imperceptible, because these areas are imperceptible, their main function is to prevent black borders and other issues when the focus changes; their rendering power and latency requirements are lower than those of peripheral areas.

[0174] In summary, the relevant computing network requirements are as follows: the demand for the area of ​​interest (the area of ​​user interest) > the demand for the surrounding area > the demand for the outer area > the demand for the area that cannot be perceived at present.

[0175] ③ For network requirement encapsulation, the same processing method as computing power requirement encapsulation can be adopted. With the goal of experience quality, network transmission requirements with different characteristics can be proposed according to the characteristics of different regions. As long as latency is the main network transmission factor, the relevant latency requirements can also meet the requirements of "areas of interest to users > surrounding areas of interest > peripheral areas > areas that cannot be perceived temporarily".

[0176] On the other hand, the specific implementation of steps S003' to S007' above should follow the following rules:

[0177] ① Regarding the computing network requirement assessment in step S003' above, the main function is to map and assess the computing network requirements for the encapsulated business requirements during the application service processing. Specifically, this can be manifested in the following aspects:

[0178] Computing power requirement assessment: Based on the corresponding graphics rendering resolution, graphics size, refresh rate, etc. in the packaging requirements, calculate the relevant computing power requirements for different areas during the packaging process. And according to the industry-standard calculation method that an overall graphics cloud application needs to achieve 4K / 120fps quality, it requires 29.6 TFLOPS of GPU computing power.

[0179] Use the region of interest (user-focused region): If this core region needs to achieve a 4K 120FPS quality level, then the graphics rendering of this part requires 29.6 TFLOPS of GPU computing power;

[0180] Focus on the surrounding area: As a secondary area, its image quality needs to reach 4K to achieve seamless connection with the core area, but its refresh rate should not be too high, such as 60FPS. In this case, the graphics rendering GPU computing power required for this area is 19.8 TFLOPS.

[0181] Peripheral area: As the outer part of the perceptible area, its image quality and refresh rate requirements are not high. For example, if it is 2K / 30FPS, then the graphics rendering GPU computing power required for this area is 3.73TFLOPS.

[0182] Temporarily unperceptible area: As an unperceptible area, to avoid black borders and other issues during rotation, the graphics quality and refresh rate are reduced. For example, at 1K / 30FPS, this area requires 1.87 TFLOPS of GPU computing power for graphics rendering.

[0183] Network Requirements Assessment: Based on the encapsulation requirements, the required bandwidth is calculated according to the image quality and size of the corresponding area, and the transmission latency requirements are estimated based on the experience quality. The corresponding bandwidth requirement is determined by calculating the relevant bitstream: Bitstream = Resolution × Frame Rate × Bit Depth / Compression Ratio. The relevant bit depth is calculated as follows: 12 bits for the center area for optimal experience, 10 bits for the periphery, and 8 bits for the outermost area, with a compression ratio of 150.

[0184] Use the area of ​​interest (i.e., the area of ​​user interest): If this core area needs to achieve 4K, 120FPS, 12bit quality, and occupies 10% of the overall screen area, then the required bitrate for this part is 4096×2160×120×12 / 150, which is 81Mbps. Considering the corresponding transmission redundancy, the required bandwidth is 162Mbps. The latency requirement for this core area is no more than 30ms.

[0185] Focus on the surrounding area: As a secondary area, the image quality needs to reach 4K, 60FPS, 10bit quality level, and this area occupies 20% of the overall screen area. Therefore, the required bit rate for this part is 30.375Mbps, and the required bandwidth is 60.75Mbps; the latency requirement for this core area is no more than 50ms.

[0186] Peripheral Area: As the outer part of the perceptible area, its image quality is 2K / 30FPS, 8bit, and its coverage is 30%. Therefore, this part requires a bitrate of 5.07Mbps and a bandwidth of 10.14Mbps. The latency requirement for this core area is no more than 80ms.

[0187] Temporarily undetectable area: As an undetectable area, with a resolution of 1K / 30FPS, 8bit, and a coverage area of ​​40%, this part requires a bitrate of 2.54Mbps and a bandwidth of 5.08Mbps. The latency requirement for this core area is no greater than 100ms.

[0188] ② Regarding the segmentation and identification of the rendering graphics task blocks in step S004' above, it is necessary to divide the corresponding rendering tasks into different rendering task blocks according to different graphics cloud application business situations and their different computing power requirements. In the embodiments of this patent, the graphics cloud application is divided into four different regions from the inside out, according to the user-focused area, the surrounding area, the outer area, and the temporarily imperceptible area mentioned above. Among them, for the user-focused area (i.e., the user-focused area), as the core area, it occupies 10% of the overall graphics and is identified as a block, which can be identified by number 1; for the surrounding area, as the secondary area, it occupies 20% of the overall graphics and is identified as a block, which can be identified by number 2; for the outer area, as the temporarily unfocused outer area, it occupies 30% of the overall graphics and is identified as a block, which can be identified by number 3; for the temporarily imperceptible area, it occupies 40% of the overall graphics and is identified as a block, which can be identified by number 4.

[0189] ③ The network requirement fusion in step S005' refers to the process of associating the segmented screen regions with the corresponding evaluated network requirements, laying the foundation for subsequent processing by different graphics processing cards within relevant computing nodes or unified computing nodes. Specifically:

[0190] Use the region of interest (user-focused region): As the rendering graphics task block number 1, its associated GPU computing power requirement is 2.96 TFLOPS; the associated network transmission bandwidth requirement is 162 Mbps, and the latency is no more than 30 ms.

[0191] Pay attention to the surrounding area: As the rendering graphics task block number 2, its associated GPU computing power requirement is 2.96 TFLOPS; the associated network transmission bandwidth requirement is 60.75 Mbps, and the latency is no more than 50 ms.

[0192] The outer region: As the rendering graphics task block numbered 3, its associated GPU computing power requirement is 1.11 TFLOPS; the associated network transmission bandwidth requirement is 10.14 Mbps, and the latency is no more than 80 ms.

[0193] Temporarily undetectable area: As the rendering graphics task block number 4, its associated GPU computing power requirement is 0.74 TFLOPS; the associated network transmission bandwidth requirement is 5.08 Mbps, and the latency is no more than 100 ms.

[0194] ④ Regarding the network transmission capacity block matching in step S006' above, this refers to matching the corresponding computing power nodes according to the network bandwidth and latency requirements corresponding to different graphics rendering task blocks. To achieve accurate matching of network transmission capabilities, firstly, computing power nodes meeting the latency requirements are found based on the coverage area corresponding to the latency requirements; then, computing power nodes that simultaneously meet both bandwidth and latency requirements are found based on the bandwidth requirements, serving as candidate computing power nodes for subsequent computing power matching. Specifically:

[0195] Matching computing power nodes to meet network latency requirements: Based on the latency information of each computing power node obtained in the pre-configuration process, and taking the business requirement initiation point as the target, computing power nodes with latency of 30ms, 50ms, 80ms and 100ms are searched from the inside out, and these are respectively used as candidate computing power nodes that meet the latency requirements for graphics rendering task blocks 1, 2, 3 and 4.

[0196] Matching computing power nodes that meet network transmission requirements: For candidate computing power nodes that meet latency requirements corresponding to graphics rendering task blocks 1, 2, 3, and 4, their current bandwidth is analyzed to ultimately find the computing power nodes that meet the overall network transmission requirements. Current bandwidth refers to the bandwidth currently available to the computing power node.

[0197] Nodes with a current bandwidth of no less than 162Mbps and a corresponding latency of 30ms are considered as candidate computing nodes for meeting the network transmission requirements of graphics rendering task block 1.

[0198] Nodes with a current bandwidth of no less than 60.75Mbps and a corresponding 50ms latency are considered as candidate computing nodes that meet the network transmission requirements for graphics rendering task block 2.

[0199] Nodes with a current bandwidth of no less than 10.14Mbps and a corresponding latency of 80ms are considered computing nodes that meet the network transmission requirements for graphics rendering task block 3.

[0200] Nodes with a current bandwidth of no less than 5.08Mbps and a corresponding latency of 100ms are considered as candidate computing nodes for meeting the network transmission requirements of graphics rendering task block 4.

[0201] ⑤ Regarding the graphics rendering computing power matching in step S007' above, it refers to matching the corresponding graphics task block rendering computing power for candidate computing power nodes that have completed network transmission capability matching, according to the following principles.

[0202] The computing power must meet the requirements of the graphics rendering task: that is, the relevant computing power nodes must meet the rendering computing power requirements of the corresponding rendering task blocks. Specifically, the rendering computing power of the computing power nodes corresponding to graphics rendering task blocks 1, 2, 3, and 4 must reach 29.6 TFLOPS, 19.8 TFLOPS, 3.73 TFLOPS, and 1.87 TFLOPS, respectively.

[0203] Equipped with equivalent computing power for seamless migration: In addition to meeting the rendering requirements of the corresponding rendering task, it also needs to have at least the remaining equivalent computing power so that when a calculation error occurs, the corresponding graphics rendering task can be migrated to the redundant computing power for processing.

[0204] Preferred computing nodes are those capable of simultaneously executing other graphics rendering tasks: that is, in addition to meeting the above two corresponding requirements, the computing node is also capable of supporting other rendering task blocks with lower requirements, and the bandwidth of the computing node can also meet the network transmission requirements of other rendering tasks.

[0205] If a computing node meets the requirements for a preferred computing node and can execute all the graphics rendering tasks in each block, and its bandwidth and latency can also support the network transmission of all rendering tasks, then all graphics rendering task blocks can be delivered to the same rendering node for execution.

[0206] Furthermore, the specific implementation of steps S009' to S010' above must follow the following rules:

[0207] First, in step S009', because there is a certain time difference between the pre-processing and actual execution, network transmission conditions may change during this period, thus requiring a re-evaluation of the network environment. For example, if the relevant network links are not faulty or congested, the current bandwidth usage of the computing node is mainly re-evaluated; that is, the current bandwidth is the total contracted outbound bandwidth of the computing node minus the bandwidth used by the current data stream. If the requirements are not met, the relevant processing operations in step S010' are performed. Conversely, if the relevant network links are faulty or congested, it is necessary to assess whether other links are available and proceed to step S010' for further processing.

[0208] Secondly, in step S010', based on the environmental assessment results of the current network, network enhancement is performed according to requirements. For example, when the network link is normal but the bandwidth is insufficient, network enhancement is performed directly by increasing the outgoing bandwidth of the computing nodes. As another example, when a network transmission link failure or congestion occurs, it is necessary to assess whether new links are available. If no new links are available, a message indicating that service cannot be provided is returned. When other links are available, it is necessary to assess whether the new connection meets the bandwidth and latency requirements. If the latency is insufficient, a message indicating that the network capacity cannot meet the requirements is returned; if the latency meets the requirements but the bandwidth is insufficient, the network transmission requirements are met by increasing the corresponding total outgoing bandwidth.

[0209] Thus, the methods for invoking computing power and network capabilities for cloud VR services described in the exemplary embodiments of this disclosure have been fully implemented. Based on the foregoing description, it can be understood that the methods for invoking computing power and network capabilities for cloud VR services described in the exemplary embodiments of this disclosure, by constructing a distributed computing network resource scheduling system based on the converged business network transmission requirements and rendering computing power requirements, achieve the matching and fusion scheduling of computing network requirements and network capabilities, and rendering computing power resources for graphics cloud applications. This has at least the following beneficial effects:

[0210] 1) The rendering computing power and network capability scheduling of user-mode graphics cloud applications have the capability to be deployed on a large scale to adapt to the existing network elements: That is, the example embodiment of this disclosure can realize flexible matching and resource scheduling of graphics rendering computing power and network resources based on business needs by building distributed computing network resource scheduling on network capabilities and computing power nodes, and solve the problem of needing to modify or even intelligently replace network elements through computing power network routing.

[0211] 2) Integrating network enhancement capabilities to flexibly adapt to the network transmission needs and changes of graphics cloud applications: That is, the inherent bandwidth acceleration and link optimization operations in related technical solutions are all passive response modes; therefore, the example embodiment of this disclosure adopts distributed computing network resource scheduling, and by connecting to network open capabilities, it can flexibly adjust based on the service needs of graphics cloud applications, and provide targeted personalized bandwidth enhancement, link optimization and other network capability enhancement services for computing power nodes.

[0212] 3) The fine-grained division of graphics rendering tasks and their distribution to different computing power nodes can effectively reduce the hardware requirements for graphics rendering and promote the reuse of existing graphics processors: In practical applications, cloud graphics rendering services require high-performance graphics processors as processing chips, but graphics processors are characterized by their high cost and are also subject to international risks. The method described in the exemplary embodiments of this disclosure can integrate computing power and network transmission capabilities, divide complex graphics rendering tasks into blocks, and reduce the requirements for graphics processors by delivering different blocks of graphics rendering tasks to different computing power nodes. Furthermore, it can leverage the capabilities of older graphics processors based on existing infrastructure, achieving cost reduction and efficiency improvement.

[0213] 4) Based on planned network transmission, avoid the impact of high-throughput graphics cloud services on the network: In practical applications, high-throughput graphics cloud applications have the characteristics of high bandwidth and low latency transmission, which can put pressure on the network, especially during peak usage periods. At the same time, the traditional elephant stream mode also impacts the network, which can easily cause network congestion, packet loss and other problems. By adopting a mode of dividing graphics rendering tasks into blocks, the network transmission data flow can be effectively organized, which can effectively reduce the impact on the network.

[0214] 5) The method described in the exemplary embodiments of this disclosure constructs a distributed computing network resource scheduling based on computing power nodes and network capabilities. With the goal of improving the experience quality of graphics cloud applications, it starts from business needs, implements fine-grained division and task allocation of graphics cloud rendering tasks based on the application layer, accurately integrates network capabilities to match computing power nodes based on different rendering graphics task blocks, and performs corresponding network capability enhancement processing according to the current network conditions. This breaks through the performance bottleneck of graphics processors, leverages the role of older equipment, reduces the impact of large data flows on network capacity, adapts to existing network element equipment, avoids the limitations of computing power routing flexibility and network element equipment upgrades and modifications. Furthermore, the method described in the exemplary embodiments of this disclosure also implements a user-level rendering computing power and network capability scheduling and capability fusion mechanism for graphics cloud applications: compared with other computing power routing solutions, it is more flexible and compatible with the existing network. By constructing a computing-network fusion resource scheduling mechanism for graphics cloud applications at the application layer, it can effectively adapt graphics cloud application rendering to efficiently integrate network transmission environment and network enhancement capabilities based on computing power nodes and network transmission capabilities. Furthermore, it also realizes flexible segmentation of graphics rendering tasks with fine granularity to solve the dual pressure of computing power and network carrying capacity: taking business experience as the starting point, graphics rendering tasks are flexibly divided into fine granularities according to different areas of interest. This not only simplifies the heavy graphics rendering tasks with fine granularity, but also facilitates the fusion and invocation of graphics processors with different processing capabilities, and reduces the dependence on the latest graphics processors by adopting a reuse model. Simultaneously, network transmission based on the block-based processing of graphics rendering task blocks changes the overall "elephant stream" pattern. By employing multiple block streams for reshaping, the impact of the "elephant stream" on network capacity is reduced, improving network stability. Furthermore, based on the current network transmission environment, an adaptive network capability enhancement mechanism for graphics rendering task blocks can be implemented. Through distributed computing network resource scheduling and network capability integration, the network transmission capacity of graphics rendering computing nodes processing different task blocks can be assessed in real time according to their current network environment. Based on business needs, the outbound bandwidth of computing nodes can be increased in real time, and congested network transmission links can be replaced, thus providing a more flexible network capability enhancement mechanism.

[0215] Finally, if the method for invoking computing power and network capabilities for cloud VR services described in the exemplary embodiments of this disclosure is to be used as a related standard or patent, the following points should be noted:

[0216] On the one hand, the method for invoking computing power and network capabilities for cloud VR services described in the example embodiments of this disclosure will be incorporated into the corresponding ITU-T cloud VR protocol-related international standards as part of cloud VR computing network resource scheduling. On the other hand, the method for invoking computing power and network capabilities for cloud VR services described in the example embodiments of this disclosure is described in the standard as part of the computing network resource scheduling process. Through the resource scheduling module of the cloud VR control layer, the fusion of computing power at the resource layer and network capabilities at the network layer is realized. The task allocation and processing of different computing power nodes with fused network capabilities are realized by dividing the graphics rendering task into blocks.

[0217] It should also be noted that the specific wording in the standard will be gradually expanded and explained in accordance with the scope of protection laid out in the exemplary embodiments of this disclosure.

[0218] The following are embodiments of the apparatus disclosed herein, which can be used to execute embodiments of the method disclosed herein. For details not disclosed in the apparatus embodiments of this disclosure, please refer to the embodiments of the method disclosed herein.

[0219] This disclosure also provides an example embodiment of a device for invoking computing power and network capabilities for cloud VR services. Specifically, refer to... Figure 9 As shown, the device for accessing computing and network capabilities for cloud VR services may include a screen area segmentation module 910, a segmented rendering request parsing module 920, and a target computing node matching module 930. Wherein:

[0220] The screen region segmentation module 910 can be used to segment the screen region of the graphics cloud application in the application layer according to the user attention area of ​​the graphics cloud application running on the cloud, through the optimization processing module of the control layer, to obtain the screen region segmentation result, and obtain the segmented rendering request based on the screen region segmentation result.

[0221] The segmented rendering request parsing module 920 can be used to respond to the segmented rendering request through the distributed computing network resource scheduling module in the control layer, parse the segmented rendering request, and obtain the screen area segmentation result and the graphics rendering computing power requirement and network transmission capability requirement corresponding to the screen area segmentation result.

[0222] The target computing power node matching module 930 can be used by the distributed computing network resource scheduling module to match the target computing power node in the resource layer for the block rendering task corresponding to the block rendering result of the screen area according to the graphics rendering computing power requirements and network transmission capability requirements.

[0223] In one exemplary embodiment of this disclosure, the screen area of ​​a graphics cloud application running in the application layer is segmented according to the user attention area of ​​the graphics cloud application to obtain a screen area segmentation result. This includes: determining the user attention area of ​​the graphics cloud application; and, based on the user attention area, determining the attention periphery area adjacent to the user attention area, the outer periphery area adjacent to the attention periphery area, and the temporarily imperceptible area adjacent to the outer periphery area; determining the area rendering precision and area refresh rate required by the graphics cloud application for the user attention area, attention periphery area, outer periphery area, and temporarily imperceptible area; determining the overall screen size of the screen area of ​​the graphics cloud application; and, based on the overall screen size and the area rendering precision and area refresh rate required in different display areas, segmenting the screen area of ​​the graphics cloud application to obtain a screen area segmentation result.

[0224] In one exemplary embodiment of this disclosure, determining the user attention area of ​​the graphics cloud application includes: determining the field of view and / or the human eye gaze position of the cloud-based graphics cloud application, and determining the user attention area based on the field of view and / or the human eye gaze position.

[0225] In one exemplary embodiment of this disclosure, obtaining a segmented rendering request based on the screen region segmentation result includes: determining the graphics rendering computing power requirement required to execute the segmented rendering task corresponding to the screen region segmentation result, and the network transmission capability requirement required to transmit the segmented rendering result corresponding to the screen region segmentation result; and generating a segmented rendering request corresponding to the screen region segmentation result based on the screen region segmentation result, the graphics rendering computing power requirement, and the network transmission capability requirement.

[0226] In one exemplary embodiment of this disclosure, matching a target computing power node in the resource layer for a segmented rendering task corresponding to the image region segmentation result, based on the graphics rendering computing power requirements and network transmission capability requirements, includes: associating the image region segmentation result, the graphics rendering computing power requirements required to execute the segmented rendering task corresponding to the image region segmentation result, and the network transmission capability requirements required to transmit the segmented rendering result corresponding to the image region segmentation result, to obtain a computing and network requirement fusion result for the segmented rendering task; determining the alternative computing power capabilities and alternative network transmission capabilities of the alternative computing power nodes included in the resource layer, and fusing the alternative computing power nodes, alternative computing power capabilities, and alternative network transmission capabilities to obtain an alternative node capability fusion result for the alternative computing power nodes; and matching a target computing power node for the segmented rendering task from the alternative computing power nodes included in the resource layer based on the computing and network requirement fusion result and the alternative node capability fusion result.

[0227] In one exemplary embodiment of this disclosure, the computational and network requirements for the segmented rendering task are correlated with the image region segmentation result, the graphics rendering computing power requirement required to execute the segmented rendering task corresponding to the image region segmentation result, and the network transmission capability requirement required to transmit the segmented rendering result corresponding to the image region segmentation result, to obtain a computational and network requirement fusion result for the segmented rendering task. This includes: sorting the image region segmentation result based on its segmentation position in the graphics cloud application to obtain a segmentation sorting result, and configuring a segmentation identifier for the image region segmentation result according to the segmentation sorting result; and correlating the image region segmentation result, the segmentation identifier, the graphics rendering computing power requirement required to execute the segmented rendering task corresponding to the image region segmentation result, and the network transmission capability requirement required to transmit the segmented rendering result corresponding to the image region segmentation result to obtain a computational and network requirement fusion result for the segmented rendering task.

[0228] In one exemplary embodiment of this disclosure, matching a target computing power node for the segmented rendering task from the candidate computing power nodes included in the resource layer, based on the network demand fusion result and the candidate node capability fusion result, includes: matching the target node capability fusion result from the candidate node capability fusion result based on the network transmission capability requirements in the network demand fusion result; and matching a target computing power node from the candidate computing power nodes for the segmented rendering task corresponding to the screen area segmentation result, based on the graphics rendering computing power requirements in the network demand fusion result and the candidate computing power capabilities in the target node capability fusion result.

[0229] In one exemplary embodiment of this disclosure, determining the alternative computing power capabilities and alternative network transmission capabilities of alternative computing power nodes included in the resource layer includes: receiving node attribute information and remaining node computing power of alternative computing power nodes reported by the resource layer, and determining the alternative computing power capabilities of the alternative computing power nodes based on the node attribute information and remaining node computing power; wherein, the node attribute information includes at least one of chip model, number of chips, memory size, and computing power; receiving the current node IP address and original egress bandwidth resources of alternative computing power nodes reported by the resource layer, and determining the alternative network transmission capabilities of the alternative computing power nodes based on the current node IP address and original egress bandwidth resources.

[0230] In one exemplary embodiment of this disclosure, determining the alternative network transmission capability of the alternative computing power node based on the current node IP address and the original egress bandwidth resources includes: evaluating the computing power capability of the alternative computing power node to obtain a computing power capability evaluation result, and determining whether the alternative computing power node can serve as a rendering task execution node based on the computing power capability evaluation result; if the alternative computing power node can serve as a rendering task execution node, then synchronizing the current node IP address and the original egress bandwidth resources of the alternative computing power node to the network layer, and receiving the alternative network transmission capability corresponding to the alternative computing power node fed back by the network layer.

[0231] In an exemplary embodiment of this disclosure, the alternative network transmission capability is obtained by the network layer in the following manner: Based on the current node IP address and the original egress bandwidth resources, the current network environment of the alternative computing node is evaluated to obtain a current network environment evaluation result; wherein, the current network environment evaluation result includes at least one of current link latency, current available egress bandwidth, and current network transmission stability; based on the current network environment evaluation result, it is determined whether the current bandwidth resources and / or current transmission links of the alternative computing node need to be enhanced; if the current bandwidth resources and / or current transmission links need to be enhanced, the current bandwidth resources are accelerated, and / or the current transmission links are optimized, and the alternative network transmission capability is determined based on the accelerated current bandwidth resources and / or optimized current transmission links.

[0232] In an exemplary embodiment of this disclosure, the first communication link between the application layer and the target computing node is obtained in the following manner: the distributed computing network resource scheduling module sends the target node IP address of the target computing node to the application layer; the application layer performs node addressing on the target computing node according to the target node IP address to establish the first communication link between the application layer and the target computing node.

[0233] In one exemplary embodiment of this disclosure, the means for invoking computing power and network capabilities further includes:

[0234] The chunked rendering task distribution module can be used to generate chunked rendering tasks by means of the distributed computing network resource scheduling module based on the screen region chunking result, the chunking identifier of the screen region chunking result, the graphics rendering computing power requirement required to execute the chunked rendering task corresponding to the screen region chunking result, the network transmission capability required to transmit the chunked rendering result corresponding to the screen region chunking result, the target computing power node, and the target computing power capability of the target computing power node, and distribute the chunked rendering task to the target computing power node through the resource layer.

[0235] In one exemplary embodiment of this disclosure, the distributed computing network resource scheduling module distributes the block rendering task to the target computing power node via the resource layer, including: the distributed computing network resource scheduling module synchronizing the target network transmission capability of the target computing power node to the network layer; the network layer evaluating the target network environment of the target computing power node based on the target network transmission capability and the network transmission capability requirements of the block rendering task, obtaining a target network environment evaluation result, and determining whether the target network transmission capability needs to be enhanced based on the target network environment evaluation result; if so, the network layer enhances the target network transmission capability and notifies the distributed computing network resource scheduling module to distribute the block rendering task; if not, it directly notifies the distributed computing network resource scheduling module to distribute the block rendering task; the distributed computing network resource scheduling module, based on the received notification, distributes the block rendering task to the target computing power node via the resource layer.

[0236] In one exemplary embodiment of this disclosure, the means for invoking computing power and network capabilities further includes:

[0237] The task execution result encapsulation module can be used to execute the block rendering task through the target computing power node, obtain the task execution result, and encapsulate the block identifier and the task execution result in the block rendering task to obtain the block rendering result.

[0238] The block rendering result feedback module can be used to feed back the block rendering result to the application layer through the target computing power node based on the first communication link;

[0239] The overall screen rendering result feedback module can be used to sequentially splice the segmented rendering results according to the segmentation identifier in the segmented rendering results through the application layer to obtain the overall screen rendering result of the graphics cloud application running on the cloud, and then feed the overall screen rendering result back to the terminal layer.

[0240] The specific details of each module in the aforementioned device for invoking computing and network capabilities for cloud VR services have been described in detail in the corresponding methods for invoking computing and network capabilities for cloud VR services, and will not be repeated here.

[0241] It should be noted that although several modules or units for the device used to perform actions have been mentioned in the detailed description above, this division is not mandatory. In fact, according to embodiments of this disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0242] Furthermore, although the steps of the method in this disclosure are described in a specific order in the accompanying drawings, this does not require or imply that the steps must be performed in that specific order, or that all the steps shown must be performed to achieve the desired result. Additional or alternative steps may be omitted, multiple steps may be combined into one step, and / or a step may be broken down into multiple steps.

[0243] In an exemplary embodiment of this disclosure, an electronic device capable of implementing the above-described method is also provided.

[0244] Those skilled in the art will understand that various aspects of this disclosure can be implemented as systems, methods, or program products. Therefore, various aspects of this disclosure can be specifically implemented in the following forms: entirely in hardware, entirely in software (including firmware, microcode, etc.), or in a combination of hardware and software, collectively referred to herein as “circuit,” “module,” or “system.”

[0245] The following reference Figure 10 To describe an electronic device 1000 according to such an embodiment of the present disclosure. Figure 10 The electronic device 1000 shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments disclosed herein.

[0246] like Figure 10 As shown, the electronic device 1000 is manifested in the form of a general-purpose computing device. The components of the electronic device 1000 may include, but are not limited to: at least one processing unit 1010, at least one storage unit 1020, a bus 1030 connecting different system components (including storage unit 1020 and processing unit 1010), and a display unit 1040.

[0247] The storage unit stores program code that can be executed by the processing unit 1010, causing the processing unit 1010 to perform the steps described in the "Exemplary Methods" section of this specification according to various exemplary embodiments of this disclosure. For example, the processing unit 1010 can perform actions such as... Figure 1Step S110: The optimization processing module of the control layer divides the screen area of ​​the graphics cloud application in the application layer into blocks according to the user attention area of ​​the graphics cloud application running on the cloud, and obtains the screen area block result, and obtains the block rendering request according to the screen area block result; Step S120: The distributed computing network resource scheduling module in the control layer responds to the block rendering request, parses the block rendering request, and obtains the screen area block result and the graphics rendering computing power requirement and network transmission capability requirement corresponding to the screen area block result; Step S130: The distributed computing network resource scheduling module matches the target computing power node in the resource layer for the block rendering task corresponding to the screen area block result according to the graphics rendering computing power requirement and network transmission capability requirement.

[0248] Storage unit 1020 may include readable media in the form of volatile storage units, such as random access memory (RAM) 10201 and / or cache memory 10202, and may further include read-only memory (ROM) 10203. Storage unit 1020 may also include a program / utility 10204 having a set (at least one) of program modules 10205, including but not limited to: an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment. Bus 1030 may represent one or more of several bus structures, including a storage unit bus or storage unit controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus using any of the various bus structures.

[0249] Electronic device 1000 can also communicate with one or more external devices 1100 (e.g., keyboard, pointing device, Bluetooth device, etc.), one or more devices that enable a user to interact with electronic device 1000, and / or any device that enables electronic device 1000 to communicate with one or more other computing devices (e.g., router, modem, etc.). This communication can be performed via input / output (I / O) interface 1050. Furthermore, electronic device 1000 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 1060. As shown, network adapter 1060 communicates with other modules of electronic device 1000 via bus 1030. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with electronic device 1000, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0250] From the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solutions according to the embodiments of this disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, external hard drive, etc.) or on a network, including several instructions to cause a computing device (such as a personal computer, server, terminal device, or network device, etc.) to execute the methods according to the embodiments of this disclosure.

[0251] In exemplary embodiments of this disclosure, a computer-readable storage medium is also provided, on which a program product capable of implementing the methods described above is stored. In some possible implementations, various aspects of this disclosure may also be implemented as a program product including program code that, when the program product is run on a terminal device, causes the terminal device to perform the steps of the various exemplary embodiments of this disclosure described in the "Exemplary Methods" section above.

[0252] The program product for implementing the above-described method according to embodiments of the present disclosure may employ a portable compact disc read-only memory (CD-ROM) and include program code, and may run on a terminal device, such as a personal computer. However, the program product of the present disclosure is not limited thereto. In this document, the readable storage medium may be any tangible medium containing or storing a program that may be used by or in conjunction with an instruction execution system, apparatus, or device.

[0253] The program product may employ any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0254] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A readable signal medium may also be any readable medium other than a readable storage medium, capable of sending, propagating, or transmitting programs for use by or in conjunction with an instruction execution system, apparatus, or device.

[0255] The program code contained on the readable medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, RF, etc., or any suitable combination thereof.

[0256] Program code for performing the operations of this disclosure can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java and C++, and conventional procedural programming languages ​​such as C or similar languages. The program code can execute entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0257] Furthermore, the above figures are merely illustrative of the processes included in the method according to exemplary embodiments of this disclosure and are not intended to be limiting. It is readily understood that the processes shown in the above figures do not indicate or limit the temporal order of these processes. Additionally, it is readily understood that these processes may be executed synchronously or asynchronously, for example, in multiple modules.

[0258] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the invention described herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not invented by this disclosure. The specification and embodiments are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the claims.

Claims

1. A method for invoking computing power and network capabilities for cloud VR services, characterized in that, include: The optimization processing module of the control layer divides the screen area of ​​the graphics cloud application in the application layer into blocks according to the user attention area of ​​the graphics cloud application running on the cloud, obtains the screen area block result, and obtains the block rendering request based on the screen area block result. The distributed computing network resource scheduling module in the control layer responds to the segmented rendering request, parses the segmented rendering request, and obtains the screen area segmentation result and the graphics rendering computing power requirement and network transmission capability requirement corresponding to the screen area segmentation result. The distributed computing network resource scheduling module matches the target computing power node in the resource layer with the block rendering task corresponding to the block rendering result of the screen area according to the graphics rendering computing power requirement and network transmission capability requirement.

2. The method for invoking computing power and network capabilities according to claim 1, characterized in that, Based on the user-focused areas of the cloud-based graphics application, the screen area of ​​the graphics application in the application layer is segmented to obtain the screen area segmentation results, including: The user attention area of ​​the graphical cloud application is determined, and based on the user attention area, the attention surrounding area, the peripheral area, and the temporarily unperceived area adjacent to the peripheral area are determined. Determine the required regional rendering precision and regional refresh rate for the graphics cloud application in the user's attention area, the surrounding area, the outer area, and the area that cannot be perceived temporarily. The overall screen size of the graphics cloud application's screen area is determined, and the screen area of ​​the graphics cloud application is divided into blocks according to the overall screen size and the area rendering precision and area refresh frequency required in different display areas, to obtain the screen area block result.

3. The method for invoking computing power and network capabilities according to claim 2, characterized in that, Determining the user's area of ​​interest in the graphics cloud application includes: Determine the field of view and / or the human eye gaze position of the cloud-based graphics application, and determine the user attention area based on the field of view and / or the human eye gaze position.

4. The method for invoking computing power and network capabilities according to claim 1, characterized in that, Based on the results of segmenting the image region, a segmented rendering request is obtained, including: Determine the graphics rendering computing power requirements required to execute the block rendering task corresponding to the block rendering result of the screen region, and the network transmission capability requirements required to transmit the block rendering result corresponding to the block rendering result of the screen region. Based on the image region segmentation results, graphics rendering computing power requirements, and network transmission capability requirements, a segmented rendering request corresponding to the image region segmentation results is generated.

5. The method for invoking computing power and network capabilities according to claim 1, characterized in that, Based on the graphics rendering computing power requirements and network transmission capability requirements, the target computing power node in the resource layer is matched for the block rendering task corresponding to the block rendering result of the screen region, including: The computational and network requirements of the segmented rendering task are correlated with the results of the segmented image region, the graphics rendering computing power requirements required to execute the segmented rendering task corresponding to the segmented image region, and the network transmission capability requirements required to transmit the segmented rendering result corresponding to the segmented image region, to obtain the computational and network requirements fusion result of the segmented rendering task. The alternative computing power capabilities and alternative network transmission capabilities of the alternative computing power nodes included in the resource layer are determined, and the alternative computing power nodes, alternative computing power capabilities, and alternative network transmission capabilities are fused to obtain the fusion result of the alternative computing power node capabilities. Based on the fusion results of the computing network requirements and the fusion results of the candidate node capabilities, a target computing node is matched for the block rendering task from the candidate computing nodes included in the resource layer.

6. The method for invoking computing power and network capabilities according to claim 5, characterized in that, The computational and network requirements for the segmented rendering task are correlated with the image region segmentation result, the graphics rendering computing power requirement required to execute the segmented rendering task corresponding to the image region segmentation result, and the network transmission capability requirement required to transmit the segmented rendering result corresponding to the image region segmentation result, to obtain the computational and network requirement fusion result of the segmented rendering task, including: Based on the position of the image region segmentation result in the image cloud application, the image region segmentation result is sorted to obtain the segmentation sorting result, and a segmentation identifier is configured for the image region segmentation result according to the segmentation sorting result. The image region segmentation result, segmentation identifier, graphics rendering computing power requirements required to execute the segmented rendering task corresponding to the image region segmentation result, and network transmission capability requirements required to transmit the segmented rendering result corresponding to the image region segmentation result are correlated to obtain the computing and network requirements fusion result of the segmented rendering task.

7. The method for invoking computing power and network capabilities according to claim 5, characterized in that, Based on the fusion results of the computing network requirements and the fusion results of the candidate node capabilities, a target computing node is matched for the block rendering task from the candidate computing nodes included in the resource layer, including: Based on the network transmission capacity requirements in the network demand fusion result, the target node capability fusion result is matched with the candidate node capability fusion result; Based on the graphics rendering computing power requirements in the network demand fusion result and the alternative computing power capabilities in the target node capability fusion result, a target computing power node is matched from the alternative computing power nodes for the block rendering task corresponding to the block division result of the screen area.

8. The method for invoking computing power and network capabilities according to claim 5, characterized in that, Determine the alternative computing power capabilities and alternative network transmission capabilities of the alternative computing power nodes included in the resource layer, including: The system receives node attribute information and remaining node computing power of candidate computing power nodes reported by the resource layer, and determines the candidate computing power capability of the candidate computing power nodes based on the node attribute information and remaining node computing power; wherein, the node attribute information includes at least one of chip model, number of chips, memory size and computing power. The system receives the current node IP address and original egress bandwidth resources of the candidate computing power nodes reported by the resource layer, and determines the candidate network transmission capabilities of the candidate computing power nodes based on the current node IP address and original egress bandwidth resources.

9. The method for invoking computing power and network capabilities according to claim 8, characterized in that, Based on the current node IP address and the original outbound bandwidth resources, determine the alternative network transmission capabilities of the candidate computing power nodes, including: The computing power of the candidate computing power nodes is evaluated to obtain the computing power evaluation results, and it is determined whether the candidate computing power nodes can be used as rendering task execution nodes based on the computing power evaluation results. If the candidate computing power node can serve as the rendering task execution node, then the current node IP address and original outbound bandwidth resources of the candidate computing power node are synchronized to the network layer, and the candidate network transmission capabilities corresponding to the candidate computing power node are received from the network layer.

10. The method for invoking computing power and network capabilities according to claim 9, characterized in that, The alternative network transmission capabilities are obtained by the network layer in the following manner: Based on the current node IP address and the original outbound bandwidth resources, the current network environment of the candidate computing power node is evaluated to obtain the current network environment evaluation result; wherein, the current network environment evaluation result includes at least one of the current link latency, the current available outbound bandwidth of the network, and the current network transmission stability; Based on the current network environment assessment results, determine whether the current bandwidth resources and / or current transmission links of the candidate computing power nodes need to be enhanced; If the current bandwidth resources and / or the current transmission links need to be enhanced, the current bandwidth resources are accelerated, and / or the current transmission links are optimized, and the alternative network transmission capabilities are determined based on the accelerated current bandwidth resources and / or the optimized current transmission links.

11. The method for invoking computing power and network capabilities according to claim 1, characterized in that, The methods for invoking computing power and network capabilities also include: The distributed computing network resource scheduling module sends the target node IP address of the target computing power node to the application layer; The application layer performs node addressing on the target computing power node based on the target node's IP address, thereby establishing a first communication link between the application layer and the target computing power node.

12. The method for invoking computing power and network capabilities according to claim 11, characterized in that, The methods for invoking computing power and network capabilities also include: The distributed computing network resource scheduling module generates a segmented rendering task based on the screen region segmentation result, the segmentation identifier of the screen region segmentation result, the graphics rendering computing power requirement required to execute the segmented rendering task corresponding to the screen region segmentation result, the network transmission capability required to transmit the segmented rendering result corresponding to the screen region segmentation result, the target computing node, and the target computing power capability of the target computing node, and then distributes the segmented rendering task to the target computing node through the resource layer.

13. The method for invoking computing power and network capabilities according to claim 12, characterized in that, The distributed computing network resource scheduling module distributes the block rendering tasks to the target computing power nodes via the resource layer, including: The distributed computing network resource scheduling module synchronizes the target network transmission capability of the target computing power node to the network layer; The network layer evaluates the target network environment of the target computing node based on the target network transmission capability and the network transmission capability requirements of the chunked rendering task, obtains the target network environment evaluation result, and determines whether the target network transmission capability needs to be enhanced based on the target network environment evaluation result. If so, the network layer enhances the target network transmission capability and notifies the distributed computing network resource scheduling module to issue the segmented rendering task; if not, it directly notifies the distributed computing network resource scheduling module to issue the segmented rendering task. Based on the received notification, the distributed computing network resource scheduling module distributes the block rendering task to the target computing power node via the resource layer.

14. The method for invoking computing power and network capabilities according to claim 12, characterized in that, The methods for invoking computing power and network capabilities also include: The target computing node executes the segmented rendering task, obtains the task execution result, and encapsulates the segment identifier and the task execution result in the segmented rendering task to obtain the segmented rendering result. The target computing node feeds back the block rendering result to the application layer based on the first communication link; The application layer sequentially splices the segmented rendering results according to the segmentation identifiers in the segmented rendering results to obtain the overall screen rendering result of the cloud-based graphics application, and then feeds the overall screen rendering result back to the terminal layer.

15. A system for invoking computing power and network capabilities for cloud VR services, characterized in that, It includes a control layer, an application layer, and a resource layer. The control layer includes an optimization processing module and a distributed computing network resource scheduling module; wherein: The optimization processing module is used to divide the screen area of ​​the graphics cloud application in the application layer into blocks according to the user attention area of ​​the graphics cloud application running in the cloud, to obtain the screen area block result, and to obtain the block rendering request based on the screen area block result. The distributed computing network resource scheduling module is used to respond to the segmented rendering request, parse the segmented rendering request, obtain the screen region segmentation result, and the corresponding graphics rendering computing power requirement and network transmission capacity requirement; and Based on the graphics rendering computing power requirements and network transmission capability requirements, target computing power nodes in the resource layer are matched for the block rendering tasks corresponding to the block rendering results of the screen area.

16. The computing power and network capability invocation system according to claim 15, characterized in that, The system for accessing computing and network capabilities also includes: The network layer, which is communicatively connected to the control layer, is used to evaluate the target network environment of the target computing power node and determine whether the target network transmission capability of the target computing power node needs to be enhanced based on the evaluation results.

17. The computing power and network capability invocation system according to claim 15, characterized in that, The resource layer includes multiple computing power nodes, which are used to execute the block rendering tasks issued by the distributed computing network resource scheduling module, obtain the block rendering results, and feed the block rendering results back to the application layer.

18. The computing power and network capability invocation system according to claim 17, characterized in that, The system for accessing computing and network capabilities also includes: The terminal layer, communicating with the application layer, is used to request the display screen of a cloud-based graphical application from the application layer; and Receive the overall screen display result corresponding to the display screen from the application layer, and display the overall screen display result.

19. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the method for invoking computing power and network capabilities for cloud VR services as described in any one of claims 1-14.

20. An electronic device, characterized in that, include: processor; as well as Memory for storing the executable instructions of the processor; The processor is configured to execute the method for invoking computing and network capabilities for cloud VR services as described in any one of claims 1-14 by executing the executable instructions.

Images