Network computing power energy sensing network, resource measurement method, device and program product

Abstract

The invention provides a network computing power energy sensing network, a resource measurement method, a resource measurement device and a program product. The method comprises the following steps: determining a current available resource corresponding to an infrastructure through a resource sensing unit; before the calculation task is executed, resources needed for executing the calculation task are measured through a measurement device based on the task type, the task scene and the task load corresponding to the calculation task, target resources are obtained, and resource configuration is carried out according to the target resources and resources corresponding to the infrastructure; in the execution process of the calculation task, generating predicted available resources corresponding to the infrastructure in the target period through a measurement device based on the current time feature and the task feature corresponding to the calculation task; performing resource configuration according to the predicted available resource, the current available resource and the target resource; the problem that a single resource measurement index is difficult to meet the requirements of task types and task scene complexity can be solved; and the requirements of task type and task scene complication can be comprehensively and accurately met.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a network computing power energy-aware network, a resource measurement method, a device, and a program product. Background Art

[0002] With the emergence of emerging services such as AR / VR, autonomous driving / Internet of Vehicles, telemedicine, and large artificial intelligence models, these computing-intensive and delay-sensitive services have put forward higher requirements on the computing and data transmission capabilities of the network. In this context, computing networks that integrate computing and storage capabilities within the network have developed rapidly. Computing networks have received great attention in China because they can connect and coordinate distributed computing nodes, optimize and efficiently utilize network and computing resources through improvements in network architecture and protocols. In response to the problem of difficulty in coordinating computing resources, computing networks rely on high-speed, mobile, secure, and ubiquitous network connections to integrate multi-level computing resources such as cloud, edge, and end. It is a digital information infrastructure that provides integrated services such as data perception, transmission, storage, and computing. At the same time, computing networks need to build a complete user service platform, perform efficient computing network resource allocation and flexible task scheduling, make full use of heterogeneous computing resources in computing networks, and better serve users. This requires promoting the design and construction of computing networks in terms of resource perception and task measurement mechanisms.

[0003] In the traditional computing power network perception network, it includes: first, periodically collecting the computing resources, storage resources and memory resources of the CPU and GPU to form computing power resources; building the topological connection relationship between routing nodes and computing nodes, and periodically perceiving the end-to-end transmission delay, link available bandwidth and packet loss rate in the network topology as three indicators for measuring network resources.

[0004] However, with the diversification and complexity of computing network service demands, the single resource measurement indicators of traditional computing network perception networks are difficult to meet the increasing demands of computing task types and task scenarios. Summary of the invention

[0005] In view of this, embodiments of the present invention provide a network computing power energy-aware network, a resource measurement method, an apparatus, and a program product to eliminate or improve one or more defects existing in the prior art.

[0006] One aspect of the present invention provides a network computing power energy perception network, comprising: a service application layer, a perception prediction layer and an infrastructure layer; the service application layer is used to perform computing tasks; the infrastructure layer includes multiple infrastructures, the infrastructure includes computing power infrastructure, network infrastructure and energy infrastructure; the perception prediction layer includes a measurement device and multiple resource perception units;

[0007] The resource sensing unit is arranged in the infrastructure, and is used to: monitor in real time the computing power resources corresponding to the computing power infrastructure, the network resources corresponding to the network infrastructure, and the energy consumption resources corresponding to the energy infrastructure, and determine the currently available resources corresponding to the infrastructure;

[0008] The measuring device is used to: before the execution of the computing task, based on the task type, task scenario and task load corresponding to the computing task, measure the resources required for executing the computing task, obtain the target resources, and configure resources according to the target resources and the resources corresponding to the infrastructure; during the execution of the computing task, based on the current time characteristics and the task characteristics corresponding to the computing task, generate the predicted available resources corresponding to the infrastructure within the target period; and configure resources according to the predicted available resources, the current available resources and the target resources.

[0009] In some embodiments of the present invention, the target resources include target network resources, target computing power resources, and target energy consumption resources; based on the task type, task scenario, and task load corresponding to the computing task, the resources required to perform the computing task are measured to obtain the target resources, including:

[0010] Obtaining a pre-established neural network fitting model; the neural network fitting model includes a computing network resource measurement model and an energy consumption resource measurement model;

[0011] Input the task type, task scenario and task load into the computing network resource measurement model, and output the target network resources and target computing power resources;

[0012] The target network resources and target computing power resources are input into the energy consumption resource measurement model, and the target energy consumption resources are output.

[0013] In some embodiments of the present invention, the target computing power resources also include fine-grained computing power features obtained based on analysis of task scenarios.

[0014] In some embodiments of the present invention, based on the current time characteristics and the task characteristics corresponding to the computing task, generating the predicted available resources corresponding to the infrastructure within the target period includes:

[0015] Based on the preset mapping relationship between the features and the forecast resources, according to the current time features and the task features, in the forecast resources, determine the target forecast resources corresponding to the current time features and the task features;

[0016] The target forecast resource is taken as the forecast available resource.

[0017] In some embodiments of the present invention, the task characteristics include energy characteristics, regional characteristics or resource characteristics corresponding to the computing task.

[0018] In some embodiments of the present invention, the measurement device is further used to: evaluate the execution progress of the computing task and the current resources corresponding to the computing task to obtain an evaluation result; and configure resources corresponding to the infrastructure based on the evaluation result.

[0019] Another aspect of the present invention provides a resource measurement method for a network computing power energy-aware network, which is applied to a network computing power energy-aware network. The network computing power energy-aware network includes a service application layer, a perception prediction layer, and an infrastructure layer; the service application layer is used to perform computing tasks; the infrastructure layer includes multiple infrastructures, the infrastructure includes computing power infrastructure, network infrastructure, and energy infrastructure; the perception prediction layer includes a measurement device and multiple resource perception units, and the resource perception units are arranged in the infrastructure; the method includes:

[0020] The resource sensing unit monitors the computing resources corresponding to the computing infrastructure, the network resources corresponding to the network infrastructure, and the energy consumption resources corresponding to the energy infrastructure in real time to determine the current available resources corresponding to the infrastructure;

[0021] Before the computing task is executed, the resources required to execute the computing task are measured by the measuring device based on the task type, task scenario and task load corresponding to the computing task, the target resources are obtained, and the resources are configured according to the target resources and the resources corresponding to the infrastructure;

[0022] During the execution of the computing task, the predicted available resources corresponding to the infrastructure within the target period are generated through the measurement device based on the current time characteristics and the task characteristics corresponding to the computing task; and resource allocation is performed according to the predicted available resources, current available resources and target resources.

[0023] Another aspect of the present invention provides a resource measurement device for a network computing power energy-aware network, the device comprising: a processor, a memory, and a computer program / instructions stored in the memory, the processor being used to execute the computer program / instructions, and when the computer program / instructions are executed, the device implements the steps of the above method.

[0024] Another aspect of the present invention provides a computer-readable storage medium having a computer program / instruction stored thereon, which implements the steps of the above method when executed by a processor.

[0025] Another aspect of the present invention provides a computer program product, comprising a computer program / instruction, which implements the steps of the above method when executed by a processor.

[0026] The network computing power energy-aware network, resource measurement method, device and program product of the present invention monitor the computing power resources corresponding to the computing power infrastructure, the network resources corresponding to the network infrastructure and the energy consumption resources corresponding to the energy infrastructure in real time through the resource sensing unit, and determine the current available resources corresponding to the infrastructure; before the computing task is executed, the measurement device measures the resources required for executing the computing task based on the task type, task scenario and task load corresponding to the computing task, obtains the target resources, and configures the resources according to the target resources and the resources corresponding to the infrastructure; during the execution of the computing task, the measurement device generates the predicted available resources corresponding to the infrastructure within the target period based on the current time characteristics and the task characteristics corresponding to the computing task; and configures the resources according to the predicted available resources, the current available resources and the resources; it can solve the problem of single resource allocation. The source measurement index is difficult to meet the needs of the complexity of computing task types and task scenarios; through multi-dimensional resource perception, task feature measurement and resource configuration, and dynamic resource prediction and configuration, it can more comprehensively and accurately meet the needs of the complexity of computing task types and task scenarios, thereby improving resource utilization and system performance, and can solve the problem that a single resource measurement index is difficult to meet the needs of the complexity of computing task types and task scenarios; through continuous monitoring by the resource perception unit, the current available resources corresponding to the infrastructure are determined, and the information corresponding to the current available resources corresponding to the infrastructure is sent to the measurement device, so that the measurement device can configure the resources provided by the foundation before and during the execution of the computing task, and can efficiently utilize various resources, thereby realizing real-time and rapid scheduling of resources and improving user experience. At the same time, by instantly understanding the status of computing resources, the availability of network resources, and the use of energy consumption resources, it is possible to optimize resource allocation to ensure that the task load requirements of different computing tasks can be met at any time, improve the flexibility of the system, and provide users with a more stable and efficient service experience.

[0027] In addition, by measuring the resources required for computing tasks through a computing task-oriented resource measurement framework, the matching between computing tasks and resources is made more accurate, thereby improving resource utilization and the efficiency of resource transactions. At the same time, through a computing task-oriented resource measurement framework, the degree of resource demand of computing tasks can be better evaluated, and resources can be dynamically adjusted and allocated according to actual conditions, thereby improving resource utilization efficiency and the overall performance of the system.

[0028] In addition, using algorithms and models to predict the supply and demand of resources within future target cycles, so as to make corresponding resource allocation and trading decisions in advance, can provide users with more accurate resource allocation and trading decision-making basis, enabling them to better grasp market trends, reduce trading risks, and optimize resource utilization efficiency. At the same time, it improves the intelligence and flexibility of the entire system.

[0029] Additional advantages, purposes, and features of the present invention will be described in part in the following description, and will become apparent to those skilled in the art after studying the following, or may be learned from the practice of the present invention. The purposes and other advantages of the present invention may be achieved and obtained by the structures specifically indicated in the specification and the accompanying drawings.

[0030] Those skilled in the art will appreciate that the objectives and advantages that can be achieved with the present invention are not limited to the above specific description, and the above and other objectives that can be achieved by the present invention will be more clearly understood from the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The drawings described herein are used to provide a further understanding of the present invention, constitute a part of the present application, and do not constitute a limitation of the present invention. In the drawings:

[0032] Figure 1 A schematic diagram of the architecture of a network computing power energy-aware network provided in one embodiment of the present invention;

[0033] Figure 2 A resource measurement framework for a network computing power energy-aware network in one embodiment of the present invention;

[0034] Figure 3 A flowchart of a method for measuring resources in a network computing power energy-aware network according to an embodiment of the present invention;

[0035] Figure 4 A block diagram of a resource measurement device for a network computing power energy-aware network provided in another embodiment of the present invention. DETAILED DESCRIPTION

[0036] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments and the accompanying drawings. Here, the illustrative embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.

[0037] It should also be noted that, in order to avoid obscuring the present invention due to unnecessary details, only structures and / or processing steps closely related to the solutions according to the present invention are shown in the accompanying drawings, while other details that are not closely related to the present invention are omitted.

[0038] It should be emphasized that the term “include / comprises” when used herein refers to the presence of features, elements, steps or components, but does not exclude the presence or addition of one or more other features, elements, steps or components.

[0039] It should also be noted that, unless otherwise specified, the term “connection” herein may refer not only to a direct connection but also to an indirect connection with an intermediate.

[0040] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. In the accompanying drawings, the same reference numerals represent the same or similar components, or the same or similar steps.

[0041] The architecture of the network computing power energy-aware network provided in the embodiment of the present application is explained in detail below.

[0042] Figure 1 Schematic diagram of the architecture of a network computing power energy-aware network provided in an embodiment of the present application. Figure 1 As shown, the network computing power energy-aware network includes a service application layer 100, an infrastructure layer 200 and a perception prediction layer 300.

[0043] The service application layer 100 is used to perform computing tasks. In the service application layer 100, users submit computing tasks or requests by accessing the system. The system will perform corresponding calculations and processing according to the user's needs, and return results or provide required services. For example, Figure 1 ,Users include computing power users, network users and energy users. ,Different types of computing tasks or requests submitted by users have different ,demands for computing power resources, network resources and energy resources.

[0044] In some embodiments of the present invention, different types of computing tasks have the same or different requirements for resources. Among them, computing tasks include but are not limited to artificial intelligence task types, big data task types, and image recognition, video processing and streaming media processing tasks. The scale of processing data of computing tasks is measured by task load. At the same time, the task quality assurance of different computing tasks, such as latency, accuracy, latency jitter, additional reliability requirements, etc., will also affect the resources required for computing tasks.

[0045] The infrastructure layer 200 includes multiple infrastructures, which are used to provide corresponding computing resources, network resources, and energy resources for computing tasks or requests submitted by users. Figure 1 As shown, the infrastructure includes computing infrastructure, network infrastructure and energy infrastructure.

[0046] In some embodiments of the present invention, the computing power infrastructure and the network infrastructure include a data center, a computing power center, and an interactive node.

[0047] Data centers include servers, computing devices, storage devices, and network devices, which are used to centrally manage and process data to meet the needs of various types of users, such as organizations or enterprises. They play a key role in supporting the information technology infrastructure and business operations of enterprises. At the same time, data centers can host various types of data, including user data, enterprise data, application data, etc.

[0048] Computing centers include servers, computing devices, storage devices, and network devices to provide computing resources. Computing centers are usually designed to be able to carry large-scale computing tasks and provide high performance and reliability. Computing centers have a wide range of applications in various fields, including scientific research, engineering design, data analysis, artificial intelligence, etc.

[0049] Interaction nodes include devices such as routers, switches, load balancers, or network optics, which are used for data exchange, communication, and coordination between different parts.

[0050] In some embodiments of the present invention, the energy infrastructure includes an energy center. The energy center is used to meet the energy needs of computing tasks and ensure the safety, reliability and efficiency of energy supply. In one example, the energy center is a power generation center, including various energy generation facilities, such as coal-fired power plants, wind power stations, solar panel arrays, etc.

[0051] It should be noted that the number of infrastructures in the infrastructure layer 200, including computing power infrastructure, network infrastructure and energy infrastructure, can be any number, and this embodiment does not limit the number of infrastructures.

[0052] In some embodiments of the present invention, the perception prediction layer 300 includes a measurement device and a plurality of resource perception units.

[0053] The resource sensing unit is set in the infrastructure to monitor the available resources in the infrastructure in real time and send them to the measurement device. The resource sensing unit includes a computing power resource sensing unit set in the computing power infrastructure, a network resource sensing unit set in the network infrastructure, and an energy consumption resource sensing unit set in the energy infrastructure.

[0054] For example: In the intelligent manufacturing scenario, it is necessary to monitor various resources on the factory production line, including equipment operating status, energy consumption, output, etc. By installing resource sensing units in the factory equipment, various resources of the factory equipment (including equipment operating status, energy consumption and output, etc.) can be monitored.

[0055] By installing sensors on equipment, monitoring various indicators and sending data to a central server for analysis, real-time monitoring and management of resources can be achieved. For example, sensors on a factory production line can be used to monitor the operating status, energy consumption, and output of equipment, and the data can be transmitted to a central server through IoT technology for real-time monitoring and analysis.

[0056] Specifically, the resource sensing unit is used to: monitor in real time the computing resources corresponding to the computing infrastructure, the network resources corresponding to the network infrastructure, and the energy consumption resources corresponding to the energy infrastructure, and determine the currently available resources corresponding to the infrastructure.

[0057] In some embodiments of the present invention, the resource sensing unit continuously monitors and determines the currently available resources corresponding to the infrastructure, and sends the information corresponding to the currently available resources corresponding to the infrastructure to the measurement device, so that the measurement device can configure the resources provided by the infrastructure before and during the execution of the computing task, and can efficiently utilize various resources, thereby realizing real-time and rapid scheduling of resources and improving user experience. At the same time, by instantly understanding the status of computing resources, the availability of network resources, and the use of energy resources, it is possible to optimize resource allocation and ensure that the task load requirements of different computing tasks can be met at any time, thereby improving the flexibility of the system and providing users with a more stable and efficient service experience.

[0058] In some embodiments of the present invention, for any computing task i, a computing task triple (c i ,s i ,d i ), where c i Indicates the task type, s i Represents task scenarios, such as image recognition, model inference training, etc. i Represents the task load, that is, the amount of data for the computing task.

[0059] Before executing a computing task, it is necessary to determine the computing power resources, network resources, and energy resources required for this computing task based on the computing task triplet. This can help with effective resource planning and allocation, optimize task performance, control costs, avoid resource waste, and improve resource efficiency, thereby ensuring that the computing task can be executed smoothly and achieve the expected results.

[0060] Specifically, the measuring device is used to: before the computing task is executed, based on the task type, task scenario and task load corresponding to the computing task, measure the resources required to execute the computing task, obtain the target resources, and configure the resources according to the target resources and the resources corresponding to the infrastructure. The target resources include target network resources, target computing power resources and target energy consumption resources.

[0061] Since the computing tasks in different task scenarios require different computing resources and network resources, in order to accurately represent the computing resources, network resources and energy resources required for different computing tasks, such as Figure 2 As shown, the present invention proposes to build a resource measurement framework for computing tasks based on knowledge graphs, including computing power resource measurement based on a computing network resource measurement model, network resource measurement, and energy consumption resource measurement based on an energy consumption resource measurement model.

[0062] Among them, the computing network resource measurement model is used to generate target network resources and target computing power resources based on task type, task scenario and task load; the energy consumption resource measurement model is used to generate target energy consumption resources based on target network resources and target computing power resources.

[0063] Specifically, based on the task type, task scenario and task load corresponding to the computing task, the resources required to execute the computing task are measured to obtain the target resources, including: obtaining a pre-established neural network fitting model; the neural network fitting model includes a computing network resource measurement model and an energy consumption resource measurement model; the task type, task scenario and task load are input into the computing network resource measurement model, and the target network resources and target computing power resources are output; the target network resources and target computing power resources are input into the energy consumption resource measurement model, and the target energy consumption resources are output.

[0064] Among them, the computing network resource measurement model is a task service model constructed by fitting and analyzing computing tasks and task scenarios, task types and task loads. It outputs the corresponding computing power resources and network resources for the computing task triples corresponding to different computing tasks to achieve universal and accurate measurement; the energy consumption resource measurement model is a task service model constructed by fitting and analyzing network resources and computing power resources. It outputs the corresponding energy consumption resources for different computing power resources and network resources to achieve universal and accurate measurement.

[0065] In some embodiments of the present invention, the computing network resource measurement model and the energy consumption resource measurement model are based on the output of the Light Gradient Boosting Machine (LightGBM), which is trained as the input of the preset neural network model. Since the LightGBM tree model has its own processing of missing and categorical variables, and has strong nonlinear fitting capabilities, there is no need to do too much processing in feature engineering, and modeling is convenient. The path of the leaf node of the tree model is extracted as a feature, and after the feature selection process is completed, the original features and the tree node path features are spliced ​​and input into the preset neural network model, and the neural network model is tuned using grid search, which can improve the accuracy of the trained model. Among them, the preset neural network model includes but is not limited to convolutional neural network (CNN), recurrent neural network (RNN), long short-term memory network (LSTM), etc.

[0066] Specifically, the training process of the computing network resource measurement model includes: obtaining training sample data, including sample task scenarios, sample task types, sample task loads, and corresponding sample computing resources and sample network resources; using the training sample data to train the preset neural network model until the preset neural network model converges to obtain the computing network resource measurement model.

[0067] The training process of the energy consumption resource measurement model includes: obtaining training sample data, including sample computing resources, sample network resources and sample energy consumption resources; using the training sample data to train a preset neural network model until the preset neural network model converges to obtain an energy consumption resource measurement model.

[0068] In actual implementation, the computing network resource measurement model and the energy consumption resource measurement model may also be models trained by machine learning methods, such as linear regression, decision tree, random forest, etc. This embodiment does not limit the implementation method of the computing network resource measurement model and the energy consumption resource measurement model.

[0069] In some embodiments of the present invention, computing power resource measurement, network resource measurement and energy consumption resource measurement are reflected in two processing processes of resource measurement tasks, namely, the quantification process from computing tasks to computing network resources and the quantification process from computing power resources and network resources to energy consumption resources. The two quantification processes are introduced below:

[0070] The first is the quantification process from computing tasks to computing resources and network resources.

[0071] In some embodiments of the present invention, the quantification process from computing tasks to computing power resources and network resources, that is, when the user side generates a computing task with computing power resource requirements and network resource requirements, it is necessary to quantify the computing power resources and network resources required for the computing task in order to provide users with more stable and suitable computing power services and network services.

[0072] In the process of quantifying tasks to computing network resources, the measurement device will evaluate the computing resource level and network resource level required for this computing task. Among them, the computing resource level includes the task type, task scenario and task load; the network resource level uses the network bandwidth required for the computing task as the main measurement indicator, which affects the communication and transfer of user data and services.

[0073] like Figure 2 As shown, in some embodiments of the present invention, the computing network resource measurement model predicts the computing power resources required for this computing task according to the task type, task scenario and task load. The unit of computing power resources is FLOPS. The computing power resource measurement definition model can be expressed as:

[0074]

[0075] Among them, t n ∈T is the task type of computing task n (such as computing-intensive, WEB service and I / O-intensive service tasks, etc.), For the task scenario of computing task n, Load n is the task load of computing task n.

[0076] In addition, the target computing power resources also include fine-grained computing power features obtained based on the analysis of the task scenario. Among them, fine-grained computing power features refer to the detailed description and division of computing power resources in order to more accurately monitor and manage computing power resources. In fields such as cloud computing, big data processing, and artificial intelligence, fine-grained computing power features can provide more flexible resource scheduling and utilization methods, thereby optimizing computing performance and cost-effectiveness. Fine-grained computing power features include but are not limited to the number of cores of the processor, the clock frequency of the processor, the cache size of the processor, or the memory size of the processor.

[0077] The network resource measurement definition model can be expressed as:

[0078]

[0079] Among them, t n ∈T is the task type of computing task n, For the task scenario of computing task n, Load n is the task load of computing task n.

[0080] After determining the computing power resources and network resources corresponding to this computing task, the computing power resources and network resources corresponding to this computing task are used as the basis through the measurement device, and the computing power resources and network resources are configured in combination with the user service level requirements. At the same time, energy consumption resources are measured based on the computing power resources and network resources corresponding to this computing task.

[0081] The second is the quantification process from computing resources, network resources to energy consumption resources.

[0082] In the quantification process from computing power resources, network resources to energy consumption resources, the measurement device outputs the energy consumption resources required to execute the computing task according to the required computing power resources and network resources and the energy consumption resource measurement model, including the operating energy consumption resources from computing power hardware, network hardware, software operating energy consumption resources and energy consumption resources for data center operation and maintenance.

[0083] Since the load of computing tasks may fluctuate over time, for example, a computing task may require more computing resources at the beginning, but as the task progresses, the demand may decrease. Therefore, during the execution of the computing task, it is necessary to adjust the computing resources, network resources and energy resources in real time so that the system can respond to the changing task load and environmental conditions more flexibly and efficiently, thereby improving resource utilization, reducing costs, optimizing performance and saving energy and protecting the environment.

[0084] In order to solve the above technical problem, in some embodiments of the present invention, the measurement device is also used to predict the available resources of the infrastructure in the future target period according to the task characteristics of the current computing task.

[0085] Specifically, the measurement device is used to: during the execution of the computing task, generate predicted available resources corresponding to the infrastructure within the target period based on the current time characteristics and the task characteristics corresponding to the computing task; and perform resource allocation according to the predicted available resources, current available resources and resources.

[0086] In order to predict the time series changes of computing power resources, network resources and energy consumption resources in the future target period, in some embodiments of the present invention, a mapping relationship is established with the current time series characteristics and the task characteristics of the computing task as input data and the predicted computing power resources, network resources and energy consumption resources as output data. Among them, the task characteristics include the energy characteristics, regional characteristics and resource characteristics corresponding to the computing task.

[0087] Among them, the current time series characteristics refer to the time characteristics of the currently monitored data, and the time accuracy of the target period corresponding to the predicted output data is to xx:xx / xx / xx; energy characteristics include wind energy, light energy, non-green electricity supply or a combination of renewable energy (green electricity) and non-renewable energy supply; geographical characteristics refer to the geographical location or region of the infrastructure node; resource characteristics refer to the available computing resources, network resources and energy consumption resources currently monitored by the infrastructure.

[0088] Specifically, based on the current time characteristics and the task characteristics corresponding to the computing task, the predicted available resources corresponding to the infrastructure within the target period are generated, including: based on a preset mapping relationship between the characteristics and the predicted resources, according to the current time characteristics and the task characteristics, in the predicted resources, the target predicted resources corresponding to the current time characteristics and the task characteristics are determined; and the target predicted resources are used as the predicted available resources.

[0089] In actual implementation, in order to achieve the best computing task execution efficiency and resource utilization, during the execution of the computing task, the measurement device also needs to evaluate the computing task based on the task progress corresponding to the computing task to obtain the evaluation result. The evaluation result is used to indicate the task completion time, resource demand adjustment, performance optimization and risk warning, etc. According to the evaluation result, the resource configuration corresponding to the infrastructure is adjusted to optimize the execution efficiency and resource utilization corresponding to the computing task. For example: in the intelligent manufacturing scenario, the production plan needs to be adjusted in real time according to the production order and equipment status. In order to achieve the best production efficiency and resource utilization, ensure that the equipment on the production line is fully utilized and meets the production needs, it is necessary to optimize the resource allocation in real time according to the current production efficiency and resource utilization.

[0090] Specifically, the measurement device is also used to: evaluate the execution progress of the computing task and the current resources corresponding to the computing task to obtain an evaluation result; and configure resources corresponding to the infrastructure based on the evaluation result.

[0091] The resource measurement method of the network computing power energy-aware network provided in the embodiment of the present application is explained in detail below.

[0092] Figure 3The present invention is a flowchart of a method for measuring resources in a network computing power energy-aware network according to an embodiment of the present invention. The method for measuring resources in a network computing power energy-aware network is applied to a network computing power energy-aware network, which includes a service application layer, a perception prediction layer, and an infrastructure layer; the service application layer is used to perform computing tasks; the infrastructure layer includes multiple infrastructures, and the infrastructure includes computing power infrastructure, network infrastructure, and energy infrastructure; the perception prediction layer includes a measurement device and multiple resource perception units, and the resource perception units are arranged in the infrastructure; the method for measuring resources in a network computing power energy-aware network may include the following steps S101 to S103:

[0093] Step S101, monitor the computing resources corresponding to the computing infrastructure, the network resources corresponding to the network infrastructure, and the energy consumption resources corresponding to the energy infrastructure in real time through the resource sensing unit to determine the current available resources corresponding to the infrastructure.

[0094] Step S102, before the computing task is executed, the resources required to execute the computing task are measured by a measuring device based on the task type, task scenario and task load corresponding to the computing task, the target resources are obtained, and the resources are configured according to the target resources and the resources corresponding to the infrastructure.

[0095] Step S103, during the execution of the computing task, the predicted available resources corresponding to the infrastructure within the target period are generated by the measurement device based on the current time characteristics and the task characteristics corresponding to the computing task; and resource allocation is performed according to the predicted available resources, current available resources and resources.

[0096] In summary, the network computing power energy-aware network and resource measurement method monitors the computing power resources corresponding to the computing power infrastructure, the network resources corresponding to the network infrastructure, and the energy consumption resources corresponding to the energy infrastructure in real time through the resource sensing unit, and determines the current available resources corresponding to the infrastructure; before the computing task is executed, the resources required to execute the computing task are measured based on the task type, task scenario and task load corresponding to the computing task through the measurement device, the target resources are obtained, and the resources are configured according to the target resources and the resources corresponding to the infrastructure; during the execution of the computing task, the predicted available resources corresponding to the infrastructure within the target period are generated through the measurement device based on the current time characteristics and the task characteristics corresponding to the computing task; the resources are configured according to the predicted available resources, the current available resources and the resources; it can solve the problem of a single resource measurement indicator. The problem that the standard is difficult to meet the needs of the complexity of computing task types and task scenarios; through multi-dimensional resource perception, task feature measurement and resource allocation, and dynamic resource prediction and configuration, it can more comprehensively and accurately meet the needs of the complexity of computing task types and task scenarios, thereby improving resource utilization and system performance, and can solve the problem that a single resource measurement indicator is difficult to meet the needs of the complexity of computing task types and task scenarios; through continuous monitoring by the resource perception unit, the current available resources corresponding to the infrastructure are determined, and the information corresponding to the current available resources corresponding to the infrastructure is sent to the measurement device, so that the measurement device can configure the resources provided by the foundation before and during the execution of the computing task, and can efficiently utilize various resources, thereby realizing real-time and rapid scheduling of resources and improving user experience. At the same time, by instantly understanding the status of computing resources, the availability of network resources, and the use of energy resources, it is possible to optimize resource allocation to ensure that the task load requirements of different computing tasks can be met at any time, improve the flexibility of the system, and also provide users with a more stable and efficient service experience.

[0097] In addition, by measuring the resources required for computing tasks through a computing task-oriented resource measurement framework, the matching between computing tasks and resources is made more accurate, thereby improving resource utilization and the efficiency of resource transactions. At the same time, through a computing task-oriented resource measurement framework, the degree of resource demand of computing tasks can be better evaluated, and resources can be dynamically adjusted and allocated according to actual conditions, thereby improving resource utilization efficiency and the overall performance of the system.

[0098] In addition, using algorithms and models to predict the supply and demand of resources within future target cycles, so as to make corresponding resource allocation and trading decisions in advance, can provide users with more accurate resource allocation and trading decision-making basis, enabling them to better grasp market trends, reduce trading risks, and optimize resource utilization efficiency. At the same time, it improves the intelligence and flexibility of the entire system.

[0099] This embodiment provides a resource measurement device for a network computing power energy-aware network, such as Figure 4 The electronic device at least includes a processor 401 and a memory 402 .

[0100] The processor 401 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 401 may be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). The processor 401 may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor 401 may be integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 401 may also include an AI (Artificial Intelligence) processor, which is used to process computing operations related to machine learning.

[0101] The memory 402 may include one or more computer-readable storage media, which may be non-transitory. The memory 402 may also include a high-speed random access memory, and a non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 402 is used to store at least one instruction, which is used to be executed by the processor 401 to implement the resource measurement method of the network computing power energy-aware network provided in the method embodiment of the present application.

[0102] In some embodiments, the electronic device may further optionally include: a peripheral device interface and at least one peripheral device. The processor 401, the memory 402 and the peripheral device interface may be connected via a bus or a signal line. Each peripheral device may be connected to the peripheral device interface via a bus, a signal line or a circuit board. Schematically, the peripheral devices include but are not limited to: a radio frequency circuit, a touch display screen, an audio circuit, and a power supply.

[0103] Of course, the resource measurement device of the network computing power energy-aware network may also include fewer or more components, which is not limited in this embodiment.

[0104] The embodiment of the present invention also provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the steps of the resource measurement method of the aforementioned network computing power energy-aware network are implemented. The computer-readable storage medium can be a tangible storage medium, such as a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a floppy disk, a hard disk, a removable storage disk, a CD-ROM, or any other form of storage medium known in the technical field.

[0105] An embodiment of the present invention also provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps of the aforementioned resource measurement method for the network computing power energy-aware network.

[0106] It should be understood by those skilled in the art that the exemplary components, systems and methods described in conjunction with the embodiments disclosed herein can be implemented in hardware, software or a combination of the two. Whether it is performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present invention are programs or code segments used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link via a data signal carried in a carrier.

[0107] It should be clear that the present invention is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present invention.

[0108] In the present invention, features described and / or illustrated for one embodiment may be used in the same or similar manner in one or more other embodiments, and / or combined with features of other embodiments or replace features of other embodiments.

[0109] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the embodiments of the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A network computing power energy-aware network, characterized in that: The network computing power energy perception network includes a service application layer, a perception prediction layer and an infrastructure layer; the service application layer is used to perform computing tasks; the infrastructure layer includes multiple infrastructures, including computing power infrastructure, network infrastructure and energy infrastructure; the perception prediction layer includes a measurement device and multiple resource perception units; The resource sensing unit is arranged in the infrastructure, and is used to: monitor in real time the computing power resources corresponding to the computing power infrastructure, the network resources corresponding to the network infrastructure, and the energy consumption resources corresponding to the energy infrastructure, and determine the currently available resources corresponding to the infrastructure; The measurement device is used to: before the execution of the computing task, based on the task type, task scenario and task load corresponding to the computing task, measure the resources required to execute the computing task, obtain the target resources, and perform resource configuration according to the target resources and the resources corresponding to the infrastructure; during the execution of the computing task, based on the current time characteristics and the task characteristics corresponding to the computing task, generate the predicted available resources corresponding to the infrastructure within the target period; and perform resource configuration according to the predicted available resources, the current available resources and the target resources.

2. The network computing power energy-aware network according to claim 1, characterized in that: The target resources include target network resources, target computing power resources and target energy consumption resources; the resources required to execute the computing task are measured based on the task type, task scenario and task load corresponding to the computing task to obtain the target resources, including: Acquire a pre-established neural network fitting model; the neural network fitting model includes a computing network resource measurement model and an energy consumption resource measurement model; Input the task type, the task scenario and the task load into the computing network resource measurement model, and output the target network resources and the target computing power resources; The target network resources and the target computing power resources are input into the energy consumption resource measurement model, and the target energy consumption resources are output.

3. The network computing power energy-aware network according to claim 2, characterized in that: The target computing power resources also include fine-grained computing power features obtained based on analysis of the task scenario.

4. The network computing power energy-aware network according to claim 1, characterized in that: The generating, based on the current time feature and the task feature corresponding to the computing task, predicted available resources corresponding to the infrastructure within the target period includes: Based on a preset mapping relationship between features and forecast resources, according to the current time feature and the task feature, in the forecast resources, determine a target forecast resource corresponding to the current time feature and the task feature; The target predicted resource is used as the predicted available resource.

5. The network computing power energy-aware network according to claim 4, characterized in that: The task characteristics include energy characteristics, regional characteristics or resource characteristics corresponding to the computing task.

6. The network computing power energy-aware network according to claim 1, characterized in that: The measurement device is also used to: evaluate the execution progress of the computing task and the current resources corresponding to the computing task to obtain an evaluation result; and configure resources corresponding to the infrastructure based on the evaluation result.

7. A resource measurement method for a network computing power energy-aware network, characterized in that: Applied to a network computing power energy-aware network, the network computing power energy-aware network includes a service application layer, a perception prediction layer and an infrastructure layer; the service application layer is used to perform computing tasks; the infrastructure layer includes multiple infrastructures, the infrastructure includes computing power infrastructure, network infrastructure and energy infrastructure; the perception prediction layer includes a measurement device and multiple resource perception units, and the resource perception unit is set in the infrastructure; the method includes: The resource sensing unit monitors in real time the computing power resources corresponding to the computing power infrastructure, the network resources corresponding to the network infrastructure, and the energy consumption resources corresponding to the energy infrastructure to determine the current available resources corresponding to the infrastructure; Before the computing task is executed, the measuring device measures the resources required for executing the computing task based on the task type, task scenario and task load corresponding to the computing task, obtains the target resources, and performs resource allocation according to the target resources and the resources corresponding to the infrastructure; During the execution of the computing task, the measurement device generates predicted available resources corresponding to the infrastructure within the target period based on the current time characteristics and the task characteristics corresponding to the computing task; and performs resource allocation according to the predicted available resources, the current available resources and the target resources.

8. A resource measurement device for a network computing power energy-aware network, comprising a processor, a memory, and a computer program / instruction stored in the memory, characterized in that: The processor is used to execute the computer program / instructions, and when the computer program / instructions are executed, the device implements the steps of the method as claimed in claim 7.

9. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: When the computer program / instructions are executed by a processor, the steps of the method as claimed in claim 7 are implemented.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the steps of the method described in claim 7 are implemented.

Images