A Resource Allocation Method for Wireless Computing Networks Based on Digital Twins

Abstract

This invention discloses a method for allocating wireless computing network resources based on digital twins, comprising: constructing a wireless computing network architecture consisting of a physical entity layer and its corresponding digital twin layer; the physical entity layer including computing nodes composed of edge servers and terminal devices; fitting the state information of the edge servers and terminal devices through the digital twin of the digital twin layer; establishing an optimization problem with the goal of minimizing the total latency and total energy consumption of the wireless computing network architecture based on the fitted state information and the computing tasks of the wireless computing network architecture; and solving the optimization problem to obtain the optimal allocation strategy for computing resources. This invention obtains the entity information of computing nodes in real time through digital twins, thereby estimating the available computing resources and bandwidth of the computing nodes in the next time slot, and realizing the decision-making of computing resource allocation and task migration under the condition of minimizing the total latency and total energy consumption of the wireless computing network architecture.

Description

Technical Field

[0001] This invention relates to the field of communication technology, and in particular to a method for allocating wireless computing network resources based on digital twins. Background Technology

[0002] The sixth generation of wireless communication aims to achieve ubiquitous intelligent connectivity in a future integrated network encompassing air, space, land, and sea. It boasts extremely low latency and enhanced global coverage, but this requires sufficient computing power. Therefore, actively exploring computing power networks to meet ubiquitous computing demands is imperative. Wireless computing power networks, due to their extensive number of computing nodes, can better utilize idle computing resources, making them ideal wireless computing environments. However, the explosive growth of IoT devices generates massive amounts of data, requiring substantial computing and communication resources. Cloud computing is a technology that provides computing resources and data storage services over the internet. It allows users and businesses to store, process, and access data on remote servers without needing to operate on local computers or personal servers. Users can access powerful computing capabilities, storage space, and various applications on demand, typically on a pay-as-you-go basis. This model improves resource utilization, reduces infrastructure costs, and allows for rapid expansion to meet changing demands. However, cloud computing suffers from unavoidable issues such as data security, bandwidth limitations, and high long-term costs. As users' demands for latency and energy efficiency become increasingly apparent, cloud computing cannot provide a satisfactory user experience. Mobile Edge Computing (MEC), as an extension of cloud computing, is a distributed computing paradigm that deploys cloud computing capabilities and IT service environments at the edge of cellular networks, close to mobile users. Its core advantages lie in reducing network congestion, lowering service delivery latency, and improving application performance. MEC allows for the rapid and flexible deployment of new applications and services by running applications and processing tasks at cellular base stations or other edge nodes. MEC has a wide range of applications, including smart cities, industrial IoT, media and entertainment, and healthcare. However, due to the high mobility, limited resources, and time-varying nature of mobile networks, MEC faces challenges in meeting the requirements of 6G networks. The implementation of artificial intelligence algorithms in edge computing consumes significant computing resources for data processing and computation, while the heterogeneous capabilities of different devices exacerbate resource shortages in MEC. Furthermore, the enhanced requirements of 6G applications render existing technologies unsuitable for latency performance.

[0003] Therefore, how to allocate and compute tasks in wireless computing networks while minimizing latency and energy consumption has become a major challenge in providing high-quality services in 6G communications. Summary of the Invention

[0004] This invention provides a method for allocating wireless computing network resources based on digital twins, which addresses the technical problem that existing technologies do not fully consider the latency requirements and energy consumption of wireless computing networks, thus making it difficult to efficiently allocate computing resources.

[0005] The first aspect of this invention provides a method for allocating wireless computing network resources based on digital twins, the method comprising:

[0006] A wireless computing network architecture is constructed, consisting of a physical entity layer and its corresponding digital twin layer; the physical entity layer includes computing nodes composed of edge servers and terminal devices;

[0007] The digital twin layer fits the state information of the edge server and terminal device through a digital twin; the state information includes the estimated value of available computing resources corresponding to the edge server and terminal device, the deviation between the actual value and the estimated value of available computing resources, location information, the DT estimate of the channel bandwidth, and the deviation between the DT estimate and the actual value;

[0008] Based on the fitted state information and the computational tasks of the wireless computing network architecture, an optimization problem is established with the goal of minimizing the total latency and total energy consumption of the wireless computing network architecture.

[0009] The optimization problem is solved using the Lyapunov function to obtain the optimal allocation strategy for computing resources.

[0010] Specifically, the physical entity layer includes multiple edge servers, multiple terminal devices, and base stations, with the edge servers deployed on the base stations; wherein the set of edge servers is represented as... The set of terminal devices is represented as Among them, computing nodes It consists of terminal devices and edge servers;

[0011] The digital twin layer includes digital twins of edge servers and digital twins of terminal devices, with the digital twins deployed on the edge servers.

[0012] Specifically,

[0013] Edge server digital twin Build as:

[0014]

[0015] In the formula: Represents edge server Estimated available computing resources Represents edge server The deviation between the actual and estimated values ​​of available computing resources; This indicates the edge server in time slot t. The position; B represents the estimated DT value of the channel bandwidth. This represents the deviation between the estimated and actual DT value of the channel bandwidth.

[0016] terminal equipment digital twin Build as:

[0017]

[0018] In the formula: Indicates terminal device Estimated available computing resources Indicates terminal device The deviation between the actual and estimated values ​​of available computing resources; This indicates the terminal device in time slot t. Location; Terminal equipment Maximum transmission power.

[0019] Specifically, the step of establishing an optimization problem based on the fitted state information and the computational tasks of the wireless computing network architecture, with the objective of minimizing the total latency and total energy consumption of the wireless computing network architecture, includes:

[0020] Based on the fitted state information and the computational tasks of the wireless computing network architecture, a transmission delay model, a computation delay model, and an energy consumption model for the wireless computing network architecture are established.

[0021] Based on the computing tasks, transmission latency model, computation latency model, and energy consumption model of the wireless computing network architecture, an optimization problem is established with the goal of minimizing the total latency and total energy consumption of the wireless computing network architecture.

[0022] Specifically, the transmission latency model includes the actual transmission latency of the computing task from the terminal device to the edge server, the actual transmission latency of the computing task from the edge server to the terminal device, and the data transmission latency between the digital twin and the edge server; the process of establishing the transmission latency model includes:

[0023] Based on the status information of edge servers and terminal devices, we establish the estimated transmission delay of computing tasks between computing nodes and determine the estimation deviation of computing tasks during transmission after the introduction of digital twins. At the same time, we establish the data transmission delay between digital twins and edge servers.

[0024] Based on the estimated transmission delay of computing tasks between computing nodes and the corresponding estimation deviation, the actual transmission delay of computing tasks from terminal devices to edge servers and the actual transmission delay of computing tasks from edge servers to terminal devices are established.

[0025] Specifically, in time slot t, the actual transmission delay of the computation task from the terminal device to the edge server. Represented as:

[0026]

[0027] In the formula: Indicates task From terminal device Transmitted to edge server The estimated transmission delay, Indicates task From terminal device Transmitted to edge server The deviation between the estimated transmission delay and the actual transmission delay;

[0028] The actual transmission delay of the computation task from the edge server to the terminal device in time slot t. Represented as:

[0029]

[0030] In the formula: Indicates task From edge server Transmitted to terminal device The estimated transmission delay, Indicates task From edge server Transmitted to terminal device The deviation between the estimated transmission delay and the actual transmission delay;

[0031] Data transmission latency between the digital twin and the edge server at time slot t. Represented as:

[0032]

[0033] In the formula: w represents the time delay required to transmit one unit of data per unit distance. For the size of the transmitted data, This refers to a regular edge server. To the edge server deployed with digital twins The shortest distance.

[0034] Specifically, the computing tasks include local computing tasks, single-node computing tasks, and multi-node computing tasks; the computing latency model includes the actual computing latency of local computing, the actual computing latency of single-node computing, and the actual computing latency of multi-node computing.

[0035] Specifically, the energy consumption model includes the energy consumption for data transmission and the energy consumption for computing tasks; the energy consumption for data transmission includes the energy consumption for data transmission between computing nodes and the energy consumption for data transmission between the digital twin and the edge server.

[0036] Among them, the energy consumption of data transmission between digital twins and edge servers Represented as:

[0037]

[0038] In the formula: This indicates that an edge server with a digital twin has been deployed. The transmission power;

[0039] Among them, the computational energy consumption of the computational task is represented for:

[0040]

[0041] In the formula: z represents the task involved in the computation. The number of computing nodes, This indicates that compute node i is executing the task. The amount of energy consumed.

[0042] Specifically, the step of introducing the Lyapunov function to solve the optimization problem and obtain the optimal allocation strategy for computing resources includes: transforming the optimization problem into an objective optimization problem without long-term constraints based on the migration cost loss queue of the Lyapunov function; solving the objective optimization problem to obtain the optimal allocation strategy for computing resources;

[0043] The objective optimization problem is expressed as:

[0044]

[0045] In the formula: Indicates in time slot The compute nodes that provide computing services; 'o' is a positive control parameter used to dynamically balance the trade-off between processing latency performance and migration cost. Indicates in time slot Total latency of the wireless computing network architecture Indicates in time slot The following migration cost loss queue Indicates the computing node in the time slot The migration cost below Indicates in time slot Total energy consumption of the wireless computing network architecture;

[0046] The constraints of the objective optimization problem include:

[0047]

[0048] In the formula: C1 represents the total latency of the wireless computing network architecture. It cannot exceed the latency requirements of the computing task. C2 represents the deviation between the mapped value of the digital twin and the corresponding entity value. The constraint; C3 indicates that the user can only select from the available range. The optimal computing node for providing computing services is selected internally. Perform the calculation.

[0049] A second aspect of the present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the wireless computing network resource allocation method based on digital twin as described above.

[0050] A third aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the wireless computing network resource allocation method based on digital twins as described above.

[0051] As can be seen from the above technical solutions, the present invention has the following advantages:

[0052] This invention provides a method for allocating wireless computing network resources based on digital twins, comprising: constructing a wireless computing network architecture consisting of a physical entity layer and its corresponding digital twin layer; the physical entity layer includes computing nodes composed of edge servers and terminal devices; fitting the state information of the edge servers and terminal devices through the digital twins of the digital twin layer; the state information includes estimated values ​​of available computing resources corresponding to the edge servers and terminal devices, the deviation between the actual and estimated values ​​of available computing resources, location information, DT estimates of channel bandwidth, and the deviation between the DT estimates and the actual values; establishing an optimization problem with the goal of minimizing the total latency and total energy consumption of the wireless computing network architecture based on the fitted state information and the computing tasks of the wireless computing network architecture; and solving the optimization problem based on the Lyapunov function to obtain the optimal allocation strategy for computing resources.

[0053] This invention comprehensively considers the latency and energy consumption of wireless computing networks. It deploys digital twins in the wireless computing network architecture to obtain the physical information of computing nodes in real time, thereby estimating the available computing resources and bandwidth of the computing nodes in the next time slot. This allows for better task migration decisions, enabling the allocation of computing resources and task migration decisions to be made while minimizing the total latency and energy consumption of the wireless computing network architecture. This makes fuller use of the computing resources of terminal devices and edge servers to handle increasingly large computing tasks, thus solving the technical problem that existing technologies do not fully consider the latency requirements and energy consumption of wireless computing networks, making it difficult to efficiently allocate computing resources. Attached Figure Description

[0054] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0055] Figure 1 A flowchart illustrating the steps of a wireless computing network resource allocation method based on digital twins, provided in an embodiment of the present invention;

[0056] Figure 2 A model diagram of the wireless computing network architecture provided in an embodiment of the present invention;

[0057] Figure 3 This is a flowchart of the computational task unloading process and result return process provided in an embodiment of the present invention. Detailed Implementation

[0058] This invention provides a method for allocating wireless computing network resources based on digital twins, which addresses the technical problem that existing technologies do not adequately consider the latency requirements and energy consumption of wireless computing networks, thus making it difficult to efficiently allocate computing resources.

[0059] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described below are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0060] Please see Figure 1The first aspect of this invention provides a method for allocating wireless computing network resources based on digital twins, the method comprising:

[0061] Step 101: Construct a wireless computing network architecture consisting of a physical entity layer and its corresponding digital twin layer; wherein, the physical entity layer includes computing nodes consisting of edge servers and terminal devices.

[0062] Please see Figure 2 The wireless computing network architecture consists of a physical entity layer and its corresponding digital twin layer. The physical entity layer includes multiple edge servers, multiple terminal devices, and base stations. The edge servers are deployed on the base stations, which serve as the data transmission medium between the terminal devices and the edge servers. The set of edge servers is represented as follows: The set of terminal devices is represented as Among them, computing nodes It consists of terminal devices and edge servers.

[0063] In the digital twin layer, the digital twin (DT) is deployed on computing nodes. The digital twin acquires information about the physical entity in real time and performs simulation to replicate the real-time state of the physical entity, providing intelligent decision support for the physical entity and optimizing task allocation strategies. In this embodiment, the digital twin (DT) can be deployed on edge servers with abundant computing resources, which can save energy consumption for data transmission between the DT and the computing nodes.

[0064] In the wireless computing network architecture provided by this invention, the computing power of edge servers and terminal devices is aggregated in a computing power pool. This allows for the arbitrary scheduling of these computing nodes within the pool, with each node capable of executing computing tasks. When a computing task requires migration, it can be migrated from the terminal device to the edge server, or vice versa. Since the terminal device and edge server can be connected via wireless communication or wired links, the digital twins deployed on the edge server can also be interconnected.

[0065] In this embodiment, the computational task generated in time slot t is defined as... , It refers to the size of the computing task, which can be divided into computing tasks executed locally and tasks that need to be migrated to other computing nodes for computing. These are the computing resources needed to complete the computing task; This refers to the computational task. The delayed demand; among which, To represent a set of time slots.

[0066] Step 102: Fit the status information of the edge server and terminal device using the digital twin of the digital twin layer.

[0067] It should be noted that the computing methods of computing tasks can be divided into local computing, single-node computing, and multi-node computing. In order to optimize the allocation strategy of computing tasks, this invention uses digital twins to estimate the available computing resources and channel bandwidth in the next time slot based on the computing resources and channel bandwidth of the current time slot, so as to make better decisions on the migration of computing tasks.

[0068] To perceive the dynamic distribution of computing power and channel bandwidth, this invention integrates digital twins into the WCPN architecture, using a lightweight digital twin model (DT model) to reflect real-time changes in computing resources. The actual available computing resources and channel bandwidth in the computing nodes are used as training data to train the digital twin model, thereby improving the estimation accuracy of the DT model.

[0069] Therefore, the digital twin of computing node j is defined as follows:

[0070]

[0071] In the formula: This represents an estimate of the computing resources available to computing node j. This represents the deviation between the actual and estimated values ​​of the computing resources available to computing node j. This indicates the position of the compute node j.

[0072] In the WCPN architecture, there are two main types of computing nodes: edge servers and terminal devices. Therefore, the dynamic distribution of information such as perceived computing power and channel bandwidth consists of two types of resources: 1) edge servers; 2) terminal devices. This embodiment uses digital twins to reflect the distribution of computing power and channel bandwidth among various computing nodes.

[0073] Based on the above equation (1), for edge servers digital twin Build as:

[0074]

[0075] In the formula: Represents edge server Estimated available computing resources Represents edge server The deviation between the actual and estimated values ​​of available computing resources; This indicates the edge server in time slot t. The position; B represents the estimated DT value of the channel bandwidth. This represents the deviation between the estimated and actual channel bandwidth (DT). In this architecture, the location of the edge server is typically fixed; therefore, the edge server in this embodiment... Location (t) is a fixed value.

[0076] terminal equipment digital twin Build as:

[0077]

[0078] In the formula: Indicates terminal device Estimated available computing resources Indicates terminal device The deviation between the actual and estimated values ​​of available computing resources; This indicates the terminal device in time slot t. Location; Terminal equipment Maximum transmission power.

[0079] It is understandable that terminal devices are mobile, so terminal devices... Location It can be represented as:

[0080]

[0081] in, Represents the position function of the terminal device. Indicates terminal device movement speed, Indicates terminal device The distance traveled.

[0082] So, for the digital twin of the WCPN architecture Represented as:

[0083]

[0084] in, It describes the state information of the computing nodes in the entire architecture. This describes the estimated returns for selecting different computing nodes. Status information includes estimates of available computing resources for edge servers and terminal devices, the deviation between the actual and estimated available computing resources, location information, DT estimates of channel bandwidth, and the deviation between the DT estimates and actual values.

[0085] This invention estimates the computing resources and bandwidth of physical entities using digital twins to optimize task offloading and resource allocation strategies that consume less latency and energy. At the physical entity layer, the optimal task decision can be selected based on the state information fitted by the digital twin layer.

[0086] Step 103: Based on the fitted state information and the computational tasks of the wireless computing network architecture, establish an optimization problem with the goal of minimizing the total latency and total energy consumption of the wireless computing network architecture.

[0087] It should be noted that this step considers not only the computation, transmission latency and energy consumption of the computation task, but also the data transmission latency and energy consumption between DT and the mapped edge server, and aims to minimize the total latency and total energy consumption in the above process, thereby obtaining the optimal task decision.

[0088] This step specifically includes:

[0089] Sub-step 1031: Based on the fitted state information and the computational tasks of the wireless computing network architecture, establish the transmission delay model, computation delay model, and energy consumption model of the wireless computing network architecture.

[0090] It should be noted that when the computing task is selected for local computing, the computing task does not involve transmission, so the transmission latency is zero. Therefore, the transmission latency model in this invention can include the actual transmission latency of the computing task from the terminal device to the edge server, the actual transmission latency of the computing task from the edge server to the terminal device, and the data transmission latency between the digital twin and the edge server.

[0091] Understandably, when a user needs to migrate a task from one compute node to another, a migration cost will be incurred, defined as follows:

[0092]

[0093] in This refers to task migration decisions, where the computing nodes of tasks in time slot t and time slot t-1 are different. Otherwise, it is 0; This represents the migration cost of moving from one compute node to another.

[0094] In this sub-step, based on the status information of the edge server and the terminal device, the estimated transmission delay of the computing task between computing nodes can be established, and the estimation deviation of the computing task during transmission caused by the introduction of digital twin can be determined. At the same time, the data transmission delay between the digital twin and the edge server can be established. Finally, based on the estimated transmission delay of the computing task between computing nodes and the corresponding estimation deviation, the actual transmission delay of the computing task from the terminal device to the edge server and the actual transmission delay of the computing task from the edge server to the terminal device can be established.

[0095] (1) The more specific process of establishing the transmission delay model is as follows:

[0096] First, based on the fitted state information of the edge server and terminal device, the estimated transmission speed between computing nodes is established. When a computing node chooses to offload computing tasks, the estimated transmission speed between computing nodes can be divided into two categories: the estimated transmission speed of computing tasks from the terminal device to the edge server, and the estimated transmission speed of computing tasks from the edge server to the terminal device.

[0097] In time slot t, the computation task originates from the terminal device. Transmitted to edge server Estimated transmission speed for:

[0098]

[0099] In time slot t, the computation task originates from the edge server. Transmitted to terminal device Estimated transmission speed for:

[0100]

[0101] In the formula: Terminal equipment Allocate to edge servers within time slot t Estimated bandwidth, Terminal equipment Transmit power at time slot t; It is an edge server Allocate time slot t to terminal devices Estimated bandwidth, It is an edge server Transmit power at time slot t; Indicates terminal device With edge servers Channel gain at time slot t; It is noise power; This is determined by the location of the terminal device. and the location of edge servers Calculated;

[0102] in,

[0103]

[0104] in, Terminal equipment To the edge server The path loss exponent between them; I is the interference from other edge servers, expressed by the formula:

[0105]

[0106] in, In addition to Other terminal devices besides those mentioned above; In addition to Other edge servers besides; Indicates the transmission power of other terminal devices; This indicates the channel gain between other terminal devices and other edge servers.

[0107] In particular, if the computing task needs to be sent outside the range of wireless communication... So edge servers Other edge servers The wired transmission rate between them is:

[0108]

[0109] In the formula: This is the distance factor.

[0110] Next, based on formulas (7), (8), and (11), the estimated transmission delay of the computation task between computing nodes is established. Wherein, the computation task... From terminal device Transmitted to edge server Estimated transmission delay for:

[0111]

[0112] In the formula: This refers to edge servers. The maximum coverage distance, for example, if a computing task needs to be migrated to edge server 'a' for computation, the transmission distance exceeds the limit. At this point, the computing task needs to be migrated to edge server b first, and then from edge server b to edge server a.

[0113] Computational tasks From terminal device Transmitted to edge server The deviation between the estimated transmission delay and the actual transmission delay for:

[0114]

[0115] So, computational task From terminal device Transmitted to edge server Actual transmission delay for:

[0116]

[0117] Among them, computing tasks From edge server Transmitted to terminal device Estimated transmission delay for:

[0118]

[0119] Computational tasks From edge server Transmitted to terminal device The deviation between the estimated transmission delay and the actual transmission delay for:

[0120]

[0121] So, computational task From edge server Transmitted to terminal device Actual transmission delay Represented as:

[0122]

[0123] To measure the consistency between the digital twin and its physical counterpart, this invention introduces a DT synchronization delay, which then determines the data transmission delay between the DT model and the mapped edge server. Represented as:

[0124]

[0125] In the formula: The latency required to transmit one unit of data per unit distance. For the size of the transmitted data, This represents the shortest distance from the edge server to the DT edge server. The DT edge server refers to an edge server deployed with a digital twin, which can obtain real-time information from other computing nodes.

[0126] To measure the accuracy of the DT model, the DT error is introduced: the deviation between the mapped value of the DT model and the actual value of the corresponding entity. :

[0127]

[0128] In the formula: It is a weighting factor. It's the packet loss rate. Represents edge server To DT edge server The number of hops for the shortest path between them.

[0129] (2) The more specific process of establishing the time delay model is as follows:

[0130] In wireless computing network architecture, computing tasks are divided into three types: local computing, single-node computing, and multi-node computing. It should be noted that local computing means the task does not require transmission and is performed directly on the local machine; single-node computing refers to a task being migrated to a single computing node for computation. If the computing task exceeds the maximum available computing resources of a single computing node, the task can be divided into multiple subtasks and migrated to multiple computing nodes for computation.

[0131] For local computation:

[0132] In the time slot The estimated computation latency for local computation of computing tasks by the terminal device. Represented as:

[0133]

[0134] in, Indicates in time slot Terminal devices that provide computing services.

[0135] Assuming the deviation between the terminal device and its DT can be obtained in advance, the calculation delay difference between the estimated and actual DT values ​​is... Represented as:

[0136]

[0137] in, Indicates in time slot Estimated available computing resources for terminal devices providing computing services Indicates in time slot The deviation between the estimated available computing resources of the terminal devices providing computing services and the actual values.

[0138] So, the actual computational latency for local computation Represented as:

[0139]

[0140] For single-node computation:

[0141] In the time slot The estimated computation latency of a computation task performed by a single computing node. Represented as:

[0142]

[0143] in, Indicates in time slot Computing nodes that provide computing services.

[0144] Assuming the deviation between the edge server and its DT can be obtained in advance, the computational latency difference between the estimated and actual DT values ​​is... Represented as:

[0145]

[0146] in, Indicates in time slot The estimated available computing resources of the computing nodes providing computing services. Indicates in time slot The deviation between the estimated available computing resources of computing nodes that provide computing services and the actual values.

[0147] So, what is the actual computational latency for single-node computation? Represented as:

[0148]

[0149] For multi-node computation:

[0150] Under the required latency constraints, when the executed task exceeds the maximum computing capacity of a single computing node, the computing task needs to be offloaded to multiple computing nodes for collaborative computation. In this case, the computing task... It can be divided into multiple subtasks So in the time slot Estimated computation latency when multiple computing nodes perform computations on a computational task It can be defined as:

[0151]

[0152] in, This represents the estimated available computing resources when executing the l-th subtask.

[0153] Among them, in time slots The difference between the estimated computation latency and the actual computation latency when a computation task is performed by multiple computing nodes is:

[0154]

[0155] So, in the time slot The actual computational latency of computing tasks performed by multiple computing nodes Represented as:

[0156]

[0157] (3) The more specific process of establishing the energy consumption model is as follows:

[0158] In a wireless computing network architecture, energy consumption includes the energy consumption for data transmission and the energy consumption for task computation. During the transmission phase, energy consumption is mainly reflected in uplink task transmission and downlink result computation transmission. Since the size of the computation result is much smaller than the size of the computation task, this embodiment only considers the energy consumption of uplink task transmission. Therefore, the transmission energy consumption between computing nodes is:

[0159]

[0160]

[0161] In the formula: and These represent the energy consumption of transmitting computing tasks from the terminal device to the edge server and the energy consumption of transmitting computing tasks from the edge server to the terminal device, respectively.

[0162] Therefore, the energy consumption for uplink task transmission is:

[0163]

[0164]

[0165] In addition to the uplink transmission energy consumption, this invention also considers the DT transmission energy consumption, that is, the data transmission energy consumption between the DT model and the mapped edge server. :

[0166]

[0167] In the formula: This indicates that an edge server with a digital twin has been deployed. The transmission power.

[0168] For compute node i, execute the task. Energy consumed for:

[0169]

[0170] in, Calculate the energy consumed by a resource per unit. It is the effective switched capacitor coefficient; due to the task The computation task can be completed by multiple computing nodes. Energy consumption can be expressed as:

[0171]

[0172] Where z is the computation task. The number of computing nodes.

[0173] Sub-step 1032: Based on the computing tasks, transmission latency model, computing latency model, and energy consumption model of the wireless computing network architecture, establish an optimization problem with the goal of minimizing the total latency and total energy consumption of the wireless computing network architecture.

[0174] The optimization objective of this invention is to minimize the total latency and total energy consumption of the above process while satisfying the latency requirements of the computing task and not exceeding the migration cost.

[0175] Among them, the total latency of the wireless computing network architecture is based on the transmission latency model and the computation latency model. Represented as:

[0176]

[0177] In the formula: c is a binary variable. When c=0, it means the task is computed locally; when c=1 and When c=1, the task is unloaded to a single compute node for computation; At that time, the task is unloaded to multiple computing nodes for computation.

[0178] Based on the energy consumption model, the total energy consumption of the wireless computing network architecture Represented as:

[0179]

[0180] Therefore, with the goal of minimizing the total delay and total energy consumption, we formulate the optimization problem P1:

[0181]

[0182] The constraints of optimization problem P1 include:

[0183]

[0184] Where C1 represents the constraint of migration cost for migrating the computing task. This represents the upper limit of the service migration rate throughout the entire migration process; C2 represents the deviation between the mapped value of the digital twin and the corresponding entity value. Constraints; C3 represents the total latency of the wireless computing network architecture. Constraints, total delay It cannot exceed the latency requirements of the computing task. C4 represents the computing node that provides computing services. Selection constraints limit users to a range of options. The optimal computing node for providing computing services is selected internally. Perform the calculation.

[0185] Step 104: Solve the optimization problem based on the Lyapunov function to obtain the optimal allocation strategy for computing resources.

[0186] Since the optimization problem P1 brings about an optimization problem due to the long-term service migration cost constraint, this step adopts the dynamic loss queue optimization method of Lyapunov function to optimize and solve it. A dynamic virtual migration cost loss queue is designed to make appropriate move and offload decisions.

[0187] The length of the migration cost loss queue can be defined as the deviation between used migration costs and available migration costs. The user's entire journey is divided into... Each time slot, the total migration cost is Then the migration cost loss queue is represented as:

[0188]

[0189] in, This represents the migration cost loss queue in time slot t. This indicates the deviation of migration cost in time slot t. This represents the upper limit of migration cost for a time slot.

[0190] Based on the designed dynamic migration cost loss queue, the Lyapunov optimization technique transforms the original optimization problem P1 into an objective optimization problem P2 without long-run constraints. The simplified objective optimization problem P2 is expressed as:

[0191]

[0192] The constraints of the objective optimization problem P2 include:

[0193]

[0194] Here, 'o' is a positive control parameter used to dynamically balance the trade-off between processing latency performance and migration cost.

[0195] Deep learning algorithms are used to solve the objective optimization problem, resulting in an optimal strategy for allocating computing resources; for details of the allocation strategy, please refer to [link / reference needed]. Figure 3 When a computing task does not need to be migrated to another node for computing, it can be computed locally; otherwise, it can use a digital twin to estimate the available computing resources and channel bandwidth for the next time slot based on the computing resources and channel bandwidth of the current time slot. In other words, it can estimate the available resource information of the physical entity in the next time slot. Under the condition of minimizing the total latency and total energy consumption, it can guide the migration strategy of computing tasks, such as single-node computing or multi-node computing, complete the optimization of computing resource allocation, improve the task cooperation capability between computing nodes, and realize the unified scheduling and management of global resources within the wireless computing network architecture.

[0196] The present invention provides a wireless computing network resource allocation method based on digital twins, which has the following advantages:

[0197] 1. Compared with traditional mobile edge computing, this invention can make fuller use of the computing resources of terminal devices and edge servers to handle increasingly large computing tasks.

[0198] 2. This invention deploys a digital twin in the wireless computing network architecture to obtain the physical information of computing nodes in real time, thereby estimating the available computing resources and bandwidth of the computing node in the next time slot, so as to obtain better task migration decisions.

[0199] 3. The actual available computing resources and channel bandwidth in the computing nodes will be used as new data to train the digital twin, thereby improving the fitting accuracy of the digital twin.

[0200] 4. In wireless computing networks that have deployed digital twins, minimizing and optimizing target latency and energy consumption takes into account the DT synchronization delay and the energy consumption generated by data transmission between the DT and computing nodes, thereby improving the real-time processing capability and resource allocation optimization capability of the wireless computing network architecture.

[0201] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described wireless computing network resource allocation method based on digital twins.

[0202] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of any of the above-mentioned wireless computing network resource allocation methods based on digital twins.

[0203] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0204] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between apparatuses or units through some interfaces, and may be electrical, mechanical, or other forms.

[0205] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0206] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0207] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0208] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for wireless computing power network resource allocation based on digital twinning, characterized in that, The method comprises: constructing a wireless computing power network architecture composed of a physical entity layer and its corresponding digital twin layer; the physical entity layer comprises computing nodes composed of edge servers and terminal devices; fitting the state information of the edge servers and the terminal devices through the digital twin layer; the state information comprises available computing resource estimation values, deviations between actual values and estimation values of available computing resources, position information, DT estimation values of channel bandwidths, and deviations between DT estimation values and actual values corresponding to the edge servers and the terminal devices; establishing an optimization problem with the minimization of the total time delay and the total energy consumption of the wireless computing power network architecture as the target according to the fitted state information and the computing tasks of the wireless computing power network architecture; solving the optimization problem based on a Lyapunov function to obtain an optimal computing resource allocation strategy.

2. The wireless computational force network resource allocation method of claim 1, wherein, The physical entity layer includes a plurality of edge servers and a plurality of terminal devices and base stations, the edge servers are deployed on the base stations; wherein a set of edge servers is represented as , a set of terminal devices is represented as , wherein a computing node is composed of a terminal device and an edge server; The digital twin layer comprises digital twins of the edge servers and digital twins of the terminal devices, and the digital twins are deployed on the edge servers.

3. The wireless computing power network resource allocation method according to claim 2, wherein Edge server Digital twin of is constructed as: wherein: represents an edge server an estimated value of available computing resources, represents an edge server a deviation of an actual value of available computing resources from the estimated value; represents a location of an edge server at time slot t; B represents a DT estimated value of channel bandwidth, represents a deviation of a DT estimated value of channel bandwidth from an actual value. Terminal device Digital twin of a wind turbine is constructed as: wherein: denotes a terminal device an estimated value of available computing resources, denotes a terminal device a deviation of an actual value of available computing resources from the estimated value; denotes a position of a terminal device at time slot t; is a maximum transmission power of a terminal device .

4. The wireless computing power network resource allocation method of claim 3, wherein, the step of establishing an optimization problem with the minimization of the total time delay and the total energy consumption of the wireless computing power network architecture as the target according to the fitted state information and the computing tasks of the wireless computing power network architecture comprises: establishing a transmission time delay model, a computing time delay model and an energy consumption model of the wireless computing power network architecture according to the fitted state information and the computing tasks of the wireless computing power network architecture; establishing an optimization problem with the minimization of the total time delay and the total energy consumption of the wireless computing power network architecture as the target based on the computing tasks, the transmission time delay model, the computing time delay model and the energy consumption model of the wireless computing power network architecture.

5. The wireless computing power network resource allocation method of claim 4, wherein, The transmission time delay model comprises actual transmission delays of computing tasks from the terminal devices to the edge servers, actual transmission delays of computing tasks from the edge servers to the terminal devices, and data transmission delays between the digital twins and the edge servers; the establishment process of the transmission time delay model comprises: based on the state information of the edge servers and the terminal devices, establishing estimated transmission delays of the computing tasks between the computing nodes and determining estimated deviations of the computing tasks in the transmission process after the introduction of the digital twins, and simultaneously establishing data transmission delays between the digital twins and the edge servers; based on the estimated transmission delays of the computing tasks between the computing nodes and the corresponding estimated deviations, establishing actual transmission delays of the computing tasks from the terminal devices to the edge servers and actual transmission delays of the computing tasks from the edge servers to the terminal devices; wherein, at a time slot t, the actual transmission delay of the computing task transmitted from the terminal device to the edge server is represented as: In the formula: represents a task from a terminal device transmitted to an edge server an estimated transmission delay, represents a task from a terminal device transmitted to an edge server a deviation of an estimated transmission delay from an actual transmission delay; In the time slot t, the actual transmission delay of the computing task transmitted from the edge server to the terminal device is represented as: wherein: represents the task from the edge server transmitted to the terminal device an estimated transmission delay, represents the task from the edge server transmitted to the terminal device a deviation of the estimated transmission delay from the actual transmission delay; Data transmission delay between digital twin and edge server at time slot t is represented as: wherein: w represents a latency required to transmit one unit of data per unit distance, for a size of data to be transmitted, represents a common edge server to an edge server in which a digital twin is deployed is the shortest distance.

6. The wireless computing power network resource allocation method of claim 4, wherein, The computing tasks comprise local computing tasks, single-node computing tasks and multi-node computing tasks; the computing time delay model comprises actual computing time delays of local computing, actual computing time delays of single-node computing and actual computing time delays of multi-node computing.

7. The wireless computational force network resource allocation method of claim 4, wherein, The energy consumption model comprises transmission energy consumption of data and computing energy consumption of computing tasks; the transmission energy consumption of data comprises transmission energy consumption of the computing tasks between the computing nodes and data transmission energy consumption between the digital twins and the edge servers; The data transmission energy consumption between the digital twin and the edge server is represented as: In the formula: indicates an edge server deployed with digital twin transmit power; wherein the computing energy consumption of the computing task represents is: wherein: z represents the number of computing nodes participating in the computing task , represents the amount of energy consumed by the computing node i to perform the task .

8. The wireless computational force network resource allocation method of claim 2, wherein, The step of solving the optimization problem by using the introduced Lyapunov function to obtain the optimal allocation strategy of the computing resources comprises: converting the optimization problem into a target optimization problem without long-term constraints based on a migration cost loss queue of the Lyapunov function; and solving the target optimization problem to obtain the optimal allocation strategy of the computing resources. The target optimization problem is expressed as: In the formula: represents the total latency of the wireless computing power network architecture under the time slot provides a computing node for computing services; o is a positive control parameter for dynamically balancing the trade-off between processing delay performance and migration cost consumption; represents the total latency of the wireless computing power network architecture under the time slot represents the migration cost loss queue under the time slot represents the migration cost of the computing node under the time slot represents the total energy consumption of the wireless computing power network architecture under the time slot ​​​​ The constraint condition of the target optimization problem comprises: In the formula: C1 represents the total latency of the wireless computing network architecture. It cannot exceed the latency requirements of the computing task. C2 represents the deviation between the mapped value of the digital twin and the corresponding entity value. The constraint; C3 indicates that the user can only select from the available range. The optimal computing node for providing computing services is selected internally. Perform the calculation. 9.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-8 when the computer program is executed by the processor. The processor executes the computer program to implement the steps of the wireless computing power network resource allocation method based on the digital twin as claimed in any one of claims 1-8.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the wireless computing power network resource allocation method based on the digital twin as claimed in any one of claims 1-8.

Images