Heterogeneous fusion computing network resource management and control method and device

Abstract

The invention provides a heterogeneous convergence computing network resource management and control method and device, relates to the technical field of communication, and aims at a heterogeneous convergence computing power network system with two communication systems of satellite communication and ground communication, adopts a hierarchical game mode, adopts a non-cooperative game between layers, and adopts a cooperative game in the layers. Through combination of the two game modes, network resources in a computing power network system can be effectively and reasonably distributed according to needs, and efficient and accurate computing power network resource management and control under a diversified communication system are realized.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a method and device for managing and controlling resources in a heterogeneous fusion computing network. Background Art

[0002] Computing network is the direction of future network evolution. When the network realizes the interconnection of distributed nodes, it also needs to have the ability to coordinate the scheduling of network resources and computing resources. To this end, a new network resource allocation method is needed to adapt to the resource allocation scenario of the current computing network, make a comprehensive trade-off between network resources and computing resources, and come up with the optimal strategy to achieve network resource management and control of the computing network. Figure 1 This is a structural diagram of a distributed computing network scenario, such as Figure 1 As shown, the client node sends a service request through the application interface, and the computing power routing node obtains the network and computing power status information according to the resource interface of each computing service node to perform computing power resource scheduling and routing addressing, reaches the target computing service node through cross-node routing, and finally completes the specific computing service according to the service interface of the target computing service node before returning.

[0003] Currently, network resource allocation methods include static allocation and dynamic allocation. Among them, static allocation is a way of pre-allocating resources, which allocates resources to tasks according to pre-set rules and policies. For example, static resource allocation is performed based on the priority of the task and the amount of resources required. Dynamic allocation is the opposite of static allocation. When the network system is running, resources are allocated according to the actual situation of the task. It can be adjusted according to the real-time needs of the task, thereby improving the utilization rate of network resources.

[0004] However, the existing network resource allocation methods only consider a single communication system. Faced with the diverse communication systems in the system, how to conduct efficient and accurate network resource management and control is the focus of current research. Summary of the invention

[0005] The purpose of the present invention is to provide a heterogeneous fusion computing network resource management method and device to achieve efficient and accurate computing network resource management under diversified communication systems.

[0006] In a first aspect, an embodiment of the present invention provides a heterogeneous fusion computing network resource management and control method, which is applied to a computing network system including a satellite link layer and a ground link layer, wherein the space-based network corresponding to the satellite link layer adopts a satellite communication system, and the ground-based network corresponding to the ground link layer adopts a ground communication system; the heterogeneous fusion computing network resource management and control method includes: Obtaining demand parameter data, current bandwidth resource data, and current task data of the computing power network system; wherein the demand parameter data is parameter data of a market demand function corresponding to the computing power network system, and the market demand function is a function that reflects the relationship between bandwidth resource demand and bandwidth resource price; According to the demand parameter data and the current bandwidth resource data, resources are allocated between the satellite link layer and the terrestrial link layer through a non-cooperative game to obtain the total bandwidth resources of the satellite link layer and the total bandwidth resources of the terrestrial link layer; According to the total bandwidth resources of the satellite link layer, the total bandwidth resources of the ground link layer and the current task data, the satellite link layer and the ground link layer are respectively allocated resources within the layer through cooperative game to obtain the node resource allocation result of the computing power network system.

[0007] Furthermore, the obtaining of the demand parameter data of the computing power network system includes: Acquire historical market data of the computing power network system within a preset time range, wherein the historical market data includes historical bandwidth demand information, price information, and cost information; According to the historical market data, the parameters of the preset market demand function are estimated through a rolling window algorithm to obtain the demand parameter data of the computing power network system.

[0008] Further, the allocating resources between the satellite link layer and the terrestrial link layer through a non-cooperative game according to the demand parameter data and the current bandwidth resource data to obtain the total bandwidth resources of the satellite link layer and the total bandwidth resources of the terrestrial link layer includes: Under the constraint condition of the given current bandwidth resource data, establishing a first utility function model of the satellite link layer and the ground link layer under a non-cooperative game; An analytical solution is obtained for the first utility function model according to the demand parameter data to obtain the total bandwidth resources of the satellite link layer and the total bandwidth resources of the ground link layer.

[0009] Furthermore, the first utility function model includes:

[0010]

[0011] in, is the node set of the satellite link layer M The corresponding utility value, is the node set of the ground link layer N The corresponding utility value, RM is the node set of the satellite link layer M The corresponding price, R N is the node set of the ground link layer N The corresponding price, C M is the node set of the satellite link layer M The corresponding cost, C N is the node set of the ground link layer N The corresponding cost, a is the intercept value in the demand parameter data, b = k / (- B ), k is the slope value in the demand parameter data, B is the current bandwidth resource data, b M is the total bandwidth resource of the satellite link layer, b N is the total bandwidth resource of the ground link layer, c M is the cost per allocated unit bandwidth of the satellite link layer, c N Cost per allocated unit bandwidth of the ground link layer.

[0012] Further, the analytical solution of the first utility function model according to the demand parameter data to obtain the total bandwidth resources of the satellite link layer and the total bandwidth resources of the ground link layer includes: Substituting the demand parameter data, the cost per allocated unit bandwidth of the satellite link layer, and the cost per allocated unit bandwidth of the terrestrial link layer into the following analytical solution of the first utility function model, the total bandwidth resources of the satellite link layer and the total bandwidth resources of the terrestrial link layer are obtained:

[0013]

[0014] in, b M is the total bandwidth resource of the satellite link layer, b N is the total bandwidth resource of the ground link layer, a is the intercept value in the demand parameter data, b = k / (- B ), k is the slope value in the demand parameter data,B is the current bandwidth resource data, c M is the cost per allocated unit bandwidth of the satellite link layer, c N is the cost per allocated unit bandwidth of the ground link layer, B The current bandwidth resource data.

[0015] Further, according to the total bandwidth resources of the satellite link layer, the total bandwidth resources of the ground link layer and the current task data, the satellite link layer and the ground link layer are respectively allocated resources within the layer through cooperative game to obtain the node resource allocation result of the computing power network system, including: For each target link layer in the satellite link layer and the terrestrial link layer, establishing a second utility function model of the target link layer under cooperative game; Taking each node in the target link layer as a single node alliance to form an initial alliance structure; According to the initial alliance structure, the total bandwidth resources of the target link layer and the current task data, with maximizing the total alliance benefit as the optimization goal, switching and merging alliances are performed to obtain a target alliance structure; wherein the alliance is formed by offloading tasks to other nodes, and the total alliance benefit is calculated based on the second utility function model; The resource allocation result under the target alliance structure is determined as the node resource allocation result of the target link layer.

[0016] Furthermore, the ground-based network corresponding to the ground link layer includes a 5G network corresponding to the 5G link layer.

[0017] In a second aspect, an embodiment of the present invention further provides a heterogeneous fusion computing network resource management and control device, which is applied to a computing network system including a satellite link layer and a ground link layer, wherein the space-based network corresponding to the satellite link layer adopts a satellite communication system, and the ground-based network corresponding to the ground link layer adopts a ground communication system; the heterogeneous fusion computing network resource management and control device includes: A data acquisition module, used to acquire demand parameter data, current bandwidth resource data and current task data of the computing power network system; wherein the demand parameter data is parameter data of a market demand function corresponding to the computing power network system, and the market demand function is a function reflecting the relationship between bandwidth resource demand and bandwidth resource price; A first allocation module is used to allocate resources between the satellite link layer and the terrestrial link layer through a non-cooperative game according to the demand parameter data and the current bandwidth resource data, so as to obtain the total bandwidth resources of the satellite link layer and the total bandwidth resources of the terrestrial link layer; The second allocation module is used to allocate resources within the satellite link layer and the ground link layer respectively through cooperative game according to the total bandwidth resources of the satellite link layer, the total bandwidth resources of the ground link layer and the current task data, so as to obtain the node resource allocation result of the computing power network system.

[0018] In a third aspect, an embodiment of the present invention further provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, and the processor implements the method of the first aspect when executing the computer program.

[0019] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the method described in the first aspect is executed.

[0020] The heterogeneous fusion computing network resource management and control method and device provided by the embodiment of the present invention are applied to a computing power network system including a satellite link layer and a ground link layer, wherein the space-based network corresponding to the satellite link layer adopts a satellite communication system, and the ground-based network corresponding to the ground link layer adopts a ground communication system; the heterogeneous fusion computing network resource management and control method comprises: obtaining demand parameter data, current bandwidth resource data and current task data of the computing power network system; wherein the demand parameter data is parameter data of a market demand function corresponding to the computing power network system, and the market demand function is a function reflecting the relationship between the bandwidth resource demand and the bandwidth resource price; according to the demand parameter data and the current bandwidth resource data, resources are allocated between the satellite link layer and the ground link layer through non-cooperative game to obtain the total bandwidth resources of the satellite link layer and the total bandwidth resources of the ground link layer; according to the total bandwidth resources of the satellite link layer, the total bandwidth resources of the ground link layer and the current task data, resources are allocated within the satellite link layer and the ground link layer respectively through cooperative game to obtain the node resource allocation result of the computing power network system. In this way, a layered game method is adopted for the heterogeneous integrated computing network system with two communication systems, satellite communication and ground communication. Non-cooperative game is adopted between layers, and cooperative game is adopted within the layer. The combination of the two game methods can effectively allocate network resources in the computing network system reasonably on demand, and realize efficient and accurate computing network resource management and control under diversified communication systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0022] Figure 1 This is a structural diagram of a distributed computing network scenario; Figure 2 Content cache map for 5G networks; Figure 3 An overall architecture diagram of a computing power network system provided by an embodiment of the present invention; Figure 4 A schematic diagram of a process flow of a heterogeneous fusion computing network resource management and control method provided by an embodiment of the present invention; Figure 5 A structural diagram of a hierarchical game in a heterogeneous fusion computing network resource management and control method provided by an embodiment of the present invention; Figure 6 An inter-layer game model in a heterogeneous fusion computing network resource management and control method provided in an embodiment of the present invention; Figure 7 A topological diagram of an intra-layer game network in a heterogeneous fusion computing network resource management and control method provided in an embodiment of the present invention; Figure 8 An overall framework diagram of a heterogeneous fusion computing network resource management and control method provided by an embodiment of the present invention; Fig. 9 A schematic diagram of the structure of a heterogeneous fusion computing network resource management and control device provided by an embodiment of the present invention; Fig.10 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0023] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0024] The network resource allocation methods in simple scenarios (i.e. the same communication system) include static allocation and dynamic allocation. With the continuous development of computing network technology, when there are multiple network nodes in the network, different network nodes have individual differences, and it is difficult to use a single model to characterize the entire system. In order to more accurately describe the actual computing network system and efficiently allocate network resources, it is necessary to comprehensively consider the heterogeneity of the communication system within the system. The game theory method can achieve this goal. The resource allocation problem in the computing network can be regarded as a game problem with global constraints. Different game methods are used to enable each network node to make its own decisions according to the situation and complete the network resource allocation control of the entire system.

[0025] By properly designing the incentive mechanism in game theory, each node can be encouraged to actively participate in resource sharing. The incentive mechanism based on utility value can give corresponding rewards according to the quantity, quality and sharing frequency of resources provided by the node. In the problem of network resource allocation, utility refers to an important evaluation indicator of user satisfaction in terms of network service quality. Define a utility function u ( x i ) represents the user (i.e., a node in the network)'s response to a certain network indicator x i The satisfaction level can be considered x i That is, the amount of resources allocated by the network to the network node. This value satisfies the condition x a ≤ x i ≤ x b . Similarly, the corresponding utility value has an upper limit. In mathematical definition, in order to ensure that the utility value can flexibly represent the user's network indicators x i The utility function needs to be twice differentiable within this limit. The trend of utility value changes with the allocated resources x i The trend of change should be uniform. In order to ensure that the more resources allocated to users, the higher the utility value of user satisfaction feedback, when the allocated resources reach a certain threshold and obtain a higher satisfaction, the change of utility value should be flat, so that the marginal utility shows a decreasing trend, that is, the utility function must be a non-decreasing function, which can be expressed as follows: (1) (2) For elastic network resources such as bandwidth, the Sigmoid curve can be used to describe the user's satisfaction with the allocated bandwidth. This curve conforms to the characteristics of twice differentiable and non-decreasing while also having a certain degree of elasticity. It is suitable for describing elastic resources in a utility-based optimization model.

[0026] By utilizing the utility function and formulating a suitable incentive mechanism to obtain the reward value, each node's own strategy selection is optimized according to the reward value, and ultimately the network resource allocation management of the entire network is realized.

[0027] Based on this, a heterogeneous fusion computing network resource management method and device provided in an embodiment of the present invention, combining non-cooperative game and cooperative game, can realize efficient and accurate computing network resource management under diversified communication systems.

[0028] To facilitate understanding of this embodiment, a heterogeneous fusion computing network resource management and control method disclosed in an embodiment of the present invention is introduced in detail below.

[0029] An embodiment of the present invention provides a method for managing and controlling resources of a heterogeneous fusion computing network, which is applied to a computing network system including a satellite link layer and a ground link layer. The space-based network corresponding to the satellite link layer adopts a satellite communication system, and the ground-based network corresponding to the ground link layer adopts a ground communication system. Optionally, the ground-based network corresponding to the ground link layer may include a 5G network corresponding to the 5G link layer. The following is an exemplary introduction using the 5G network corresponding to the 5G link layer as an example. The method can be executed by an electronic device with data processing capabilities, which can be a server device of an edge computing platform (i.e., a virtualization platform) added to the computing network system.

[0030] It should be noted that the ground-based network corresponding to the ground link layer is not limited to the 5G network corresponding to the 5G link layer. Other networks, such as 6G networks, may also be used in other embodiments.

[0031] In order to meet the actual situation of the current computing power network multi-communication system, the resource management method based on the new game model proposed in this embodiment sets the specific application scenario as a multi-communication system computing power network scenario with a combination of satellite communication and 5G link communication. Two communication systems are adopted: satellite communication and 5G link communication. Due to its advantages such as low latency, low cost, wide coverage, and large bandwidth, the satellite communication system is suitable for application in a computing power network environment where the network traffic is constantly expanding and the scale of Internet terminals is increasing massively. Figure 2 This is the content cache structure diagram of the 5G network. As a new type of mobile communication network, the communication nodes in the 5G network have storage capabilities and can meet the needs of coordinated computing power scheduling in the computing network.

[0032] like Figure 3 As shown in the figure, the system is generally divided into two layers. The upper layer is the satellite link layer. Due to the limited resources of satellite nodes, the user terminal devices in the computing network can directly unload traffic with satellite nodes by adding an edge computing platform (i.e., a virtualization platform). The user terminal devices can be any computing network user terminals connected to the system, such as ground-based users, air-based users, space-based users, and sea-based users. The lower layer is the 5G link layer. The ground 5G network is composed of an access network (i.e., a 5G base station), a bearer network, and a core network based on a virtualization platform.

[0033] Figure 3 The satellite link layer and the 5G link layer can interact between layers, and adjacent nodes in each layer can interact within the layer. Figure 3It can be seen that each node participating in the game in the network does not need to transmit data to the central node, but only needs to exchange information with adjacent nodes, thus saving more communication costs. At the same time, there is no single point of failure problem, which greatly improves the robustness of the system. Since the entire game optimization process does not rely on the central node, the scalability of the network is enhanced. It is applied to computing power network scenarios and has the characteristics of fast optimization convergence speed and less computing resource occupation.

[0034] See also Figure 4 The flowchart of a heterogeneous fusion computing network resource management method shown in FIG. 1 mainly includes the following steps S410 to S430: Step S410, obtaining demand parameter data, current bandwidth resource data and current task data of the computing power network system; wherein the demand parameter data is parameter data of the market demand function corresponding to the computing power network system.

[0035] The demand parameter data, current bandwidth resource data and current task data of the computing power network system can be obtained in real time to realize real-time dynamic resource allocation. Of course, the protection scope of the present invention is not limited to this. In other embodiments, data acquisition and resource allocation can also be performed according to a preset frequency.

[0036] The above-mentioned demand parameter data can be used to calculate the utility function of the satellite link layer and the ground link layer during the non-cooperative game between the layers. The demand parameter data is the parameter data of the market demand function corresponding to the computing power network system. The market demand function is a function that reflects the relationship between the demand for bandwidth resources and the price of bandwidth resources. A possible market demand function is a linear function. In this case, the demand parameter data may include a slope value and an intercept value. The demand parameter data can be estimated based on historical data, or it can be updated regularly or in real time.

[0037] In the computing power network, different computing power providers determine the total computing power supply that can be provided. When building the market demand function, it is necessary to define market parameters, and estimate and update the parameters of the market demand function through historical market data and rolling window algorithms. Based on this, the demand parameter data can be obtained in the following ways: obtain the historical market data of the computing power network system within a preset time range, and the historical market data includes historical bandwidth demand information, price information and cost information; according to the historical market data, estimate the parameters of the preset market demand function through the rolling window algorithm to obtain the demand parameter data of the computing power network system.

[0038] The above-mentioned preset time range can be set according to actual needs, and this embodiment does not limit this. For example, the preset time range can be the last six months or the last year. In a possible implementation, in order to improve the accuracy of the demand parameter data, before performing parameter estimation, the historical market data can be cleaned and feature extracted to obtain the target features corresponding to the historical bandwidth demand information, price information and cost information. When performing parameter estimation, the data in the rolling window is used. For example, the data of the last 30 days can be selected as the rolling window, and the window data is used for linear regression to obtain the demand parameter data.

[0039] The above-mentioned current bandwidth resource data may be the total bandwidth resources of the system. When allocating resources between layers, the sum of the total bandwidth resources of the satellite link layer and the total bandwidth resources of the ground link layer cannot exceed the current bandwidth resource data. The above-mentioned current task data may include a task identifier, a task type, and a task priority, etc. Among them, the task identifier is used to identify specific task information, and the task type is divided into two types: computing tasks and storage tasks. The task priority is used to determine the priority of using resources after the task is assigned to a specific node. The current task data can be used for switching and merging alliances during cooperative games within the layer, where the alliance is formed by offloading tasks to other nodes.

[0040] Step S420, according to the demand parameter data and the current bandwidth resource data, resources are allocated between the satellite link layer and the terrestrial link layer through non-cooperative game to obtain the total bandwidth resources of the satellite link layer and the total bandwidth resources of the terrestrial link layer.

[0041] Non-cooperative games are generally composed of participants, strategy sets, cost functions, and equilibrium points. In this embodiment, the participants are the node sets of the satellite link layer and the node sets of the ground link layer. The strategy set refers to the set of possible bandwidth resource allocation results. The cost function is also the utility function. The equilibrium point refers to a stable strategy combination, that is, a stable bandwidth resource allocation result. In non-cooperative games, all participants are selfish and rational, and their cost functions often conflict. The parties to the game adopt the best strategy to meet their needs under the given game environment and rules. In this embodiment, under the constraint of the total bandwidth resources of the given system, the two parties to the game construct a cost function, seek the Nash equilibrium point, and allocate the network resources of the entire system between layers.

[0042] Since the total bandwidth resources of the system are certain, the inter-layer game can regard the two-layer network as two entities in a competitive relationship, and the Cournot game is a classic oligopoly competition game, which is applicable to the inter-layer game scenario in this embodiment. In the two-layer network, each layer of the network acts as a game participant and decides its bandwidth resource allocation strategy. Based on this, in some possible embodiments, step S420 may include: under the constraint of the given current bandwidth resource data, establishing a first utility function model of the satellite link layer and the ground link layer under a non-cooperative game; according to the demand parameter data, the first utility function model is analyzed and solved to obtain the total bandwidth resources of the satellite link layer and the total bandwidth resources of the ground link layer.

[0043] In one possible implementation, the market demand function is linear and the price is , i = M , N , the first utility function model includes:

[0044]

[0045] in, is the node set of the satellite link layer M The corresponding utility value, is the node set of the ground link layer N The corresponding utility value, R M is the node set of the satellite link layer M The corresponding price, R N is the node set of the ground link layer N The corresponding price, C M is the node set of the satellite link layer M The corresponding cost, C N is the node set of the ground link layer N The corresponding cost, a is the intercept value in the demand parameter data, b = k / (- B ), k is the slope value in the demand parameter data, B is the current bandwidth resource data, b M is the total bandwidth resource of the satellite link layer, b N is the total bandwidth resource of the ground link layer, c M is the cost per allocated unit bandwidth of the satellite link layer,c N Cost per allocated unit bandwidth of the ground link layer.

[0046] In a possible implementation, in order to prevent the system from not converging, the inter-layer game is solved analytically, and the utility function is respectively b M and b N Take the derivative and set the derivative to zero to calculate the optimal solution. Then, the demand parameter data, the cost per allocated unit bandwidth of the satellite link layer, and the cost per allocated unit bandwidth of the ground link layer are substituted into the following analytical solution of the first utility function model to obtain the total bandwidth resources of the satellite link layer and the total bandwidth resources of the ground link layer:

[0047]

[0048] in, b M is the total bandwidth resource of the satellite link layer, b N is the total bandwidth resource of the ground link layer, a is the intercept value in the demand parameter data, b = k / (- B ), k is the slope value in the demand parameter data, B is the current bandwidth resource data, c M is the cost per allocated unit bandwidth of the satellite link layer, c N Cost per allocated unit bandwidth of the ground link layer.

[0049] Through the above solution steps, the network bandwidth resources allocated to each layer by the system in the inter-layer game can be obtained. This result is used as the total bandwidth resources within the layer to impose conditional constraints on the subsequent intra-layer cooperative game optimization problem.

[0050] Step S430, according to the total bandwidth resources of the satellite link layer, the total bandwidth resources of the ground link layer and the current task data, the satellite link layer and the ground link layer are respectively allocated intra-layer resources through cooperative game to obtain the node resource allocation result of the computing power network system.

[0051] The satellite link layer and the ground link layer have the same intra-layer game mode. Taking the satellite link layer as an example, for the intra-layer inter-node game, the ultimate optimization goal is to solve the optimal resource allocation strategy between the computing network nodes. The strategy selection of each communication node in the network is the best decision made for the strategies of other nodes in the layer, and has nothing to do with the lower-level participants (i.e., the ground link layer). At the same time, any satellite node in the layer needs to tend to an equilibrium point. Cooperative game refers to a game in which some participants form an alliance and cooperate. Taking the satellite link layer as an example, the participants refer to the nodes in the satellite link layer. The alliance is formed by offloading tasks to other nodes, that is, the nodes with task offloading relationships constitute an alliance. When using cooperative game to study the resource allocation problem between satellite communication link nodes, it is first necessary to establish a node utility function model and set the optimization goal as the maximum benefit of the intra-layer alliance.

[0052] In some possible embodiments, step S430 may include: for each target link layer in the satellite link layer and the ground link layer, establishing a second utility function model of the target link layer under a cooperative game; wherein the target link layer is a satellite link layer or a ground link layer; taking each node in the target link layer as a single-node alliance to form an initial alliance structure; according to the initial alliance structure, the total bandwidth resources of the target link layer and the current task data, with maximizing the total alliance benefit as the optimization goal, performing alliance switching and merging operations to obtain a target alliance structure; wherein the alliance is formed by offloading tasks to other nodes, and the total alliance benefit is calculated based on the second utility function model; and determining the resource allocation result under the target alliance structure as the node resource allocation result of the target link layer.

[0053] When switching and merging alliances, it is necessary to satisfy that the union of each alliance set is the node set within the layer, and the intersection of any two alliances is empty. Alliance switching and merging operations can be performed according to the set rules: randomly select a computing power node within the layer, select and join an alliance within its communication range, which is called the first alliance, and then randomly select an optional alliance, which is called the second alliance. The total system utility value corresponding to the computing power node in the two alliances is calculated respectively. If the total system utility value corresponding to the first alliance is greater than the total system utility value corresponding to the second alliance, it means that the alliance selected by the node is correct. Otherwise, the node is added to the second alliance, and the current alliance structure is updated. The above process is repeated until the alliance structure that satisfies the game to reach a balanced state is found.

[0054] In this embodiment, maximizing the alliance revenue is used as the system optimization goal, and after the game iteration convergence, a stable alliance structure is obtained. Since cooperative games are used within the layer, collaboration and information sharing between computing nodes are required, and the particle swarm optimization algorithm (PSO for short) can be used for solving. Based on the above intra-layer game model, each particle represents a computing node within the layer and its resulting bandwidth resources, and the optimal resource allocation strategy is found through iteration. After continuous iterations, each iteration will adjust the bandwidth resources obtained by the computing nodes within the layer to improve the overall efficiency of the system.

[0055] The heterogeneous fused computing network resource management and control method provided in the embodiment of the present invention adopts a layered game method for a heterogeneous fused computing network system with two communication systems, satellite communication and ground communication. Non-cooperative game is adopted between layers, and cooperative game is adopted within the layer. The combination of the two game methods can effectively allocate network resources in the computing network system reasonably on demand, and realize efficient and accurate computing network resource management and control under diversified communication systems.

[0056] For ease of understanding, the following takes the ground-based network corresponding to the ground link layer as the 5G network corresponding to the 5G link layer as an example to introduce the above-mentioned heterogeneous fusion computing network resource management and control method in detail.

[0057] This embodiment can solve the following problems: 1. In the current computing network resource management and control methods, the communication system considered is single, without combining with the current development and changes. The communication system in the computing network will face diversification. Satellite-ground integration is the future network development trend. The communication system involved in the network is no longer single, but diversified. Faced with the diversified communication systems in the system, how to conduct efficient and accurate computing network resource management and control is the focus of current research.

[0058] 2. Traditional computing network resource allocation methods have many shortcomings. Among them, static allocation cannot dynamically adjust resources for allocation, which will lead to low resource utilization; while dynamic allocation can improve resource utilization, it requires higher management overhead and more complex algorithms to support it. In the computing network resource management method based on game theory, different game methods used will also have a certain impact on the results. For example, when a single non-cooperative game is used, since each participant in the non-cooperative game does not care whether the selected strategy will have an adverse impact on the benefits of other game participants, the result may not be the global optimal solution to the optimization problem.

[0059] This embodiment comprehensively considers the two communication systems in the computing network, combines satellite communication with 5G communication, and accurately characterizes the complex computing network system by constructing a new game model of multiple communication systems. After the inter-layer game and the intra-layer game work together, the inter-layer game regards the satellite link layer and the 5G link layer as a whole, and a non-cooperative game is played between the two to obtain the allocation of network resources to each layer in the entire system. After the inter-layer game is completed, the intra-layer game is carried out. The network nodes in each layer adopt a cooperative game. Under given constraints, find the equilibrium state that maximizes the benefits for each node in the layer. Record the network resources allocated to each node in the system in this state, which is the optimal solution to the resource allocation optimization problem of the computing network system. By comprehensively considering the distribution of network resources and the network status in the system, the online scheduling and allocation of computing network resources under multiple communication systems can be achieved.

[0060] The new hierarchical game model proposed in this embodiment is described in detail below. Figure 5 As shown, in this embodiment, the satellite communication network nodes are selected as the upper-layer game subjects, and the ground 5G link communication network nodes are selected as the lower-layer game subjects, taking into account the resource allocation strategy selection of nodes in the two different communication system networks. The inter-layer game under the two different communication systems and the game between nodes in each layer adopt different forms. The links under the two communication systems share the resources of the entire system. First, the upper network (i.e., the space-based network corresponding to the satellite link layer) and the lower network (i.e., the ground-based network corresponding to the ground link layer) conduct inter-layer non-cooperative games to obtain the relevant network bandwidth resources allocated in each layer. After that, the communication nodes in each layer of the network conduct cooperative games, and under given constraints, find the equilibrium state that maximizes the benefits for each node in the layer. Each node in the layer is a single-node alliance to form an initial alliance structure. Each node searches for a set of communicable nodes within its communication range. Since the network node devices in the system do not generate bandwidth resource requirements at every moment, other nodes in the system can independently choose to offload tasks to other idle nodes within their communication range, and relieve the pressure of a single node through this collaboration. In the process of intra-layer game, the alliance is formed by offloading tasks to other nodes. The alliance is switched and merged according to the set rules, and after repeated iterations, it finally converges to a stable state, thereby improving the resource allocation and control efficiency between the entire computing power network system.

[0061] 1) Non-cooperative game between layers like Figure 6The inter-layer game model shown first analyzes the inter-layer non-cooperative game. The non-cooperative game is generally composed of participants, strategy sets, cost functions, equilibrium points, etc. In non-cooperative games, all participants are selfish and rational, and their cost functions often conflict. The parties to the game adopt the best strategy to meet their needs under the given game environment and rules. In this embodiment, under the constraint of the total bandwidth resources of the given system, the two parties to the game construct a cost function, seek the Nash equilibrium point, and allocate the network resources of the entire system between layers.

[0062] Since the total bandwidth resources of a given system are certain, the inter-layer game can regard the two-layer network as two entities in a competitive relationship, and the Cournot game is a classic oligopoly competition game, which is applicable to the inter-layer game scenario in this embodiment. In the two-layer network, each layer of the network acts as a game participant and decides its bandwidth resource allocation strategy.

[0063] Participants are decision-makers, represented by symbols ,in M represents the set of nodes in the space-based network, N Represents the node set in the 5G network. M and N The allocated bandwidth resources are used b M and b N Indicates that the total bandwidth resource of the entire system is defined as To obtain the utility function of the participants, assume that the market demand function is linear and the price is , a and b It is a parameter related to external factors such as the actual network topology (i.e., demand parameter data), as shown in formula (3), which is the mathematical representation of the utility function. c M and c N Represents the cost per allocated unit bandwidth in satellite network and 5G network respectively.

[0064] (3) In order to prevent the system from not converging, the inter-layer game is solved here, and the utility function is respectively b M and b N Take the derivative and set the derivative to zero to calculate the optimal solution. The analytical solution can be obtained as shown in formula (4) and formula (5).

[0065] (4) (5) Through the above solution steps, the network bandwidth resources allocated to each layer by the system in the inter-layer game can be obtained. This result is used as the total bandwidth resources within the layer to impose conditional constraints on the subsequent intra-layer cooperative game optimization problem.

[0066] 2) Intra-layer cooperative game The satellite link layer and the 5G link layer have the same intra-layer game mode. Taking the satellite link layer as an example, for the intra-layer inter-node game, the ultimate optimization goal is to solve the optimal resource allocation strategy between the computing network nodes. The strategy selection of each communication node in the network is the best decision made for the strategies of other nodes in the layer, which has nothing to do with the participants in the lower layer. At the same time, any satellite node in the layer needs to tend to an equilibrium point. When using cooperative game to study the resource allocation problem between satellite communication link nodes, it is first necessary to establish a node utility function model and set the optimization goal as the maximum benefit of the intra-layer alliance.

[0067] like Figure 7 The intra-layer game network topology diagram shown in the figure shows that in the cooperative game in each layer, taking the intra-layer game of the 5G network as an example, it is assumed that there are n users (i.e., computing nodes) in the layer competing for the network resources in the layer. These computing nodes can cooperate with each other, and the constraint protocol is set to ensure that each user node in the layer can reasonably and efficiently allocate system network resources after the cooperative game. When all nodes comply with the protocol, a cooperative game alliance is formed. Through the constraints of the protocol, different communication nodes in the layer adopt strategies according to the rules, thereby ensuring the maximization of system benefits. Define a cooperative alliance game as ,in U Represents the set of game participants, that is, the set of all computing nodes in the layer, V represents the coalition utility, which is defined in A real-valued map on , where Represents the set of all possible subsets of participants. Set U Any subset of can be viewed as a union, including the empty set and the full set.

[0068] For the Alliance The characteristic function represents and The maximum utility of the alliance in the game can be expressed by formula (6). V i It can be calculated using formula (7), which represents the i Computing nodes u i The utility function of Set to a fixed value related to the specific network topology to represent the utility value brought by unit bandwidth; is the penalty coefficient. When the actual network resources allocated to the node do not meet the bandwidth requirements, a corresponding penalty will be incurred. b i Indicates the amount of bandwidth actually allocated to the node. d i Indicates the bandwidth demand of the node.

[0069] (6) (7) In order to determine whether the nodes form an alliance, a discriminant variable is defined It is used to indicate whether there is a task offloading relationship between computing power nodes. A value of 1 indicates that there is task offloading between the two nodes, and a value of 0 indicates that there is no task offloading.

[0070] At any time after the intra-layer game begins, there will be an alliance structure divided into , which means that the node set U Divide into alliances, including represents the set of all possible alliance structures, n represents the number of nodes, and must satisfy the following formula (8). In the initial stage of the game, a single-node alliance (i.e., each node acts as an alliance) is used to obtain the initial alliance structure: .

[0071] (8) After obtaining the initial single-node alliance structure, the alliance switching and merging operations are performed according to the set rules: Since the alliance state is not stable yet, a computing power node in the layer is randomly selected and selected and joined in an alliance within its communication range. , and then randomly select an optional alliance , respectively calculate the computing power node in the alliance Alliance The corresponding total system utility is If the node is not in the correct alliance, then the node is added to the alliance. , update the current alliance structure, and repeat the above process until an alliance structure that satisfies the game equilibrium is found.

[0072] This intra-layer game model takes the maximization of alliance benefits as the system optimization goal, and after the game iteration converges, a stable alliance structure is obtained. Therefore, the total benefit of the intra-layer game alliance can be described as formula (9).

[0073] (9) The optimization problem can be further expressed as shown in formula (10): (10) in, b N is the total bandwidth resource allocated to the 5G link layer after inter-layer game. There are three constraints in the optimization problem: r 1 means that for bandwidth resource allocation tasks, the system processing time t Cannot exceed the set maximum time t max ,The processing time of the system refers to the time that will be recorded directly after the system is running, excluding the time required for the non-cooperative game between layers.,This constraint constrains the time allocated for a game to meet the requirements; r 2 indicates that the network bandwidth resource value allocated to each computing power node in the layer is non-negative; r 3 means that the total network bandwidth resources allocated to all computing nodes does not exceed the total bandwidth resources allocated to the 5G link layer after inter-layer game.

[0074] Since cooperative games are used within the layer, collaboration and information sharing between computing nodes are required, and the particle swarm optimization algorithm can be used to solve it. Based on the above intra-layer game model, each particle represents an intra-layer computing node and its resulting bandwidth resources, and the optimal resource allocation strategy is found through iteration. After continuous iteration of the algorithm, each iteration will adjust the bandwidth resources obtained by the computing nodes within the layer to improve the overall efficiency of the system.

[0075] The embodiments of the present invention have the following technical effects: Compared with the existing computing power network resource management and control technology, the embodiment of the present invention adopts a new game model that is more in line with the actual computing network environment. Taking into account the current trend of satellite-ground integrated network, satellite communication and ground 5G link communication are introduced into the computing power network system. A layered game model is constructed, and non-cooperative games are adopted between layers, and cooperative federal games are adopted within the layers. The combination of the two game modes can effectively allocate network resources in the computing network system reasonably on demand.

[0076] like Figure 8 As shown in the figure, the computing network user terminal sends and unloads tasks to the satellite link layer and the 5G link layer. After the two-layer network receives the task information, it first conducts a non-cooperative game between the layers to allocate the total network resources of each layer. After that, the two-layer network simultaneously conducts an intra-layer cooperative federal game. According to the above process, the PSO algorithm is used to perform secondary computing network resource allocation between the computing nodes in the layer. By comprehensively considering the distribution of network resources and the network status in the system, the online scheduling and allocation of computing network resources under the multi-communication system is finally realized.

[0077] Corresponding to the above-mentioned heterogeneous fusion computing network resource management and control method, the embodiment of the present invention also provides a heterogeneous fusion computing network resource management and control device, which is applied to a computing network system including a satellite link layer and a ground link layer. The space-based network corresponding to the satellite link layer adopts a satellite communication system, and the ground-based network corresponding to the ground link layer adopts a ground communication system. Fig. 9 The structure diagram of a heterogeneous fusion computing network resource management and control device shown in the figure includes: The data acquisition module 901 is used to acquire the demand parameter data, current bandwidth resource data and current task data of the computing power network system; wherein the demand parameter data is the parameter data of the market demand function corresponding to the computing power network system, and the market demand function is a function reflecting the relationship between the bandwidth resource demand and the bandwidth resource price; A first allocation module 902 is used to allocate resources between the satellite link layer and the terrestrial link layer through a non-cooperative game according to the demand parameter data and the current bandwidth resource data, so as to obtain the total bandwidth resources of the satellite link layer and the total bandwidth resources of the terrestrial link layer; The second allocation module 903 is used to allocate resources within the satellite link layer and the ground link layer respectively through cooperative game according to the total bandwidth resources of the satellite link layer, the total bandwidth resources of the ground link layer and the current task data, so as to obtain the node resource allocation result of the computing power network system.

[0078] The heterogeneous fused computing network resource management and control device provided in the embodiment of the present invention adopts a layered game method for a heterogeneous fused computing network system with two communication systems, satellite communication and ground communication. Non-cooperative game is adopted between layers, and cooperative game is adopted within the layer. The combination of the two game methods can effectively allocate network resources in the computing network system reasonably on demand, and realize efficient and accurate computing network resource management and control under diversified communication systems.

[0079] Furthermore, the above-mentioned data acquisition module 901 is specifically used to: obtain the historical market data of the computing power network system within a preset time range, and the historical market data includes historical bandwidth demand information, price information and cost information; according to the historical market data, the parameters of the preset market demand function are estimated through a rolling window algorithm to obtain the demand parameter data of the computing power network system.

[0080] Furthermore, the above-mentioned first allocation module 902 is specifically used to: establish a first utility function model of the satellite link layer and the ground link layer under a non-cooperative game under the constraint condition of the given current bandwidth resource data; and obtain an analytical solution for the first utility function model according to the demand parameter data to obtain the total bandwidth resources of the satellite link layer and the total bandwidth resources of the ground link layer.

[0081] Furthermore, the first utility function model includes:

[0082]

[0083] in, is the node set of the satellite link layer M The corresponding utility value, is the node set of the ground link layer N The corresponding utility value, R M is the node set of the satellite link layer M The corresponding price, R N is the node set of the ground link layer N The corresponding price, C M is the node set of the satellite link layer M The corresponding cost, C N is the node set of the ground link layer N The corresponding cost, a is the intercept value in the demand parameter data, b = k / (- B ), k is the slope value in the demand parameter data, B is the current bandwidth resource data, b M is the total bandwidth resource of the satellite link layer, b N is the total bandwidth resource of the ground link layer, c M is the cost per allocated unit bandwidth of the satellite link layer, c N Cost per allocated unit bandwidth of the ground link layer.

[0084] Further, the first allocation module 902 is further used to: bring the demand parameter data, the cost per allocated unit bandwidth of the satellite link layer, and the cost per allocated unit bandwidth of the terrestrial link layer into the following analytical solution of the first utility function model to obtain the total bandwidth resources of the satellite link layer and the total bandwidth resources of the terrestrial link layer:

[0085]

[0086] in,b M is the total bandwidth resource of the satellite link layer, b N is the total bandwidth resource of the ground link layer, a is the intercept value in the demand parameter data, b = k / (- B ), k is the slope value in the demand parameter data, B is the current bandwidth resource data, c M is the cost per allocated unit bandwidth of the satellite link layer, c N is the cost per allocated unit bandwidth of the ground link layer, B The current bandwidth resource data.

[0087] Furthermore, the second allocation module 903 is specifically used to: for each target link layer in the satellite link layer and the ground link layer, establish a second utility function model of the target link layer under a cooperative game; regard each node in the target link layer as a single-node alliance to form an initial alliance structure; according to the initial alliance structure, the total bandwidth resources of the target link layer and the current task data, with maximizing the total alliance benefit as the optimization goal, perform alliance switching and merging operations to obtain a target alliance structure; wherein the alliance is formed by offloading tasks to other nodes, and the total alliance benefit is calculated based on the second utility function model; and determine the resource allocation result under the target alliance structure as the node resource allocation result of the target link layer.

[0088] Furthermore, the ground-based network corresponding to the above-mentioned ground link layer includes a 5G network corresponding to the 5G link layer.

[0089] The heterogeneous fusion computing network resource control device provided in this embodiment has the same implementation principle and technical effects as those of the aforementioned heterogeneous fusion computing network resource control method embodiment. For the sake of brief description, for matters not mentioned in the embodiment of the heterogeneous fusion computing network resource control device, reference may be made to the corresponding contents in the aforementioned heterogeneous fusion computing network resource control method embodiment.

[0090] like Fig.10 As shown, an electronic device 1000 provided by an embodiment of the present invention includes: a processor 1001, a memory 1002 and a bus, the memory 1002 stores a computer program that can be run on the processor 1001, when the electronic device 1000 is running, the processor 1001 and the memory 1002 communicate through the bus, and the processor 1001 executes the computer program to implement the above-mentioned heterogeneous fusion computing network resource management and control method.

[0091] Specifically, the memory 1002 and the processor 1001 can be general-purpose memories and processors, which are not specifically limited here.

[0092] 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 heterogeneous fusion computing network resource management and control method described in the previous method embodiment is executed. The computer-readable storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a RAM, a magnetic disk or an optical disk, and other media that can store program codes.

[0093] The term "and / or" herein is only a description of the association relationship of the associated objects, indicating that there may be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the term "at least one" herein represents any combination of at least two of any one or more of a plurality of. For example, including at least one of A, B, and C can represent including any one or more elements selected from the set consisting of A, B, and C.

[0094] In all examples shown and described herein, any specific values ​​should be interpreted as merely exemplary and not as limiting, and thus other examples of the exemplary embodiments may have different values.

[0095] The flow charts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the devices, methods and computer program products according to multiple embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flow chart, and the combination of the boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0096] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0097] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0098] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0099] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for managing and controlling heterogeneous fusion computing network resources, characterized in that: Applicable to a computing network system including a satellite link layer and a ground link layer, wherein the space-based network corresponding to the satellite link layer adopts a satellite communication system, and the ground-based network corresponding to the ground link layer adopts a ground communication system; The heterogeneous fusion computing network resource management and control method includes: Obtaining demand parameter data, current bandwidth resource data, and current task data of the computing power network system; wherein the demand parameter data is parameter data of a market demand function corresponding to the computing power network system, and the market demand function is a function that reflects the relationship between bandwidth resource demand and bandwidth resource price; According to the demand parameter data and the current bandwidth resource data, resources are allocated between the satellite link layer and the terrestrial link layer through a non-cooperative game to obtain the total bandwidth resources of the satellite link layer and the total bandwidth resources of the terrestrial link layer; According to the total bandwidth resources of the satellite link layer, the total bandwidth resources of the ground link layer and the current task data, the satellite link layer and the ground link layer are respectively allocated resources within the layer through cooperative game to obtain the node resource allocation result of the computing power network system.

2. The heterogeneous fusion computing network resource management and control method according to claim 1 is characterized in that: The obtaining of the demand parameter data of the computing power network system includes: Acquire historical market data of the computing power network system within a preset time range, wherein the historical market data includes historical bandwidth demand information, price information, and cost information; According to the historical market data, the parameters of the preset market demand function are estimated through a rolling window algorithm to obtain the demand parameter data of the computing power network system.

3. The heterogeneous fusion computing network resource management and control method according to claim 1 is characterized in that: The allocating resources between the satellite link layer and the terrestrial link layer through non-cooperative game according to the demand parameter data and the current bandwidth resource data to obtain the total bandwidth resources of the satellite link layer and the total bandwidth resources of the terrestrial link layer includes: Under the constraint condition of the given current bandwidth resource data, establishing a first utility function model of the satellite link layer and the ground link layer under a non-cooperative game; An analytical solution is obtained for the first utility function model according to the demand parameter data to obtain the total bandwidth resources of the satellite link layer and the total bandwidth resources of the ground link layer.

4. The heterogeneous fusion computing network resource management and control method according to claim 3 is characterized in that: The first utility function model includes: in, is the node set of the satellite link layer M The corresponding utility value, is the node set of the ground link layer N The corresponding utility value, R M is the node set of the satellite link layer M The corresponding price, R N is the node set of the ground link layer N The corresponding price, C M is the node set of the satellite link layer M The corresponding cost, C N is the node set of the ground link layer N The corresponding cost, a is the intercept value in the demand parameter data, b = k / (- B ), k is the slope value in the demand parameter data, B is the current bandwidth resource data, b M is the total bandwidth resource of the satellite link layer, b N is the total bandwidth resource of the ground link layer, c M is the cost per allocated unit bandwidth of the satellite link layer, c N Cost per allocated unit bandwidth of the ground link layer.

5. The heterogeneous fusion computing network resource management and control method according to claim 3 is characterized in that: The step of obtaining an analytical solution for the first utility function model according to the demand parameter data to obtain the total bandwidth resources of the satellite link layer and the total bandwidth resources of the ground link layer includes: Substituting the demand parameter data, the cost per allocated unit bandwidth of the satellite link layer, and the cost per allocated unit bandwidth of the terrestrial link layer into the following analytical solution of the first utility function model, the total bandwidth resources of the satellite link layer and the total bandwidth resources of the terrestrial link layer are obtained: in, b M is the total bandwidth resource of the satellite link layer, b N is the total bandwidth resource of the ground link layer, a is the intercept value in the demand parameter data, b = k / (- B ), k is the slope value in the demand parameter data, B is the current bandwidth resource data, c M is the cost per allocated unit bandwidth of the satellite link layer, c N Cost per allocated unit bandwidth of the ground link layer.

6. The heterogeneous fusion computing network resource management and control method according to claim 1 is characterized in that: The method of allocating resources within the satellite link layer and the ground link layer respectively according to the total bandwidth resources of the satellite link layer, the total bandwidth resources of the ground link layer and the current task data through cooperative game to obtain the node resource allocation result of the computing power network system includes: For each target link layer in the satellite link layer and the terrestrial link layer, establishing a second utility function model of the target link layer under cooperative game; Taking each node in the target link layer as a single node alliance to form an initial alliance structure; According to the initial alliance structure, the total bandwidth resources of the target link layer and the current task data, with maximizing the total alliance benefit as the optimization goal, switching and merging alliances are performed to obtain a target alliance structure; wherein the alliance is formed by offloading tasks to other nodes, and the total alliance benefit is calculated based on the second utility function model; The resource allocation result under the target alliance structure is determined as the node resource allocation result of the target link layer.

7. The heterogeneous fusion computing network resource management and control method according to any one of claims 1 to 6, characterized in that: The ground-based network corresponding to the ground link layer includes a 5G network corresponding to the 5G link layer.

8. A heterogeneous fusion computing network resource management and control device, characterized in that: Applicable to a computing network system including a satellite link layer and a ground link layer, wherein the space-based network corresponding to the satellite link layer adopts a satellite communication system, and the ground-based network corresponding to the ground link layer adopts a ground communication system; The heterogeneous fusion computing network resource management and control device includes: A data acquisition module, used to acquire demand parameter data, current bandwidth resource data and current task data of the computing power network system; wherein the demand parameter data is parameter data of a market demand function corresponding to the computing power network system, and the market demand function is a function reflecting the relationship between bandwidth resource demand and bandwidth resource price; A first allocation module is used to allocate resources between the satellite link layer and the terrestrial link layer through a non-cooperative game according to the demand parameter data and the current bandwidth resource data, so as to obtain the total bandwidth resources of the satellite link layer and the total bandwidth resources of the terrestrial link layer; The second allocation module is used to allocate resources within the satellite link layer and the ground link layer respectively through cooperative game according to the total bandwidth resources of the satellite link layer, the total bandwidth resources of the ground link layer and the current task data, so as to obtain the node resource allocation result of the computing power network system.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, characterized in that: When the processor executes the computer program, the heterogeneous fusion computing network resource management and control method described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the heterogeneous fusion computing network resource management and control method described in any one of claims 1 to 7 is executed.

Images