Spatial computing automatic networking methods, devices, equipment, media and products

Abstract

This application discloses a method, apparatus, device, medium, and product for automatic spatial computing networking, relating to the field of computer technology. The method includes: constructing a super-object model; performing cluster analysis on at least two devices based on the super-object model to obtain an initial network topology; obtaining the current device state of the devices corresponding to the initial network topology; inputting the current device state into a pre-trained reinforcement learning policy model and outputting a first policy; updating the initial network topology according to the first policy to obtain a first network topology; determining the network performance of the first network topology; and if the network performance reaches a preset performance threshold, determining the first network topology as the target network topology. This application improves the overall network performance and resource utilization efficiency.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to methods, apparatus, equipment, media and products for automatic networking of spatial computing. Background Art

[0002] With the rapid development of IoT technology, the number of IoT devices and their application scenarios are experiencing explosive growth, posing unprecedented challenges to network management. Traditional static networking methods, based on a fixed network architecture, pre-defined device connection relationships, and resource allocation strategies, are no longer sufficient to meet the demands of today's complex and ever-changing network environments. Static networking determines the network topology, device connection relationships, and resource allocation scheme at the initial network design stage, and these settings typically remain unchanged during network operation. However, with the continuous increase and movement of IoT devices, and the constant changes in business requirements, the limitations of static networking methods are becoming increasingly apparent. Their fixed and immutable nature prevents the network from responding promptly and effectively to dynamic changes such as device additions or removals, location changes, or adjustments in business requirements, thus affecting the overall network performance and reliability. These problems make static networking methods inadequate for dealing with today's complex and dynamic environments, failing to meet the needs of the rapid growth of IoT devices and the diversification of application scenarios, resulting in poor network performance in dynamic IoT networking environments. Summary of the Invention

[0003] The main objective of this application is to provide a method, apparatus, device, medium, and product for automatic spatial computing networking, which aims to solve the technical problem of low network performance in the dynamic networking environment of Internet of Things devices.

[0004] To achieve the above objectives, this application proposes an automatic spatial computing network formation method, comprising:

[0005] Construct a super-object model, and perform cluster analysis on at least two devices based on the super-object model to obtain the initial network topology. The super-object model includes the hardware specifications, dynamic states, and spatial locations of all preset devices.

[0006] Obtain the current device state of the device corresponding to the initial network topology, input the current device state into the pre-trained reinforcement learning policy model, and output the first policy. The reinforcement learning policy model is a policy model built and trained based on the super object model and the initial network topology.

[0007] The initial network topology is updated according to the first strategy to obtain the first network topology.

[0008] Determine the network performance of the first network topology. If the network performance reaches a preset performance threshold, then determine the first network topology as the target network topology.

[0009] In one embodiment, the steps of constructing a super-object model include:

[0010] Generate attribute vectors based on the device's hardware specifications and dynamic status;

[0011] Generate a three-dimensional vector based on the spatial location information of the device;

[0012] Define metadata fields for each device, and generate metadata vectors based on these metadata fields;

[0013] A super object model is constructed based on attribute vectors, 3D vectors, metadata vectors, and a pre-defined time series. The time series represents the data differences that occur when the attribute vectors, 3D vectors, and metadata vectors change over time.

[0014] In one embodiment, the step of performing cluster analysis on at least two devices based on the super-object model to obtain the initial network topology includes:

[0015] Based on the super object model, a preset clustering algorithm is used to group at least two devices to obtain at least one subnet. The clustering algorithm groups devices with similar functions or spatial proximity into the same subnet according to the super object model.

[0016] For each subnet, the first connection relationship between all devices belonging to the subnet is determined, and the subnet structure is optimized based on the first connection relationship. The initial network topology includes each subnet after the structure optimization.

[0017] In one embodiment, the step of updating the initial network topology according to a first strategy to obtain a first network topology includes:

[0018] If the initial network topology has multiple subnets, determine whether there is a policy conflict between the first policies corresponding to each subnet;

[0019] If there is no policy conflict between the first policies corresponding to each subnet, then for each subnet, the network topology of the subnet is updated according to the first policy corresponding to each subnet to obtain the first network topology.

[0020] In one embodiment, after determining whether there is a policy conflict between the first policies corresponding to each subnet, the method further includes:

[0021] If at least two subnets have conflicting first policies, then the subnets with conflicting policies are determined to be conflicting subnets.

[0022] Obtain the global policy model sent by the preset central control unit, wherein the central control unit determines the global network state based on the subnet state of all subnets, and constructs the global policy model based on the global network state;

[0023] The second decision for determining the conflict subnet is determined based on the global strategy model. The network topology of the conflict subnet is then updated based on the second decision to obtain the updated conflict subnet.

[0024] For subnets without policy conflicts, the network topology is updated according to the first policy corresponding to the subnets without policy conflicts to obtain the second subnet;

[0025] The first network topology is determined based on the updated conflict subnet and / or second subnet.

[0026] In one embodiment, the automatic spatial computing networking method further includes:

[0027] The super object model at the preset initial moment is defined as a state vector, the action vector is defined according to the adjustment strategy between devices, and the strategy model is defined based on the mapping relationship between the state vector and the action vector.

[0028] Determine the network performance under the state vector and action vector, and define a reward function based on the network performance. The reward function is used to evaluate the network performance under the state vector.

[0029] The policy model is trained using a learning algorithm based on the reward function to obtain a reinforcement learning policy model.

[0030] Furthermore, to achieve the above objectives, this application also proposes an automatic spatial computing network device, which includes:

[0031] The super-object model generation module constructs a super-object model and performs cluster analysis on at least two devices based on the super-object model to obtain an initial network topology. The super-object model includes the hardware specifications, dynamic status, and spatial location of all preset devices.

[0032] The policy generation module obtains the current device state of the device corresponding to the initial network topology, inputs the current device state into the pre-trained reinforcement learning policy model, and outputs the first policy. The reinforcement learning policy model is a policy model built and trained based on the super object model and the initial network topology.

[0033] The network topology generation module updates the initial network topology structure according to the first strategy to obtain the first network topology structure;

[0034] A performance detection module determines the network performance of the first network topology. If the network performance reaches a preset performance threshold, the first network topology is determined to be the target network topology. Furthermore, to achieve the above objectives, this application also proposes an automatic spatial computing networking device. The device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. The computer program is configured to implement the steps of the automatic spatial computing networking method described above.

[0035] In addition, to achieve the above objectives, this application also proposes a medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the spatial computing automatic networking method described above.

[0036] In addition, to achieve the above objectives, this application also provides a product, which is a computer program product. The computer program product includes a computer program, which, when executed by a processor, implements the steps of the spatial computing automatic networking method described above.

[0037] One or more technical solutions proposed in this application have at least the following technical effects:

[0038] This application constructs a super-object model, which includes the hardware specifications, dynamic states, and spatial locations of all preset devices. By abstracting and integrating device information, this application forms a virtual resource pool, providing basic data support for subsequent network optimization and decision-making. Based on the super-object model, cluster analysis is performed on at least two devices to obtain an initial network topology, thereby optimizing network performance and resource utilization. The current device states of the devices corresponding to the initial network topology are obtained and input into a pre-trained reinforcement learning policy model, outputting a first policy that enables the model to select the optimal action based on the current network state to maximize long-term rewards, such as improving network performance and reducing resource consumption. The reinforcement learning policy model is a policy model constructed and trained based on the super-object model and the initial network topology. The initial network topology is updated according to the first policy to obtain a first network topology. The network performance of the first network topology is determined. If the network performance reaches a preset performance threshold, the first network topology is determined as the target network topology. Through continuous optimization and adjustment of the reinforcement learning policy model, a network topology that meets the network performance requirements is gradually found. In this way, this application can significantly improve the overall network performance and resource utilization efficiency. Attached Figure Description

[0039] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0040] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 This is a flowchart illustrating an embodiment of the automatic networking method for spatial computing in this application.

[0042] Figure 2 This is a flowchart illustrating Embodiment 2 of the spatial computing automatic networking method of this application;

[0043] Figure 3 This is another flowchart illustrating Embodiment 2 of the spatial computing automatic networking method of this application;

[0044] Figure 4 This is a schematic diagram of the module structure of the automatic spatial computing networking device according to an embodiment of this application;

[0045] Figure 5 This is a schematic diagram of the device structure of the hardware operating environment involved in the automatic networking method for spatial computing in the embodiments of this application.

[0046] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0047] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0048] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0049] It should be noted that the executing entity in this embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or an electronic device or terminal system capable of performing the above functions. The following description uses a system as an example to illustrate this embodiment and the subsequent embodiments.

[0050] Based on this, this embodiment provides an automatic spatial computing network formation method, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the automatic spatial computing network formation method of this application. The automatic spatial computing network formation method includes steps S10 to S40:

[0051] Step S10: Construct a super-object model, perform cluster analysis on at least two devices based on the super-object model, and obtain the initial network topology. The super-object model includes the hardware specifications, dynamic status, and spatial location of all preset devices.

[0052] It's important to note that the Super-Object Model is an abstract model that integrates the hardware specifications, dynamic states, and spatial location information of all pre-defined devices in the network. This model provides fundamental data support for automatic network formation, enabling network construction and management to be optimized based on the capabilities of these devices. Hardware specifications refer to the physical configuration of the devices, including the number of CPU cores, memory size, and storage capacity; these are fundamental parameters for device performance. Dynamic states refer to the operational status of the devices, such as current load and available resources; these states change over time and are crucial for real-time network adjustments. Spatial location refers to the precise physical location of the devices, including geographical location and physical altitude, which directly impacts the network's geographical layout and signal coverage. The initial network topology is obtained based on the Super-Object Model using a pre-defined clustering algorithm. It defines the initial connection methods of the devices in the network.

[0053] First, it is necessary to collect the hardware specifications, dynamic status, and spatial location information of all preset devices. This information is integrated into an abstract model, namely the super-object model. This model includes not only the static information of the devices (such as hardware specifications) but also dynamic information (such as current load and available resources) and spatial information such as the geographical location of the devices. Based on the device information in the super-object model, a preset clustering algorithm (such as K-means clustering) is used to generate an initial network topology.

[0054] Step S20: Obtain the current device state of the device corresponding to the initial network topology, input the current device state into the pre-trained reinforcement learning policy model, and output the first policy. The reinforcement learning policy model is a policy model constructed and trained based on the super object model and the initial network topology.

[0055] It should be noted that the current device state refers to the real-time status information of the device at a certain moment, including but not limited to the actual usage of its hardware specifications and dynamic state changes. The reinforcement learning policy model is a machine learning model that makes decisions by continuously trying and learning to maximize a certain reward signal. In this embodiment, it can select the optimal network topology based on the device's current state, and is used to optimize network configuration based on the device state and network topology. The first strategy is a network configuration scheme output by the reinforcement learning policy model based on the current device state and the initial network topology, used to guide the updating of the network topology.

[0056] The system obtains the current device states of the devices corresponding to the initial network topology. This state information includes real-time operating data of the devices, such as power consumption and temperature, as well as their relative positions in the network. This information is then input into a pre-trained reinforcement learning policy model. Based on the input device state information and the initial network topology, the reinforcement learning policy model outputs a first policy. This policy is derived by the model through continuous learning and experimentation, aiming to optimize network configuration and improve network performance.

[0057] Step S30: Update the initial network topology according to the first strategy to obtain the first network topology;

[0058] Step S40: Determine the network performance of the first network topology. If the network performance reaches a preset performance threshold, then determine the first network topology as the target network topology.

[0059] It should be noted that the first network topology is the updated network topology based on the reinforcement learning policy model and state vectors. It represents an intermediate state in the network optimization process. Network performance refers to the various network performance characteristics exhibited by the network system under specific conditions (such as given state and action vectors), including latency, throughput, and resource utilization. These network performance metrics are key parameters for measuring network efficiency and effectiveness. The performance threshold is a preset performance standard used to determine whether the network topology meets the performance requirements. If the network performance of the first network topology reaches this threshold, it will be determined as the final target network topology.

[0060] The generated first policy is used to update the initial network topology. This process involves adjusting the connection relationships or resource allocation strategies between devices based on the action vectors selected by the policy model to generate the first network topology. Next, the network performance of the first network topology is determined. This step is accomplished by measuring key performance indicators such as network throughput and latency. If the network performance reaches a preset performance threshold, the first network topology is determined as the target network topology. This means that the network configuration has been optimized to meet the requirements and can be put into use. If it does not, it may be necessary to further adjust the first policy, update the weight parameters of the first policy, and re-update and evaluate the network topology.

[0061] This embodiment constructs a super-object model, which includes the hardware specifications, dynamic states, and spatial locations of all preset devices. By abstracting and integrating device information, this embodiment forms a virtual resource pool, providing basic data support for subsequent network optimization and decision-making. Based on the super-object model, cluster analysis is performed on at least two devices to obtain an initial network topology to optimize network performance and resource utilization. The current device state of the devices corresponding to the initial network topology is obtained and input into a pre-trained reinforcement learning policy model, outputting a first policy that enables the model to select the optimal action based on the current network state to maximize long-term rewards, such as improving network performance and reducing resource consumption. The reinforcement learning policy model is a policy model constructed and trained based on the super-object model and the initial network topology. The initial network topology is updated according to the first policy to obtain a first network topology. The network performance of the first network topology is determined. If the network performance reaches a preset performance threshold, the first network topology is determined as the target network topology. Through continuous optimization and adjustment of the reinforcement learning policy model, a network topology that meets the network performance requirements is gradually found. In this way, this application can significantly improve the overall network performance and resource utilization efficiency.

[0062] Based on Embodiment 1 of this application, in Embodiment 2 of this application, the content that is the same as or similar to that in Embodiment 1 can be referred to the above description, and will not be repeated hereafter. Based on this, refer to... Figure 2 Step S10 further includes steps A10 to A40:

[0063] Step A10: Generate attribute vectors based on the device's hardware specifications and dynamic status;

[0064] It should be noted that the attribute vector is a vector composed of the device's hardware specifications (such as the number of CPU cores, memory size, network bandwidth, etc.) and dynamic state (such as current load, available resources, etc.), used to describe the device's hardware resources and dynamic state.

[0065] Collect hardware specifications and dynamic status information of the devices. This information includes hardware specifications such as the number of CPU cores, memory size, and network bandwidth, as well as dynamic status information such as current load and available resources. This information is standardized and combined into a vector, namely an attribute vector. This vector comprehensively describes the device's hardware resources and dynamic status.

[0066] Step A20: Generate a three-dimensional vector based on the spatial location information of the device;

[0067] It should be noted that a three-dimensional vector is a vector composed of the spatial location information of a device (such as geographical location, physical location, etc.), and usually includes three coordinate values: x, y, and z, which are used to represent the position of the device in space.

[0068] Collect the spatial location information of the equipment, including geographical location (such as latitude and longitude) and physical spatial location (such as floor, room number, etc.). Encode this information into a three-dimensional vector, where the x, y, and z coordinates represent the equipment's position in different dimensions of space. This three-dimensional vector can accurately represent the equipment's spatial location information.

[0069] Step A30: Define metadata fields for each device and generate metadata vectors based on the metadata fields;

[0070] It should be noted that the metadata vector is a vector composed of additional information fields (such as device type, manufacturer, firmware version, etc.) defined for each device, used to extend the device's descriptive information.

[0071] Define additional information fields for each device, such as device type, manufacturer, firmware version, etc. Combine this information into a vector, namely the metadata vector. This vector can expand the device's descriptive information and provide additional device characteristics.

[0072] Step A40: Construct a super object model based on the attribute vector, 3D vector, metadata vector, and preset time series. The time series represents the data differences generated by the attribute vector, 3D vector, and metadata vector as they change over time.

[0073] It should be noted that time series is a data sequence that represents the change of device status over time, used to record the historical changes and trends of device status.

[0074] Data on device status changes over time is recorded, including historical load and resource utilization. This data is organized into time series to represent historical changes and trends in device status. Time series provide dynamic information about device status, aiding in subsequent network optimization and decision-making. The Super Object Model integrates attribute vectors, time series, 3D vectors, and metadata vectors to form a virtual resource pool that comprehensively describes the device's hardware resources, dynamic status, spatial location, and additional information. In subsequent network optimization and decision-making processes, the Super Object Model will serve as a crucial data foundation, supporting the implementation of automated networking methods.

[0075] For example, the hardware specifications of the device (such as the number of CPU cores, memory size, network bandwidth, etc.), the dynamic state of the device (such as current load, available resources, etc.), and the spatial attributes of the device (such as geographical location, physical location, etc.) can be used as input dependencies. The hardware specifications and dynamic state of the device are collected in a standardized manner and described as an attribute vector D = [d_1, d_2, ..., d_n], where d_i represents the value of the i-th attribute. The spatial attributes of the device are encoded as a three-dimensional vector S = [s_x, s_y, s_z]. Additional metadata fields are defined for each device, such as device type, manufacturer, firmware version, etc., forming a metadata vector M = [m_1, m_2, ..., m_k]. Time series data T = [t_1, t_2, ..., t_m] is used to represent the changes in device state over time, and the device state is monitored in real time, with data in the super-object model being updated periodically. Finally, the attribute vector, three-dimensional vector, metadata vector, and time series data are combined into a super-object model O = {D, S, M, T}, which serves as the basis for subsequent network optimization and decision-making.

[0076] This embodiment successfully constructed a super-object model, providing comprehensive data support for subsequent network optimization and decision-making. This process demonstrates the complete workflow from device information collection to model construction, ensuring the accuracy and practicality of the super-object model.

[0077] In one feasible implementation, refer to Figure 3 Step S10 further includes steps B10 to B20:

[0078] Step B10: Based on the super object model, use a preset clustering algorithm to group at least two devices to obtain at least one subnet. The clustering algorithm groups devices with similar functions or spatial proximity into the same subnet according to the super object model.

[0079] It's important to note that clustering algorithms divide a dataset into multiple clusters or groups, ensuring that data points within the same cluster are similar, while data points between different clusters are significantly different. In this step, devices are grouped into multiple subnets. Within a network, a subnet is a logical grouping of devices that communicate within the subnet through specific connections.

[0080] First, based on the device information (including hardware specifications, dynamic states, spatial attributes, etc.) in the super-object model, clustering algorithms (such as K-means clustering) are used to group the devices. The clustering algorithm divides the devices into different clusters based on their similarity (such as functional attributes, spatial location, etc.), with each cluster representing a subnet. The clustering algorithm groups similar devices into the same subnet by calculating the distance or similarity between them, facilitating subsequent network optimization and management. After clustering, multiple subnets are obtained, each containing a group of similar devices.

[0081] Step B20: For each subnet, determine the first connection relationship between all devices belonging to the subnet, and optimize the subnet structure based on the first connection relationship. The initial network topology includes each subnet after structural optimization.

[0082] Within each subnet, pre-defined connection algorithms (such as shortest path algorithms and spanning tree algorithms) are used to generate connections between devices. These algorithms consider factors such as device hardware resources, dynamic states, and spatial locations to determine the optimal connection paths and methods. The connection algorithms calculate the connection costs and benefits between devices, selecting the optimal connection paths and methods to ensure minimal latency and maximized resource complementarity within the subnet. Within each subnet, the connections between devices are generated, forming the initial network structure. All the resulting subnets (including their internal connections) are then aggregated to form the initial network topology. This topology describes the connections between devices and the layout of the subnets.

[0083] This embodiment uses a clustering algorithm based on a super-object model to group devices into multiple subnets. Then, within each subnet, a connectivity algorithm is used to generate connections between devices. Finally, all subnets are aggregated to form the initial network topology. This structure provides the foundation for subsequent network optimization and decision-making.

[0084] Based on Embodiment 1 or Embodiment 2 of this application, the content in Embodiment 3 of this application that is the same as or similar to Embodiment 1 or Embodiment 2 can be referred to the above description, and will not be repeated hereafter. Step S30 is followed by steps C10 to C20:

[0085] Step C10: If the initial network topology has multiple subnets, determine whether there is a policy conflict between the first policies corresponding to each subnet;

[0086] It should be noted that in this embodiment, each subnet will generate its own first policy. The first policies of different subnets may conflict. A policy conflict occurs when two or more policies are applied at the same time, and their goals, methods or results are contradictory, resulting in the inability to satisfy or achieve them at the same time.

[0087] First, determine if multiple subnets exist in the initial network topology. In large networks, for ease of management and optimization, the network is typically divided into multiple subnets. If multiple subnets do exist in the initial network topology, then conflict detection needs to be performed on the first policy corresponding to each subnet. The specific process of conflict detection is to compare the first policies of each subnet and check whether they contradict each other or cannot be implemented simultaneously. For example, if the first policies of two subnets both require configuring a specific device, but the configuration parameters conflict with each other, then these two policies conflict.

[0088] Furthermore, when there are multiple subnets in the initial network topology, a multi-agent cooperation mechanism can be introduced. An agent is established in each subnet, with different agents responsible for managing their respective subnets. These agents possess the ability to autonomously perceive the environment, process data, and make decisions. They can share state information and policies, and are used to collect subnet state information in real-time or periodically and send it to a central node (such as a central control unit). The central node integrates the state information of all subnets and determines whether there are policy conflicts between different subnets.

[0089] Step C20: If there is no policy conflict between the first policies corresponding to each subnet, then for each subnet, update the network topology of the subnet according to the first policy corresponding to each subnet to obtain the first network topology.

[0090] If, after testing, no conflicts are found between the first policies corresponding to each subnet, then the network topology of each subnet can be updated according to its corresponding first policy. The update process may include adjusting device connectivity, changing link bandwidth, and optimizing device configuration parameters. These operations are performed based on the optimization suggestions in the first policy, aiming to improve network performance, stability, and security.

[0091] This embodiment effectively solves the problem of potential policy conflicts between multiple subnets in the initial network topology by introducing a policy conflict detection mechanism and a multi-agent cooperation mechanism. When no policy conflict is detected, the network topology of each subnet can be accurately updated according to its first policy, thereby optimizing network performance and enhancing stability and security. This process not only avoids contradictions between policies but also ensures the rationality and efficiency of network configuration, providing strong support for the optimized management of large-scale networks. In a feasible implementation, steps C10 may be followed by steps D10 to D30:

[0092] Step D10: If there are at least two subnets whose first policies conflict, then the subnets with conflicting policies are determined to be conflicting subnets.

[0093] It should be noted that subnets with conflicting policies are called conflicting subnets. In a network of subnets, if at least two subnets have conflicting first policies, then these subnets are considered conflicting subnets.

[0094] Step D20: Obtain the global policy model sent by the preset central control unit, wherein the central control unit determines the global network state based on the subnet state of all subnets and constructs the global policy model based on the global network state;

[0095] It should be noted that the central control unit is a device or system component responsible for monitoring and managing the entire network system. It can collect subnet status information from all subnets, determine the global network status based on this information, and build a global policy model to optimize the performance and stability of the entire network.

[0096] Each subnet obtains (or, if a multi-agent cooperation mechanism is introduced, obtains the model from the agents within the subnet) a pre-defined global policy model sent by the central control unit. This model is constructed by the central control unit based on the global network state determined by the subnet state information of all subnets. It considers the needs and states of all subnets and aims to achieve globally optimal network performance.

[0097] Step D30: Determine the second decision for the conflict subnet based on the global policy model, and update the network topology of the conflict subnet based on the second decision to obtain the updated conflict subnet.

[0098] It should be noted that the global policy model is a policy model built upon the global network state, used to guide the configuration and optimization of the entire network. It considers the state and requirements of all subnets, aiming to achieve globally optimal network performance. The second decision is a new network configuration or optimization scheme formulated based on the global policy model for conflicting subnets. This decision aims to resolve policy conflicts, enabling conflicting subnets to coexist harmoniously with other subnets and optimizing the overall network performance.

[0099] The second decision is made based on the global policy model to determine the conflicting subnet. This decision is specifically designed for the conflicting subnet to resolve policy conflicts and maintain a consistent network configuration with other subnets. This decision may include adjusting device connectivity, changing link bandwidth, and optimizing device configuration parameters. Then, the network topology of the conflicting subnet is updated according to the second decision, resulting in the updated conflicting subnet. This subnet has resolved its policy conflicts and maintains a harmonious network configuration with other subnets.

[0100] Step D40: For subnets without policy conflicts, update the network topology according to the first policy corresponding to the subnets without policy conflicts to obtain the second subnet;

[0101] It should be noted that the second subnet is a subnet obtained by updating the network topology according to the corresponding first policy for subnets that do not have policy conflicts. These subnets maintain the original network configuration and together with other subnets (including the updated conflicting subnets) constitute the new network topology.

[0102] Step D50: Determine the first network topology based on the updated conflict subnet and / or second subnet.

[0103] Simultaneously, for subnets without policy conflicts, the network topology is updated according to their corresponding first policy. Finally, the first network topology is determined based on the updated conflicting subnets and / or the second subnet. This new network topology is the result of applying the global policy model, which considers the needs and states of all subnets and achieves globally optimal network performance.

[0104] This embodiment effectively solves the problem of policy conflicts between subnets by introducing a global policy model. When a policy conflict is detected, the global policy model constructed by the central control unit is used to formulate a second decision for the conflicting subnet and update its network topology accordingly, ensuring harmonious coexistence among subnets. Simultaneously, for subnets without policy conflicts, updates are performed according to their first policy, ultimately obtaining the globally optimal first network topology, improving network performance and stability.

[0105] In one feasible implementation, the automatic networking method for spatial computing may include steps E10 to E30:

[0106] Step E10: Define the super object model at the preset initial moment as a state vector, define the action vector according to the adjustment strategy between devices, and define the strategy model based on the mapping relationship between the state vector and the action vector.

[0107] It should be noted that a state vector refers to a set of parameters describing the current state of each device in the network environment. In this embodiment, it refers to the current state of each device at a preset initial moment, including but not limited to the device's hardware resources (such as CPU and memory usage), software functional state, dynamic state (such as current load and available resources), and spatial attributes (such as geographical location and physical altitude). The initial moment can be a point in time series, not just the moment the super-object model is created. An action vector refers to a set of operations or decisions that the reinforcement learning model can perform in a given state, such as adjusting the connection relationship between devices or reallocating resources. The selection of action vectors aims to optimize network performance, such as reducing latency, increasing throughput, or improving resource utilization. The policy model is a model defined based on the mapping relationship between state vectors and action vectors, used to decide the adjustment strategy of the network. It is trained through a learning algorithm to optimize network performance.

[0108] The super-object model at its initial moment is defined as a state vector. This state vector contains all the key information of the super-object model at the initial moment, such as the device's state and attribute values. Simultaneously, action vectors are defined according to the adjustment strategy between devices. These action vectors represent a set of actions or operations that a device can take. Next, a policy model is defined based on the mapping relationship between the state vector and the action vector. This policy model describes which actions the devices should take in different states to achieve optimal network performance. To evaluate the effectiveness of these actions, the network performance of the state vector and action vector needs to be determined. This network performance can be obtained by measuring key indicators such as network transmission efficiency, latency, and packet loss rate.

[0109] Step E20: Determine the network performance under the state vector and action vector, and define a reward function based on the network performance. The reward function is used to evaluate the network performance of the action vector under the state vector.

[0110] It's important to note that network performance refers to key metrics such as transmission efficiency, latency, and packet loss rate at a given moment. These metrics are used to evaluate the network's operational status and device performance. The reward function, defined based on network performance, is used to evaluate the network performance of an action vector within a state vector. It provides the reward values ​​obtainable by taking different actions in different states, guiding the learning and optimization of the policy model.

[0111] A reward function is defined based on network performance. This reward function gives the reward value obtained by taking different actions in different states. The larger the value of the reward function, the better the network performance of that action in that state. This reward function is used to guide the learning and optimization of the policy model.

[0112] Step E30: Train the policy model using a learning algorithm based on the reward function to obtain a reinforcement learning policy model.

[0113] The policy model is trained using a learning algorithm based on the reward function to obtain a reinforcement learning policy model. During the training process, the policy model continuously tries different action vectors and adjusts its decisions based on feedback from the reward function. Through repeated training and learning, the policy model gradually finds the optimal action vector to maximize network performance. The final reinforcement learning policy model can select the optimal action vector based on the current state vector, thereby optimizing the state and performance of the entire network.

[0114] For example, when training a reinforcement learning policy model, a state space is first defined, which concretizes the current state of each device (the current state of the super-object model) as a state vector S_t containing time step t, where S_t = [D_t, S_t, M_t, T_t] and t represents the time step. Then, an action space A is defined, containing various possible actions, such as adjusting the connection relationships between devices and optimizing resource allocation strategies, specifically represented as A = {a_1, a_2, ..., a_p}. To evaluate the effectiveness of these actions, a reward function R(S_t, A_t) is designed, which is calculated based on key network performance metrics (such as latency, throughput, and resource utilization). Next, reinforcement learning algorithms such as deep Q-networks or policy gradients are used to iteratively train the policy model through multiple rounds, gradually optimizing it to enable it to select the optimal action based on different states. Finally, a trained reinforcement learning policy model π_θ is output, which can make dynamic decisions in a real-time network environment to optimize network performance and resource utilization.

[0115] For example, when generating a network topology using a reinforcement learning policy model, the reinforcement learning policy model π_θ dynamically selects the optimal action A_t based on the state vector S_t. This action A_t belongs to the action set A and aims to adjust the connection relationships between devices or reallocate resources to optimize the network. Subsequently, action A_t is executed, updating the current network topology G_t (where t represents a time step), thereby generating a new network structure G_{t+1}. Next, based on the performance metrics (such as latency, throughput, resource utilization, etc.) of the new network structure G_{t+1}, the reward value R(S_t, A_t) is calculated. This reward value reflects the impact of action A_t on network performance. Based on the obtained reward value, the parameters of the policy model π_θ are updated to make it more accurate and efficient in future decisions. This process of dynamic decision-making, action implementation, reward feedback, and policy model updating is repeated until the network topology and performance metrics reach the preset optimization conditions. Finally, the optimized network topology G is output, which not only meets the current device status but also conforms to business requirements, thereby improving network performance.

[0116] This embodiment evaluates network performance using state vectors and action vectors, and then designs a reward function to train and optimize the policy model. Finally, a reinforcement learning policy model that can intelligently adjust the network structure and resource allocation strategy is obtained. The reinforcement learning policy model can be applied to real-world network systems, intelligently selecting the optimal action vector based on the real-time state vector to optimize network performance.

[0117] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the automatic networking method for spatial computing in this application. Any simple modifications based on this technical concept are within the protection scope of this application.

[0118] This application also provides an automatic networking device for space computing; please refer to [reference needed]. Figure 4 The automatic networking device for spatial computing includes:

[0119] The super-object model generation module 10 constructs a super-object model and performs cluster analysis on at least two devices based on the super-object model to obtain an initial network topology. The super-object model includes the hardware specifications, dynamic status, and spatial location of all preset devices.

[0120] The policy generation module 20 obtains the current device state of the device corresponding to the initial network topology, inputs the current device state into the pre-trained reinforcement learning policy model, and outputs the first policy. The reinforcement learning policy model is a policy model constructed and trained based on the super object model and the initial network topology.

[0121] The network topology generation module 30 updates the initial network topology structure according to the first strategy to obtain the first network topology structure;

[0122] The performance detection module 40 determines the network performance of the first network topology. If the network performance reaches a preset performance threshold, the first network topology is determined to be the target network topology.

[0123] The spatial computing automatic networking device provided in this application, employing the spatial computing automatic networking method described in the above embodiments, can solve the technical problem of low network performance in a dynamic networking environment of IoT devices. Compared with the prior art, the beneficial effects of the spatial computing automatic networking device provided in this application are the same as those of the spatial computing automatic networking method described in the above embodiments, and other technical features in the spatial computing automatic networking device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0124] This application provides a space computing automatic networking device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the space computing automatic networking method in Embodiment 1 above.

[0125] The following is for reference. Figure 5 This document illustrates a structural schematic diagram of a spatial computing automatic networking device suitable for implementing embodiments of this application. The spatial computing automatic networking device in the embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), and in-vehicle terminals (e.g., in-vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. Figure 5 The spatial computing automatic networking device shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0126] like Figure 5As shown, the spatial computing automatic networking device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the spatial computing automatic networking device. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the space computing automated networking device to communicate wirelessly or wiredly with other devices to exchange data. Although the figure shows a space computing automated networking device with various systems, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems can be implemented alternatively.

[0127] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0128] The spatial computing automatic networking device provided in this application, employing the spatial computing automatic networking method described in the above embodiments, can solve the technical problem of low network performance in a dynamic networking environment of IoT devices. Compared with the prior art, the beneficial effects of the spatial computing automatic networking device provided in this application are the same as those of the spatial computing automatic networking method described in the above embodiments, and other technical features of this spatial computing automatic networking device are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0129] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0130] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0131] This application provides a medium, which is a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, which are used to execute the spatial computing automatic networking method in the above embodiments.

[0132] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems or devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0133] The aforementioned computer-readable storage medium may be included in the space computing automatic networking device; or it may exist independently and not be assembled into the space computing automatic networking device.

[0134] The aforementioned computer-readable storage medium carries one or more programs, which, when executed by the space computing automatic networking device, cause the space computing automatic networking device to:

[0135] Construct a super-object model, and perform cluster analysis on at least two devices based on the super-object model to obtain the initial network topology. The super-object model includes the hardware specifications, dynamic states, and spatial locations of all preset devices.

[0136] Obtain the current device state of the device corresponding to the initial network topology, input the current device state into the pre-trained reinforcement learning policy model, and output the first policy. The reinforcement learning policy model is a policy model built and trained based on the super object model and the initial network topology.

[0137] The initial network topology is updated according to the first strategy to obtain the first network topology.

[0138] Determine the network performance of the first network topology. If the network performance reaches a preset performance threshold, then determine the first network topology as the target network topology.

[0139] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0140] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0141] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0142] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described automatic spatial computing networking method, which can solve the technical problem of low network performance in the dynamic networking environment of Internet of Things devices. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as the beneficial effects of the automatic spatial computing networking method provided in the above embodiments, and will not be repeated here.

[0143] This application also provides a product, which is a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described automatic spatial computing networking method.

[0144] The computer program product provided in this application can solve the technical problem of low network performance in the dynamic networking environment of Internet of Things (IoT) devices. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the spatial computing automatic networking method provided in the above embodiments, and will not be repeated here.

[0145] The above are only some embodiments of this application and do not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A method for automatic spatial computing network formation, characterized in that, The automatic networking method for spatial computing includes: Construct a super-object model, and perform cluster analysis on at least two devices based on the super-object model to obtain an initial network topology. The super-object model includes the hardware specifications, dynamic states, and spatial locations of all preset devices. The current device state of the device corresponding to the initial network topology is obtained, and the current device state is input into the pre-trained reinforcement learning policy model to output the first policy. The reinforcement learning policy model is a policy model constructed and trained based on the super object model and the initial network topology. The initial network topology is updated according to the first strategy to obtain the first network topology. Determine the network performance of the first network topology; if the network performance reaches a preset performance threshold, then determine the first network topology as the target network topology. The steps for constructing the super-object model include: Generate attribute vectors based on the hardware specifications and dynamic state of the device; Generate a three-dimensional vector based on the spatial location information of the device; Define metadata fields for each device, and generate a metadata vector based on the metadata fields; A super object model is constructed based on the attribute vector, three-dimensional vector, metadata vector, and a preset time series, wherein the time series represents the data differences generated by the attribute vector, three-dimensional vector, and metadata vector as they change over time.

2. The automatic spatial computing networking method as described in claim 1, characterized in that, The step of performing cluster analysis on at least two devices based on the super-object model to obtain the initial network topology includes: Based on the super-object model, a preset clustering algorithm is used to group the at least two devices to obtain at least one subnet. The clustering algorithm groups devices with similar functions or spatial proximity into the same subnet according to the super-object model. For each subnet, a first connection relationship is determined among all devices belonging to the subnet, and the subnet is structurally optimized based on the first connection relationship, wherein the initial network topology includes each subnet after structural optimization.

3. The automatic spatial computing networking method as described in claim 1, characterized in that, The step of updating the initial network topology according to the first strategy to obtain the first network topology includes: If the initial network topology has multiple subnets, determine whether there is a policy conflict between the first policies corresponding to each subnet; If there is no policy conflict between the first policies corresponding to each subnet, then for each subnet, the network topology of the subnet is updated according to the first policy corresponding to each subnet to obtain the first network topology.

4. The automatic spatial computing networking method as described in claim 3, characterized in that, After the step of determining whether there is a policy conflict between the first policies corresponding to each subnet, the method further includes: If at least two subnets have conflicting first policies, then the subnets with conflicting policies are determined to be conflicting subnets. Obtain a global policy model sent by a preset central control unit, wherein the central control unit determines the global network state based on the subnet state of all subnets, and constructs a global policy model based on the global network state; The second decision for the conflict subnet is determined based on the global policy model, and the network topology of the conflict subnet is updated based on the second decision to obtain the updated conflict subnet. For subnets without policy conflicts, the network topology is updated according to the first policy corresponding to the subnets without policy conflicts to obtain the second subnet; The first network topology is determined based on the updated conflict subnet and / or the second subnet.

5. The automatic spatial computing networking method as described in any one of claims 1 to 4, characterized in that, The automatic networking method for spatial computing also includes: The super object model at a preset initial moment is defined as a state vector, the action vector is defined according to the adjustment strategy between the devices, and the strategy model is defined based on the mapping relationship between the state vector and the action vector. Determine the network performance of the state vector and the action vector, and define a reward function based on the network performance, wherein the reward function is used to evaluate the network performance of the action vector under the state vector; The policy model is trained using a learning algorithm based on the reward function to obtain a reinforcement learning policy model.

6. An automatic spatial computing networking device, characterized in that, The automatic networking device for spatial computing includes: A super-object model generation module constructs a super-object model and performs cluster analysis on at least two devices based on the super-object model to obtain an initial network topology. The super-object model includes the hardware specifications, dynamic states, and spatial locations of all preset devices. The super-object model generation module is also used to generate attribute vectors based on the hardware specifications and dynamic states of the devices; generate three-dimensional vectors based on the spatial location information of the devices; define metadata fields for each device and generate metadata vectors based on the metadata fields; and construct the super-object model based on the attribute vectors, three-dimensional vectors, metadata vectors, and a preset time series, wherein the time series represents the data differences generated by the attribute vectors, three-dimensional vectors, and metadata vectors over time. The policy generation module obtains the current device state of the device corresponding to the initial network topology, inputs the current device state into the pre-trained reinforcement learning policy model, and outputs the first policy. The reinforcement learning policy model is a policy model constructed and trained based on the super object model and the initial network topology. The network topology generation module updates the initial network topology structure according to the first strategy to obtain the first network topology structure; The performance detection module determines the network performance of the first network topology. If the network performance reaches a preset performance threshold, the first network topology is determined to be the target network topology.

7. An automatic networking device for spatial computing, characterized in that, The automatic spatial computing networking device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the automatic spatial computing networking method as described in any one of claims 1 to 5.

8. A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the spatial computing automatic networking method as described in any one of claims 1 to 5.

9. A computer program product comprising a computer program that, when executed by a processor, implements the steps of the spatial computing automatic networking method as described in any one of claims 1 to 5.

Images