Method and apparatus for a network node to access a network

Abstract

The present disclosure provides a method and device for a network node to access a network, and relates to the technical field of communication. The specific implementation scheme is: applied to a to-be-accessed node to be accessed to a preset network, comprising: broadcasting a detection message to the preset network, wherein the detection message comprises a preset function; receiving and establishing physical layer synchronization with the preset network according to the operation parameters in the response information fed back by a first node; establishing a communication channel through the first node and a second node, wherein the to-be-accessed node can receive models on the first node and the second node; in response to successful authentication, obtaining the models from the first node and / or the second node through the communication channel, and learning and training the obtained models, and in response to the learned and trained models meeting the model evaluation conditions of the preset network, joining the preset network through the first node and the second node. The present disclosure improves the access performance of the to-be-accessed node in the preset network.

Description

Technical Field

[0001] This invention relates to the field of communication technology, and in particular to a method and apparatus for a network node to access a network. Background Technology

[0002] In the future Internet of Everything, network nodes are trending towards intelligence. This intelligence leads to a rapid expansion of the information space, even resulting in a dimensional disaster. This exacerbates the difficulty of representing the information carrying space, making it difficult for traditional network service capabilities to match the high-dimensional information space. The excessive amount of data transmitted in communications makes information service systems unable to continuously meet people's complex, diverse, and intelligent information transmission needs. However, using artificial intelligence models to encode, propagate, and decode business information can significantly reduce the amount of data transmitted in communication services, greatly improving information transmission efficiency. These models are relatively stable and possess reusability and propagation capabilities. The propagation and reuse of these models will help enhance network intelligence while reducing overhead and resource waste, forming a highly intelligent and simplified network.

[0003] As intelligent and simplified networks gradually evolve, in response to new network nodes wanting to join the existing network, suitable propagation methods are still needed to ensure the performance of these new nodes. Summary of the Invention

[0004] This disclosure provides a method and apparatus for a network node to access a network.

[0005] According to a first aspect of this disclosure, a method for a network node to access a network is provided, wherein the method is applied to a node to be accessed in a preset network, and the method includes:

[0006] Broadcast probe messages to a preset network, wherein the probe messages include preset functions of the node to be connected;

[0007] The system receives response information sent by the first node in the preset network based on the probe information. The response information includes the operation parameters of the first node in the preset network, and the node to be connected can receive the model on the first node.

[0008] Establish physical layer synchronization with the preset network based on the operating parameters;

[0009] The first node searches for the second node in the preset network and establishes a communication channel with the second node, wherein the node to be connected is able to receive the model on the second node.

[0010] Upon receiving successful authentication information from the first node and the second node, the model is retrieved from the first node and / or the second node via the communication channel;

[0011] The acquired model is trained. In response to the model being trained meeting the model evaluation criteria of the preset network, it is added to the preset network through the first node and the second node.

[0012] According to a second aspect of this disclosure, a method for a network node to access a network is provided, wherein the method is applied in a preset network and includes:

[0013] Receive probe information broadcast by the node to be connected, wherein the probe message includes the preset functions of the node to be connected;

[0014] The first node that can send response information to the node to be accessed is found based on the detection information. The model on the first node can be received by the node to be accessed. The response information includes the operation parameters of the first node in the preset network. The operation parameters are used to enable the node to be accessed and the preset network to establish physical layer synchronization.

[0015] The first node finds the second node and controls the second node to establish a communication channel with the node to be connected. The model on the second node can be received by the node to be connected.

[0016] After receiving the authentication success information from the first node and the second node, control the first node and the second node to send the model to the node to be connected through the communication channel, so that the node to be connected can learn and train according to the obtained model.

[0017] The trained models in the nodes to be accessed are evaluated according to the model evaluation conditions. In response to the evaluation being passed, the nodes to be accessed are added through the first node and the second node.

[0018] According to a third aspect of this disclosure, an apparatus for a network node to access a network is provided, wherein the apparatus, applied to a node to be accessed in a preset network, includes:

[0019] The broadcast unit is used to broadcast probe messages to a preset network, wherein the probe messages include preset functions of the node to be accessed;

[0020] The receiving unit is used to receive response information sent by the first node in the preset network based on the probe information. The response information includes the operation parameters of the first node in the preset network, and the node to be connected can receive the model on the first node.

[0021] The synchronization unit is used to establish physical layer synchronization with the preset network according to the operating parameters;

[0022] The channel establishment unit is used to find the second node in the preset network through the first node and establish a communication channel with the second node, wherein the node to be accessed can receive the model on the second node.

[0023] The acquisition unit, in response to receiving successful authentication information from the first node and the second node, acquires the model from the first node and the second node respectively through the communication channel;

[0024] The added unit is used to learn and train the acquired model. In response to the model being learned and trained meeting the model evaluation conditions of the preset network, it is added to the preset network through the first node and the second node.

[0025] According to a fourth aspect of this disclosure, an apparatus for a network node to access a network is provided, wherein the apparatus is applied in a preset network and includes:

[0026] The receiving unit is used to receive probe information broadcast by the node to be accessed, wherein the probe message includes the preset functions of the node to be accessed;

[0027] The lookup unit is used to find the first node that can send response information to the node to be accessed based on the detection information. The model on the first node can be received by the node to be accessed. The response information includes the operation parameters of the first node in the preset network. The operation parameters are used to enable the node to be accessed and the preset network to establish physical layer synchronization.

[0028] The first control unit locates the second node through the first node and controls the second node to establish a communication channel with the node to be connected, wherein the model on the second node can be received by the node to be connected.

[0029] After receiving the authentication success information from the first node and the second node, the second control unit controls the first node and the second node to send the model to the node to be connected through the communication channel, so that the node to be connected can learn and train based on the obtained model.

[0030] Add a unit by evaluating the trained model in the node to be accessed according to the model evaluation conditions. In response to the evaluation being passed, add the node to be accessed through the first node and the second node.

[0031] According to a fifth aspect of this disclosure, an electronic device is provided, comprising:

[0032] At least one processor; and

[0033] A memory that is communicatively connected to at least one processor; wherein,

[0034] The memory stores instructions that can be executed by at least one processor, such that at least one processor can perform the methods described above.

[0035] According to a sixth aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are used to cause a computer to perform the methods described above.

[0036] The beneficial effects of the technical solution provided in this disclosure are:

[0037] The solution provided in this embodiment establishes a communication channel between the node to be accessed and a network node in a preset network. The node to be accessed obtains the model from the network node through the communication channel. At this time, the node to be accessed does not join the preset network. In response to the model capability of the node to be accessed meeting the model evaluation conditions of the preset network, it joins the preset network, thereby improving the access performance of the node to be accessed in the preset network.

[0038] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0039] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0040] Figure 1 This is a flowchart illustrating a method for a network node to access a network according to Embodiment 1 of this disclosure. Figure 1 ;

[0041] Figure 2 This is a flowchart illustrating a method for a network node to access a network according to Embodiment 1 of this disclosure. Figure 2 ;

[0042] Figure 3 This is a schematic diagram of node A to be accessed not being joined to the preset network W according to Embodiment 1 of this disclosure;

[0043] Figure 4 This is a schematic diagram of a probe message broadcast by node A to be accessed according to Embodiment 1 of this disclosure;

[0044] Figure 5 This is a schematic diagram illustrating the establishment of a communication channel between node A to be accessed and network node B6 according to Embodiment 1 of this disclosure;

[0045] Figure 6 This is a schematic diagram illustrating the establishment of a communication channel between node A to be accessed and network nodes B5 and B7 according to Embodiment 1 of this disclosure.

[0046] Figure 7 This is a schematic diagram of model propagation according to Embodiment 1 of this disclosure;

[0047] Figure 8 This is a schematic diagram of node A to be accessed joining a preset network W according to Embodiment 1 of this disclosure;

[0048] Figure 9This is a flowchart illustrating a method for a network node to access a network according to Embodiment 2 of this disclosure;

[0049] Figure 10 This is a schematic diagram of the structure of a device for a network node to access a network according to Embodiment 3 of this disclosure;

[0050] Figure 11 This is a schematic diagram of the structure of a network node accessing a network according to Embodiment 4 of this disclosure. Detailed Implementation

[0051] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0052] In the intelligent and simplified network, business information is primarily disseminated through artificial intelligence (AI) models. By using AI models to compress the initial business information into a second business information related to the AI ​​model, the data communication volume in the network is significantly reduced, with compression efficiency far exceeding traditional compression algorithms. Specifically, the sending device uses a pre-configured first model to extract the initial business information and obtain the second business information to be transmitted; the sending device then transmits the second business information to the receiving device. The receiving device receives the second business information and uses the pre-configured second model to recover the third business information. The third business information recovered from the second model has slight quality differences compared to the original first business information, but the content is identical, providing a virtually indistinguishable user experience. Before the sending device transmits the second business information to the receiving device, an update module determines whether the receiving device needs to update the second model. If an update is required, the module transmits the pre-configured third model to the receiving device, which then uses the third model to update the second model. By processing business information through a pre-trained AI model, the data transmission volume in communication services can be significantly reduced, greatly improving information transmission efficiency. These models are relatively stable and possess reusability and propagation capabilities. Model propagation and reuse help enhance network intelligence while reducing overhead and resource waste. A model can be divided into several model slices according to different segmentation rules. These model slices can also be transferred between different network nodes, and the slices can be assembled into a model. Model slices can be stored distributed across multiple network nodes. When a network node discovers that it is missing or needs to update a certain model or model slice, it can request it from nearby nodes that may possess that slice.

[0053] The transmission of service information and models both occur within the communication network, based on network protocols. The network nodes traversed along the transmission paths include intelligent routers. The functions of intelligent routers include, but are not limited to, service information transmission, model transmission, model self-updating, and security protection. The transmission function of intelligent routers involves transmitting service information or models from source nodes to destination nodes, with multiple paths existing between them. The model transmission function of intelligent routers can transmit model slices, improving transmission speed by strategically arranging model slices to travel along multiple paths.

[0054] Example 1

[0055] Figure 1 This disclosure illustrates a method for a network node to access a network, applied to a node seeking to access a preset network, such as... Figure 1 As shown, the method includes:

[0056] Step S101: Broadcast a probe message to a preset network, wherein the probe message includes the preset functions of the node to be accessed;

[0057] Step S102: Receive response information sent by the first node in the preset network based on the probe information. The response information includes the operation parameters of the first node in the preset network, and the node to be connected can receive the model on the first node.

[0058] Step S103: Establish physical layer synchronization with the preset network according to the operation parameters;

[0059] Step S104: The first node searches for the second node in the preset network and establishes a communication channel with both the first and second nodes, wherein the node to be connected is able to receive the model on the second node.

[0060] Step S105: In response to receiving successful authentication information from the first node and the second node, the model is obtained from the first node and / or the second node through the communication channel.

[0061] Step S106: The obtained model is trained. In response to the model being trained meeting the model evaluation conditions of the preset network, it is added to the preset network through the first node and the second node.

[0062] This disclosure establishes a communication channel between the node to be accessed and a network node in a preset network, and the node to be accessed obtains the model from the network node through the communication channel. At this time, the node to be accessed does not join the preset network. When the model capability in the node to be accessed meets the model evaluation conditions of the preset network, it joins the preset network, thereby improving the access performance of the node to be accessed in the preset network.

[0063] Specifically, such as Figure 2 As shown, in response to a node seeking to join a preset network, the node broadcasts a probe message to the preset network. If no network node in the preset network responds to the probe message, the node continues to broadcast probe messages until it receives a response from a network node in the preset network. Based on the operation parameters in the response message, the node establishes physical layer synchronization with the preset network. The network node that provides the response is designated as the first node. Subsequently, the first node finds a second node in the preset network and establishes a communication channel with the second node. The first and second nodes authenticate the node seeking to join. If authentication fails, the node seeking to join... The node continues to broadcast probe messages to the preset network. If authentication is successful, the preset network obtains the model from the first and second nodes and transmits the obtained model to the node to be joined through the communication channel. At this time, the node to be joined does not participate in routing. The node to be joined learns and trains the received model. It is necessary to judge the model capability of the node to be joined. If the model trained in the node to be joined meets the model evaluation conditions of the preset network, the node to be joined joins the preset network through the first and second nodes. If the model trained in the node to be joined does not meet the model evaluation conditions of the preset network, the process returns to the step of obtaining the model from the first and second nodes.

[0064] It should be noted that the steps of the corresponding methods are not necessarily performed in the order shown and described in this specification in other embodiments. In some other embodiments, the methods may include more or fewer steps than those described in this specification. Furthermore, a single step described in this disclosure may be broken down into multiple steps in other embodiments; and multiple steps described in this specification may be combined into a single step in other embodiments.

[0065] Specifically, the aforementioned communication channel can be a logical channel;

[0066] It should be noted that a logical channel is a channel that transmits different types of information over a physical channel.

[0067] Specifically, in a network, all nodes are interconnected. When a new node needs to join an existing preset network, it only needs to establish connections with several nodes already connected to the preset network and perform the corresponding network configuration to add the node to the preset network.

[0068] For example, a schematic diagram of the node to be connected, A, and the preset network W is shown below. Figure 3 As shown, the preset network W includes multiple network nodes: B1-B8. Node A to be connected can broadcast a probe message to the preset network W. When network node B6 in the preset network responds, node A can establish a communication channel with neighboring network nodes B5 and B7 through network node B6. It should be noted that, in order to ensure the transmission of the model, node A to be connected must have the ability to receive the model from network nodes B5 and B7. After subsequent network configuration, node A to be connected can join the preset network W through the established connections with network nodes B5, B6, and B7.

[0069] It should be noted that, for ease of explanation, this disclosure selects a preset number of network nodes that can provide feedback response information as the first node, and sets the network nodes that are adjacent to the first node and can be received by the model of the node to be accessed A as the second node. The second node can also provide feedback response information. Therefore, in the example above, network node B6 is the first node, and network nodes B5 and B7 are both second nodes. The preset number can be customized. For example, the preset number can be set to one.

[0070] This disclosure provides a possible implementation, wherein the first node, the second node, and the node to be connected include, but are not limited to, cloud server smart nodes, small base station smart nodes, drone smart nodes, smartphone smart nodes, laptop smart nodes, and intelligent transportation smart nodes.

[0071] Specifically, new nodes—nodes to be connected—and existing network nodes in the preset network include, but are not limited to, intelligent nodes such as cloud servers, small base stations, drones, smartphones, laptops, and intelligent vehicles.

[0072] This disclosure provides a possible implementation in which the first node determines, based on a preset function in the received probe information, that the node to be accessed has the ability to receive the model on the first node.

[0073] This disclosure provides a possible implementation method, wherein the preset functions include: computing power and storage power.

[0074] For example, network nodes in a pre-defined network can determine whether a node to be connected can store the model transmitted by that network node based on its storage capacity.

[0075] For example, network nodes in a pre-defined network can determine whether a node to be connected can run the model transmitted by that network node based on its computing power.

[0076] This disclosure provides a possible implementation method in which a preset network includes multiple network nodes. The network nodes receive probe information. In response to the network nodes determining, based on the preset function of the received probe information, that the node to be accessed has the ability to receive the model on the network node, a preset number of network nodes are selected as the first node according to the shortest route principle. The first node sends response information to the node to be accessed in response to the probe message. The preset number can be customized, for example, it can be set to one.

[0077] In this disclosure, the shortest route principle is used to reduce the energy consumption of the propagation model when providing feedback response information.

[0078] Specifically, when a node to be connected wants to join a preset network, it needs to broadcast probe information to query whether any network nodes in the preset network have responded to the probe information. The probe information includes the preset functions of the node to be connected. Existing network nodes in the preset network receive the probe information, and the network nodes determine whether the preset functions of the node to be connected can meet the various capability requirements of the model transmitted by this node based on the probe message:

[0079] If satisfied, a preset number of network nodes are selected as the first node according to the shortest route principle, and the nodes respond to the probe messages. The preset number can be customized; for example, it can be set to one.

[0080] If the conditions are not met, the access node will continue to probe.

[0081] For example, the preset network W includes multiple network nodes: B1-B8. For node A to be connected, node A is unaware of the routing relationships between the network nodes within the preset network W. Therefore, node A can broadcast a probe message to the preset network W. This probe message includes the preset functions of node A. Each network node within the preset network W determines whether node A can accept the model on its node based on the probe message. A schematic diagram of node A and the preset network W is shown below. Figure 3 As shown, network nodes B3-B7 receive probe messages and determine whether the node to be connected, A, can receive the model on the network node based on the probe messages.

[0082] If satisfied, a preset number of network nodes are selected as the first node according to the shortest route principle, and the first node sends a response message to the probe message to the node A to be connected.

[0083] For example, if node A to be connected can receive models from network nodes B5, B6 and B7, then according to the shortest route principle, the network node with the shortest route among network nodes B5, B6 and B7 is selected, that is, network node B6 is selected as the first node. At this time, network node B6 sends a response information to the probe message to node A to be connected.

[0084] If the conditions are not met, the probe will continue until access node A is reached.

[0085] This disclosure provides a possible implementation method in which the node to be accessed establishes physical layer synchronization with a preset network according to operating parameters, specifically including:

[0086] The network nodes in the preset network store operation parameters for a preset time period;

[0087] The first node sends a response containing operation parameters to the node to be connected, enabling the node to analyze the operation parameters within a preset time period, thereby achieving physical layer synchronization between the node and the preset network based on the analysis results.

[0088] Specifically, each network node in the preset network stores the operation parameters of the preset network within a preset time period;

[0089] When a network node establishes a connection with a node to be connected, that is, the network node sends a response message to the node to be connected, the network node at this time is called the first node, and the response message includes the operation parameters of the first node. The node to be connected analyzes the operation parameters and establishes physical layer synchronization with the preset network.

[0090] For example, such as Figure 3 As shown, the preset network W includes multiple network nodes: network nodes B1-B8, and network node B6 is the first node. Network node B6 establishes a connection with node A to be connected. Network node B6 sends response information containing its own operation parameters to node A to be connected. Node A to be connected analyzes the operation parameters and establishes physical layer synchronization with the preset network W.

[0091] This disclosure provides a possible implementation method in which the second node obtains the detection information through the first node and determines, based on the preset functions in the detection information, that the node to be accessed has the ability to receive the model on the second node.

[0092] This disclosure provides a possible implementation in which the node to be accessed searches for neighboring network nodes in a preset network through a first node, wherein the neighboring network nodes can receive probe information;

[0093] In response to the network node obtaining information from the first node and determining, based on the preset functions in the probe information, that the node to be accessed has the ability to receive the model on the network node, the network node is set as the second node.

[0094] Specifically, after the node to be connected establishes physical layer synchronization with the preset network, the node to be connected searches for neighboring network nodes in the preset network through the first node with which it has already established a connection. The network node can determine whether the preset functions of the node to be connected can meet the various capability requirements of receiving the model transmitted by this node based on the probe messages:

[0095] If the conditions are met, set the network node as the second node and establish a communication channel between the second node and the node to be connected.

[0096] If the conditions are not met, the node to be connected continues to search through the first node with which a connection has already been established, until a second node can be found or all network nodes in the preset network have been searched.

[0097] For example, the preset network W includes multiple network nodes: B1-B8. Node A, which is to be connected, has already established a connection with network node B6, which is the first node. Node A then searches for neighboring nodes—network nodes B5 and B7—in the preset network through network node B6. Network nodes B5 and B7 then determine, based on the probe messages, whether the preset functions of the node to be connected meet the various capability requirements of the model receiving the data transmitted by this node.

[0098] If the conditions are met, set the network node as the second node, and establish a communication channel between the second node and the node to be connected, such as... Figure 6 As shown;

[0099] If the conditions are not met, the node to be connected continues to search through the first node with which a connection has already been established, until a second node can be found or all network nodes in the preset network have been searched.

[0100] This disclosure provides a possible implementation, wherein, in response to the first node sending response information to the node to be accessed, the node to be accessed establishes a communication channel with the first node;

[0101] The communication channel between the node to be connected and the first node is defined as the first channel, and the communication channel between the node to be connected and the second node is defined as the second channel.

[0102] The transmission capacity of the first channel is determined based on the size of the model to be transmitted in the first node;

[0103] The transmission capacity of the second channel is determined based on the size of the model to be transmitted in the second node.

[0104] It should be noted that the response from the first node to the node to be connected indicates that a communication channel has been established between the node to be connected and the first node.

[0105] Specifically, for ease of explanation, the communication channel between the node to be connected and the first node can be defined as the first channel, and the communication channel between the node to be connected and the second node can be defined as the second channel;

[0106] During the establishment of the first channel, the node to be accessed, A, negotiates the basic transmission capacity of the first channel with the first node to ensure that any one or more models on the first node can transmit.

[0107] During the establishment of the second channel, the node to be accessed, A, negotiates the basic transmission capacity of the second channel with the second node to ensure that any one or more models on the second node can transmit.

[0108] For example, the preset network W includes multiple network nodes: B1-B8. Among them, the node to be connected, A, has already established a connection with network node B6, which is the first node. Furthermore, the node to be connected, A, establishes a communication channel with network nodes B5 and B7 through network node B6, which are the second nodes.

[0109] like Figure 6 As shown, there is a communication channel b between node A to be connected and network node B6;

[0110] There is a communication channel a between node A to be connected and network node B5;

[0111] There is a communication channel c between node A to be connected and network node B7;

[0112] At this point, it is necessary to determine the basic transmission capacity of communication channel a based on the size of the model on network node B5, so that any one or more models on network node B5 can be transmitted to the node A to be accessed through communication channel a.

[0113] The basic transmission capacity of communication channel b needs to be determined based on the size of the model on network node B6, so that any one or more models on network node B6 can be transmitted to the node A to be accessed through communication channel b.

[0114] The basic transmission capacity of communication channel c needs to be determined based on the size of the model on network node B7, so that any one or more models on network node B7 can be transmitted to the node A to be connected through communication channel c.

[0115] This disclosure provides a possible implementation method in which the node to be connected establishes a list of neighboring nodes, the list of neighboring nodes including node information of the first node and node information of the second node.

[0116] Specifically, the neighboring node list includes node information of all nodes that have established communication channels with the node to be connected; thus, the efficiency of the subsequent propagation model is improved by using the neighboring node list.

[0117] This disclosure provides a possible implementation method in which a first node and a second node authenticate the node to be accessed, and the first node and the second node send authentication response information to the node to be accessed, the authentication response information including authentication success information and authentication failure information.

[0118] Specifically, both the first and second nodes need to authenticate the node to be accessed, thereby verifying whether the model in this node can be passed to the node to be accessed in terms of security and privacy, thus improving security.

[0119] For example, the preset network W includes multiple network nodes: B1-B8. Among them, the node to be connected, A, has already established a communication channel with network nodes B6, B5, and B7. At this time, network node B6 is the first node, and network nodes B5 and B7 are the second nodes.

[0120] Network nodes B5, B6, and B7 authenticate the node to be accessed, determine whether transmitting the model to the node to be accessed will infringe on the security and privacy of their own nodes, and protect the information and data privacy of their own nodes during the transmission process.

[0121] This disclosure provides a possible implementation method in which the authentication response information of the first node and the second node are both authentication success information, and the preset network controls the first node and / or the second node to transmit the corresponding model to the node to be accessed through the communication channel according to the model version function.

[0122] If at least one authentication failure message is found among all the authentication response messages received by the node to be accessed, the node to be accessed will continue to broadcast probe information to the preset network.

[0123] Specifically, after successful authentication, the preset network, designated as the node to be connected (A), selects a corresponding model from its neighboring nodes based on the model version and functionality. Each model contributes only a portion of its capabilities to prevent GAN mode collapse when a single model becomes too influential. The selected model undergoes knowledge distillation and transfer learning, and is then transmitted to the node to be connected (A) via a communication channel. The node to be connected (A) then trains itself using the acquired model. Specifically, the first and second nodes can be assigned to a list of neighboring nodes, and at least one network node from this list can be selected as a neighboring node.

[0124] It should be noted that authentication is considered successful only when both the first and second nodes send authentication response messages indicating successful authentication.

[0125] For example, such as Figure 7 As shown, the preset network W includes multiple network nodes: B1-B8. Among them, the node to be connected, A, has already established a communication channel with network nodes B6, B5, and B7. At this time, network node B6 is the first node, and network nodes B5 and B7 are the second nodes.

[0126] After network nodes B5, B6, and B7 successfully authenticate the access node A, the preset network can select at least one model from network nodes B5, B6, and B7 according to model propagation and model version functionality. Here, the preset network W selects the corresponding model from network nodes B5, B6, and B7 for the access node A. Each selected model contributes only a portion of the capabilities and propagates the model through communication channels a, b, and c, respectively. After knowledge distillation and transfer learning, the preset network transmits the new model to the access node A, which then receives the new model and performs learning and training.

[0127] This disclosure provides a possible implementation method in which a preset network uses a distributed real-time model evaluation framework to evaluate the trained model. When the trained model meets the model evaluation criteria, the node to be added to the preset network is added through a first node and a second node, such as... Figure 8 As shown;

[0128] When the learned and trained model does not meet the model evaluation conditions, the preset network continues to propagate the model and selects the model in the first and / or second nodes according to the model version function. After model distillation, the model is passed to the node to be connected, so that the node to be connected can continue to learn and train until the learned and trained model of the node to be connected meets the model evaluation conditions.

[0129] The distributed real-time model evaluation framework is built on the network nodes of a pre-defined network, where nodes collaborate to deploy, train, evaluate, optimize, and update models.

[0130] For example, the preset network W includes multiple network nodes: B1-B8, wherein the process of node A joining the preset network W is as follows: Figure 3-8 As shown, the specific steps are as follows:

[0131] At this point, the node to be connected has not yet established a connection with the preset network. A schematic diagram of the node to be connected A and the preset network W is shown below. Figure 3 As shown;

[0132] For node A to be connected, it does not know the routing relationships between network nodes within the preset network W. Therefore, node A can broadcast probe messages to the preset network W, such as... Figure 4 As shown, the probe message has a preset propagation range, where, Figure 4 The dashed sector in the diagram represents a probe message; at this time, only network nodes B3-B7 received the probe information.

[0133] The probe information includes the preset functions of the node to be accessed, A. Each network node in the preset network W determines whether the node to be accessed, A, can accept the model on its own node based on the probe message. Network nodes B3-B7 receive the probe message, and network nodes B5-7 all determine from the probe message that the node to be accessed, A, can accept the model on network nodes B5-7. Based on the shortest route principle, they select network node B6 from network nodes B5-7 to send a response message to the node to be accessed. At this point, the node to be accessed, A, and network node B6 establish a communication channel. Figure 5 As shown;

[0134] Node A, the node to be connected, searches for neighboring nodes—network nodes B5 and B7—in the preset network through network node B6. At this point, network nodes B5 and B7, based on the probe messages, determine that the preset functions of the node to be connected meet the capability requirements of the receiving node's transmission model. Network nodes B5 and B7 are then set as second nodes, and a communication channel is established between the second nodes and the node to be connected, such as... Figure 6 As shown;

[0135] The node to be connected, A, negotiates the basic transmission capacity of the communication channel with network nodes B5, B6, and B7 to ensure that any one or more models on the network nodes can transmit.

[0136] Network node B5, network node B6 and network node B7 authenticate the node to be accessed, determine whether transmitting the model to the node to be accessed will infringe on the security and privacy of their own nodes, and protect the information and data privacy of their own nodes during the transmission process.

[0137] After successful authentication, if Figure 7 As shown, the preset network can select at least one model from network nodes B5, B6, and B7 based on model propagation and model version functionality. Here, the preset network W is the network to be accessed by node A, which has selected the corresponding model from network nodes B5, B6, and B7. Each selected model contributes only a portion of the capabilities and propagates the model through communication channels a, b, and c, respectively. After knowledge distillation and transfer learning, the preset network transmits the new model to node A, which then receives the new model and performs learning and training.

[0138] The pre-defined network employs a distributed real-time model evaluation framework to assess the trained model. When the trained model meets the model evaluation criteria, the node to be added is added to the pre-defined network through the first and second nodes, such as... Figure 8 As shown.

[0139] Example 2

[0140] Figure 9 This disclosure illustrates a method for a network node to access a network, applied in a preset network, such as... Figure 9 As shown, the method includes:

[0141] Step S201: Receive probe information broadcast by the node to be connected, wherein the probe message includes the preset functions of the node to be connected;

[0142] Step S202: Based on the detection information, a first node that can send response information to the node to be accessed is found. The model on the first node can be received by the node to be accessed. The response information includes the operation parameters of the first node in the preset network. The operation parameters are used to enable the node to be accessed and the preset network to establish physical layer synchronization.

[0143] Step S203: The first node finds the second node and controls the second node to establish a communication channel with the node to be connected, wherein the model on the second node can be received by the node to be connected.

[0144] Step S204: After receiving the authentication success information from the first node and the second node, control the first node and the second node to send the model to the node to be accessed through the communication channel, so that the node to be accessed can learn and train the obtained model.

[0145] Step S205: Evaluate the trained model in the node to be accessed according to the model evaluation conditions. In response to the evaluation being passed, add the node to be accessed through the first node and the second node.

[0146] This disclosure establishes a communication channel between the node to be accessed and a network node in a preset network, and the node to be accessed obtains the model from the network node through the communication channel. At this time, the node to be accessed does not join the preset network. In response to the model capability in the node to be accessed meeting the model evaluation conditions of the preset network, it joins the preset network, thereby improving the access performance of the node to be accessed in the preset network.

[0147] This disclosure provides a possible implementation, wherein the first node, the second node, and the node to be connected include, but are not limited to, cloud server smart nodes, small base station smart nodes, drone smart nodes, smartphone smart nodes, laptop smart nodes, and intelligent transportation smart nodes.

[0148] This disclosure provides a possible implementation method in which the first node determines, based on a preset function in the received probe information, that the node to be accessed has the function of receiving the model on the first node.

[0149] This disclosure provides a possible implementation method, wherein the preset functions include: computing power, storage capacity, and the ability to implement intelligent applications.

[0150] This disclosure provides a possible implementation method in which a preset network includes multiple network nodes. The network nodes receive probe information. In response to the network nodes determining, based on the preset function of the received probe information, that the node to be accessed has the function of receiving the model on the network node, the network node is selected as the first node according to the shortest route principle. The first node sends response information to the probe message to the node to be accessed.

[0151] This disclosure provides a possible implementation method in which the node to be accessed establishes physical layer synchronization with a preset network according to operating parameters, specifically including:

[0152] The network nodes in the preset network store operation parameters for a preset time period;

[0153] The first node sends a response containing operation parameters to the node to be connected, enabling the node to analyze the operation parameters within a preset time period, thereby achieving physical layer synchronization between the node and the preset network based on the analysis results.

[0154] This disclosure provides a possible implementation method in which the second node obtains the detection information through the first node, and determines that the node to be accessed has the function of receiving the model on the second node based on the preset function in the detection information.

[0155] This disclosure provides a possible implementation method in which the node to be connected searches for neighboring network nodes in a preset network through a first node;

[0156] In response to the network node obtaining information from the first node and determining, based on the preset functions in the probe information, that the node to be accessed has the function of receiving the model on the network node, the network node is set as the second node.

[0157] This disclosure provides a possible implementation, wherein, in response to the first node sending response information to the node to be accessed, the node to be accessed establishes a communication channel with the first node;

[0158] The communication channel between the node to be connected and the first node is defined as the first channel, and the communication channel between the node to be connected and the second node is defined as the second channel.

[0159] The transmission capacity of the first channel is determined based on the size of the model to be transmitted in the first node;

[0160] The transmission capacity of the second channel is determined based on the size of the model to be transmitted in the second node.

[0161] This disclosure provides a possible implementation method in which the node to be connected establishes a list of neighboring nodes, the list of neighboring nodes including node information of the first node and node information of the second node.

[0162] This disclosure provides a possible implementation method in which a first node and a second node authenticate the node to be accessed, and the first node and the second node send authentication response information to the node to be accessed, the authentication response information including authentication success information and authentication failure information.

[0163] This disclosure provides a possible implementation method in which the authentication response information of the first node and the second node are both authentication success information, and the preset network controls the first node and / or the second node to transmit the corresponding model to the node to be accessed through the communication channel according to the model version function.

[0164] If at least one authentication failure message is found among all the authentication response messages received by the node to be accessed, the node to be accessed will continue to broadcast probe information to the preset network.

[0165] This disclosure provides a possible implementation method in which the preset network uses a distributed real-time model evaluation framework to evaluate the learned and trained model.

[0166] The beneficial effects achieved by the embodiments of this disclosure are the same as those described above for the method of network nodes accessing the network in nodes to be accessed in a preset network, and will not be repeated here.

[0167] Example 3

[0168] Figure 10 This invention discloses an apparatus 30 for a network node to access a network, which is applied to a node to be accessed in a preset network. Figure 10 As shown, it includes:

[0169] The broadcast unit 301 is used to broadcast a probe message to a preset network, wherein the probe message includes preset functions of the node to be accessed;

[0170] The receiving unit 302 is used to receive response information sent by the first node in the preset network according to the detection information. The response information includes the operation parameters of the first node in the preset network, and the node to be accessed can receive the model on the first node.

[0171] Synchronization unit 303 is used to establish physical layer synchronization with a preset network according to operating parameters;

[0172] The channel establishment unit 304 is used to find the second node in the preset network through the first node and establish a communication channel with the second node, wherein the node to be accessed is able to receive the model on the second node.

[0173] The acquisition unit 305, in response to receiving successful authentication information from the first node and the second node, acquires the model from the first node and the second node respectively through the communication channel.

[0174] Unit 306 is added to learn and train the acquired model. In response to the model being learned and trained meeting the model evaluation conditions of the preset network, it is added to the preset network through the first node and the second node.

[0175] This disclosure establishes a communication channel between the node to be accessed and a network node in a preset network, and the node to be accessed obtains the model from the network node through the communication channel. At this time, the node to be accessed does not join the preset network. In response to the model capability in the node to be accessed meeting the model evaluation conditions of the preset network, it joins the preset network, thereby improving the access performance of the node to be accessed in the preset network.

[0176] The beneficial effects achieved by the embodiments of this disclosure are the same as those described above for the method embodiments of network nodes accessing the network in nodes to be accessed in a preset network, and will not be repeated here.

[0177] Example 4

[0178] Figure 11This disclosure illustrates a network node access device 40 provided in an embodiment of the present disclosure, applied in a preset network, such as... Figure 11 As shown, the device 40 includes:

[0179] The receiving unit 401 is used to receive probe information broadcast by the node to be accessed, wherein the probe message includes the preset functions of the node to be accessed;

[0180] The lookup unit 402 is used to find the first node that can send response information to the node to be accessed based on the detection information. The model on the first node can be received by the node to be accessed. The response information includes the operation parameters of the first node in the preset network. The operation parameters are used to enable the node to be accessed and the preset network to establish physical layer synchronization.

[0181] The first control unit 403 searches for the second node through the first node and controls the second node to establish a communication channel with the node to be connected, wherein the model on the second node can be received by the node to be connected.

[0182] After receiving the authentication success information from the first node and the second node, the second control unit 404 controls the first node and the second node to send the model to the node to be accessed through the communication channel, so that the node to be accessed can learn and train the obtained model.

[0183] Add unit 405 to evaluate the trained model in the node to be accessed according to the model evaluation conditions. In response to the evaluation being passed, add the node to be accessed through the first node and the second node.

[0184] This disclosure establishes a communication channel between the node to be accessed and a network node in a preset network, and the node to be accessed obtains the model from the network node through the communication channel. At this time, the node to be accessed does not join the preset network. In response to the model capability in the node to be accessed meeting the model evaluation conditions of the preset network, it joins the preset network, thereby improving the access performance of the node to be accessed in the preset network.

[0185] The beneficial effects achieved by the embodiments of this disclosure are the same as those described above for the method embodiments of network nodes accessing the network in nodes to be accessed in a preset network, and will not be repeated here.

[0186] The acquisition, storage, and application of user personal information involved in the technical solution disclosed herein comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0187] According to embodiments of this disclosure, this disclosure also provides an electronic device and a readable storage medium.

[0188] The electronic device includes:

[0189] At least one processor; and

[0190] A memory that is communicatively connected to at least one processor; wherein,

[0191] The memory stores instructions that can be executed by at least one processor, such that at least one processor can perform the methods described above.

[0192] The electronic device establishes a communication channel between the node to be accessed and a network node in the preset network. The node to be accessed obtains the model from the network node through the communication channel. At this time, the node to be accessed does not join the preset network. In response to the model capability of the node to be accessed meeting the model evaluation conditions of the preset network, it joins the preset network, thereby improving the access performance of the node to be accessed in the preset network.

[0193] The non-transitory computer-readable storage medium stores computer instructions for causing a computer to perform the methods provided in embodiments of this disclosure.

[0194] The readable storage medium establishes a communication channel between the node to be accessed and a network node in the preset network. The node to be accessed obtains the model from the network node through the communication channel. At this time, the node to be accessed does not join the preset network. In response to the model capability in the node to be accessed meeting the model evaluation conditions of the preset network, it joins the preset network, thereby improving the access performance of the node to be accessed in the preset network.

[0195] The electronic device includes a processor, a communication interface, a memory, and a communication bus. The communication bus facilitates communication between the processor, the communication interface, and the memory. The communication interface facilitates communication between the electronic device and other devices. The memory stores a computer program. The processor executes the computer program stored in the memory to implement the method for a network node to access the network in any of the above embodiments. A computer-readable storage medium stores a computer program. When executed by a processor, the computer program implements the method for a network node to access the network in any of the above embodiments. The computer-readable storage medium can be any medium that a computer can store or access, including but not limited to electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof.

[0196] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A method for a network node to access a network, characterized in that, The method, applied to nodes seeking to access a preset network, includes: Broadcast a probe message to a preset network, wherein the probe message includes preset functions of the node to be accessed; The node to be connected receives response information sent by a first node in the preset network according to the probe message, wherein the response information includes the operation parameters of the first node in the preset network, and the node to be connected is able to receive the model on the first node. Establish physical layer synchronization with the preset network according to the operating parameters; The first node searches for a second node in the preset network and establishes a communication channel with the second node, wherein the node to be connected is able to receive the model on the second node; Upon receiving successful authentication information from the first node and the second node, the model is obtained from the first node and / or the second node through the communication channel; The obtained model is trained, and in response to the trained model meeting the model evaluation conditions of the preset network, it is added to the preset network through the first node and the second node.

2. The method for a network node to access a network as described in claim 1, wherein, The first node determines, based on the preset function in the received probe message, that the node to be connected has the ability to receive the model on the first node.

3. The method for a network node to access a network as described in claim 1 or 2, wherein, The preset network includes multiple network nodes. The network nodes receive the probe message. In response to the network node determining, based on the preset function of the received probe message, that the node to be accessed has the ability to receive the model on the network node, the network node is selected as the first node according to the shortest route principle. The first node sends the response information of the probe message to the node to be accessed.

4. The method for a network node to access a network as described in claim 1, wherein, The second node obtains the probe message through the first node and determines, based on the preset function in the probe message, that the node to be connected has the ability to receive the model on the second node.

5. The method for a network node to access a network as described in claim 1 or 4, wherein, The node to be connected searches for neighboring network nodes in the preset network through the first node, wherein the neighboring network nodes can receive the probe message; In response to the network node obtaining through the first node and determining, based on the preset function in the probe message, that the node to be accessed has the ability to receive the model on the network node, the network node is set as the second node.

6. The method for a network node to access a network as described in claim 1, 2, or 4, wherein, In response to the first node sending the response information to the node to be accessed, the node to be accessed establishes a communication channel with the first node; The communication channel between the node to be accessed and the first node is defined as the first channel, and the communication channel between the node to be accessed and the second node is defined as the second channel. The transmission capacity of the first channel is determined based on the size of the model in the first node; The transmission capacity of the second channel is determined based on the size of the model in the second node.

7. The method for a network node to access a network as described in claim 1, wherein, The first node and the second node authenticate the node to be accessed, and send authentication response information to the node to be accessed. The authentication response information includes authentication success information and authentication failure information.

8. The method for a network node to access a network as described in claim 7, wherein, The authentication response information of the first node and the second node are both authentication success information. The preset network controls the first node and / or the second node to transmit the corresponding model to the node to be accessed through the communication channel according to the model version function. When at least one authentication failure message is found among all the authentication response messages received by the node to be accessed, the node to be accessed continues to broadcast probe messages to the preset network.

9. The method for a network node to access a network as described in claim 1, wherein, The preset network uses a distributed real-time model evaluation framework to evaluate the learned and trained model.

10. A method for a network node to access a network, characterized in that, When applied to a preset network, the method includes: Receive a probe message broadcast by the node to be connected, wherein the probe message includes the preset functions of the node to be connected; The first node capable of sending response information to the node to be accessed is found based on the probe message. The model on the first node can be received by the node to be accessed. The response information includes the operation parameters of the first node in the preset network. The operation parameters are used to enable the node to be accessed and the preset network to establish physical layer synchronization. The first node is used to find the second node, and the second node is controlled to establish a communication channel with the node to be connected, wherein the model on the second node can be received by the node to be connected; After receiving the authentication success information from the first node and the second node, the system controls the first node and the second node to send the model to the node to be accessed through the communication channel, so that the node to be accessed can learn and train the obtained model. The trained models in the nodes to be connected are evaluated according to the model evaluation conditions. In response to the evaluation being passed, the nodes to be connected are added through the first node and the second node.

11. An apparatus for a network node to access a network, characterized in that, Applied to nodes seeking to access a preset network, including: A broadcast unit is used to broadcast a probe message to a preset network, wherein the probe message includes preset functions of the node to be accessed; A receiving unit is configured to receive response information sent by a first node in the preset network according to the probe message, wherein the response information includes the operation parameters of the first node in the preset network, and the node to be accessed is capable of receiving the model on the first node. A synchronization unit is used to establish physical layer synchronization with the preset network according to the operating parameters; A channel establishment unit is used to locate a second node in the preset network through the first node and establish a communication channel with both the first node and the second node, wherein the node to be accessed is able to receive the model on the second node. The acquisition unit, in response to receiving successful authentication information from the first node and the second node, acquires the model from the first node and the second node respectively through the communication channel; The joining unit is used to learn and train the acquired model. In response to the model being learned and trained meeting the model evaluation conditions of the preset network, it is added to the preset network through the first node and the second node.

12. An apparatus for a network node to access a network, characterized in that, Applied to a pre-defined network, including: A receiving unit is configured to receive a probe message broadcast by a node to be accessed, wherein the probe message includes preset functions of the node to be accessed; The lookup unit is used to find a first node that can send response information to the node to be accessed based on the probe message, wherein the model on the first node can be received by the node to be accessed, and the response information includes the operation parameters of the first node in the preset network, and the operation parameters are used to enable the node to be accessed and the preset network to establish physical layer synchronization. The first control unit locates the second node through the first node and controls the second node to establish a communication channel with the node to be accessed, wherein the model on the second node can be received by the node to be accessed; After receiving the authentication success information from the first node and the second node, the second control unit controls the first node and the second node to send the model to the node to be accessed through the communication channel, so that the node to be accessed can learn and train the obtained model. The addition unit evaluates the trained model in the node to be connected according to the model evaluation conditions, and adds the node to be connected through the first node and the second node after the evaluation is passed.

13. An electronic device, comprising: At least one processor; as well as A memory that is in communication with the at least one processor; wherein, The memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1-9 or 10.

14. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-9 or 10.

Images