Adaptive mesh network architecture construction method for industrial internet of things hybrid networking

Abstract

The application relates to the technical field of computer network communication, and discloses an adaptive Mesh network architecture construction method for industrial Internet of Things hybrid networking, comprising the following steps: when a node in a network receives a data packet to be forwarded, a data quantum state label encapsulated in the data packet is acquired, the data quantum state label is used for representing potential demand intensity of the data packet on different network service quality dimensions; the node calculates a comprehensive cost for forwarding the data packet for each next-hop neighbor node; the comprehensive cost is calculated in combination with matching degrees between the data quantum state label and path capabilities related to the next-hop neighbor node. Through the accurate and differentiated matching of network resources to service demands, a solid foundation is provided for guaranteeing the service quality of key services in complex industrial scenarios, and accurate and selective avoidance of related data flow is realized, so that the load balancing capability and overall throughput of the network are effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer network communication, and particularly to an adaptive Mesh network architecture construction method for industrial Internet of Things hybrid networking. BACKGROUND

[0002] With the deepening of industrial 4.0 and intelligent manufacturing, the industrial Internet of Things has become a key infrastructure for connecting physical production equipment and digital information systems. In complex industrial environments such as large factories, mining areas or automated warehouses, due to wiring difficulties, strong equipment mobility and other factors, wireless Mesh networks with self-organizing and self-repairing characteristics have become a widely used deployment solution as the core bearing technology. Wireless Mesh networks can economically and efficiently achieve wide-range and high-density device coverage due to their flexible topology and high network redundancy.

[0003] However, data communication in modern industrial application scenarios presents unprecedented complexity and diversity. A variety of types of heterogeneous data streams are transmitted in the network at the same time, such as device closed-loop control instructions requiring millisecond-level response, high-definition video monitoring streams occupying a large amount of bandwidth, critical safety alarm signals requiring extremely high transmission success rate, and periodic reporting of non-real-time device status data, etc. These different data streams of business have different requirements for the quality of service of the network, including different emphasis on transmission delay, available bandwidth, and data reliability.

[0004] The existing mainstream wireless Mesh network routing technology usually relies on relatively single or simple physical layer and link layer metrics when making path selection decisions. For example, some protocols take minimizing transmission hops as the core goal, while others use received signal strength indication or estimated transmission times as link quality indicators to construct routing. These methods are universal in design, but their fundamental flaw is that they are "ignorant" of the inherent business needs of the data packets they carry. The network cannot distinguish between an urgent control instruction and a regular log file when making forwarding decisions, but treats them equally and only forwards them according to a universal "best" path standard.

[0005] This kind of "one size fits all" routing mechanism exposes its limitations in the increasingly complex industrial environment. It often leads to the mismatch of network resources, for example, a critical data packet that is extremely sensitive to delay may be routed to a path that is reliable but has higher latency due to a slight advantage of a comprehensive link quality indicator, thus missing the best response opportunity. Conversely, a file transfer task that requires large bandwidth may be indiscriminately directed to an area composed of low-power, narrow-bandwidth nodes, not only inefficient in its own transmission, but also likely to cause congestion and interference to other low-latency services that need to be protected in the area. In addition, the existing technology is often passive and rough in its adjustment mechanism when facing network congestion, lacking precise analysis of the causes of congestion and differentiated guidance ability for data flow, making it difficult to achieve efficient load balancing.

[0006] Therefore, the existing technology urgently needs a new industrial Internet of Things network construction method that can overcome the limitations of traditional physical metrics, realize deep perception of data service intent, and on this basis, establish a self-adaptive network operation mechanism that can dynamically learn, intelligently decide and accurately adjust, to truly meet the diversified and high-performance requirements of network communication for future intelligent industrial applications. SUMMARY

[0007] In view of the shortcomings of the prior art, the present application provides an adaptive Mesh network architecture construction method for industrial Internet of Things hybrid networking, which solves the problem that the routing decision of the existing industrial Internet of Things Mesh network lacks perception of the internal needs of heterogeneous data services, thereby leading to blind routing selection and low matching degree of network resources and service demand.

[0008] To achieve the above purpose, the present application is implemented by the following technical scheme: an adaptive Mesh network architecture construction method for industrial Internet of Things hybrid networking, comprising the following steps:

[0009] S1, when a node in the network receives a data packet to be forwarded, a data quantum state label encapsulated in the data packet is obtained, the data quantum state label is used to represent the potential demand intensity of the data packet for different network service quality dimensions;

[0010] S2, the node calculates a comprehensive cost for forwarding the data packet for each next-hop neighbor node; the calculation of the comprehensive cost combines the matching degree between the data quantum state label and the path ability related to the next-hop neighbor node, the path affinity value based on historical transmission experience, and the dissipation field signaling based on the congestion state of the next-hop neighbor node;

[0011] S3, the node selects the next-hop neighbor node with the lowest cost according to the calculated comprehensive cost for forwarding the data packet.

[0012] Preferably, the data quantum state label is a multi-dimensional vector, whose dimensions include:

[0013] a latency potential, for characterizing the sensitivity of the data packet to transmission latency;

[0014] a bandwidth potential, for characterizing the demand of the data packet to transmission bandwidth;

[0015] a reliability potential, for characterizing the strictness of the data packet to transmission success rate;

[0016] an aggregation potential, for characterizing the path stability demand of the data packet as a part of data flow.

[0017] Preferably, the method further comprises a step of updating the path affinity value:

[0018] when the data packet successfully arrives at the destination node, sending back a confirmation frame along the original path;

[0019] when receiving the confirmation frame, the node on the path increases the path affinity value stored internally and corresponding to the intended type of the data packet.

[0020] Preferably, the step of updating the path affinity value further comprises:

[0021] before increasing the path affinity value, judging whether the number of data packets flowing to the same next-hop neighbor node and having similar data quantum state labels satisfies an affinity resonance condition within a preset time window;

[0022] if the affinity resonance condition is satisfied, a larger preset increment is used to increase the path affinity value, so as to accelerate the memory of the efficient path.

[0023] Preferably, the method further comprises a step of generating the dissipation field signaling:

[0024] the node monitors its own node capability matrix in real time, which contains the processor load and forwarding queue length of the node;

[0025] when an index in the node capability matrix exceeds a preset congestion threshold, the data quantum state label of the data packet in its forwarding queue is analyzed, a dissipation field signaling is generated, which can represent the dominant intention of the current congestion cause, and is broadcasted to its neighbor nodes.

[0026] Preferably, the dissipation field signaling is a dissipation field vector consistent with the dimensions of the data quantum state label, and each component value of the dissipation field vector is determined by the average distribution of the data quantum state labels of all data packets in the forwarding queue analyzed by the node.

[0027] Preferably, the step of calculating the comprehensive cost is specifically calculating by the following formula:

[0028] ;

[0029] wherein, is the comprehensive cost, is the current node, is the next-hop neighbor node, is the data packet; is the mismatch metric, is the correction term based on the path affinity value, is the correction term based on the dissipation field signaling.

[0030] Preferably, the dissipation field correction in the comprehensive cost is calculated by the following formula:

[0031] ;

[0032] wherein, is the dissipation influence factor, is the data quantum state label of the data packet, is the dissipation field vector corresponding to the dissipation field signaling broadcast by the next-hop neighbor node, is a function for calculating the similarity between and

[0033] Preferably, the node capability matrix of the node further comprises: the residual energy of the node and the link state information to each neighbor node, and the link state information contains available bandwidth, average delay and link reliability.

[0034] The adaptive Mesh network architecture construction system of the industrial Internet of Things hybrid networking comprises:

[0035] To solve the above technical problems, the present application provides a novel adaptive Mesh network architecture construction method and system of the industrial Internet of Things hybrid networking. The scheme introduces a mechanism capable of representing the intrinsic demand of data, and establishes a distributed intelligent decision-making framework combining historical experience learning and real-time congestion avoidance, so that the network can self-organize and adaptively match the optimal transmission path for different data.

[0036] The present application provides an adaptive Mesh network architecture construction method of the industrial Internet of Things hybrid networking.

[0037] ​​​​In one embodiment, the method comprises: when a node in the network receives a data packet to be forwarded, first obtaining a data quantum state label encapsulated in the data packet. This data quantum state label is not simply a priority label, but a structured data used to finely characterize the potential demand intensity of the data packet for different network quality of service dimensions (such as delay, bandwidth, reliability, etc.).

[0038] In a preferred embodiment, the data quantum state label is a multi-dimensional vector, the dimensions of which can include:

[0039] a delay potential used to characterize the sensitivity of the data packet to transmission delay;

[0040] a bandwidth potential used to characterize the demand of the data packet for transmission bandwidth;

[0041] a reliability potential used to characterize the strictness of the data packet to transmission success rate;

[0042] an aggregation potential used to characterize the path stability demand of the data packet as part of a data flow.

[0043] After obtaining the data quantum state label, the node will calculate a comprehensive cost for forwarding the data packet for each of its reachable next-hop neighbor nodes . This comprehensive cost is a multi-dimensional, dynamic evaluation result, and its calculation process breaks through the limitations of traditional reliance on single physical indicators, and innovatively integrates three-dimensional considerations:

[0044] the matching degree of intention and ability: evaluating the matching degree between the demand carried by the data quantum state label of the data packet and the actual physical ability provided by the path from the current node to the next-hop neighbor node.

[0045] the memory of historical transmission experience: using a path affinity value distributed among nodes. This value is the memory of the network to historical successful transmission experience, and reflects the effectiveness of a path in history for meeting specific types of data demand.

[0046] avoidance of real-time congestion state: introducing a dissipation field signaling actively broadcasted by a possibly congested downstream node. This signaling can intelligently indicate the type of data causing congestion, so that the upstream node can make a pre-judgment and avoidance.

[0047] In one specific embodiment, the comprehensive cost is calculated by the following formula:

[0048] ;

[0049] wherein, is the current node, is the next-hop neighbor node, is the data packet; is the mismatch metric, is the correction term based on the path affinity value, is the correction term based on the dissipative field signaling.

[0050] Finally, the node selects the neighbor node with the lowest cost value as the best next-hop according to the comprehensive cost calculated by all next-hop neighbor nodes, and transmits the data packet out.

[0051] In order to realize the self-organizing learning of the network, the method of the application further comprises a mechanism for dynamically updating the path affinity value. When the data packet successfully reaches the destination node, an acknowledgement frame will be returned along the original path. After receiving the acknowledgement frame, the nodes on the path will increase the path affinity value corresponding to the intended type of the data packet stored in the internal storage.

[0052] In order to further improve the learning efficiency, in the embodiment, the updating mechanism further introduces the concept of affinity resonance. The node will judge whether the number of data packets flowing to the same next-hop neighbor node and having similar data quantum state labels satisfies an affinity resonance condition within a preset time window. If it satisfies, a larger preset increment is used to increase the path affinity value, so as to realize the rapid convergence and memory reinforcement of the efficient and stable transmission path.

[0053] In order to realize the adaptive adjustment and congestion avoidance of the network, the method of the application further comprises a generation and application mechanism of the dissipative field signaling. In the embodiment, each node will monitor its node capability matrix in real time. When the index in the node capability matrix exceeds the preset congestion threshold, the node will analyze the statistical distribution of the data quantum state labels of the data packets in its congestion queue, thereby generating a dissipative field signaling capable of representing the dominant intention of the current congestion cause, and broadcasting it to its neighbor nodes.

[0054] In a specific embodiment, the dissipative field correction term in the comprehensive cost is calculated by the following formula to realize intelligent avoidance:

[0055] ;

[0056] wherein, is the dissipative influence factor, is the data quantum state label of the data packet, is the data quantum state label of the data packet, a dissipative field vector corresponding to the broadcast dissipative field signaling, is a function for calculating and the similarity between them. This mechanism makes the forwarding cost of a data packet dynamically increase due to its own intention and the relevance of the congestion cause of the downstream node, thereby achieving accurate and selective avoidance of congestion rather than indiscriminate path suppression.

[0057] The second aspect of the application provides an adaptive Mesh network architecture construction system for industrial Internet of Things hybrid networking.

[0058] The system is designed to perform the steps of the aforementioned method, which can include:

[0059] a data encapsulation module configured to encapsulate a data quantum state label in the data packet;

[0060] a routing decision module configured to calculate the comprehensive cost of forwarding to each next-hop neighbor node when the node in the system receives the data packet;

[0061] a forwarding selection module configured to select the next-hop neighbor node with the lowest cost for forwarding according to the calculated comprehensive cost.

[0062] When calculating the comprehensive cost, the routing decision module processes the data quantum state label of the data packet, the path affinity value stored by the node, and the dissipative field signaling received from the neighbor node.

[0063] The application provides an adaptive Mesh network architecture construction method for industrial Internet of Things hybrid networking. The method has the following beneficial effects:

[0064] 1. The application introduces a data quantum state label that can finely represent the internal demand of data services, and performs matching calculation with the physical capabilities of the path, so that the network node has the ability to deeply perceive the data intention for the first time. Compared with the traditional technology that only relies on physical layer indicators for blind forwarding, the application can distinguish services with different quality of service requirements from the source, and evaluate the matching degree of intention and capability at each hop, thereby realizing accurate and differentiated matching of network resources to service demand, and providing a solid foundation for guaranteeing the quality of service of key services in complex industrial scenarios.

[0065] 2. The application establishes a negative feedback regulation mechanism based on dissipative field signaling, realizes an active and accurate congestion avoidance, and the upstream node can calculate the similarity between its data packet intention and the dissipative field intention when making routing decisions, to judge the risk of exacerbating congestion, thereby realizing accurate and selective avoidance of related data flow, and effectively improving the load balancing capability and overall throughput of the network.

[0066] 3、The application constructs a self-organizing network without central controller and with high robustness by distributing the core intelligence such as routing decision, path memory and congestion regulation to each network node and relying on local interaction rules to emerge macroscopic adaptive behavior, which eliminates the inherent single point failure risk and performance bottleneck of centralized scheme, enables the network to adapt to dynamic events such as node failure, link quality fluctuation and traffic load change autonomously and flexibly, and exhibits stronger environmental adaptability and self-repairing ability. BRIEF DESCRIPTION OF DRAWINGS

[0067] Figure 1 is a method flowchart of the application;

[0068] Figure 2 is a system architecture diagram of the application. DETAILED DESCRIPTION

[0069] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the application.

[0070] Embodiment:

[0071] Please refer to the accompanying drawings Figure 1 The embodiment of the application provides an adaptive Mesh network architecture construction method of industrial Internet of Things hybrid networking, including the following steps:

[0072] S1, when a node in the network receives a data packet to be forwarded, a data quantum state label encapsulated in the data packet is obtained, and the data quantum state label is used to represent the potential demand intensity of the data packet for different network service quality dimensions;

[0073] In this embodiment, in order to construct an adaptive Mesh network capable of accurately perceiving its own state and network environment, S1, node initialization and state representation, as a basic step of the method of the application, its core is to establish a dynamic and real-time self-state description mechanism for each forwarding node in the network.

[0074] In an embodiment, when a physical device, such as a sensor gateway or wireless router, is first powered on and intends to join the mesh network built by the present application, it will perform a network initialization procedure. The procedure can include, but is not limited to, finding and connecting to one or more designated network coordinators through a broadcast discovery protocol, completing identity authentication, obtaining a globally unique internal network identification ID, and synchronizing basic network configuration parameters, etc. This initialization step ensures that the node has the basic qualifications to be uniquely identified and managed in the network, and is the prerequisite for its participation in all subsequent network activities.

[0075] After the initialization is completed, the core of the method enters the stage of continuously and dynamically characterizing the node's own capabilities. Specifically, each node will monitor and maintain a structured data set locally in real time, which is defined as the node capability matrix in the present application. The node capability matrix is not a static configuration parameter, but a self-cognition basis for the node to make distributed and intelligent decisions, and the accuracy and timeliness of its information is directly related to the final effect of the adaptive adjustment of the entire network.

[0076] In a specific embodiment, the node capability matrix , for example, can include the following multiple dimension key performance indicators, each of which reflects the current state of the node from a specific side: First, in order to characterize the node's computing resource occupation, the node capability matrix includes processor load

[0077] . This indicator reflects the busy degree of the node's internal microcontroller unit or central processing unit. In a specific implementation, this value can be periodically obtained by monitoring the CPU utilization statistical information provided by the node's operating system kernel. A higher processor load means that the node may not be able to process complex protocol stack calculations or high-rate packet forwarding in a timely manner, which is crucial for evaluating its ability to handle delay-sensitive services. Second, in order to intuitively reflect the congestion degree of the node's data forwarding, the node capability matrix includes forwarding queue length

[0078] . This indicator directly quantifies how many data packets are currently queued in the node's network interface buffer waiting to be sent. This value can be obtained by querying the status information of the network interface driver. Third, for nodes deployed in special workstations and powered by batteries or other limited energy sources, the node capability matrix preferably includes remaining energy

[0079] . This indicator reflects the current energy level of the node's power supply, which is crucial for evaluating its ability to handle energy-constrained services. This index is of great importance to the energy efficiency balance of the whole network and the life cycle of specific nodes. In the process of routing decision, actively avoiding nodes with low energy can effectively prevent communication interruption caused by unexpected offline of nodes. For nodes powered by stable external power supply, this index can be set to a constant maximum value.

[0080] In addition, in order to comprehensively evaluate the pros and cons of a potential path, the node capability matrix must also include link state information about connection quality . This information is not a single value, but a set that details the dynamic characteristics of the wireless or wired link from the current node to each of its neighbor nodes. Taking the link to a neighbor node as an example, its link state information can be subdivided into:

[0081] Available bandwidth : This index is used to evaluate the data carrying capacity of the link. In a preferred embodiment, it can be actively estimated by sliding average calculation of the data throughput of recent successful transmissions, or by sending a lightweight sequence of probe packets during idle periods, thus obtaining a value closer to reality than the theoretical bandwidth.

[0082] Average delay : This index is a key measure of link response speed. It can be accurately obtained by measuring the round-trip time of acknowledgment frames, or by one-way delay estimation through the embedding of timestamps in data packets.

[0083] Link reliability : This index is used to evaluate the stability of the link. It can be quantified by statistics of the data packet success rate in the recent time window, i.e. subtracting the packet loss rate. High reliability links are the first choice for transmission of critical control or status data.

[0084] To ensure that the node capability matrix can truly reflect the rapidly changing industrial environment, the node will perform a complete update of all the above-mentioned indexes at a preset, relatively high frequency. This continuous, high-frequency self-state representation ensures that the path capability information relied upon in the subsequent S2 step of comprehensive cost calculation is highly timely and accurate.

[0085] S2, the node calculates a comprehensive cost for forwarding data packets for each next-hop neighbor node. The calculation of the comprehensive cost combines the degree of matching between the data quantum state label and the path capability of the next-hop neighbor node, the path affinity value based on historical transmission experience, and the dissipation field signaling based on the congestion state of the next-hop neighbor node.

[0086] In this embodiment, when the method of the present invention is executed to step S2: the node receives the data packet to be forwarded and makes routing decisions and forwards it based on the data intent, historical experience and real-time congestion status, it demonstrates the core intelligence of the present invention that distinguishes it from traditional routing mechanisms.

[0087] Specifically, when a forwarding node in the network Upon receiving a packet Pk to be forwarded, the node immediately initiates a distributed, multi-dimensional routing decision-making process. The goal of this process is to find the most adaptive forwarding path for the packet among all possible next-hop neighbors.

[0088] The first step in this decision-making process is to parse and obtain the data quantum state tags encapsulated within the data packet Pk. The data quantum state tag is a key technical feature for achieving intent awareness in this invention. It is assigned at the source of data generation, carrying the inherent and potential needs of the data packet for network services. In a preferred embodiment, the tag is a standardized four-dimensional vector. Each component is a logical starting point for all subsequent calculations, based on four dimensions: latency, bandwidth, reliability, and aggregation.

[0089] After obtaining the quantum state label of the data, the node The core task is to provide each of its active next-hop neighbor nodes. Calculate a comprehensive cost for evaluating the quality of forwarding. This comprehensive cost calculation abandons the limitations of a single physical metric, innovatively integrating data intent, route capability, historical experience, and real-time congestion warnings to form a comprehensive and dynamic basis for decision-making.

[0090] In one specific embodiment, the overall cost is calculated according to the following formula:

[0091] ;

[0092] The three components of this formula each embody different decision-making considerations, and their detailed principles are explained below:

[0093] First, there is the mismatch in the measurement items. This item aims to address the issue from the most fundamental physical level. To calculate this, nodes... First, it is necessary to determine the neighbor nodes obtained in step S1. Node capability matrix In addition to the link state information between the two, a path capability vector consistent with the dimension of the data quantum state label is constructed. For example, the delay capability component of the path capability vector. is according to neighbor nodes The average delay reported in its NCM , after normalization, whose value is inversely proportional to the delay size. Similarly, the bandwidth capacity component is directly proportional to the available bandwidth .

[0094] After constructing the path capacity vector, the mismatch metric term can be calculated by the following formula:

[0095] ;

[0096] In this formula, is a system preset, configurable weight vector, which allows network managers to assign different importance to different dimensions of mismatch according to the overall strategy. For example, for a network whose main business is real-time control, the weight of the delay dimension can be assigned a higher value. The calculation result of this term intuitively reflects to what extent the internal demand of node cannot be met if the data packet is sent to it.

[0097] Secondly, the path affinity correction term , which enables the network to remember and prefer paths that have been proven to be effective for a specific type of service in history. In the embodiment, its calculation formula is:

[0098] ;

[0099] In this formula, represents the path affinity value for a specific intent type maintained and shared by neighbor nodes . The intent type may be preferably defined as the type corresponding to the component with the largest value in the data packet QST vector. is an adjustable affinity influence factor. This mechanism is a direct manifestation of the positive feedback learning effect in subsequent step S3.

[0100] Thirdly, the dissipation field correction term , which is the core mechanism for realizing active and intelligent congestion avoidance in the present application, and gives the network a forward-looking early warning ability. Its calculation formula is as follows:

[0101] ;

[0102] In this formula, is the dissipation field value of neighbor nodes The dissipative field vector corresponding to the dissipative field signaling actively broadcasted, the generation of which will be described in detail in step S4, has the core function of announcing the dominant data intention type causing congestion to the upstream node. is an adjustable dissipative influence factor.

[0103] Cosine similarity function The application here is particularly crucial, it is not simply to punish all paths leading to the congested node, but to accurately calculate the intention vector of the current data packet Consistency between the dominant intention vector causing downstream congestion This means that if the demand of a data packet is completely different from the reason for the congestion of the downstream node, the value calculated by this item will be low, and the forwarding obstacle of the data packet will be small. On the contrary, if the intention of the data packet is highly consistent with the reason for the congestion, the value of this item will be significantly increased, thereby effectively guiding the data packet to avoid the congestion point.

[0104] When the node calculates the comprehensive cost of all reachable next-hop neighbor nodes, it will perform a deterministic selection operation:

[0105] ;

[0106] That is, the neighbor node with the lowest comprehensive cost is selected as the best next hop for this forwarding. Finally, the node sends the data packet Pk through the network interface connected to the node , thereby completing a complete routing and forwarding operation based on multi-dimensional intelligent decision-making.

[0107] S3, the node selects the next-hop neighbor node with the lowest cost according to the calculated comprehensive cost, for forwarding the data packet.

[0108] In this embodiment, in order to enable the adaptive mesh network to have the ability to learn from successful experience and form effective memory, the method of the present application further performs step S3: the node updates the network path memory in a positive feedback manner containing resonance mechanism according to the successful transmission experience. This step is the core positive feedback link for realizing network self-organization and path optimization in the present application, which enables the network to dynamically identify, consolidate and preferentially use those transmission paths that have been proven to be effective for specific business demands.

[0109] In a specific implementation, the positive feedback learning mechanism is triggered in conjunction with a successful end-to-end data transmission event. Specifically, when a data packet, after passing through the routing decision process in step S2, is successfully delivered to its final destination node, the destination node generates and sends an acknowledgment frame as a success signal. This acknowledgment frame does not involve rerouting within the network; instead, it is strictly limited to hop-by-hop forwarding along the path traversed by the original data packet, returning to the data source. This design ensures that every intermediate forwarding node contributing to a successful transmission receives this positive feedback signal without omission, thus providing them with the opportunity to participate in the construction of network memory.

[0110] When any intermediate node on the path Upon receiving this acknowledgment frame, it will immediately initiate an update process for its internally stored path affinity value. The path affinity value is not a general evaluation of the next-hop node, but rather a fine-grained memory tightly coupled to the packet's intent type. Specifically, the node... The intent type of the original data packet will first be parsed from the acknowledgment frame. Subsequently, the node This will be specifically targeted at this type of intent. And the next hop direction for generating the confirmation frame, to perform affinity enhancement operations.

[0111] In this embodiment, the affinity value is not updated using a static linear accumulation method, but rather a more intelligent affinity resonance mechanism that senses flow patterns. This mechanism allows the affinity value to be updated more dynamically. It is dynamic and changes, and its core is that the node will perform a resonance condition judgment before updating.

[0112] The logic for determining the affinity resonance condition is as follows: node It will check within the most recent preset time window Within this node, has the number of data packets forwarded to the same next-hop neighbor node that generated the acknowledgment frame, and which have similar data quantum state tags, exceeded a preset resonance threshold? .

[0113] Based on the results of the resonance condition determination, the increment of the affinity value It will be determined in the following way:

[0114] ;

[0115] In this formula, It is a regular, small base increment used to provide a basic reward for all successful transmissions. is a much larger resonance increment. The technical purpose of this design is that when a stable, high-flow transmission channel for a specific service is formed in the network, the method of the present application can identify the formation of such consensus and, by using a larger increment , so that it can quickly stand out in subsequent routing competition.

[0116] To ensure the long-term adaptability of the network and prevent it from being unable to adapt to new changes in the network environment due to the solidification of historical path information, the affinity update mechanism also includes an essential affinity decay link. Specifically, all affinity values stored in all nodes will be periodically decayed over time at a fixed, slower rate.

[0117] Therefore, in a complete update cycle, the final state of an affinity value will be determined by the increment accumulation and time decay, and its complete update formula can be expressed as:

[0118] ;

[0119] wherein, is the affinity value before updating, is the value after updating, is a small numerical decay factor preset by the system. This decay mechanism ensures that the affinity advantage of any path is not permanent. Once a path is no longer used or its performance declines, its affinity value will gradually be forgotten, thus providing opportunities for the emergence of new, better paths.

[0120] Through the execution of the above step S3, the present application constructs a complete, closed-loop positive feedback learning system. A successful transmission will enhance the affinity of a specific path for a specific intent through this step; and this enhanced affinity value will directly affect the comprehensive cost calculation of subsequent data packets in step S2, increasing the probability of the path being selected again. This iterative reinforcement process enables the network to automatically and without human intervention to emerge optimized virtual transmission channels for different service types.

[0121] Please refer to the accompanying Figure 2 , another embodiment of the present application provides an adaptive mesh network architecture construction system for industrial Internet of Things hybrid networking, comprising the following steps:

[0122] A data encapsulation module configured to encapsulate a data quantum state label in the data packet, the data quantum state label being used to represent the potential demand intensity of the data packet for different network service quality dimensions;

[0123] a routing decision module configured to calculate, for each next-hop neighbor node of a node in the system, a composite cost for forwarding when the node receives a data packet, the composite cost being calculated based on a match between a data quantum state label and a path capability associated with the next-hop neighbor node, a path affinity value based on historical transmission experience, and a dissipation field signaling based on a congestion state of the next-hop neighbor node;

[0124] a forwarding selection module configured to select, based on the composite cost calculated by the routing decision module, a next-hop neighbor node with a lowest cost for forwarding the data packet.

[0125] The system of the embodiment can be used to perform the method embodiments described above, and has similar principles and technical effects, which will not be described here again.

Claims

1. A method for constructing an adaptive mesh network architecture for hybrid industrial IoT networking, characterized in that, Includes the following steps: S1. When a node in the network receives a data packet to be forwarded, it acquires a data quantum state tag encapsulated in the data packet. The data quantum state tag is used to characterize the potential demand intensity of the data packet for different network service quality dimensions. S2. The node calculates the total cost for forwarding the data packet for each of its next-hop neighbor nodes; The calculation of the overall cost combines the matching degree between the data quantum state label and the path capability associated with the next-hop neighbor node, the path affinity value obtained based on historical transmission experience, and the dissipation field signaling is obtained based on the congestion state of the next-hop neighbor node. S3. The node selects the next-hop neighbor node with the lowest cost based on the calculated comprehensive cost, and uses it to forward the data packet; The method further includes the step of generating the dissipation field signaling: The node monitors its own node capability matrix in real time, which includes the node's processor load and forwarding queue length; When the index in the node capability matrix exceeds the preset congestion threshold, the data quantum state label of the data packets in its forwarding queue is analyzed, a dissipation field signaling that can characterize the dominant intention of the current congestion cause is generated, and it is broadcast to its neighboring nodes. The dissipation field signaling is a dissipation field vector with the same dimension as the data quantum state label. The values ​​of each component of the dissipation field vector are determined by the node analyzing the average distribution of the data quantum state labels of all data packets in its forwarding queue. The specific steps for calculating the comprehensive cost are as follows: The cost is calculated using the following formula: ； in, The total cost is... For the current node, For the next-hop neighbor node, For data packets; For a mismatched metric, This is a correction term based on the path affinity value. This is a correction term based on the dissipation field signaling.

2. The method for constructing an adaptive Mesh network architecture for industrial IoT hybrid networking according to claim 1, characterized in that, The data quantum state label is a multi-dimensional vector, whose dimensions include: The delay potential is used to characterize the sensitivity of the data packet to transmission delay; Bandwidth potential is used to characterize the bandwidth requirement of the data packet. The reliability potential is used to characterize how stringent the data packet is to the success rate of transmission. The aggregation potential is used to characterize the path stability requirement of the data packet as part of the data stream.

3. The method for constructing an adaptive mesh network architecture for industrial IoT hybrid networking according to claim 1, characterized in that, The method further includes the step of updating the path affinity value: Once the data packet successfully reaches the destination node, an acknowledgment frame is returned along the original path. Upon receiving the acknowledgment frame, a node on the path increases its internally stored path affinity value, which corresponds to the intent type of the data packet. The step of updating the path affinity value further includes: Before increasing the path affinity value, it is determined whether the number of data packets with similar data quantum state tags flowing to the same next-hop neighbor node within a preset time window satisfies the affinity resonance condition. If the affinity resonance condition is met, a preset increment is used to increase the path affinity value in order to accelerate the memorization of efficient paths.

4. The method for constructing an adaptive Mesh network architecture for industrial IoT hybrid networking according to claim 1, characterized in that, Dissipation field correction in the overall cost The calculation is performed using the following formula: ； in, As a dissipative influence factor, The data quantum state tag of the data packet, For the next-hop neighbor node The dissipation field vector corresponding to the dissipation field signaling of the broadcast. For use in calculation and A function of similarity between them.

5. The method for constructing an adaptive Mesh network architecture for industrial IoT hybrid networking according to claim 1, characterized in that, The node capability matrix of the node also includes: the node's remaining energy and the link state information to each neighboring node, wherein the link state information includes available bandwidth, average latency and link reliability.

6. An adaptive mesh network architecture construction system for hybrid industrial IoT networking, comprising the adaptive mesh network architecture construction method for hybrid industrial IoT networking according to any one of claims 1-5, characterized in that, include: A data encapsulation module is configured to encapsulate a data quantum state tag within a data packet, the data quantum state tag being used to characterize the potential demand intensity of the data packet for different network service quality dimensions; The routing decision module is configured to calculate a comprehensive cost for forwarding for each next-hop neighbor node when a node in the system receives the data packet. The calculation of the comprehensive cost combines the matching degree between the data quantum state label and the path capability associated with the next-hop neighbor node, the path affinity value obtained based on historical transmission experience, and the dissipation field signaling is obtained based on the congestion state of the next-hop neighbor node. The forwarding selection module is configured to select the next-hop neighbor node with the lowest cost based on the comprehensive cost calculated by the routing decision module, for forwarding the data packet.

Images