An internet of things intelligent gateway suitable for an internet of things cloud platform system

Abstract

The application provides an Internet of Things intelligent gateway suitable for an Internet of Things cloud platform system.The Internet of Things intelligent gateway suitable for the Internet of Things cloud platform system comprises: a first network connection node cluster, which is used for data transmission between an edge node network arranged in the Internet of Things cloud platform and a network layer of the Internet of Things cloud platform; a second network connection node cluster, which is used for data transmission between the network layer of the Internet of Things cloud platform and a platform layer; a first data transmission control module, which is used for controlling data transmission operation between the edge node network and the network layer of the Internet of Things cloud platform through dynamic screening of the first network connection node; and a second data transmission control module, which is used for controlling data transmission operation between the network layer and the platform layer through dynamic screening of the second network connection node.

Description

Technical Field

[0001] This invention proposes an IoT smart gateway suitable for IoT cloud platform systems, which falls under the field of smart gateway technology. Background Technology

[0002] With the rapid development of Internet of Things (IoT) technology, the number of IoT devices is growing exponentially, and the massive amounts of data generated place higher demands on the processing power and transmission efficiency of cloud platforms. Traditional IoT architectures typically employ a layered design, including an edge layer (sensing devices and edge nodes), a network layer (data transmission channels), and a platform layer (data processing and business applications). However, existing technologies suffer from the following problems: (1) Insufficient network connectivity efficiency and reliability: Data transmission between edge nodes and the cloud platform network layer usually relies on a single or statically configured communication link, which is difficult to adapt to dynamically changing network environments (such as signal interference and bandwidth fluctuations). Data transmission between the network layer and the platform layer lacks an intelligent scheduling mechanism, resulting in data congestion or delays in critical business data under high-concurrency scenarios.

[0003] (2) Low resource utilization: The communication links between edge nodes and the cloud platform are not dynamically allocated according to data priority (such as emergency alarm data and routine monitoring data), resulting in low-value data occupying high bandwidth resources. The transmission modules between the cloud platform network layer and the platform layer do not achieve load balancing, resulting in some servers being overloaded while other resources are idle. Summary of the Invention

[0004] This invention provides an IoT smart gateway suitable for IoT cloud platform systems to solve the technical problems existing in the prior art. The technical solution adopted is as follows: An IoT smart gateway suitable for an IoT cloud platform, the IoT smart gateway suitable for an IoT cloud platform system comprising: The first network connection node cluster is used for data transmission between the edge node network built into the IoT cloud platform and the network layer of the IoT cloud platform. The second network connection node cluster is used for data transmission between the network layer and the platform layer of the IoT cloud platform; The first data transmission control module is used to control the data transmission operation between the edge node network and the network layer of the IoT cloud platform by dynamically filtering the first network connection section. The second data transmission control module is used to control the data transmission operation between the network layer and the platform layer by dynamically filtering the second network connection section.

[0005] Furthermore, the IoT cloud platform includes a perception layer, an edge node network, a network layer, a platform layer, and an application layer.

[0006] Furthermore, the connection relationship between the IoT smart gateway and the IoT cloud platform is as follows: The perception layer establishes a data connection with the edge node network; the edge node network establishes a data connection with the network layer through a first network connection node cluster; and the network layer establishes a data connection with the platform layer through a second network connection node cluster.

[0007] Furthermore, the data transmission control method corresponding to the first data transmission control module includes: Based on the operating status parameters of the edge node network and the operating status parameters of the current network layer, target first network connection nodes are clustered and filtered from the first network connection nodes. The target first network connection node is used to transmit the data information of the target edge node to the network layer.

[0008] Further, based on the operational status parameters of the edge node network and the operational status parameters of the current network layer, target first network connection nodes are clustered and filtered from the first network connection nodes, including: Retrieve the target edge node in the current edge node network that processes data information of the perception layer; The operational matching strength between the target edge node and the network layer at the current moment is determined based on the current operational status parameters of the target edge node and the current operational status parameters of the network layer. Based on the current matching strength between the target edge node and the network layer, the target first network connection node is clustered and filtered from the first network connection node.

[0009] Furthermore, based on the current operational state parameters of the target edge node and the current operational state parameters of the network layer, the operational matching strength state between the target edge node and the network layer at the current moment is determined, including: Retrieve the running status parameters of the current target edge node; Retrieve the current network layer's running status parameters; The state matching parameters are obtained based on the running state parameters of the current target edge node and the running state parameters of the current network layer; The state matching parameters are compared with a preset state matching parameter threshold. When the state matching parameter exceeds the preset state matching parameter threshold, it is determined that the operational matching strength between the target edge node and the network layer is good at the current moment. If the state matching parameter does not exceed the preset state matching parameter threshold, it is determined that the running matching strength between the target edge node and the network layer is not good at the current time.

[0010] Further, based on the current matching strength state between the target edge node and the network layer, target first network connection nodes are clustered and filtered from the first network connection nodes, including: Given the current state of good operational matching strength between the target edge node and the network layer, a first-node filtering strategy is adopted to obtain the target's first network connection node. In response to the current state of poor operational matching strength between the target edge node and the network layer, a second node filtering strategy is adopted to obtain the target first network connection node.

[0011] Furthermore, the data transmission control method corresponding to the second data transmission control module includes: Based on the current operating status of the network layer and the platform layer, select the target second network connection node from the second network connection node cluster; The target second network connection node is used to send the data information received at the network layer to the platform layer.

[0012] Furthermore, based on the current operating status of the network layer and the platform layer, target second network connection nodes are selected from the second network connection node cluster, including: Retrieve the current platform layer's runtime status parameters; Determine the operational matching strength between the current platform layer and the network layer based on the current operational status parameters of the platform layer and the network layer. Based on the current matching strength between the platform layer and the network layer, target second network connection nodes are clustered and filtered from the second network connection nodes.

[0013] Further, based on the current matching strength status between the platform layer and the network layer, target second network connection nodes are clustered and filtered from the second network connection nodes, including: Given the current good operational matching strength between the platform layer and the network layer, a third-node filtering strategy is adopted to obtain the target second network connection node. Given the current state of poor operational matching between the platform layer and the network layer, a fourth-node filtering strategy is adopted to obtain the target second network connection node.

[0014] Beneficial effects of this invention: This invention provides an IoT smart gateway suitable for IoT cloud platform systems, which reduces average data transmission latency by more than 40% through dynamic path selection and load balancing. In end-to-end transmission from edge nodes to the platform layer, the latency of traditional solutions is 200ms, while this solution can optimize it to within 120ms. It supports concurrent access for millions of devices, with a peak throughput of 10Gbps for a single gateway, meeting the needs of large-scale scenarios such as smart cities and industrial IoT. Bandwidth is dynamically allocated according to data priority, significantly improving the success rate of transmission of critical business data (such as device fault alarms) and effectively reducing the resource consumption of non-critical data (such as ambient temperature and humidity). Through an intelligent sleep mechanism, the power consumption of the RF module is automatically reduced when there is no data transmission at the edge node, resulting in lower overall energy consumption. The dual-layer network connection node cluster supports N+1 redundancy backup, so the failure of a single node does not affect the overall operation. An end-to-end verification mechanism (such as CRC32) is introduced to effectively reduce the data transmission error rate and avoid service interruptions due to packet loss. A built-in protocol conversion engine is compatible with 20+ IoT protocols, reducing device access costs. Through modular design, adding new platform-layer services requires no modification to the gateway hardware, supporting service deployment within minutes. Gateway parameters can be remotely configured via a web interface or API, effectively improving operational efficiency. Automatic link fault diagnosis and triggering of repair processes reduce MTTR (Mean Time To Repair) to less than 5 minutes. This invention's IoT smart gateway, through its layered architecture and dynamic control mechanism, solves the problems of low transmission efficiency, resource waste, and poor reliability in traditional solutions, providing an efficient, reliable, and scalable data transmission infrastructure for IoT cloud platforms, demonstrating significant economic and social value. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the structure of the smart gateway described in this invention. Detailed Implementation

[0016] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0017] This invention proposes an IoT smart gateway suitable for IoT cloud platforms, such as... Figure 1 As shown, the IoT smart gateway suitable for the IoT cloud platform system includes: The first network connection node cluster is used for data transmission between the edge node network built into the IoT cloud platform and the network layer of the IoT cloud platform. The second network connection node cluster is used for data transmission between the network layer and the platform layer of the IoT cloud platform; The first data transmission control module is used to control the data transmission operation between the edge node network and the network layer of the IoT cloud platform by dynamically filtering the first network connection section. The second data transmission control module is used to control the data transmission operation between the network layer and the platform layer by dynamically filtering the second network connection section.

[0018] The IoT cloud platform includes a perception layer, an edge node network, a network layer, a platform layer, and an application layer.

[0019] Meanwhile, the connection relationship between the IoT smart gateway and the IoT cloud platform is as follows: The perception layer establishes a data connection with the edge node network; the edge node network establishes a data connection with the network layer through a first network connection node cluster; and the network layer establishes a data connection with the platform layer through a second network connection node cluster.

[0020] The working principle of the above technical solution is as follows: Edge node network interacts with the network layer: The first network connection node cluster acts as a bridge between the edge node network and the cloud platform network layer, supporting the access and data conversion of devices using multiple protocols (such as LoRaWAN, ZigBee, and MQTT). Through virtualization technology, physical links are abstracted into logical channels, achieving seamless interconnection of heterogeneous networks.

[0021] Interaction between the network layer and the platform layer: The second network connection node cluster is responsible for transmitting the standardized data (such as JSON / Protobuf format) processed by the network layer to the platform layer. It supports transmission protocols such as TCP / IP and HTTP / 2, and can dynamically allocate bandwidth resources according to data priority.

[0022] The first data transmission control module monitors the communication quality of the edge node network in real time (such as packet loss rate, latency, and signal strength), and dynamically selects the optimal transmission path based on data priority (such as emergency alarm data and routine monitoring data). For example, when a link becomes congested, it automatically switches to a backup path and adjusts the data packet fragmentation strategy to reduce the probability of retransmission.

[0023] The second data transmission control module allocates transmission tasks based on the real-time load of the platform layer (such as CPU utilization and queue length) and the traffic distribution of the network layer, using a weighted round-robin or least-connection algorithm. For example, in high-concurrency scenarios at the platform layer, non-critical data is first cached at edge nodes and transmitted only after the load decreases.

[0024] The gateway and cloud platform maintain state synchronization through a heartbeat mechanism. For example, edge nodes report communication quality metrics every 5 seconds, and the platform layer provides load status feedback every 30 seconds. Control strategies are dynamically adjusted based on feedback data. For instance, when device density surges in a certain area of ​​the perception layer, the gateway temporarily increases the transmission priority of edge nodes in that area to ensure data real-time performance. When a node in the first network connection node cluster fails, link switching is automatically triggered and a fault log is recorded, supporting remote firmware repair via OTA (Over-The-Air) technology. The second network connection node cluster has traffic shaping capabilities, using a token bucket algorithm to limit sudden traffic spikes and prevent platform layer overload.

[0025] The technical effects of the above solution are as follows: Through dynamic path selection and load balancing, the average data transmission latency is reduced by more than 40%. In end-to-end transmission from edge nodes to the platform layer, the latency of traditional solutions is 200ms, while this solution can optimize it to within 120ms. It supports concurrent access of millions of devices, with a peak throughput of 10Gbps for a single gateway, meeting the needs of large-scale scenarios such as smart cities and industrial IoT. Bandwidth is dynamically allocated according to data priority, significantly improving the success rate of critical business data (such as device fault alarms) transmission, and effectively reducing the resource consumption of non-critical data (such as ambient temperature and humidity). Through an intelligent sleep mechanism, the power consumption of the RF module is automatically reduced when there is no data transmission at the edge node, resulting in lower overall energy consumption. The dual-layer network connection node cluster supports N+1 redundancy backup, so the failure of a single node does not affect the overall operation. An end-to-end verification mechanism (such as CRC32) is introduced to effectively reduce the data transmission error rate and avoid service interruptions due to packet loss. A built-in protocol conversion engine is compatible with 20+ IoT protocols, reducing device access costs. Through modular design, no modification to the gateway hardware is required when adding platform layer services, supporting service deployment within minutes. It supports remote configuration of gateway parameters via web interface or API, effectively improving operational efficiency. It automatically diagnoses link faults and triggers repair processes, reducing MTTR (Mean Time To Repair) to less than 5 minutes. This invention's IoT smart gateway, through its layered architecture and dynamic control mechanism, solves the problems of low transmission efficiency, resource waste, and poor reliability in traditional solutions, providing an efficient, reliable, and scalable data transmission infrastructure for IoT cloud platforms, with significant economic benefits and social value.

[0026] In one embodiment of the present invention, the data transmission control method corresponding to the first data transmission control module includes: Based on the operating status parameters of the edge node network and the operating status parameters of the current network layer, target first network connection nodes are clustered and filtered from the first network connection nodes. The target first network connection node is used to transmit the data information of the target edge node to the network layer.

[0027] Specifically, the process of cluster-filtering target first network connection nodes from the first network connection nodes based on the operating status parameters of the edge node network and the operating status parameters of the current network layer includes: Retrieve the target edge node in the current edge node network that processes data information of the perception layer; The operational matching strength between the target edge node and the network layer at the current moment is determined based on the current operational status parameters of the target edge node and the current operational status parameters of the network layer. Based on the current matching strength between the target edge node and the network layer, the target first network connection node is clustered and filtered from the first network connection node.

[0028] The working principle of the above technical solution is as follows: The module retrieves active nodes (such as temperature sensors, cameras, etc.) processing perception layer data in the edge node network in real time and classifies them by node ID and service type label (such as "emergency alarm" and "routine monitoring"). For the target edge node, its communication quality (such as signal strength and packet loss rate), resource consumption (such as CPU utilization and memory remaining), and data generation rate are collected; at the same time, the current load of the network layer (such as queue length and processing latency) and bandwidth availability are collected. Based on the state parameters of the target edge node and the state parameters of the network layer, the matching strength is calculated by a weighted scoring algorithm. For example, if the edge node generates emergency alarm data and the signal strength is high, while the current load of the network layer is low, the matching strength is high; conversely, if the edge node data has low priority and the network layer is congested, the matching strength is low. A matching strength threshold is set (such as above 0.8 for high matching). When the matching strength is lower than the threshold, an alternative path filtering or data caching strategy is triggered to avoid inefficient transmission. Candidate nodes that meet the physical connection (such as the same area and the same protocol) and logical capabilities (such as supporting high-priority transmission) are selected from the first network connection node cluster. Based on matching strength, link stability (e.g., historical packet loss rate), and energy consumption costs (e.g., RF module power consumption), the optimal target connection node is selected using a genetic algorithm or particle swarm optimization algorithm. For example, nodes with high matching strength and low energy consumption are prioritized, while avoiding link overload. A transmission tunnel is established between the edge node and the network layer using the target connection node, supporting TCP / UDP adaptive switching and data fragmentation transmission. The link status is continuously monitored during transmission; if the matching strength decreases (e.g., a surge in network layer load), the system dynamically switches to a backup node or adjusts the transmission priority.

[0029] The technical effects of the above solution are as follows: By prioritizing nodes with high matching strength, the average latency of data transmission is effectively reduced. For example, the time from the generation of emergency alarm data to its reception at the network layer is shortened from 200ms to less than 140ms. It supports the transmission of thousands of data packets per second per node, meeting the needs of large-scale scenarios such as industrial IoT and smart cities. Dynamically allocating bandwidth based on matching strength effectively improves the success rate of high-priority data (such as equipment fault alarms), thereby reducing the resource consumption of low-priority data (such as ambient temperature and humidity). By selecting low-power connection nodes, the overall power consumption of the edge node RF module is reduced, extending the device's battery life. It supports multi-path redundant transmission, automatically switching to a backup path when the matching strength of the primary path decreases. The introduction of end-to-end verification and retransmission mechanisms effectively reduces the data transmission error rate and avoids service interruptions due to packet loss. The module is adaptable to different protocols (such as LoRaWAN and ZigBee) and device types; no hardware modification is required when adding edge nodes, supporting service deployment within minutes. Through standardized interfaces and network layer interaction, the dependence of cloud platform upgrades on edge nodes is reduced, improving the overall flexibility of the system. It supports remote configuration of transmission policies via web interface or API, effectively improving operation and maintenance efficiency. It automatically diagnoses link faults and triggers repair processes, reducing MTTR (Mean Time To Repair) to less than 3 minutes and minimizing the need for manual intervention. The first data transmission control module of this invention, through multi-dimensional state perception and dynamic matching algorithms, solves the problems of fixed transmission paths, inefficient resource allocation, and poor reliability in traditional solutions, providing efficient, reliable, and scalable data transmission control capabilities for IoT cloud platforms, with significant economic benefits and social value.

[0030] In one embodiment of the present invention, determining the operational matching strength state between the target edge node and the network layer at the current moment based on the operational state parameters of the current target edge node and the operational state parameters of the current network layer includes: Retrieve the current target edge node's operating status parameters; wherein, the target edge node's operating status parameters include the remaining service resource redundancy and the cache queue congestion coefficient; specifically, the remaining service resource redundancy = the proportion of the edge node's current idle computing resources (CPU + memory) to the total resources × the data processing concurrency capability redundancy coefficient, and the data processing concurrency capability redundancy coefficient = idle concurrency / maximum concurrency; the cache queue congestion coefficient = edge node's pending data cache queue length ÷ cache queue maximum threshold; Retrieve the current network layer's operational status parameters; these parameters include link protocol compatibility and dynamic load margin. Specifically, link protocol compatibility = the matching success rate between the network layer's supported transmission protocols and the edge node's data protocols × protocol conversion delay optimization coefficient, and protocol conversion delay optimization coefficient = standard delay / actual conversion delay; dynamic load margin = the proportion of the network layer's currently available bandwidth to the total bandwidth / the number of currently connected nodes in the network layer. Furthermore, dynamic load margin reflects the network layer's redundancy for carrying newly added edge nodes, distinct from the traditional "bandwidth utilization rate." The state matching parameter M = (R × S) - (α × Q / L) is obtained based on the operating state parameters of the current target edge node and the operating state parameters of the current network layer; where R represents the remaining service resource redundancy of the current target edge node; S represents the link protocol adaptability of the current network layer; α represents the adjustment coefficient, with a value of 0 < α < 0.5; Q represents the buffer queue congestion coefficient of the current target edge node; and L represents the dynamic load margin of the current network layer. The state matching parameters are compared with a preset state matching parameter threshold. When the state matching parameter exceeds the preset state matching parameter threshold, it is determined that the operational matching strength between the target edge node and the network layer is good at the current moment. If the state matching parameter does not exceed the preset state matching parameter threshold, it is determined that the running matching strength between the target edge node and the network layer is not good at the current time.

[0031] The working principle of the above technical solution is as follows: The above technical solution takes dynamic adaptability as the core requirement, and retrieves two types of key operating parameters of the target edge node (remaining service resource redundancy and cache queue congestion coefficient) and two types of key operating parameters of the network layer (link protocol adaptability and dynamic load margin). Each parameter is calculated through specific quantitative formulas (such as the remaining service resource redundancy combined with the idle ratio of computing resources and the concurrency redundancy, and the dynamic load margin combined with the available bandwidth ratio and the number of connected nodes), to ensure that the parameters can accurately reflect the real-time operating status of the node and the network layer. Real-time comparison and judgment: The calculated matching parameters are compared with the threshold. If the matching parameter is greater than or equal to the threshold, it is judged as "good matching strength"; If the matching parameter is less than the threshold, it is determined as "poor matching strength".

[0032] Meanwhile, the aforementioned state matching parameters enhance the synergistic effect between the remaining service capacity of edge nodes and the adaptability of network layer protocols through the positive correlation product term (R×S), quantify the constraint relationship between edge node buffer congestion and network layer load margin through the negative correlation division term (α×Q / L), and balance the influence weight of negative factors through the adjustment coefficient α (0<α<0.5) to avoid a single factor from overdoing the matching result. Finally, the output is the state matching parameter M, which comprehensively reflects the degree of adaptability between the two. The technical effects of the above solution are as follows: It abandons commonly used parameters such as static bandwidth and fixed latency in existing technologies, and instead selects parameters focusing on resource redundancy, dynamic bearer capacity, and protocol adaptation corresponding to operational status. This accurately captures the dynamic correlation between the service potential of edge nodes and the elasticity of network layer capacity, solving the problem that traditional parameters are difficult to reflect real-time adaptation requirements. Furthermore, the formula is constructed using only basic arithmetic operations, resulting in a simple and easy-to-implement structure. By combining collaborative enhancement with constraint balancing, it highlights the core role of node service capabilities and network layer adaptability while effectively suppressing the impact of negative factors such as cache congestion and insufficient load, thus significantly improving the accuracy of matching goodness judgment. On the other hand, parameter calculation is centered on actual operational adaptation requirements. The state matching parameter M truly reflects the collaborative operational potential between the node and the network layer. Combined with threshold judgment rules, it can quickly and accurately output matching strength results, thereby effectively improving matching judgment efficiency and providing a reliable basis for subsequent selection of target network connection nodes.

[0033] One embodiment of the present invention involves cluster-filtering target first network connection nodes from the first network connection nodes based on the matching strength state between the target edge node and the network layer at the current moment, including: Given the current state of good operational matching strength between the target edge node and the network layer, a first-node filtering strategy is adopted to obtain the target's first network connection node. In response to the current state of poor operational matching strength between the target edge node and the network layer, a second node filtering strategy is adopted to obtain the target first network connection node.

[0034] Specifically, the first node's filtering strategy is as follows: Prioritize selecting the first network connection node that is "source-compatible" with the target edge node's data protocol (requiring no protocol conversion or only minimal conversion) to match the core advantage of link protocol compatibility S; If there are multiple compatible nodes from the same source, they are sorted in ascending order by the network layer link load fluctuation coefficient. The network layer link load fluctuation coefficient = maximum load value - minimum load value within 1 minute, which is used to reflect load stability and adapt to the redundancy characteristics of dynamic load margin L. Select any one of the top 3 nodes in the ranking as the target node to avoid excessive filtering complexity and balance efficiency.

[0035] Meanwhile, the second node's filtering strategy is as follows: Prioritize selecting the first network connection node with "cache unloading function" to directly take over the backlog of cached data from edge nodes, thus specifically alleviating Q's shortcomings; If there are multiple nodes with cache offloading function, they are sorted in descending order of protocol conversion adaptation redundancy, where protocol conversion adaptation redundancy = number of protocol types supported by the node - number of protocol types currently carried, which is used to make up for the problem of insufficient protocol adaptation S of the matching link. If there are no nodes with cache offloading capabilities, directly select the node with the highest dynamic load margin L at the network layer to quickly increase load redundancy.

[0036] The data transmission control method corresponding to the second data transmission control module includes: Based on the current operating status of the network layer and the platform layer, select the target second network connection node from the second network connection node cluster; The target second network connection node is used to send the data information received at the network layer to the platform layer.

[0037] Simultaneously, based on the current operating status of the network layer and the platform layer, target second network connection nodes are selected from the second network connection node cluster, including: Retrieve the current platform layer's runtime status parameters; Determine the operational matching strength between the current platform layer and the network layer based on the current operational status parameters of the platform layer and the network layer. Based on the current matching strength between the platform layer and the network layer, target second network connection nodes are clustered and filtered from the second network connection nodes.

[0038] Specifically, based on the current matching strength status between the platform layer and the network layer, target second network connection nodes are clustered and filtered from the second network connection nodes, including: Given the current good operational matching strength between the platform layer and the network layer, a third-node filtering strategy is adopted to obtain the target second network connection node. Given the current state of poor operational matching between the platform layer and the network layer, a fourth-node filtering strategy is adopted to obtain the target second network connection node.

[0039] Specifically, the third-node filtering strategy is as follows: Prioritize filtering data aggregation nodes with the same source that are preset in the platform layer. These nodes can upload data from edge nodes of the same type to the platform layer in batches, reducing the processing pressure on the platform layer and echoing the adaptation advantages between the network layer and the platform layer. If there are multiple data aggregation nodes from the same source, sort them in descending order of platform layer interface response stability. The stability of platform layer interface response is calculated as the number of successful responses in the last 5 minutes / the total number of requests, to avoid high-frequency fluctuations affecting transmission. Select the first node in the sorted list as the target node (balancing aggregation efficiency and stability).

[0040] The fourth node filtering strategy is as follows: Prioritize selecting second network connection nodes with protocol adaptive conversion plugins. These plugins can dynamically adapt to the protocol differences between the network layer and the platform layer, making up for the adaptation shortcomings. If there are multiple nodes with adaptive plugins, sort them in descending order of the availability of the backup links from the nodes to the platform layer. The availability is calculated as backup link bandwidth / main link bandwidth. This is to address the main link congestion caused by high network layer load. If there are no nodes with adaptive plugins, directly filter the nodes with the most "currently unoccupied" interfaces in the platform layer to quickly release the platform layer's capacity.

[0041] The working principle of the above technical solution is as follows: First, in the first data transmission control phase, from the edge node to the network layer: First, retrieve the core operating parameters (remaining service resource redundancy R, cache queue congestion coefficient Q) of the target edge node and the key operating parameters of the network layer (link protocol adaptability S, dynamic load margin L) to ensure that the parameters can accurately reflect the dynamic operating status of both. The state matching parameters are calculated by M=(R×S)-(α×Q÷L) and compared with the preset threshold to quantitatively determine the strength of the running matching. Then, based on the judgment results, differentiated screening is performed: when the match is good, the first node screening strategy is adopted (prioritize same-source compatibility → sort by load fluctuation coefficient → select the top 3 nodes) to strengthen the synergistic advantages of protocol adaptation and load stability; when the match is not good, the second node screening strategy is adopted (prioritize cache unloading function → sort by protocol conversion adaptation redundancy → select the node with the highest L if none) to specifically make up for the shortcomings of cache congestion and insufficient protocol adaptation. By utilizing the selected target first network connection nodes, the data transmission from the edge nodes to the network layer is completed.

[0042] The second data transmission control phase, from the network layer to the platform layer: Retrieve real-time operating status parameters of the network layer and platform layer, and determine the operational matching strength (good / bad) of the "platform layer-network layer" through parameter correlation analysis, referring to the judgment logic of edge node-network layer. Differential filtering is performed based on the judgment results: when the match is good, the third-node filtering strategy is adopted (prioritizing nodes of the same source data aggregation → sorting by interface response stability → selecting the first node), which reduces the data processing pressure of the platform layer by relying on the adaptation advantage; when the match is not good, the fourth-node filtering strategy is adopted (prioritizing protocol adaptive conversion plugins → sorting by backup link availability → selecting the node with the most unoccupied interfaces if none are found), which dynamically solves problems such as protocol differences and link congestion. By utilizing the selected target second network connection node, the transmission of network layer data to the platform layer is completed.

[0043] The above technical solution links two transmission stages, with the transmission status of the previous stage providing the basis for the next stage. The selection strategies are strongly correlated with the operating parameter characteristics and matching strength shortcomings of the corresponding level, ensuring that the data transmission of the entire link is adapted to the real-time operating status.

[0044] The effects of the above technical solution are as follows: By adopting the aforementioned dynamic adaptability parameters, the problem that traditional static parameters (such as fixed bandwidth and absolute latency) cannot reflect real-time status is avoided; at the same time, by using four strategies that all follow the principle of strengthening advantages when well matched and compensating for shortcomings when poorly matched, the blind screening logic of existing technologies based on a single dimension (such as maximum bandwidth and minimum latency) is abandoned, thereby effectively improving the accuracy and efficiency of target node screening; furthermore, the first strategy highlights protocol adaptation and load stability, the second strategy focuses on cache offloading and protocol compatibility, the third strategy emphasizes data aggregation and interface stability, and the fourth strategy solves protocol adaptation and link redundancy, achieving accuracy in target node screening under different states, thereby improving transmission adaptability.

[0045] On the other hand, the above-mentioned technical solutions can effectively improve both the efficiency and reliability of the entire link transmission. Specifically, designs such as source compatibility and adaptive protocol conversion reduce the overhead of cross-level protocol conversion, cache offloading and backup link utilization alleviate data congestion, and data aggregation reduces the processing pressure on the platform layer. Screening dimensions such as load fluctuation coefficient and interface response stability reduce the risk of transmission fluctuation, and designs such as unoccupied interface optimization and dynamic load margin adaptation improve link carrying redundancy, thereby reducing the failure rate of the entire link transmission and significantly improving data transmission efficiency.

[0046] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. An IoT smart gateway suitable for IoT cloud platform systems, characterized in that, The IoT smart gateway suitable for IoT cloud platform systems includes: The first network connection node cluster is used for data transmission between the edge node network built into the IoT cloud platform and the network layer of the IoT cloud platform. The second network connection node cluster is used for data transmission between the network layer and the platform layer of the IoT cloud platform; A first data transmission control module is used to control data transmission between the edge node network and the network layer of the IoT cloud platform by dynamically filtering the first network connection node cluster. The dynamic filtering control includes retrieving target edge nodes processing data information of the perception layer in the edge node network at the current moment; determining the operational matching strength state between the target edge node and the network layer at the current moment from the first network connection node cluster using a state matching parameter M between the operational state parameters of the target edge node and the operational state parameters of the current network layer; filtering and obtaining target first network connection nodes using the operational matching strength state; and transmitting the data information of the target edge node to the network layer using the target first network connection node. The state matching parameter M is obtained using the following formula: M=(R×S)-(α×Q / L) Where R represents the remaining service resource redundancy of the current target edge node; S represents the link protocol adaptability of the current network layer; α represents the adjustment coefficient, with a value of 0 < α < 0.5; Q represents the buffer queue congestion coefficient of the current target edge node; L represents the dynamic load margin of the current network layer; where, the remaining service resource redundancy = the proportion of the edge node's current idle computing resources to the total resources × the data processing concurrency redundancy coefficient, and the data processing concurrency redundancy coefficient = idle concurrency / maximum concurrency; the buffer queue congestion coefficient = the length of the edge node's pending data buffer queue ÷ the maximum threshold of the buffer queue; the link protocol adaptability = the matching success rate between the transmission protocol supported by the network layer and the edge node's data protocol × the protocol conversion delay optimization coefficient, and the protocol conversion delay optimization coefficient = standard delay / actual conversion delay; the dynamic load margin = the proportion of the network layer's current available bandwidth to the total bandwidth / the number of currently connected nodes in the network layer; The second data transmission control module is used to control the data transmission between the network layer and the platform layer by dynamically filtering the second network connection node cluster.

2. The IoT smart gateway for IoT cloud platform systems according to claim 1, characterized in that, The IoT cloud platform includes a perception layer, an edge node network, a network layer, a platform layer, and an application layer.

3. The IoT smart gateway suitable for IoT cloud platform systems according to claim 1 or 2, characterized in that, The connection relationship between the IoT smart gateway and the IoT cloud platform is as follows: The perception layer establishes a data connection with the edge node network; the edge node network establishes a data connection with the network layer through a first network connection node cluster; and the network layer establishes a data connection with the platform layer through a second network connection node cluster.

4. The IoT smart gateway for IoT cloud platform systems according to claim 1, characterized in that, Determine the current state of the operational matching strength between the target edge node and the network layer, including: The state matching parameters are compared with a preset state matching parameter threshold. When the state matching parameter exceeds the preset state matching parameter threshold, it is determined that the operational matching strength between the target edge node and the network layer is good at the current moment. If the state matching parameter does not exceed the preset state matching parameter threshold, it is determined that the running matching strength between the target edge node and the network layer is not good at the current time.

5. The IoT smart gateway for an IoT cloud platform system according to claim 1, characterized in that, The target first network connection node is obtained by using the running match strength state filtering, including: Given the current state of good operational matching between the target edge node and the network layer, a first node filtering strategy is adopted to obtain the target first network connection node; wherein, the first node filtering strategy is as follows: Prioritize selecting the first network connection node that is compatible with the target edge node's data protocol to match the core advantage of link protocol compatibility S; If there are multiple compatible nodes from the same source, sort them in ascending order according to the network layer link load fluctuation coefficient, where the network layer link load fluctuation coefficient = the maximum load value within 1 minute - the minimum load value; Select any one of the top 3 nodes in the ranking as the target first network connection node; To address the current situation where the operational matching strength between the target edge node and the network layer is poor, a second node filtering strategy is employed to obtain the target's first network connection node. The second node filtering strategy is as follows: Prioritize selecting the first network connection node with cache unloading functionality; If there are multiple nodes with cache offloading function, sort them in descending order of protocol conversion adaptation redundancy, where protocol conversion adaptation redundancy = number of protocol types supported by the node - number of protocol types currently carried. If there are no nodes with cache offloading capabilities, directly select the node with the highest dynamic load margin L at the network layer to quickly increase load redundancy.

6. The IoT smart gateway for an IoT cloud platform system according to claim 1, characterized in that, The data transmission control methods corresponding to the second data transmission control module include: Based on the current operating status of the network layer and the platform layer, select the target second network connection node from the second network connection node cluster; The target second network connection node is used to send the data information received at the network layer to the platform layer.

7. The IoT smart gateway for an IoT cloud platform system according to claim 6, characterized in that, Based on the current operating status of the network layer and the platform layer, target second network connection nodes are selected from the second network connection node cluster, including: Retrieve the current platform layer's runtime status parameters; Obtain the state matching parameter M based on the current platform layer's running state parameters and the network layer's running state parameters; The operational matching strength state between the current platform layer and the network layer is determined based on the comparison result between the state matching parameter M and the preset state matching parameter threshold. Based on the current operational matching strength between the platform layer and the network layer, target second network connection nodes are selected from the second network connection node cluster.

Images