Internet of Things networking method and system based on LoRa Mesh

Abstract

The invention provides an Internet of Things networking method and system based on LoRa Mesh, and relates to the technical field of Internet of Things networking. The system is composed of a gateway node with LoRa and cellular communication capability, a relay node supporting routing forwarding and low-power-consumption standby and multiple types of terminal nodes, and comprises a radio frequency identification unit, a data acquisition unit and a sensing unit. The method comprises the following steps: establishing a data channel between a local Mesh network and a public network through a gateway; the relay node expands network coverage; the terminal node is dormant in a non-communication period, and is awakened and reported at an acquisition moment or when an event is triggered; the Mesh routing dynamically selects a path according to link quality, energy consumption and hop count, and supports self-healing reconstruction when a node fails; the gateway receives the data and then uploads the data through a cellular network, and a downlink instruction is transmitted in a terminal wake-up window hop by hop to complete interaction. According to the invention, low-power-consumption operation, wide-area coverage and high-reliability data transmission are realized, and unified access of multiple types of terminals is supported.

Description

Technical Field

[0001] This invention relates to the field of Internet of Things (IoT) networking technology, specifically to an IoT networking method and system based on LoRa Mesh. Background Technology

[0002] In low-power wide-area networks (LPWANs) for the Internet of Things (IoT), LoRa technology is widely used in environmental monitoring, industrial control, and other scenarios due to its long communication distance and low power consumption. Traditional LoRa networks using a star topology communicate directly between terminal devices and the gateway, playing a crucial role in terms of simple deployment and stable connections.

[0003] However, as the scale of applications expands, this approach faces significant limitations: on the one hand, in complex environments or wide-area deployments, a single gateway is difficult to provide comprehensive coverage, which can easily lead to communication blind spots and seriously affect data integrity; on the other hand, if coverage is extended by adding gateways or traditional relays, it will lead to a surge in energy consumption, which cannot meet the core requirements of terminal devices for long-term battery life; it is difficult to achieve wide-area coverage and high-reliability access for multiple types of terminals without sacrificing power consumption. Summary of the Invention

[0004] To achieve the above objectives, this invention provides an IoT networking method and system based on LoRa Mesh to solve the aforementioned technical problems.

[0005] In a first aspect, the present invention provides an IoT networking method based on LoRa Mesh, comprising: A data channel between a local LoRa Mesh network and the public network is established through a gateway node equipped with both LoRa and cellular communication modules. Extend network communication range and maintain coverage by using relay nodes with routing forwarding and low-power standby capabilities; Multiple types of terminal nodes, including radio frequency identification units, data acquisition units and sensing units, are connected to the LoRa Mesh network. The terminal nodes enter sleep mode during non-communication cycles and are woken up to report data when the acquisition time or a threshold event is triggered. The Mesh routing module dynamically selects forwarding paths based on link quality, energy consumption status, and hop count, and automatically reconstructs them when a node fails. After receiving data, the gateway node uploads it to the remote platform via the cellular network. When receiving control commands, it completes the interaction within the wake-up window of the terminal node through hop-by-hop transmission, thereby achieving low-power operation, wide-area coverage and highly reliable data transmission under the condition of unified access of multiple types of terminals.

[0006] As a further technical solution, a dynamic neighbor table maintenance mechanism is used during the operation of relay nodes. Specifically, the relay node actively collects the signal strength, packet loss rate, and energy consumption status of neighboring nodes within a preset detection period and stores this data in its local neighbor table. The neighbor table not only serves as an input parameter for path selection but also triggers fast rerouting. When the continuous packet loss rate of a link exceeds a threshold or the signal strength falls below a set threshold, the relay node immediately executes the fast rerouting mechanism. This mechanism first generates a priority list containing candidate forwarding nodes, then sorts them based on their energy consumption margin and historical transmission success rate, and selects the optimal path to update the local routing table. This update process uses differential broadcasting to reduce network load, sending update information only to affected nodes to avoid excessive overhead from network-wide broadcasting. Through this design, relay nodes can quickly achieve self-healing when local links fail, maintaining network stability and continuous communication.

[0007] As a further technical solution, dynamic duty cycle adjustment logic is introduced during the sleep and wake-up process of terminal nodes. The gateway node generates a time distribution curve based on the historical data collection frequency and the probability of sudden event triggering, and sends it to the terminal nodes. During periods of high event probability, the terminal nodes actively shorten the sleep cycle and increase the listening window to ensure that critical events can be reported in a timely manner. During periods of low event probability, the sleep cycle is extended to reduce energy waste caused by ineffective listening. At the same time, the terminal nodes make secondary adjustments based on their remaining power: if the power is insufficient, even during periods of high event probability, the sleep cycle will be extended and the collection frequency will be reduced to prioritize the continuous operation of the device. This adaptive duty cycle logic enables the terminal nodes to extend battery life while ensuring the real-time performance of critical events, and solves the contradiction between excessive energy consumption and insufficient real-time performance caused by fixed-cycle wake-up.

[0008] As a further technical solution, the Mesh routing module introduces cross-layer joint decision logic during path selection. While collecting traditional parameters such as link quality, energy consumption status, and hop count, the routing module additionally obtains information on the physical layer received signal strength and the MAC layer queue length. The system constructs a comprehensive cost function, sums the parameters from different layers according to preset weights, and performs local simulation calculations to predict the transmission delay and energy consumption of candidate paths. When a certain indicator shows abnormal fluctuations, other layer parameters can be used as compensation, thereby avoiding frequent route switching due to a single parameter anomaly and ensuring the stability and long-term effectiveness of path selection. The final generated routing table must be verified to be stable through simulation before it can be updated to prevent frequent updates from causing system jitter.

[0009] As a further technical solution, the comprehensive cost function is as follows: Wherein, C(P) is the comprehensive cost of path P, the smaller the value, the better the path; this value serves as the direct basis for path selection; LQ(P) is the link quality, calculated by periodically measuring packet loss rate and Received Signal Strength Indication (RSSI) at relay nodes, its value is (0,1], the higher the value, the more reliable the link; E(P) is the energy consumption overhead, the average power consumption of all nodes on the path in the current forwarding task, derived from node energy monitoring, its value range is [0,1], 1 represents high energy consumption, 0 represents low energy consumption; H(P) is the hop count, i.e. the number of relay nodes included in the path, this value is obtained directly from the routing table, and its value is a positive integer; RSSI(P) is the Received Signal Strength Indication, measured in real time by the physical layer, its absolute value is... The smaller the value, the stronger the signal, used to severely penalize congested paths; QL(P) is the queue length, detected by the MAC layer as the number of queued data packets in the send buffer, with a value range of non-negative integers, reflecting the degree of path congestion, and a larger value indicates more severe delay; ΔLQ(P) is the link quality change rate, calculated by the routing module based on historical link quality records, reflecting link stability, with positive values ​​indicating a declining trend and negative values ​​indicating an increasing trend; α, β, γ, δ, θ are weighting parameters used to balance the importance of different indicators. Each parameter is adaptively adjusted by the gateway node according to the application scenario and real-time status. For example, when the overall network energy consumption is high, the gateway automatically increases the value of the energy consumption weight β, and increases γ or δ in emergency monitoring tasks.

[0010] As a further technical solution, a hierarchical caching and confirmation feedback system is introduced in the gateway node. When the cellular link is interrupted or the public network bandwidth decreases, the gateway node divides the data into critical data and ordinary data according to their importance and stores them in different cache areas. Critical data includes emergency alarms, identity authentication information, etc., while ordinary data includes periodically collected data. When the link is restored, the gateway node prioritizes uploading critical data and waits for confirmation from the remote platform at each level. Only after successful confirmation will it upload ordinary data. If the cache space is close to the limit, the gateway node will issue a control command to the terminal node to dynamically adjust its collection cycle to reduce the data generation speed and prevent cache overflow.

[0011] As a further technical solution, a hop-by-hop transmission window synchronization calibration mechanism is designed during the transmission of downlink control commands. After receiving the downlink signal from the upstream node, each relay node opens its receiving window sequentially according to the delay sequence preset by the gateway and listens for downlink commands within that window. When setting the delay sequence, the gateway node considers the clock deviation of neighboring nodes and the link propagation delay, ensuring that the receiving windows of each node in the hop-by-hop transmission do not overlap or have gaps. If a node fails to receive the command in time due to deviation, the system can detect the anomaly through the confirmation feedback from the next hop node and immediately trigger a retransmission. This mechanism ensures that the downlink command can accurately reach the wake-up window of the terminal node, improving the reliability and real-time performance of remote control.

[0012] As a further technical solution, differentiated access and routing strategies are adopted for different types of terminal nodes. When the radio frequency identification unit enters the network, it prioritizes the allocation of low-latency channels and establishes direct or short-hop paths to ensure immediate response to identification requests. The periodic data acquisition unit selects the energy-efficient path based on the remaining power and delays reporting when the path is congested to extend the node's lifespan. When performing critical monitoring tasks, the sensing unit triggers multi-path parallel forwarding and generates redundant copies to improve data delivery rate. All differentiated strategies are executed collaboratively by relay nodes and gateway nodes, and parameters are corrected in real time on the remote platform to ensure that all types of terminal nodes can obtain the most suitable quality of service under the same network architecture, thereby meeting the diversified and differentiated needs of services under a unified network architecture.

[0013] Secondly, the present invention provides an IoT networking system based on LoRa Mesh, comprising: At least one gateway node with LoRa communication and cellular communication modules is used to establish a data channel between the local LoRaMesh network and the public network and perform remote management; multiple relay nodes with routing forwarding and low-power standby functions are used to extend the communication range and enhance coverage; multiple types of terminal nodes, including radio frequency identification units, data acquisition units and sensing units, are provided, and the terminal nodes have low-power sleep and threshold event wake-up mechanisms; the system's built-in Mesh routing module is used to perform dynamic path selection based on link quality, energy consumption status and hop count, and to achieve self-healing reconstruction when a node fails; the gateway node is configured with hierarchical caching and hop-by-hop delivery mechanisms to achieve low-power operation, wide-area coverage and highly reliable data transmission under the condition of unified access for multiple types of terminals.

[0014] Further, it includes a hardware and software co-optimization module. The relay node is configured with a neighbor table maintenance module and a fast rerouting unit at the hardware level to periodically detect the quality of neighbor links and generate candidate path sequences locally when anomalies occur. The gateway node has a built-in hierarchical cache controller and priority scheduling unit. When the public network is interrupted, it stores key data locally and uploads it first after recovery. At the same time, it sends periodic adjustment commands to the terminal nodes to control the data generation rate. The terminal nodes embed duty cycle adjustment logic and receive window timing circuit, so that they can open the window synchronously according to the hop-by-hop timing when receiving downlink control.

[0015] This invention provides an IoT networking method and system based on LoRa Mesh, which has the following beneficial effects: This invention addresses the problems of LoRa networking relying on the central node to rebuild paths when links fail, leading to communication interruptions and excessive network-wide broadcast overhead, by introducing a dynamic neighbor table maintenance and fast rerouting mechanism in the relay node. Through this mechanism, the relay node can quickly generate candidate paths and locally update routes using differential broadcast when a link anomaly is detected, thereby significantly reducing the network-wide load and achieving decentralized dynamic repair.

[0016] This invention introduces adaptive duty cycle adjustment logic into the terminal node, combining the event probability distribution issued by the gateway node with the remaining power status of the terminal node, to achieve dynamic optimization of the sleep and wake-up cycle; it solves the contradiction between energy waste and critical event reporting delay under the fixed cycle mechanism, enabling the terminal node to maintain high real-time performance during high event probability periods, reduce invalid listening to save energy during low event probability periods, and further extend the sleep cycle to ensure operational life when the power is insufficient.

[0017] This invention introduces cross-layer joint decision logic into Mesh routing, comprehensively considering multi-dimensional parameters such as link quality, energy consumption status, hop count, physical layer received signal strength, and MAC layer queue length, constructs a comprehensive cost function, and combines it with local simulation calculation to achieve path selection; it solves the problem that relying on a single indicator for routing decision-making can easily lead to frequent path switching and instability, thereby significantly improving the robustness and long-term effectiveness of path selection. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the process of the present invention; Figure 2 This is a schematic diagram of the fast rerouting process for relay nodes in this invention; Figure 3 This is a schematic diagram of the dynamic duty cycle adjustment process of the terminal node in this invention. Detailed Implementation

[0019] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] refer to Figures 1-3 Example 1: This example provides an IoT networking system based on LoRa Mesh. The system consists of a 4G gateway device, a relay device, an RFID device, a DTU device, and a gas sensor device. These devices form a self-organizing mesh topology within a local area through their built-in LoRa communication modules and Mesh routing protocol. A low-power operation mechanism enables unified access and reliable transmission for multiple types of terminals across different scenarios. This example comprehensively discloses the system composition, terminal access, routing selection, low-power mechanism, cache control, and downlink synchronization process in accordance with the limitations of the claims.

[0021] During system deployment, a 4G gateway equipped with both LoRa and cellular communication modules serves as the core node, responsible for establishing a local LoRa network. The Mesh network provides a data channel to the public network, enabling remote uploading and centralized management of local data. The gateway's cellular module supports 4G communication and has a reserved interface for internet access, allowing reliable transmission of collected data to cloud platforms or remote servers. Simultaneously, the gateway features a tiered caching and priority scheduling unit. When the public network link is interrupted, it can categorize and store collected data according to its criticality to prevent data loss. For example, in a test scenario, the gateway's cache capacity is configured to 256MB, with 64MB allocated to the critical data area for storing alarm and authentication information, and 192MB allocated to the general data area for storing periodic environmental monitoring data. When the public network link is interrupted, the gateway first stores gas concentration data exceeding the threshold detected by the gas sensor in the critical cache area and immediately uploads it to the remote platform after the link is restored, ensuring the real-time nature of important information. If the cache space utilization exceeds 80%, the gateway will issue control commands to adjust the terminal's data acquisition cycle, such as extending the gas sensor's sampling cycle from 10 seconds to 30 seconds, to reduce new data generation and avoid cache overflow.

[0022] As intermediate nodes, relay devices possess routing and forwarding capabilities as well as low-power standby functions, extending communication range and enhancing network coverage. During operation, relay nodes execute a dynamic neighbor table maintenance mechanism, actively collecting signal strength, packet loss rate, and energy consumption status of neighboring nodes within a preset detection period and recording them in the local neighbor table. For example, a relay node measures the received signal strength of surrounding nodes one by one with a detection period of 60 seconds. When the signal strength is below -120dBm or the continuous packet loss rate exceeds 15%, the link quality is deemed unreliable, triggering a fast rerouting mechanism. Under this mechanism, the relay node generates a priority list containing candidate forwarding nodes, sorted according to energy reserves and historical transmission success rates. For example, if a candidate node has 80% remaining power and a historical transmission success rate exceeding 95%, this node has a higher priority and is more likely to be selected as a new forwarding path. The rerouting results are only notified to affected neighboring nodes through differential broadcasting, thereby avoiding network-wide broadcasting, reducing bandwidth and energy consumption, and ensuring the high efficiency of local self-healing.

[0023] Terminal equipment includes RFID devices, DTU devices, and gas sensor devices. RFID devices are mainly used for tag information collection, supporting low-power wake-up and intermittent communication. They prioritize low-latency channels during network access and establish direct or single-hop paths where possible to ensure immediate identification responses. DTU devices are used to collect serial port data or 485 bus data from industrial sites and transmit it via LoRa. The mesh network forwards data to the gateway, employing an energy-efficient data caching and batch upload mechanism. For example, data collected by the DTU in the field is cached every five minutes. When the accumulated data reaches ten records, it is uploaded in batches to the relay node, which then forwards it to the gateway via multiple hops to reduce energy consumption from multiple short messages. Gas sensor devices monitor ambient gas concentrations and feature low-power sleep and timed wake-up mechanisms. They can proactively report data when the gas concentration exceeds a threshold. For example, when the carbon dioxide concentration exceeds 1000 ppm, the sensor immediately wakes up and generates an alarm message, which is then uploaded to the gateway first. In non-event-triggered scenarios, the sensor samples and reports according to a preset cycle, with a default cycle of 10 seconds. When the battery level is below 20%, the sampling cycle is automatically extended to 60 seconds to extend battery life.

[0024] The sleep and wake-up process of terminal nodes incorporates dynamic duty cycle adjustment logic. The gateway node generates a time distribution curve based on the historical data collection frequency and the probability of sudden events, and sends it to the terminal nodes. For example, in environmental monitoring tasks, the probability of a gas leak event occurring at night is only 2%, while the probability during daytime working hours is as high as 15%. Based on this distribution curve, the terminal node actively shortens the sleep cycle to 5 seconds and increases the listening time during the daytime to ensure that critical events can be reported in a timely manner, while extending the sleep cycle to 60 seconds during the low-probability period at night to reduce invalid listening. At the same time, the terminal node also makes a secondary correction based on its own battery level. If the remaining battery level is less than 20%, even during high-probability periods, the sleep cycle will be extended to 30 seconds and the sampling frequency will be reduced to ensure the device's endurance in long-term operation scenarios.

[0025] The Mesh routing module introduces cross-layer joint decision logic during path selection, comprehensively considering multi-dimensional parameters such as link quality, energy consumption status, hop count, physical layer received signal strength, and MAC layer queue length; the system constructs a comprehensive cost function. Where LQ(P) is the link quality, ranging from 0 to 1, with higher values ​​indicating higher reliability; E(P) is the power consumption, ranging from 0 to 1, with 1 indicating high power consumption and 0 indicating low power consumption; H(P) is the hop count, a positive integer; RSSI(P) is the received signal strength indicator, expressed in dBm, with smaller absolute values ​​indicating stronger signals; QL(P) is the queue length, a non-negative integer, representing the number of data packets buffered at the MAC layer. For example, in three candidate paths, the first path has LQ of 0.8, E of 0.2, H of 2, and RSSI of... The first path has a cost of -110dBm and a QL of 5; the second path has a cost of 0.7, an efficiency of 0.1, an efficiency of 1, an RSSI of -120dBm and a QL of 2; the third path has a cost of 0.9, an efficiency of 0.3, an efficiency of 3, an RSSI of -150dBm and a QL of 8. After calculating the comprehensive cost function, the first path has the lowest cost value, and the system selects it as the optimal path. To avoid frequent route switching caused by an abnormality in a single indicator, other indicators can provide compensation when a certain parameter fluctuates in the short term, thereby ensuring the stability and long-term effectiveness of path selection.

[0026] During downlink control command transmission, the system employs a hop-by-hop in-window synchronization calibration mechanism. After receiving the downlink signal from the upstream node, each relay node opens its receiving window sequentially according to the delay sequence issued by the gateway and listens for downlink commands within the window. For example, in a four-hop link consisting of a gateway, two relay nodes, and one terminal node, the gateway first sends the downlink command and sets Relay 1 to open its receiving window after 30 milliseconds, Relay 2 after 60 milliseconds, and the terminal node after 90 milliseconds. By dynamically adjusting the delay sequence, the system ensures that the receiving windows of each node do not overlap and have no gaps. If a node fails to receive the command due to clock deviation, the confirmation feedback from the next node will immediately report the anomaly, triggering a retransmission mechanism to ensure that the control command is reliably delivered to the terminal.

[0027] This system achieves a unified network architecture and differentiated quality of service in heterogeneous device access scenarios. RFID devices prioritize low-latency paths, DTU devices adopt the most energy-efficient path and extend their lifespan through batch uploading, while gas sensor devices trigger multi-path parallel forwarding to generate redundant copies at critical moments, improving data delivery rate. For example, in a gas leak test, the gas sensor device simultaneously reported alarm information through two different relay paths, ultimately achieving a 100% success rate on the cloud platform, while the success rate in single-path mode was only 95%. This enhances the reliability of multi-path redundancy in critical tasks.

[0028] In summary, this embodiment achieves the goals of low-power operation, wide-area coverage, and high-reliability data transmission by introducing mechanisms such as dynamic neighbor table maintenance, fast rerouting, adaptive duty cycle, cross-layer routing decision, hierarchical caching, and hop-by-hop synchronization into gateway nodes, relay nodes, and terminal nodes.

[0029] Example 2: This example provides an application process for an IoT networking method based on LoRa Mesh. This example describes the data transmission, cache management, and downlink interaction aspects, enabling the system to maintain low power consumption, wide coverage, and high reliability data interaction capabilities even when facing public network outages, link fluctuations, and diverse needs of multiple terminals. The system consists of gateway nodes equipped with LoRa communication modules and cellular communication modules, relay nodes supporting routing forwarding and low-power standby, and various types of terminal nodes, including RFID devices, DTU devices, and gas sensor devices. All nodes form a LoRa Mesh network with wide coverage and self-healing capabilities through a self-organizing mesh topology, and the gateway node ultimately completes the interconnection with the public network.

[0030] During data reporting, terminal nodes generate data frames through event triggering or timed sampling. RFID devices immediately send request data when reading tag information. Due to the high real-time requirements of identification services, the system allocates low-latency channels for them, prioritizing single-hop or few-hop paths to reach the gateway, thus ensuring data transmission within 100 milliseconds. DTU devices are mainly responsible for serial port data acquisition in industrial settings. Their data volume is relatively large, but the real-time requirements are moderate, so a batch caching and time-sharing upload mechanism is adopted. For example, in the test scenario, the DTU device caches ten records every five minutes and uploads them to the relay node when the local timer expires, and the relay node forwards them hop-by-hop to the gateway. Gas sensor devices use a combination of periodic acquisition and threshold triggering. By default, they detect carbon dioxide concentration every ten seconds. When the concentration exceeds the preset threshold of 1000 ppm, alarm data is immediately reported. This alarm data is copied during routing and forwarded in parallel through two paths, ensuring that alarm information still reaches the gateway 100% even when the link fluctuates.

[0031] After receiving data from the terminal, the gateway node needs to upload it to the remote platform via the cellular module. Due to potential interruptions or bandwidth degradation in the public network link, this embodiment introduces a hierarchical caching and acknowledgment feedback mechanism for the gateway node. Specifically, the gateway divides data into critical data and ordinary data according to their importance and stores them in different cache areas. In the experimental configuration, the total cache capacity of the gateway is 256MB, with 64MB for critical data and 192MB for ordinary data. When a cellular link interruption is detected, the concentration alarm information exceeding the threshold uploaded by the gas sensor is immediately stored in the critical cache area, while the DTU acquisition... Periodic data is then sent to the general buffer. After the link is restored, the gateway first schedules the data in the critical buffer for uploading, and waits for confirmation from the remote platform after each batch of data is uploaded. Only after the confirmation signal is returned can the gateway continue to upload the next batch of general data. Through step-by-step confirmation, it is ensured that critical data will not be lost due to link jitter. If the buffer space utilization exceeds 80%, the gateway will send a control command to the terminal node to dynamically adjust the sampling period to reduce the new data generation rate. For example, when the buffer utilization reaches 90%, the gateway can extend the gas sensor period from 10 seconds to 30 seconds to prevent buffer overflow before the link is restored.

[0032] During downlink control command transmission, this embodiment employs a hop-by-hop delivery and window-based synchronization calibration mechanism. When the gateway node needs to send control commands to the remote platform, it generates downlink data frames and sets a delay sequence, causing each relay node to sequentially open its receiving window and listen to upstream data. For example, in a four-hop path consisting of a gateway, two relay nodes, and one terminal node, the gateway sets relay one to open its receiving window after 30 milliseconds, relay two after 60 milliseconds, and the terminal node after 90 milliseconds when sending commands. This hierarchical delay method ensures that the receiving windows neither overlap nor create gaps, thereby avoiding the risk of commands being lost during transmission. If relay two fails to receive the command at the correct time due to clock deviation, the terminal node will report an anomaly in the next hop feedback, and the system will immediately trigger a retransmission mechanism to ensure that the command reaches the target terminal. In actual testing, even when there is a clock error of five milliseconds in the relay nodes, this mechanism can still ensure that the success rate of downlink command transmission remains above 99%.

[0033] While performing route forwarding, relay nodes also undertake neighbor table maintenance and fast rerouting tasks. In this embodiment, the relay node's neighbor detection cycle is set to 60 seconds, and the signal strength and packet loss rate of neighbor nodes are measured in each cycle. When the signal strength of a link is detected to be below -120dBm or the continuous packet loss rate exceeds 15%, the system immediately determines that the link is unavailable and initiates the fast rerouting mechanism. The rerouting mechanism generates a priority list of candidate nodes and sorts them according to their remaining power consumption and historical transmission success rate. For example, if candidate node A has 80% remaining power and a historical transmission success rate of 95%, while candidate node B has only 40% remaining power and a historical success rate of 80%, the system will prioritize node A as the new forwarding path. The routing table update process uses differential broadcasting, sending update information only to affected neighbor nodes, avoiding bandwidth waste caused by network-wide broadcasting. In an experimental network covering two square kilometers, this mechanism can complete local path repair within three seconds, ensuring the continuity of data transmission.

[0034] The terminal node's operating mechanism in this embodiment also embodies dynamic duty cycle adjustment logic. The gateway node generates a time distribution curve based on historical event statistics and sends it to the terminal to guide it in selecting an appropriate sleep cycle during different time periods. For example, in industrial field statistics, the probability of gas leak events between 8:00 AM and 6:00 PM is 15%, while it is only 2% between midnight and 6:00 AM. Therefore, the gas sensor shortens the sleep cycle to five seconds during the day to improve the real-time performance of event detection, while extending it to sixty seconds at night to reduce ineffective energy consumption. When the sensor's battery power is below 20%, even during periods of high event probability, the sleep cycle will be extended to thirty seconds and the number of samplings will be reduced to ensure that the device can continue to operate and avoid monitoring blind spots due to excessive power consumption. In a week-long operation experiment, this mechanism extended the sensor's battery life by approximately 30% while still ensuring the detection and reporting of critical events.

[0035] In summary, this embodiment, while maintaining a unified network architecture, introduces hierarchical caching and step-by-step confirmation mechanisms at the gateway, neighbor table maintenance and fast rerouting mechanisms at the relay nodes, and dynamic duty cycle adjustment mechanisms at the terminal nodes. This enables the entire system to achieve stable and reliable data interaction even under complex conditions such as public network interruptions, link fluctuations, and insufficient power.

[0036] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An IoT networking method based on LoRa Mesh, characterized in that, Includes the following steps: A data channel between a local LoRa Mesh network and the public network is established through a gateway node equipped with both LoRa and cellular communication modules. Extend network communication range and maintain coverage by using relay nodes with routing forwarding and low-power standby capabilities; The LoRaMesh network connects multiple types of terminal nodes, including radio frequency identification units, data acquisition units, and sensing units. The terminal nodes enter sleep mode during non-communication cycles and are woken up to report data when the acquisition time or a threshold event is triggered. The Mesh routing module dynamically selects forwarding paths based on link quality, power consumption status, and hop count, and automatically reconstructs them when a node fails. After receiving data, the gateway node uploads it to the remote platform via the cellular network, and when receiving control commands, it completes the interaction within the wake-up window of the terminal node by hop-by-hop transmission.

2. The IoT networking method based on LoRa Mesh according to claim 1, characterized in that: When the relay node performs forwarding, a dynamic maintenance process of the neighbor table is introduced. That is, each relay node collects the signal strength, packet loss rate and energy consumption status of the neighbor nodes within a preset detection period and records them in the neighbor table. When a relay node detects that the continuous packet loss rate of a link exceeds a threshold or the signal strength is lower than a set threshold, the relay node triggers a fast rerouting mechanism. First, it generates a priority list containing multiple candidate forwarding nodes. Then, it selects the optimal path based on the neighbor's energy reserve and historical transmission success rate and updates the local routing table. This update is only notified to the affected nodes through differential broadcast.

3. The IoT networking method based on LoRa Mesh according to claim 2, characterized in that: The sleep and wake-up process of the terminal node introduces dynamic duty cycle adjustment logic. Specifically, the gateway node generates a time distribution curve based on the historical data collection frequency and the probability of sudden events, and sends the curve to the terminal node. During periods of high event probability, the terminal node actively shortens the sleep cycle and increases the listening time to ensure that critical events are reported in real time. During periods of low event probability, the sleep cycle is extended to reduce invalid listening. At the same time, the terminal node adjusts the duty cycle a second time based on its remaining power. If the power is insufficient, the sleep cycle is automatically extended, and the collection frequency is reduced even during high-probability periods.

4. The IoT networking method based on LoRa Mesh according to claim 3, characterized in that: The Mesh routing module introduces cross-layer joint decision logic during the path selection process, that is, while collecting link quality, energy consumption status and hop count, it additionally obtains the physical layer received signal strength and MAC layer queue length. When calculating candidate paths, the routing module first constructs a comprehensive cost function, summing the parameters of multiple layers according to their weights, and then calculates the transmission delay and energy consumption of different paths through local simulation. When a single indicator fluctuates, other level indicators can provide compensation to avoid frequent route switching due to abnormalities in a single indicator.

5. The IoT networking method based on LoRa Mesh according to claim 4, characterized in that: The gateway node has a hierarchical caching and confirmation feedback mechanism during the data upload process: when the cellular link is interrupted or the public network bandwidth decreases, the gateway node divides the data into critical data and ordinary data according to the importance of the data and stores them in different cache areas; when the link is restored, the gateway first uploads the critical data and waits for the remote platform to confirm it step by step, and then uploads the ordinary data step by step after receiving the confirmation. If the cache space is close to the limit, the gateway will dynamically adjust the terminal collection cycle through downlink control commands to reduce the generation of new data and prevent cache overflow.

6. The IoT networking method based on LoRa Mesh according to claim 5, characterized in that: The hop-by-hop transmission of downlink control commands employs an in-window synchronous calibration process: after receiving the downlink signal from the upstream node, each relay node sequentially opens its receiving window according to a preset delay sequence and listens for downlink commands within the window; this delay sequence is dynamically adjusted by the gateway based on neighbor clock deviation and link propagation delay, thereby ensuring that the receiving windows of each node do not overlap and have no gaps during hop-by-hop transmission; if a node fails to receive the command in time due to an error, the system will detect the anomaly in the confirmation feedback from the next hop node and immediately trigger a retransmission.

7. The IoT networking method based on LoRa Mesh according to claim 6, characterized in that: The method employs differentiated access and routing strategies for different types of terminals: the RFID unit prioritizes allocating low-latency channels and establishing direct or short-hop paths upon network access to ensure immediate response for identity recognition; the periodic acquisition unit selects the energy-efficient path based on remaining battery power and delays reporting when paths are congested to extend lifespan; the sensing unit triggers multi-path parallel forwarding during critical monitoring tasks, generating redundant copies to improve data delivery rate; various strategies are executed collaboratively by relays and gateways, and parameters are corrected in real time on a remote platform to ensure that different terminals obtain the most suitable quality of service under the same network architecture.

8. A system for an IoT networking method based on LoRa Mesh according to any one of claims 1-7, characterized in that, include: At least one gateway node with LoRa communication and cellular communication modules is used to establish a data channel between the local LoRa Mesh network and the public network and perform remote management; multiple relay nodes with routing forwarding and low-power standby functions are used to extend the communication range and enhance coverage; multiple types of terminal nodes, including radio frequency identification units, data acquisition units and sensing units, are provided, and the terminal nodes have low-power sleep and threshold event wake-up mechanisms; the system's built-in Mesh routing module is used to perform dynamic path selection based on link quality, energy consumption status and hop count, and to achieve self-healing reconstruction when a node fails; the gateway node is configured with hierarchical caching and hop-by-hop delivery mechanisms to achieve low-power operation, wide-area coverage and highly reliable data transmission under the condition of unified access for multiple types of terminals.

9. An IoT networking system based on LoRa Mesh according to claim 8, characterized in that, The relay node is configured with a neighbor table maintenance module and a fast rerouting unit at the hardware level to periodically detect the quality of neighbor links and generate candidate path sequences locally when anomalies occur; the gateway node has a built-in hierarchical cache controller and a priority scheduling unit, which stores key data locally when the public network is interrupted and uploads it first after recovery, while sending periodic adjustment commands to the terminal to control the data generation rate. The terminal node embeds duty cycle adjustment logic and a receiving window timing circuit, so that it opens the window synchronously according to the hop-by-hop timing sequence when receiving downlink control.

Images