Deterministic data polling method and system based on LoRa-MESH ad hoc network

Abstract

The invention relates to the field of wireless ad hoc network communication, and provides a deterministic data polling method and system based on a LoRa-MESH ad hoc network, and the method comprises the steps: analyzing a radio frequency carrier signal to generate a whole network physical feature set, resolving a node impedance mapping table, and constructing a global weighted topological matrix; an atomic transmission time slot reference is determined according to a service data frame structure and radio frequency communication parameters, a node time slot mapping set is generated in combination with a global weighted topological matrix, and whole-network synchronization is achieved through channel broadcasting; responding to a polling cycle to trigger and send a physical layer trigger signal to generate a local data payload, and executing differential transmission and aggregation to generate a cascade aggregation data packet; a missing node list is generated based on a whole network equipment shadow table, a complete service data set is generated by directionally additionally recording a time domain hole resource pool, and cloud service data flow is generated through heterogeneous protocol conversion and safe reporting. According to the method, global topology perception and deterministic time domain scheduling are fused, and a high-reliability data polling and integrity governance closed-loop mechanism is constructed.

Description

Technical Field

[0001] This invention relates to the field of wireless ad hoc network communication, and particularly to a deterministic data polling method and system based on LoRa-MESH ad hoc networks. Background Technology

[0002] Currently, the massive number of edge terminal accesses has created an urgent need for wide-area coverage and reduced deployment costs for wireless communication networks. How to overcome the uncertainty bottleneck of traditional communication mechanisms and construct a data polling mechanism that is low-cost, highly reliable, and time-deterministic in a constrained network environment with low bandwidth and multi-hop cascading has become a key problem that urgently needs to be solved in the current Internet of Things (IoT) field.

[0003] Chinese patent application CN111372215A discloses a single-channel synchronous information acquisition system and method based on LoRa. This system is used to synchronously acquire data from various sensors within a power system, including one LoRa gateway and at least one LoRa sensor. Both the LoRa gateway and the LoRa sensor employ single-channel node-type LoRa chips. The LoRa gateway triggers the synchronous acquisition of the LoRa sensor and receives the sampled data uploaded by each sensor. The LoRa sensor receives the synchronous acquisition command from the LoRa gateway, performs data acquisition, and sends the sampled data back to the gateway. The synchronous acquisition method includes: utilizing the pre-wake-up mechanism provided by LoRa to trigger synchronous sampling of each LoRa sensor, with each sensor sequentially uploading data according to predetermined rules.

[0004] However, current technology still faces many challenges. In large-scale multi-hop ad hoc network applications such as smart city lighting management, existing LoRa-MESH networks mostly use contention-based access mechanisms for data interaction. Faced with deep network environments with hundreds of cascaded nodes, if the control center needs to initiate high-frequency real-time status polling or synchronous switching commands to a massive number of street light controllers, high-concurrency channel contention between nodes can lead to data collisions and network congestion. In this situation, the transmission mechanism based on a random backoff strategy cannot guarantee reliable delivery of commands within a deterministic time window, resulting in an exponential deterioration in network transmission latency. If the system cannot complete the distribution of commands across the entire network within milliseconds or fails to capture fault alarms from deep nodes in a timely manner, asynchronous delays will occur when street light groups are turned on, or in emergency situations such as circuit overload, communication blockage may prevent the timely triggering of fuse protection, thereby reducing the quality of urban public lighting services. Summary of the Invention

[0005] To achieve the above objectives, this invention provides a deterministic data polling method based on LoRa-MESH self-organizing networks, the specific technical solution of which is as follows: The analog radio frequency carrier signals fed back by the nodes to be connected to the network are analyzed to aggregate and generate a set of physical features of the entire network. Based on the set of physical features of the entire network, the node impedance mapping table is calculated, and a hierarchical secondary sorting operation is performed according to the node impedance mapping table to construct a global weighted topology matrix. Based on the preset service data frame structure parameters and radio frequency communication parameters, the atomic transmission time slot reference is determined. The spatial topology-temporal resource mapping is performed in combination with the global weighted topology matrix to generate the node time slot mapping set. The nodes to be scheduled are then driven to generate the network-wide synchronization state via downlink channel broadcast. In response to the polling cycle trigger signal and the network-wide synchronization status, a physical layer trigger signal is sent. The node to be scheduled is triggered to generate a local data payload by matching the interrupt through the physical layer trigger signal or the hardware timer task scheduler. Differentiated data transmission and aggregation logic is then performed on the local data payload to generate cascaded aggregated data packets. Based on the shadow table of all network devices, reverse unpacking and state mapping logic is performed on the cascaded aggregated data packets to generate a list of missing nodes. Based on the list of missing nodes, targeted micro-supplementation logic is performed on the time-domain hole resource pool of the node time slot mapping to generate a complete business dataset for the entire network. Finally, cloud business data stream is generated through heterogeneous protocol conversion and security reporting logic.

[0006] Furthermore, the method for constructing the global weighted topology matrix includes: In response to the initial network discovery command, a network broadcast beacon is sent to the radio frequency channel, the analog radio frequency carrier signal fed back by the node to be joined is parsed, physical feature signals and power management unit status are extracted to generate a single-point physical feature set, and an aggregation operation is performed on all single-point physical feature sets to generate a full network physical feature set. Based on the physical feature set of the entire network, the topology impedance component, attenuation impedance component and interference impedance component are extracted respectively. The topology impedance component, attenuation impedance component and interference impedance component are fused to generate a single-point communication transmission impedance. The mapping relationship between the single-point communication transmission impedance and the node to be connected to the network is established to generate a node impedance mapping table. Extract the logical network depth of each node to be added to the network from the node impedance mapping table as the first sorting key value to perform first-level grouping sorting, extract the single-point communication transmission impedance of each node to be added to the network as the second sorting key value to perform second-level priority sorting, and generate a global weighted topology matrix.

[0007] Furthermore, the method for extracting the topological impedance component includes: reading the time delay factor weight in the preset scenario adaptation weighting coefficient, obtaining the logical network depth of the node to be connected to the network, calculating the product of the time delay factor weight and the logical network depth, and generating the topological impedance component. The method for extracting the attenuation impedance component includes: reading the attenuation factor weight in the scenario adaptation weighting coefficient, obtaining the absolute value of the received signal strength indication of the node to be connected to the network, calculating the first ratio of the absolute value to the preset reference signal strength reference constant, and generating the attenuation impedance component based on the product of the attenuation factor weight and the first ratio. The method for extracting the interference impedance component includes: reading the interference factor weight in the scene adaptation weighting coefficient, obtaining the signal-to-noise ratio of the node to be connected to the network, calculating a second ratio with a preset noise reference constant as the numerator and the sum of the signal-to-noise ratio and the noise reference constant as the denominator, and performing a product operation between the interference factor weight and the second ratio to generate the interference impedance component. The scenario adaptation weighting coefficient consists of a delay factor weight, an attenuation factor weight, and an interference factor weight, and the sum of the three weights is 1. The single-point physical feature set includes a received signal strength indicator, which is the absolute value of the power level of the analog radio frequency carrier signal received by the radio frequency front-end of the node to be connected to the network.

[0008] Furthermore, the method for generating the network-wide synchronization state includes: Based on the preset service data frame structure parameters and radio frequency communication parameters, the sensor acquisition delay, main control processing delay, physical layer radiation time consumption and radio frequency propagation margin of the node to be scheduled are measured respectively, and linear time domain superposition is performed to generate an atomic transmission time slot reference. Based on the global weighted topology matrix and the atomic transmission time slot reference, the spatial topology-temporal resource mapping logic is executed, and the absolute start transmission time is calculated and generated in combination with the preset hierarchical protection interval, thus generating the node time slot mapping set; Differential compression coding is performed on the node time slot mapping set to construct the timing configuration broadcast frame. The timing configuration broadcast frame is then broadcast and distributed through the downlink physical channel to trigger the nodes to be scheduled to perform non-volatile write and sleep-wake mode switching, and generate and output the network-wide synchronization status.

[0009] Furthermore, the execution method of the spatial topology-temporal resource mapping logic includes: Based on the logical network depth in the global weighted topology matrix, the entire network data polling cycle is divided into multiple non-overlapping hierarchical transmission cycles. For each hierarchical transmission cycle, the product of the extracted intra-layer impedance sorting index and the atomic transmission time slot reference is calculated to obtain the discrete time slot offset of each node to be scheduled within the hierarchical transmission cycle. The intra-layer impedance sorting index is a sequence number generated based on the magnitude of the single-point communication transmission impedance. A preset hierarchical protection interval is configured between two adjacent hierarchical transmission cycles, and a timing superposition calculation is performed to generate an absolute start transmission time. The logic of the timing superposition calculation includes: obtaining the reference clock source value of the current polling cycle, the total duration consumed by all hierarchical transmission cycles, the discrete time slot offset, and the cumulative hierarchical protection interval duration, and performing a summation operation to obtain the absolute start transmission time. The total duration consumed by all hierarchical transmission cycles is the product of the logical network depth of the node to be scheduled minus one and the preset maximum hierarchical transmission cycle. The cumulative hierarchical protection interval duration is the product of the logical network depth and the hierarchical protection interval. The absolute start time of emission is calculated and the atomic transmission time slot reference is used to define the end time of emission. The absolute start time of emission and the end time of emission are associated and stored to construct a node time slot mapping set.

[0010] Furthermore, the non-volatile writing step includes: receiving a timing configuration broadcast frame, using the network interface identifier of the node to be scheduled as the index key value, extracting the absolute start transmission time and transmission end time belonging to the network interface identifier from the timing configuration broadcast frame; and writing the absolute start transmission time and transmission end time into the non-volatile storage unit of the node to be scheduled. The steps for switching between sleep and wake-up modes include: in response to the completion of a non-volatile write operation, controlling the node to be scheduled to enter a deep sleep state with only real-time clock counting; executing logic to compare the real-time clock count value with the absolute start transmission time stored in the non-volatile memory unit; in response to the matching result that the real-time clock count value and the absolute start transmission time value are consistent, driving the node to be scheduled to switch to the data transmission state; and controlling the node to be scheduled to re-enter the deep sleep state when the real-time clock count value reaches the transmission end time.

[0011] Furthermore, the method for generating the cascaded aggregated data packets includes: In response to the polling cycle trigger signal generated by the hardware timer task scheduler when the accumulated pulse count value reaches the preset polling cycle value and the start of data polling in the network synchronization state, the absolute time base reference value and service instruction type of the entire network are read, an aggregated trigger frame is constructed using a bit-by-bit concatenation splicing algorithm, and the aggregated trigger frame is mapped to a physical layer trigger signal; In response to physical layer trigger signals or hardware timer task scheduler matching interrupts, dual-mode wake-up activation and parallel acquisition operations are started to obtain node local sensor interface signals. The node local sensor interface signals are mapped to local sensing vectors and local data payloads are generated through payload encapsulation operations. The global weighted topology matrix is ​​parsed to identify whether the node to be scheduled is a leaf node or a relay node. Differentiated data transmission and aggregation logic is performed on the local data payload to construct cascaded aggregated data packets. In response to the absolute start time of transmission, the cascaded aggregated data packets are transmitted to the next higher logical parent node.

[0012] Furthermore, the method for executing the differentiated data transmission and aggregation logic includes: For the scheduled nodes identified as leaf nodes, a standard physical layer frame header is constructed and a cyclic redundancy check code is calculated. The standard physical layer frame header, local data payload, and cyclic redundancy check code are concatenated to generate a concatenated aggregated data packet. The concatenated aggregated data packet is configured as the uplink data packet of the next higher logical parent node. For nodes identified as relay nodes to be scheduled, the receive listening window is activated to buffer uplink data packets. The uplink data packets are unpacked and stripped to extract the valid data payload. Based on the node time slot mapping set, the local data payload and the valid data payload are aggregated and spliced ​​to construct a composite payload sequence. The composite payload sequence is then encapsulated with a standard physical layer frame header and a cyclic redundancy check code to generate a concatenated aggregated data packet.

[0013] Furthermore, the method for generating the cloud service data stream includes: The node time slot mapping set is used to perform reverse unpacking operation on the concatenated aggregated data packets to extract the composite payload sequence. The composite payload sequence is processed by binary stream fixed-length slicing to separate node data blocks. The node data blocks are mapped to the network device shadow table according to the global weighted topology matrix. The network device shadow table is traversed to calculate the network data integrity index to generate a list of missing nodes. Traverse the node time slot mapping set to extract the time domain hole resource pool, map the missing node list to the time domain hole resource pool to build a temporary supplementary recording scheduling table, perform silent directional retransmission operation based on the temporary supplementary recording scheduling table to generate supplementary recording service data packets, and update the supplementary recording service data packets to the network-wide device shadow table to generate a complete network-wide service dataset. The system parses the complete business dataset across the entire network and performs heterogeneous protocol format mapping to generate application layer payloads. It then establishes a transport layer secure tunnel to encrypt and encapsulate the application layer payloads to generate secure transmission data packets. Finally, it maps the secure transmission data packets to cloud-based business data streams and pushes them to the cloud-based IoT platform.

[0014] A deterministic data polling system based on LoRa-MESH ad hoc networks is used to implement the aforementioned deterministic data polling method based on LoRa-MESH ad hoc networks. The system includes a topology management module, a resource allocation module, a data polling module, and a data delivery module. The topology management module is used to parse the analog radio frequency carrier signals fed back by the nodes to be added to the network to aggregate and generate a set of physical features of the entire network, calculate the node impedance mapping table based on the set of physical features of the entire network, and perform a hierarchical secondary sorting operation based on the node impedance mapping table to construct a global weighted topology matrix. The resource allocation module is used to determine the atomic transmission time slot reference based on the preset service data frame structure parameters and radio frequency communication parameters, perform spatial topology-temporal resource mapping in combination with the global weighted topology matrix to generate a node time slot mapping set, and drive the nodes to be scheduled to generate a network-wide synchronization state via downlink channel broadcast. The data polling module is used to respond to the polling cycle trigger signal and the physical layer trigger signal sent by the network-wide synchronization status. It triggers the node to be scheduled to generate a local data payload by matching the interrupt through the physical layer trigger signal or the hardware timer task scheduler, and performs differentiated data transmission and aggregation logic on the local data payload to generate cascaded aggregated data packets. The data delivery module generates a list of missing nodes by performing reverse unpacking and state mapping logic on cascaded aggregated data packets based on the shadow table of all network devices. Based on the list of missing nodes, it performs targeted micro-supplementation logic on the time-domain void resource pool of the node time slot mapping set to generate a complete business dataset for the entire network. Finally, it generates a cloud-bound business data stream through heterogeneous protocol conversion and security reporting logic.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention analyzes the radio frequency physical characteristics of nodes to be connected to the network to calculate the single-point communication transmission impedance, and performs a hierarchical secondary sorting of nodes in combination with the logical network depth. This transforms the disordered physical link state into a global weighted topology matrix with deterministic scheduling meaning, thus avoiding the problems of blind topology construction and nondeterministic routing decisions caused by channel quality fluctuations in large-scale deep network environments.

[0016] This invention performs spatial topology-temporal resource mapping on the global weighted topology matrix based on atomic transmission time slot reference, transforming the random access behavior of nodes into strictly orthogonal absolute transmission windows on the time axis. This eliminates channel congestion and data collisions caused by traditional contention-based communication in large-scale deep network environments, and achieves deterministic, conflict-free access to physical layer media.

[0017] This invention performs physical layer protocol header stripping and effective payload reassembly on multi-source uplink data by executing time-series splicing cascading aggregation logic at relay nodes. This transforms discrete, fragmented interactions in deep networking into single burst cascading transmissions, reducing accumulated protocol overhead in multi-hop environments and solving the problems of wasted air interface resources and accumulated transmission delays caused by traditional step-by-step forwarding modes.

[0018] This invention verifies the integrity of cascaded data by constructing a shadow table of all network devices at the edge gateway and performing targeted micro-supplementation on missing nodes using a temporal hole resource pool in deterministic time slot mapping. This achieves zero-conflict closed-loop repair of all network data without increasing additional channel bandwidth, solving the problems of high packet loss rate and low retransmission efficiency caused by unreliable multi-hop links in large-scale ad hoc networks. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating the principle of the deterministic data polling method based on LoRa-MESH self-organizing network of the present invention. Figure 2 This is a schematic diagram of the first-level grouping sorting topology structure based on logical network depth of the present invention; Figure 3 This is a schematic diagram illustrating the principle of identity recognition of nodes to be scheduled in a globally weighted topology matrix according to the present invention. Figure 4 This is a functional block diagram of the deterministic data polling system based on LoRa-MESH self-organizing network of the present invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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.

[0022] Example 1: Please see Figure 1 As shown, this embodiment provides a deterministic data polling method based on LoRa-MESH ad hoc networks, including: Step S1000: Analyze the analog radio frequency carrier signals fed back by the nodes to be added to the network to aggregate and generate a set of physical features of the entire network. Based on the entire network's physical feature set The node impedance mapping table is obtained by solving. And according to the node impedance mapping table Perform hierarchical secondary sorting operations to construct a globally weighted topology matrix. .

[0023] Specifically, this step is intended in response to the initial network discovery instruction. This approach uses discretely distributed nodes in a LoRa-MESH network, operating under dynamic electromagnetic conditions, as radio frequency sensing targets. By leveraging the multipath attenuation of radio frequency signals during spatial propagation and the differences in background noise, it maps the potential physical feature set in the current communication link that is prone to packet loss into a single-point communication transmission impedance. The numerical scalar signals in the data are used to deterministically sort nodes within the same logical level based on their impedance values, thereby generating a globally weighted topology matrix that includes the hierarchical relationships and link weights of the entire network. This provides quantitative topology data input containing link quality information for subsequent steps.

[0024] Further, step S1000 includes: Step S1100, respond to the initial network discovery command Sending network broadcast beacons to the wireless radio frequency channel The analog radio frequency carrier signal fed back by the node to be connected to the network is analyzed, and physical feature signals and power management unit status are extracted to generate a single-point physical feature set. An aggregation operation is performed on all single-point physical feature sets to generate a full network physical feature set. .

[0025] Specifically, this step is intended in response to the initial network discovery instruction. And issue network broadcast beacons This approach takes discretely distributed nodes in a time-varying electromagnetic environment within a large-scale, wide-area, multi-hop ad hoc network as radio frequency sensing objects. Utilizing the multipath effect and energy coupling characteristics of LoRa spread spectrum modulation signals during spatial propagation, it maps the potential, invisible analog radio frequency carrier signals and node micro-power states in the current wireless radio frequency channel into digital channel state information and energy weight vectors within a single-point physical feature set. Finally, it aggregates the dispersed single-point physical feature sets into a comprehensive network physical feature set. This enables the digital quantification of link physical quality and communication capabilities at the source of network topology construction, providing underlying characteristic data input for subsequent steps.

[0026] In the specific implementation process, in order to solve the problem of blind topology construction during the initialization phase of large-scale LoRa-MESH networks, this step responds to the initial network discovery command from the upper-layer business logic. The smart gateway controls its LoRa radio frequency communication unit to initiate a network broadcast beacon to the wireless radio frequency channel in the designated operating frequency band. Any node seeking to join the network located within the wireless signal coverage area of ​​the smart gateway can capture the network broadcast beacon at its radio frequency front end. Or the network broadcast beacon forwarded from the previous level relay node. At this time, instead of immediately triggering the regular handshake network access request, it forcibly enters the physical layer parameter sampling mode. This includes sending the network broadcast beacon. The smart gateways or relay nodes are collectively referred to as uplink beacon transmitters.

[0027] In the physical layer parameter sampling mode, the node to be added to the network utilizes its built-in baseband processing chip to process the network broadcast beacon. The analog radio frequency carrier signal is analyzed in the time and frequency domain to extract physical feature signals characterizing the quality of the single-hop physical link between the node to be connected to the network and the uplink beacon transmitter, and the power management unit status of the node to be connected to the network is read simultaneously. Subsequently, the node to be connected to the network performs digital mapping and vector combination of the parsed physical feature signals and the power management unit status to generate a single-point physical feature set that can characterize the current node's communication capabilities in multiple dimensions. The node to be connected to the network encapsulates the generated single-point physical feature set into a network response frame and sends it back to the smart gateway through the wireless radio frequency channel, completing the conversion of physical world signal features into digital space topology parameters.

[0028] The single-point physical feature set is a four-dimensional feature vector set containing all physical layer state information of the node to be connected to the network and its corresponding wireless communication module. It is used to quantify the communication capabilities of nodes distributed in different geographical locations at the network edge. Specifically, this single-point physical feature set is composed of logical network depth, received signal strength indication, signal-to-noise ratio, and energy weighting coefficients. The logical network depth is the value of the node to be connected when it receives the network broadcast beacon. The relative hierarchical value calculated based on the beacon frame counter, ranging from an integer of 1 to 15, is used to quantify the cumulative delay risk and reference time slot offset faced by uplink data packets during multi-hop transmission. The received signal strength indication is the absolute value of the power level measured by the radio frequency front-end of the node to be connected to the network when receiving the analog radio frequency carrier signal. Its value is typically a real number ranging from -140dBm to -30dBm, used to characterize the spatial path loss and power attenuation of the single-hop physical link in the complex electromagnetic environment of a smart city. The signal-to-noise ratio (SNR) is the ratio of the effective power of the analog radio frequency carrier signal to the current channel background noise power, and its value ranges from -20dB to 15dB. It is used to characterize the demodulation margin and anti-interference robustness of the LoRa spread spectrum modulation signal under the current electromagnetic interference environment. The energy weighting coefficient is a normalized scalar reflecting the power supply endurance of the node to be connected to the network. Its value ranges from 0 to 1 and it is used to solve the lifespan balancing problem of heterogeneous energy nodes when networking, such as street light controllers powered by mains power and geomagnetic detectors powered by batteries.

[0029] Finally, the intelligent gateway performs an aggregation operation on all received single-point physical feature sets to generate a network-wide physical feature set. The specific execution logic of the aggregation operation is as follows: The intelligent gateway collects the single-point physical feature sets fed back by all nodes waiting to join the network, and uses the set union operation rules to aggregate the discrete single-point physical feature sets into a unified set of physical features for the entire network. This aggregation operation maps the previously invisible, multipath-affected analog radio environment and node micro-energy states in large-scale wide-area multi-hop ad hoc network application scenarios into a computable and sortable digital network link state matrix, providing underlying physical data support for subsequent steps.

[0030] Step S1200, based on the entire network physical feature set The topological impedance component, attenuation impedance component, and interference impedance component are extracted separately, and then fused to generate a single-point communication transmission impedance. Establish single-point communication transmission impedance The mapping relationship between the nodes to be added to the network is used to generate a node impedance mapping table. .

[0031] Specifically, this step is designed to respond to the set of physical characteristics of the entire network from step S1100. This approach treats large-scale, wide-area, multi-hop ad hoc network applications as an equivalent distributed parametric physical circuit network. Utilizing the principle of physical-information isomorphic mapping, it integrates the physical feature set of the entire network. The multidimensional physical characteristic parameters of each node to be added to the network are used as input variables and loaded into the preset communication link cost evaluation logic. The logic calculates a digital scalar that characterizes the obstruction effect of the single-hop physical link between each node to be added to the uplink beacon source on the transmission of discrete service data units (SDUs), i.e., the single-point communication transmission impedance. This generates a node impedance mapping table. At the source of routing decisions, the nondeterministic wireless environment quality is transformed into a computable deterministic routing cost metric.

[0032] In the specific implementation process, the smart gateway receives the entire network's physical feature set. Then, the communication transmission impedance calculation process is initiated. This calculation process abandons the routing mechanism in existing technologies that solely relies on the minimum hop count (Min-Hop) or the maximum signal strength (Max-RSSI), and instead traverses the entire network's physical characteristic set. For each node awaiting network access, a communication link cost evaluation logic based on multi-dimensional weighting is established. This evaluation logic is based on the principle of physical-information isomorphism mapping: that is, in large-scale wide-area multi-hop ad hoc network applications, the transmission impairment vector experienced by the Discrete Service Data Unit (SDU) flowing through the radio frequency channel in a specified operating frequency band is physically equivalent to the impedance experienced by current in a non-ideal circuit. Therefore, the queuing delay at the topology level, the path loss at the physical layer, and the noise interference from the electromagnetic environment are defined as the three orthogonal impedance components hindering communication, and are linearly weighted and superimposed to form the single-point communication transmission impedance. .

[0033] The single-point communication transmission impedance It is a dimensionless routing scalar that characterizes the degree to which a single-hop physical link hinders data transmission for a node seeking to join the network. It is used at the routing decision level to transform the nonlinear wireless environment quality into a linear numerical ranking criterion. Its numerical calculation logic is as follows: the delay factor weight is multiplied by the logical network depth, defined as the topology impedance component, used to quantify the queuing delay at the topology level; this is then added to the attenuation factor weight multiplied by the absolute value of the received signal strength indication and the reference signal strength constant. The product of these ratios is defined as the attenuation impedance component, used to quantify the path loss of the physical layer; this is then added to the interference factor weight multiplied by the noise reference constant. Signal-to-noise ratio and noise reference constant The product of the ratios of these sums is defined as the interference impedance component, used to quantify noise interference in the electromagnetic environment. The sum of the topological impedance component, attenuation impedance component, and interference impedance component constitutes the final measure of link quality, i.e., the single-point communication transmission impedance. The delay factor weight, attenuation factor weight, and interference factor weight are collectively referred to as the scenario adaptation weighting coefficient. The delay factor weight is the first component of the scenario adaptation weighting coefficient, the attenuation factor weight is the second component, and the interference factor weight is the third component. These are dimensionless floating-point numbers whose sum equals 1, used for software-defined network configuration based on different quality of service requirements. For example, in a latency-sensitive smart city street light synchronization control scenario, increasing the delay factor weight penalizes multi-hop paths; in a photovoltaic power station data acquisition scenario with strong electromagnetic interference, increasing the interference factor weight avoids high-noise nodes. The reference signal strength baseline constant... It is a preset normalized denominator, typically set to 120, corresponding to a sensitivity limit of -120 dBm, used to eliminate differences in the physical dimensions of signal strength indication; the noise reference constant It is a preset correction factor, usually set to 20, used as a denominator correction term. When the signal-to-noise ratio value approaches zero or is negative, it ensures that the denominator is always greater than zero, maintains the mathematical convergence of the calculation results, and prevents the impedance value from diverging due to denominator singularity.

[0034] After the calculation is completed, the smart gateway will calculate the single-point communication transmission impedance. Establish a one-to-one mapping relationship with the unique network interface identifier of the corresponding node to be added to the network, and generate a node impedance mapping table. .

[0035] Step S1300: Extract the node impedance mapping table The logical network depth of each node to be added to the network is used as the first sorting key to perform first-level grouping and sorting, and the single-point communication transmission impedance of each node to be added to the network is extracted. The second sorting key is used to perform a secondary priority sort, generating a globally weighted topology matrix. .

[0036] Specifically, this step is intended to respond to the node impedance mapping table from step S1200. In the large-scale, wide-area, multi-hop ad hoc network application scenario of smart city lighting and new energy photovoltaic management, the set of discretely distributed nodes in the entire network awaiting network access in an uncontrolled competitive access state is taken as the topology reconstruction object. This is achieved by utilizing logical network depth and single-point communication transmission impedance. The dual constraint law maps the node impedance table. The unordered data characterizing the physical link quality is mapped to a globally weighted topology matrix. The structured timing priority sequence in the medium access control layer transforms the simple physical layer link state parameters into a topology data structure with deterministic scheduling meaning before the time division multiple access time slot allocation action occurs.

[0037] In the specific implementation process, the smart gateway calls its built-in topology management unit to process the input node impedance mapping table. Perform a hierarchical secondary sorting operation. This hierarchical secondary sorting operation aims to construct a structured scheduling benchmark from a network-wide perspective. Its specific execution logic is as follows: First, the smart gateway performs a first-level grouping and sorting. It then extracts the node impedance mapping table. The logical network depth of each node to be added to the network is used as the first sorting key value to divide the entire set of nodes into several concentric circle logical hierarchy subsets centered on the smart gateway. This partitioning operation establishes the macroscopic time-division multiplexing cycle for data transmission. That is, nodes with smaller logical hierarchy values ​​are physically closer to the gateway, and the data aggregation priority of these near-end nodes has a stricter constraint in the time domain. They must be prioritized for data forwarding processing over far-end nodes with larger logical hierarchy values ​​to conform to the data aggregation rules of multi-hop networks. Subsequently, within each determined concentric circle logical hierarchy subset, the smart gateway performs a secondary priority sorting. The single-point communication transmission impedance of all nodes to be added to the network within the concentric circle logical hierarchy subset is extracted. As the second sorting key value, based on the single-point communication transmission impedance The size of the nodes in the same level within the concentric circle logic hierarchy subset is used to sort them in ascending order. If the single-point communication transmission impedance of a node to be added to the network is... The smaller the value, the better the physical link quality, and the higher the sequence index position of the node to be added to the network within the current concentric circle logical hierarchy subset will be placed.

[0038] Further, please refer to Figure 2 As shown, Figure 2 This is a schematic diagram of the first-level grouping and sorting topology based on logical network depth according to the present invention. For example... Figure 2 As shown, a smart gateway located at the center of the topology is the core, surrounded by several concentric logical hierarchical subsets centered on the smart gateway. These concentric rings are spatially hierarchically divided according to the extracted logical network depth values. The innermost region, closest to the smart gateway, contains near-end nodes with lower logical level values, represented by blue nodes in the diagram; the middle region extending outwards contains nodes awaiting network entry with medium logical level values, represented by green nodes; and the outermost region, farthest from the smart gateway, contains far-end nodes with higher logical level values, represented by yellow nodes. The outward arrows in the diagram indicate a gradually increasing logical network depth. This hierarchical division based on logical network depth establishes the macroscopic time-division multiplexing cycle and priority order for data transmission in large-scale wide-area multi-hop ad hoc networks.

[0039] Through the dual sorting logic combining the first-level grouping sort and the second-level priority sort, the smart gateway generates a four-dimensional feature matrix containing node identity, hierarchical relationship, and sorting weight, i.e., a global weighted topology matrix. The globally weighted topological matrix It is composed of several vertically stacked feature row vectors, where each feature row vector corresponds to a node to be joined in a large-scale wide-area multi-hop ad hoc network application scenario. Its mathematical structure is expressed as follows: The feature row vector consists of a network interface identifier. Uplink relay identifier Logical network depth and single-point communication transmission impedance The four elements are arranged in an ordered manner. This is the globally weighted topological matrix. The row sequence index directly maps to the media access priority in subsequent steps, achieving a deterministic mapping from physical layer link quality to link layer access timing. The network interface identifier... The uplink relay identifier represents the medium access control address or network short address of the node to be added to the network corresponding to the feature row vector, used to uniquely index the corresponding physical device in the subsequently generated deterministic discrete time slot scheduling table; It represents the ID of the next-level relay node in the routing topology tree corresponding to the node to be added to the network, which is used to clarify the hop-by-hop forwarding path of uplink and downlink data and ensure the connectivity of the network tree structure.

[0040] Step S2000: Based on the preset service data frame structure parameters And radio frequency communication parameters determine the atomic transmission time slot reference Combined with the global weighted topology matrix Perform spatial topology-temporal resource mapping to generate node time slot mapping sets. And through downlink channel broadcast, it drives the nodes to be scheduled to generate a network-wide synchronization state. .

[0041] Specifically, this step is intended to respond to the global weighted topology matrix from step S1300. In large-scale wide-area multi-hop ad hoc network application scenarios, nodes to be scheduled are treated as time-domain resource allocation objects, utilizing the business data frame structure parameters. And the smallest granularity of physical layer transmission for determining radio frequency communication parameters, i.e., the atomic transmission time slot reference. and combined with a reference clock source The physical-temporal mapping algorithm is used to weight the global topology matrix. The spatial hierarchy and impedance sorting are mapped to absolute emission windows on the time axis, generating a node time slot mapping set. It drives all network nodes into a distributed local synchronization state through a broadcast distribution mechanism, thereby eliminating co-frequency interference and data collisions caused by random competition at the source of physical layer media access control.

[0042] Further, step S2000 includes: Step S2100, based on the preset service data frame structure parameters The sensor acquisition delay of the node to be scheduled was measured using radio frequency communication parameters. Main control processing latency Physical layer radiation time and radio frequency propagation margin The four components measured above are linearly superimposed in the time domain to generate an atomic transport time slot reference. .

[0043] Specifically, this step is intended to respond to the global weighted topology matrix from step S1300. The global weighted topology matrix has been incorporated into large-scale wide-area multi-hop ad hoc network application scenarios. The node to be scheduled, i.e., the aforementioned node to be connected to the network, performs a complete physical process of reporting service data once, which is taken as the object of time-domain quantization. Using the link differential decomposition logic, the pre-set service data frame structure parameters in the intelligent gateway configuration unit are parsed. With radio frequency communication parameters, the entire chain of operations, from physical environment sensing, analog-to-digital signal conversion, digital logic processing to radio frequency signal radiation, is mapped to a unique digital time measurement unit, namely the atomic transmission time slot reference. This provides a minimum transmission time interval measurement benchmark for subsequent steps, eliminating channel resource waste or transmission collision risks caused by time slot allocation deviations at the source of resource allocation.

[0044] In the specific implementation process, the intelligent gateway calls the preset service data frame structure parameters in its non-volatile memory. In accordance with the radio frequency communication parameters, the data transmission process of the node to be scheduled is segmented and the delay is measured. The specific logic of the segmented delay measurement is as follows: First, the intelligent gateway performs segmented delay measurement based on the service data frame structure parameters. The sensor hardware parameters defined in the standard are used to measure the inherent hardware time required for the sensor interface front-end of the node to be scheduled to complete physical quantity acquisition and analog-to-digital conversion, i.e., the sensing acquisition latency. Secondly, based on the microprocessor performance indicators of the node to be scheduled, the instruction cycle time required for its computing core to perform data encryption, cyclic redundancy check, and protocol stack encapsulation is measured, i.e., the main control processing latency. Next, based on the aforementioned radio frequency communication parameters, the physical layer radiation time required for the radio frequency communication unit of the node to be scheduled to modulate the encapsulated baseband digital signal stream into an analog radio frequency carrier signal and complete air transmission is calculated. Finally, a protection time window, i.e., radio frequency propagation margin, is reserved to compensate for electromagnetic wave spatial propagation delay and hardware transceiver state switching. The intelligent gateway performs a linear time-domain superposition of the independent delay components of the four dimensions mentioned above to calculate the minimum physical quantity of time representing the channel resources occupied by any node to be scheduled, i.e., the atomic transmission time slot reference. The atomic transport time slot reference The value is equal to: the sensing acquisition delay With the main control processing delay The sum, plus the physical layer packet length With air transmission rate The ratio, i.e., physical layer radiation time, plus the radio frequency propagation margin. The sum of .

[0045] Wherein, the physical layer data packet length It originates from the structural parameters of the business data frame. The parsing result represents the number of bits in a complete data frame, including the physical layer preamble, synchronization word, payload, and checksum, in bits, and is used to define the data capacity boundary of a single transmission; the air transmission rate It is a physical layer indicator derived from radio frequency communication parameters. It represents the physical layer equivalent bit rate of the LoRa radio frequency module under the current setting of spreading factor, signal bandwidth and coding rate. Its unit is bits per second and it is used to characterize the transmission capability of the channel.

[0046] Step S2200, based on the global weighted topology matrix With atomic transport time slot reference Execute spatial topology-temporal resource mapping logic, combined with preset hierarchical protection intervals. Calculate and generate absolute start launch time Generate node time slot mapping set .

[0047] Specifically, this step is intended to respond to the global weighted topology matrix from step S1300. Atomic transport time slot reference from step S2100 and reference clock source In the large-scale wide-area multi-hop ad hoc network application scenario of smart city lighting and new energy photovoltaic management, the nodes to be scheduled are mapped to a continuous time axis, and the spatial topology-temporal resource mapping logic is executed to generate the global weighted topology matrix. The logical network depth, which characterizes spatial attributes, is transformed into hierarchical transmission cycles on the time axis, and the intra-layer impedance, which characterizes physical quality, is sorted by index. This is converted into discrete time slot offsets of hierarchical transmission cycles, thereby calculating the absolute start transmission time of each node to be scheduled. Generate a node time slot mapping set containing the entire network time series plan. At the physical layer media access control source, orthogonal collision-free access of all network nodes is achieved.

[0048] In the specific implementation process, the smart gateway executes the spatial topology-temporal resource mapping logic, which includes the following three time-related sub-steps: First, the smart gateway is based on the global weighted topology matrix. The logical network depth in the algorithm linearly divides the entire network data polling cycle into several non-overlapping hierarchical transmission cycles on the time axis, ensuring that the sets of nodes with the same number of hops at different logical network depths are physically isolated in the time domain. That is, the set of nodes at the current logical network depth and the set of nodes at the next logical network depth work in completely different time windows, thereby avoiding hidden terminal interference.

[0049] Secondly, within each defined hierarchical transmission cycle, the intelligent gateway uses a global weighted topology matrix. In-layer impedance sorting index Combined with atomic transmission time slot reference Discrete time slot offsets are allocated to all nodes to be scheduled within the current level transmission cycle.

[0050] Finally, to eliminate clock errors accumulated due to crystal oscillator temperature drift or multipath effects in large-scale deep networking environments, the smart gateway enforces a preset insertion level protection interval between two adjacent transmission cycles. This allows for the calculation and generation of the absolute start launch time. The layer impedance sorting index is mentioned above. Based on single-point communication transmission impedance The sequence number generated by the size is an integer starting from 0; single-point communication transmission impedance. The lower the value of a high-quality node, the smaller its sort index value, and the more accurate its calculated absolute start emission time. The earlier a position is, the higher its priority in launching.

[0051] The absolute start launch time It is the time anchor point at which a node to be scheduled is authorized to activate its RF power amplifier for data reporting within the current network-wide data polling cycle. It is the starting boundary of the physical layer transmission time interval, with its unit being milliseconds. It is used to strictly define the node's transmission behavior in the time domain, and its numerical calculation logic is as follows: Absolute start transmission time Equal to the reference clock source Add the difference between the logical network depth and 1, and the maximum transmission cycle of the layer. The product of these two values, i.e., the total time consumed by all preceding layers, plus the intra-layer impedance sorting index. With atomic transport time slot reference The product of the discrete time slot offset within the current layer's transmission cycle, plus the logical network depth and the layer protection interval. The sum of the products. Wherein, the reference clock source... This represents the midnight of the current polling cycle and is the time reference origin for network-wide synchronization. Its value is typically provided by the gateway's GPS / BeiDou timing module or a high-precision local clock; the maximum transmission cycle of the specified level... This represents the preset maximum time window required to complete data transmission for all scheduled nodes with the same hop count within a logical network depth. Its value must be greater than the maximum number of nodes within that logical network depth and the atomic transmission time slot reference. The product of these is used to achieve time isolation between node sets with different logical network depths, preventing cross-layer interference; the hierarchical protection interval This indicates the preset inter-layer silence duration, which is typically between 10ms and 20ms. It is used to absorb the cumulative timing deviations caused by hardware crystal oscillator errors or electromagnetic multipath propagation, and to prevent cross-layer collisions of data from node sets at adjacent logical network depths.

[0052] Ultimately, the smart gateway will calculate the absolute start launch time. and the corresponding launch end time Combine to generate a node time slot mapping set The launch end time is mentioned above. Numerically equal to the absolute start launch time and atomic transport time slot reference sum.

[0053] Step S2300: Map the node time slot set Differential compression coding is performed to construct timing configuration broadcast frames. These frames are then broadcast and distributed via downlink physical channels to trigger scheduled nodes to perform non-volatile writes and switch between sleep and wake-up modes, generating and outputting the network-wide synchronization status. .

[0054] Specifically, this step is intended to respond to the node slot mapping set from step S2200. The node time slot mapping set generated on the edge computing gateway side in a large-scale wide-area multi-hop ad hoc network application scenario that integrates smart city lighting and new energy photovoltaic management. The data is then distributed to each node awaiting scheduling. This step utilizes a differential compression coding algorithm and a downlink physical channel broadcast distribution mechanism to distribute the abstract node time slot mapping set. The timing configuration parameters are converted into device-side timing configuration parameters in the local non-volatile memory of the node to be scheduled, and the time-triggered discontinuous reception mode is activated accordingly, thereby outputting the network-wide synchronization status. This mechanism enables a shift in communication behavior from random competitive eavesdropping to deterministic on-demand wake-up on the physical terminal side.

[0055] In the specific implementation process, the smart gateway executes timing distribution and synchronization activation logic, which includes four consecutive actions: differential compression coding of data, downlink physical channel broadcast distribution, non-volatile writing, and sleep / wake-up mode switching. The specific execution logic is as follows: First, the smart gateway invokes the protocol stack encoding unit to map the node time slots. Perform Delta Compression Encoding (DCE) operation. Due to the payload limitations of the LoRa physical channel, and to reduce air interface signaling overhead, the smart gateway extracts the node time slot mapping set. absolute start launch time The time value of the timing start node with the smallest value is used as the global time base reference value, and the time parameters of the remaining subsequent timing nodes are converted into incremental offset values ​​relative to the preceding node of the timing adjacent node, thereby encapsulating and constructing a timing configuration broadcast frame that maximizes the effective payload.

[0056] Secondly, the smart gateway is configured to broadcast and distribute LoRa downlink physical channels. Using a preset maximum transmit power and spreading factor, it performs a timing configuration broadcast frame transmission operation to the large-scale wide-area multi-hop ad hoc network with a preset number of retransmissions within a preset synchronization window. The maximum transmit power refers to the peak signal strength output by the RF front-end amplifier, used to ensure that the downlink signal can penetrate obstacles to cover edge nodes; the spreading factor refers to the number of chips that are spread for each information bit in LoRa modulation, used to improve the signal-to-noise ratio gain and anti-interference capability of the receiver.

[0057] Furthermore, in a large-scale wide-area multi-hop ad hoc network, each node to be scheduled initiates a distributed local resolution procedure upon receiving the timing configuration broadcast frame. The node to be scheduled, upon receiving the timing configuration broadcast frame, determines its resolution based on its own network interface identifier. The absolute start transmission time belonging to the node to be scheduled is decompressed and extracted from the timing configuration broadcast frame. With the end of launch And extract the absolute start launch time With the end of launch Non-volatile write operations are performed on the non-volatile memory unit of the currently scheduled node to prevent the loss of timing configuration due to unexpected power failure and reset of the currently scheduled node.

[0058] Finally, after the non-volatile write operation is completed, all scheduled nodes with fixed parameters automatically enter a time-triggered discontinuous reception state. This means the currently scheduled node cuts off the main power supply to its RF transceiver front-end and enters a deep sleep mode with microampere-level power consumption, retaining only the low-power real-time clock. This occurs if and only if the local low-power real-time clock count of the scheduled node matches the absolute start transmission time in the non-volatile memory. During matching, the currently scheduled node is momentarily woken up by a hardware timed task scheduler interrupt to perform data reporting at the RF transceiver front-end, and the data reporting continues until the end of the transmission period. The radio frequency circuit is then immediately shut down, returning to sleep mode.

[0059] Through the aforementioned distribution process of the timing-configured broadcast frame and the switching process between discontinuous reception and operating modes, the smart gateway enters the synchronization window waiting phase. During this phase, the smart gateway starts its internal downlink synchronization protection timer to execute countdown logic. The duration of this downlink synchronization protection timer is not arbitrarily set, but strictly follows the following physical layer parameter relationship: its value is equal to the physical layer transmission duration of the timing-configured broadcast frame. With preset retransmission count The product of these factors, plus the maximum propagation delay margin. The smart gateway determines at the physical layer that all nodes to be scheduled have completed parameter fixing and entered deep sleep mode only when the downlink synchronization protection timer experiences a count overflow interrupt. In response to this count overflow interrupt signal, the smart gateway's main control logic unit sets the network-wide synchronization flag in the system status register to an active level, thereby generating and outputting the network-wide synchronization status. This is the synchronized status across the entire network. As a state machine transition instruction of the system's media access control layer, it signifies that the large-scale wide-area multi-hop ad hoc network has completed the smooth migration from the topology construction and resource allocation stage to the deterministic data polling stage. The radio frequency front-ends of all nodes to be scheduled are in a ready and silent state, waiting for the reference clock to trigger the first round of data reporting.

[0060] The physical layer transmission duration of the timing configuration broadcast frame is specified. This refers to the single-frame timing configuration of broadcast frames within the current radio frequency communication parameters, such as spreading factor and bandwidth, with the air flight time determined by the physical characteristics of the LoRa modem; the preset retransmission count... This refers to the redundant transmission count pre-set by the smart gateway to ensure downlink broadcast reliability, used to cover packet loss that may occur due to random channel fading; the maximum propagation delay margin It is a composite time margin, the value of which is equal to the sum of the spatial propagation delay of electromagnetic waves to the node with the maximum logical network depth and the hardware write cycle time of the non-volatile memory of the node to be scheduled. It is used to physically ensure that even the node at the very end of the network has completed configuration and power off its RF power.

[0061] Step S3000, respond to the polling cycle trigger signal Synchronized with the entire network Send physical layer trigger signal Trigger signal through physical layer Alternatively, a hardware timer scheduler may match an interrupt to trigger the node to be scheduled to generate a local data payload. and local data payload Perform differentiated data transmission and aggregation logic to generate cascaded aggregated data packets. .

[0062] Specifically, this step is intended to respond to the network-wide synchronization status from step S2300. and polling cycle trigger signal The generation process uses discretely distributed, passively waiting nodes in large-scale wide-area multi-hop ad hoc network application scenarios as data acquisition sources, utilizing the local sensor interface signals of the nodes. The physical environment characteristics are acquired, and the discrete local data payload is aggregated through the cascading aggregation mechanism of relay nodes. Reconstructed into cascaded aggregated data packets containing all network service data. This enables the physical splicing and one-time backhaul of multi-node data at the data link layer, solving the problem of air interface time fragmentation caused by frequent handshakes in deep networking environments.

[0063] Further, step S3000 includes: Step S3100: In response to the polling cycle trigger signal generated by the hardware timed task scheduler when the accumulated pulse count value reaches the preset polling cycle value. Synchronized with the entire network Initiate data polling to read the absolute time base reference value of the entire network. and business instruction types An aggregated trigger frame is constructed using a bit-by-bit concatenation algorithm. and aggregate trigger frames Mapped to physical layer trigger signal .

[0064] Specifically, this step is designed to respond to the polling cycle trigger signal generated by the hardware timer scheduler inside the smart gateway. and the network-wide synchronization status from step S2300 In large-scale wide-area multi-hop ad hoc network application scenarios, nodes in deep sleep or ready-to-go states are used as command recipients to execute multicast polling triggering and time base calibration logic. Through four core steps with temporal causal relationships—trigger signal generation, logical frame construction, physical signal modulation, and calibration execution—the absolute time base reference value of the entire network is generated. and business instruction types Aggregated trigger frames Modulated as a physical layer trigger signal Simultaneously, it is delivered to the scheduling node with a logical network depth of 1, namely the first-layer relay node, and performs dynamic time base calibration while issuing service collection instructions, thereby eliminating the cumulative clock drift caused by long-cycle operation.

[0065] In the specific implementation process, the smart gateway executes multicast polling triggering and time base calibration logic. This logic includes four core steps with temporal causal relationships: trigger signal generation, logical frame construction, physical signal modulation, and calibration execution. The specific execution logic is as follows: First, the main control logic unit of the smart gateway runs a hardware timer task scheduler, which counts the hardware clock based on a preset polling period value stored in the system configuration register. When the current accumulated pulse count value of the hardware timer task scheduler reaches the preset polling period value, the scheduler generates a polling period trigger signal. The smart gateway responds to the trigger signal during this polling cycle. And verify the network-wide synchronization status in the system status register. Once the signal level is active, the data polling process is initiated.

[0066] Secondly, the smart gateway accesses the satellite timing module interface via its internal bus, such as a timing unit integrating GPS / BeiDou / PTP protocols, to read the current absolute time base reference value for the entire network. It also retrieves the business instruction type corresponding to this polling task from the internal task scheduling queue. Among them, the absolute time base reference value of the entire network. It is a 64-bit precision timestamp originating from the satellite timing module interface inside the smart gateway. Its value represents the number of milliseconds accumulated from the starting time reference point of Coordinated Universal Time (UTC) to the current time, serving as a reference clock source; the business instruction type This is a functional attribute identifier for this polling task. For example, 0x01 indicates full data collection, and 0x02 indicates only alarm data reporting. This is used to instruct the node to be scheduled to call the corresponding sensor driver subroutine. Subsequently, the smart gateway uses a bit-by-bit concatenation algorithm to construct a special aggregated trigger frame. The aggregated trigger frame The binary bit sequence structure is equal to: multicast frame header The bit sequence, and the absolute time base reference value of the entire network. The bit sequence is concatenated bit by bit and then combined with the business instruction type. The bit sequence is concatenated and finally combined with the cyclic redundancy check code. A continuous data stream formed by concatenating bit sequences. The multicast frame header is... This is the physical link layer addressing control field, whose value is preset to a universal group address identifier, such as 0xFFFF, pre-installed in the communication protocol stack of all nodes to be scheduled in the entire network. It is used to instruct the physical layer decoder at the receiving end to bypass its internal unicast address filtering logic based on the device's unique identifier, forcibly waking up nodes to be scheduled that are in a discontinuous reception sleep state and whose logical network depth is 1; the cyclic redundancy check code... It is a 16-bit error control sequence calculated based on a specific generator polynomial, used to ensure the data integrity of downlink commands in environments with strong electromagnetic interference.

[0067] Next, the smart gateway configures the LoRa RF front-end, reads the multicast-specific spreading factor and downlink center frequency pre-stored in the RF configuration register, and aggregates the trigger frame. As baseband data, forward error correction coding and linear frequency modulation spread spectrum modulation are performed to generate physical layer trigger signals that propagate to the entire network's air interface. The physical layer trigger signal It radiates in the form of electromagnetic waves in the physical space of a large-scale, wide-area, multi-hop self-organizing network, effectively covering the nodes to be scheduled that are within the single-hop radio frequency coverage range of the smart gateway.

[0068] Finally, the physical layer trigger signal is received. The nodes to be scheduled with a logical network depth of 1 are demodulated to restore the aggregate trigger frames. The node to be scheduled parses the service instruction type. To determine the data acquisition task and simultaneously extract the absolute time base reference value of the entire network. It compares this value with the current value of the local low-power real-time clock register and uses the network-wide absolute time base reference value. By directly rewriting the local low-power real-time clock register, the local time is forced to be aligned to the gateway reference, thus completing dynamic time base calibration.

[0069] Step S3200, respond to the physical layer trigger signal Alternatively, a hardware timer scheduler can be matched with an interrupt to initiate dual-mode wake-up activation and parallel acquisition operations to obtain local sensor interface signals from the node. , will the node's local sensor interface signal Mapped to local sensing vectors A local data payload is generated through payload encapsulation operations. .

[0070] Specifically, this step is intended to respond to the physical layer trigger signal from step S3100. Or generated by the local hardware timed task scheduler based on the absolute start launch time from step S2200. In response to hardware matching interruption events, the system uses nodes connected to industrial-grade physical sensors in large-scale wide-area multi-hop ad hoc network applications as data acquisition sources, and employs a pre-awakening parallel acquisition mechanism to acquire node local sensor interface signals characterizing the power grid operating status and environmental physical properties. and the node's local sensor interface signal Mapped to standardized local sensing vectors This allows for the execution of a headerless encapsulation strategy to avoid the physical layer addressing overhead in the standard frame structure, encapsulating the data into a clean local payload. This allows for the standardization and lightweighting of perceived data at the data source, providing the smallest effective data unit for subsequent cascading aggregation and transmission.

[0071] In the specific implementation process, the intelligent gateway and the node to be scheduled collaboratively execute the deterministic perception data acquisition and encapsulation logic. This logic includes four core steps with physical timing constraints: dual-mode pre-awakening activation, parallel sampling, vector mapping, and payload encapsulation. The specific execution logic is as follows: First, the node to be scheduled performs a wake-up activation operation based on dual guarantees of active time-domain and passive spatial-domain operation. The node to be scheduled, in a time-triggered discontinuous reception state, continuously runs its low-power real-time clock and simultaneously activates the periodic channel activity detection mode of the RF front-end. This wake-up activation operation aims to drive the main control microprocessor of the node to be scheduled to switch from deep sleep mode to full-speed operation mode, specifically including two parallel wake-up paths: Path 1: Deterministic active wake-up based on absolute time. When the count value of the local low-power real-time clock reaches a preset wake-up time threshold, the hardware timer scheduler is triggered to match an interrupt, and the node to be scheduled automatically exits deep sleep mode. The wake-up time threshold is numerically equal to the absolute start transmission time. Subtract RF link warm-up time The radio frequency link warm-up time These are constants that are pre-determined and fixed based on the electrical characteristics of the hardware of the node to be scheduled. Their values ​​encompass the sum of the microprocessor's external crystal oscillator start-up and stabilization time, the phase-locked loop locking time, and the analog front-end capacitor charging time. They are designed to provide the necessary physical layer startup buffer period to ensure that the absolute start-up time is met. The system was already in full-speed, ready state when it arrived.

[0072] Path Two: Event-based Passive Wake-up Based on Physical Signals. When the radio frequency front-end of the node to be scheduled detects a physical layer preamble sequence in the air that matches its spreading factor and bandwidth parameters in channel activity detection mode, i.e., a physical layer trigger signal... At this time, the RF module sends an RF preamble detection interrupt signal to the main control microprocessor, triggering the node to be scheduled to immediately exit sleep mode. This logic mainly serves as a redundancy compensation mechanism for path one, used to respond to non-periodic real-time polling commands initiated by the smart gateway, or to achieve passive synchronization acquisition when the local clock and absolute time are out of sync due to crystal oscillator temperature drift, ensuring data acquisition based on the node time slot mapping set. Even within a time window where the time triggering mechanism fails or is not covered, the system still possesses the ability to respond to emergencies.

[0073] Triggered by either of the two paths mentioned above, the node to be scheduled completes the power domain switch and enters the full-speed running ready state, providing a stable clock source and voltage reference for subsequent parallel sampling operations.

[0074] Secondly, after being woken up, the node to be scheduled invokes its internal local sensor interface driver to perform parallel sampling operations on the physical sensor devices connected to the node through an analog-to-digital converter, thereby acquiring the node's local sensor interface signals. The parallel sampling operation refers to simultaneously reading data from multiple physical sensor devices within the same clock cycle or a very short time window to ensure the consistency of multi-dimensional data such as voltage and current across time slices and avoid phase drift caused by serial sampling.

[0075] Next, the data processing unit of the node to be scheduled processes the collected local sensor interface signals of the node. Standardization quantization and vector mapping are performed to generate local sensing vectors. The local sensing vector The digital set representing the physical state of the nodes currently to be scheduled is composed of grid voltage sampling values. Loop current sampling value Instantaneous active power Cumulative electricity consumption And a multidimensional ordered array composed of auxiliary environmental parameters. Among them, the grid voltage sample value... These values ​​are obtained by collecting data from voltage transformers and calculating using the root mean square algorithm, ranging from 0V to 380V, and are used to characterize the steady-state voltage quality of the power supply line; the circuit current sampling value... It is the real-time current value collected by a current transformer, with a value range of 0-16A, used to monitor the load operating status and short-circuit anomalies; the instantaneous active power The cumulative energy consumption is calculated by integrating the product of synchronously sampled instantaneous voltage and current values ​​over the power frequency cycle. The instantaneous active power The time integral value is used for carbon emission measurement; the auxiliary environmental parameters include the carbon dioxide concentration obtained by a non-dispersive infrared sensor. The operating temperature of the equipment is obtained through a thermistor. and ambient light intensity obtained through a photosensitive sensor .

[0076] Finally, the node to be scheduled performs a payload encapsulation operation. The node to be scheduled encapsulates the local sensing vector... After converting it to a binary byte stream and removing the source medium access control address, destination MAC address, and physical layer preamble, it is directly encapsulated as the local data payload. This payload encapsulation operation utilizes the node time slot mapping set generated in step S2200. This means that the parent node located one layer above the logical topology can uniquely identify the data source solely by the absolute moment of the received wireless signal on the time axis, without the need to resolve additional header address overhead.

[0077] Step S3300: Analyze the global weighted topology matrix. To identify whether the node to be scheduled is a leaf node or a relay node, the local data payload is adjusted accordingly. Perform differentiated data transmission and aggregation logic to construct cascaded aggregated data packets. and respond to the absolute start launch time Cascaded aggregate data packets Transmit to the next higher logical parent node.

[0078] Specifically, this step is intended to respond to the local data payload from step S3200. Uplink data packets The arrival and the absolute start launch time from step S2200 The timing-triggered events will enable large-scale wide-area multi-hop ad hoc network applications based on the global weighted topology matrix. The nodes to be scheduled in the constructed logical topology tree structure serve as deterministic concatenated transmission channels for data flow. Leveraging the deterministic timing advantage established at the physical layer, the nodes to be scheduled, identified as relay nodes, execute a concatenated aggregation mechanism. This mechanism strips the physical layer protocol headers, extracts the payloads, and concatenates the timing sequences of discrete sensing data streams originating from different bottom-level leaf nodes at the data link layer, constructing concatenated aggregated data packets with physical length increasing inversely with each layer. This enables hierarchical aggregation of multi-source data streams and single-frame burst transmission at the data link layer, reducing physical frame header overhead and channel contention in deep network environments.

[0079] In the specific implementation process, this step is based on the global weighted topology matrix from step S1300. The topology hierarchy identity of the node to be scheduled is parsed, and differentiated data transmission and aggregation logic is executed accordingly.

[0080] First case: If the node to be scheduled is in the globally weighted topology matrix If no child node index exists in the topology tree, meaning the node to be scheduled is at the end of the tree, then it is marked as a leaf node. This leaf node does not execute multi-source data aggregation logic but instead executes the endpoint direct connection transmission strategy. When the count value of the local low-power real-time clock reaches the absolute start transmission time... At that time, the leaf node directly calls the LoRa radio front end to handle the local data payload. Perform physical layer encapsulation operations.

[0081] The execution logic of the physical layer encapsulation operation is as follows: Construct a standard physical layer frame header. And a 32-bit cyclic redundancy check code for full-frame error detection. Subsequently, the leaf node uses a bitwise concatenation algorithm to assemble the standard physical layer frame header. Local data payload and Cyclic Redundancy Check (CRC) codes The data packets are concatenated to generate cascaded aggregated data packets. Specifically, the cascaded aggregated data packets In the first scenario, it's the basic form with an aggregation degree of 1, but the protocol definition remains consistent with the output of subsequent relay nodes. The concatenated aggregated data packets are then processed using a preset LoRa uplink communication frequency channel. Directed transmission to the global weighted topology matrix The logical parent node mapped in the middle. This cascaded aggregated data packet. After reaching the parent node, the uplink data packet will be used as the parent node to execute the second scenario of this step. It participates in the subsequent cascaded aggregation process.

[0082] Among them, the standard physical layer frame header This represents the physical link layer control field, whose value contains the unique device identifier of the current leaf node, i.e., the network interface identifier. As the source address and the device identifier of the logical parent node, i.e., the uplink relay identifier. The destination address is used to establish the routing direction of data within a single-hop physical link; the cyclic redundancy check code It is a 32-bit check sequence calculated based on a specific generator polynomial, and its check range covers the area from the standard physical layer frame header. From the start bit to the local data payload The termination bits of all data are ensured to guarantee that the parent node can pass the integrity check during reception and filter out erroneous frames caused by channel noise.

[0083] The second scenario: If the node to be scheduled is in the globally weighted topology matrix If at least one valid direct child node index exists, it is identified as a relay node. This relay node performs a concatenated aggregation transmission mode, and its specific logic includes four core steps with data link layer processing characteristics: receive buffering, unpacking and stripping, concatenated splicing, and encapsulation forwarding, as detailed below: First, the node to be scheduled, identified as a relay node, at its absolute start transmission time... Before arrival, open the receive listening window. The opening period of this receive listening window is based on the node time slot mapping set generated in step S2200. Calculations show that its duration strictly covers the weighted topology matrix of the relay node in the global domain. The transmit time slots of all direct child nodes mapped in the middle are used to receive uplink data packets uploaded by the direct child nodes in sequence. .

[0084] Secondly, the relay node receives and demodulates each uplink data packet successfully within its receiving and listening window. Perform the unpacking and stripping operation. The specific execution logic of the unpacking and stripping operation is as follows: After the relay node verifies the data integrity using the physical layer decoder, it removes the uplink data packet according to the preset physical layer frame structure protocol. Standard physical layer frame header at the start position Only the valid data payload is extracted from the physical layer preamble used for radio frequency synchronization. It is then stored in the local aggregation cache, thereby eliminating the redundant addressing and synchronization overhead in single-hop physical link transmission.

[0085] Next, the relay node performs the aggregation and splicing operation. The relay node then loads its local data payload. As per the agreement, it is placed first or last, and then according to the node time slot mapping set. The order in which direct child node transmission slots are defined determines the effective data payload of all child nodes within the local aggregation buffer. The binary bit streams are tightly concatenated to form a composite payload sequence with dynamically variable physical length.

[0086] Finally, the relay node performs a re-encapsulation and forwarding operation. The relay node adds a unique standard physical layer frame header to the composite payload sequence. Cyclic Redundancy Check Code Constructing cascaded aggregated data packets If and only if the count value of the local low-power real-time clock reaches the absolute start transmit time. At that time, the node to be scheduled transmits the cascaded aggregated data packets through the LoRa radio front end. One-time burst transmission to the global weighted topology matrix The logical parent node of the mapping.

[0087] Further, please refer to Figure 3 As shown, Figure 3 This is a schematic diagram illustrating the principle of identity recognition of nodes to be scheduled in a globally weighted topology matrix according to the present invention. Figure 3 As shown, a logical topology tree is constructed with the smart gateway as the core. The circles in the diagram represent nodes to be scheduled, and these nodes are distributed at different levels of the logical topology tree. The diagram clearly shows the identification results of nodes in two specific positions: nodes located at the end of the logical topology tree without downlink connections are identified as leaf nodes; while nodes located in the middle level of the logical topology tree with at least one downlink connection are identified as relay nodes.

[0088] Step S4000, based on the shadow table of all network devices Cascaded aggregation data packets Execute reverse unpacking and state mapping logic to generate a list of missing nodes. Based on the list of missing nodes Node slot mapping set Mid-time domain void resource pool Execute targeted micro-recording logic to generate a complete business dataset for the entire network. The cloud-based business data stream is generated through heterogeneous protocol conversion and security reporting logic. .

[0089] Specifically, this step is intended to respond to the smart gateway receiving a cascaded aggregated data packet from the first-layer relay node with a logical network depth of 1 in step S3300. and the time slot mapping set from node S2200 in step S2200 and the global weighted topology matrix from step S1300 The call uses the smart gateway as the aggregation and governance center for network-wide data, utilizing the shadow tables of all network devices. The state update mechanism aggregates the cascaded data packets of distributed, unscheduled nodes in the physical world. Mapping to a full network state mirror in digital space to verify data integrity, and mining node time slot mapping sets. Unallocated temporal void resource pool Establish a conflict-free directional radio frequency recording link, perform directional micro-recording on the target missing nodes identified during integrity verification, and finally generate a topology matrix containing the global weighted matrix. Cloud business data stream containing complete status data of all registered physical nodes. This allows for the maximization of both transmission reliability and channel resource utilization at the physical level without increasing additional channel overhead, thus enabling a shift from probabilistic transmission to a highly reliable deterministic delivery communication mode.

[0090] Further, step S4000 includes: Step S4100, using the node time slot mapping set Cascaded aggregation data packets Perform reverse unpacking to extract the composite payload sequence, then perform binary stream fixed-length slicing on the composite payload sequence to separate node data blocks, based on the global weighted topology matrix. Map node data blocks to the network-wide device shadow table. Traverse the shadow tables of all network devices Calculate the data integrity index of the entire network To generate a list of missing nodes .

[0091] Specifically, this step is intended to respond to the smart gateway receiving the cascaded aggregated data packets from step S3300, which are aggregated via LoRa-MESH self-organizing network multi-hop aggregation. The weighted topology matrix from step S1300 is stored in the local non-volatile memory on the edge side. With the time slot mapping set from node S2200 As the basis for logical decoding, the smart gateway reverses the tightly concatenated composite payload sequence at the physical layer into independent node data blocks at the logical layer, and maps them to the network-wide device shadow table in the edge computing domain. This allows for the decoupling of cascading coupling at the physical transmission layer and the decoupling of independent attributes at the logical service layer, while simultaneously quantifying the data transmission quality of the LoRa-MESH self-organizing network in real time and outputting a network-wide data integrity index. and list of missing nodes This provides an index basis for subsequent targeted micro-supplementation.

[0092] In the specific implementation process, the smart gateway executes a shadow table based on all network devices. The reverse unpacking and state mapping logic comprises three core components with a logically progressive relationship: reverse unpacking, device shadow mapping, and integrity measurement, as detailed below: First, the protocol parsing unit of the smart gateway receives the cascaded aggregated data packets from the first-layer relay node in the LoRa-MESH ad hoc network. Then, a reverse unpacking operation is performed. The specific execution logic of the reverse unpacking operation is as follows: The smart gateway first verifies the cyclic redundancy check code. Confirm the cascaded aggregation data packet Bit integrity during physical channel transmission. If the check passes, the standard physical layer frame header is stripped. To extract the composite payload sequence. Secondly, the smart gateway calls the locally stored global weighted topology matrix. and node time slot mapping set Retrieve the cascaded aggregated data packet The sequence of logical subtree nodes contained therein. The logical subtree node sequence refers to the sequence based on the globally weighted topology matrix. The defined topology connection relationship includes all data packets belonging to the first-layer relay node, i.e., cascaded aggregation data packets. The set consisting of the next-level direct child nodes of the transmitting source. This is because the transmitting end uses a node-based time slot mapping set in step S3300. The cascading aggregation mechanism uses a smart gateway as the receiving end, based on the node time slot mapping set. The predefined child node time slot allocation order and preset single-node payload byte length are specified in the code. The composite payload sequence is processed by binary stream fixed-length slicing, separating the continuous composite payload sequence into multiple independent node data blocks. Each separated node data block is then mapped back to the network interface identifier of the corresponding current leaf node according to the child node time slot allocation order. This completes the identity restoration of node data blocks without explicit addressing information. The child node time slot allocation order is the effective data payload of the relay node to each direct child node in step S3300. The physical arrangement order during binary ordered concatenation; the single-node payload byte length. It is based on the local sensing vector in step S3200 The data type and bit width are fixed constant values.

[0093] Secondly, the smart gateway performs device shadow mapping operations. The smart gateway maintains a network-wide device shadow table in its local non-volatile memory. This table is based on the global weighted topology matrix. Includes network interface identifiers of all nodes to be added to the network. A pre-initialized and constructed digital mirror storage space is used as the edge computing side for all nodes to be scheduled in the LoRa-MESH self-organizing network. The smart gateway processes and separates each node's data block via fixed-length slices, based on the local sensing vector defined in step S3200. The data format is parsed into grid voltage sample values. Loop current sampling value and instantaneous active power Standard physical quantity values, etc. Subsequently, the smart gateway calls the node time slot mapping set. Obtain the relative position index of the currently processed node data block in the composite payload sequence. Map the node time slots based on the relative position index. Find the corresponding network interface identifier in the middle The parsed standard physical quantity values ​​are then written into the shadow table of all network devices. The network interface identifier The corresponding node status register. This device's shadow mapping operation enables the mapping of physical layer binary data streams without addressing information to data streams with explicit network interface identifiers. Orthogonalization transformation of logical layer business objects.

[0094] Finally, the smart gateway executes a shadow table based on all network devices. The system performs integrity measurements and builds missing indexes. The smart gateway traverses the shadow tables of all network devices that have undergone state write operations. The number of valid response nodes that successfully updated their node status register values ​​within the current polling cycle is counted, and this number is then compared with the global weighted topology matrix. The total number of registered physical nodes recorded in the middle Compare and calculate the data integrity index of the entire network. The network-wide data integrity index The specific calculation logic is as follows: The smart gateway calculates the weighted topology matrix across the entire domain. Each node to be scheduled is substituted into its indicator function for logical judgment. If the judgment result is 1, it is added to the cumulative sum of the number of valid response nodes; otherwise, the node is considered to have missing data. Subsequently, the smart gateway sums the output values ​​of the indicator functions of all nodes to be scheduled to obtain the number of valid response nodes that actually successfully uploaded data in this round, and divides this number of valid response nodes by the total number of registered physical nodes. The resulting quotient is the data integrity index of the entire network. The total number of registered physical nodes. It is a global weighted topological matrix The total capacity of the nodes to be scheduled is defined by the row or column dimension, and its value is a positive integer greater than 0; the substitution characteristic function is a binary function that maps the logical truth value of the node affiliation relationship to a computable value, and its value is determined by the global weighted topology matrix. The network interface identifier of a node to be scheduled When a node is a valid response node, the function value is 1; otherwise, the function value is 0. This function is used to convert discrete, present, or missing node states into summable statistical values.

[0095] For scheduled nodes that fail to match data updates, i.e., those with an indicator function output value of 0, such as those whose data failed to converge to the gateway due to a broken intermediate link in the LoRa-MESH network or channel interference, the smart gateway will set its network interface identifier. Extract and store in the list of missing nodes , which serves as the target object for performing targeted micro-recording in step S4200.

[0096] Step S4200: Traverse the node time slot mapping set Extracting the Temporal Hollow Resource Pool List of missing nodes Mapped to the temporal void resource pool To construct a temporary supplementary scheduling table According to the temporary supplementary scheduling table Perform a silent retransmission operation to generate supplementary service data packets, and update the shadow tables of all network devices with these supplementary service data packets. To generate a complete business dataset for the entire network. .

[0097] Specifically, this step is intended to respond to the missing node list in step S4100. In the non-empty trigger state, the node time slot mapping set from step S2200 will be... And the full network device shadow table after updating the status in step S4100. As the input basis, the node time slot mapping set The discretely distributed time-domain void resource pool is generated due to the uneven depth of the branch node hierarchy in the logical topology tree of the multi-hop ad hoc network, resulting in short branch idle time or preset system-level protection intervals. As a resource to compensate for available time slots in time-division multiplexing mode, it utilizes the full-band electromagnetic silence window formed by non-target missing nodes that have completed data reporting and have no missing records automatically entering deep sleep mode for the remainder of this round of polling. To the list of missing nodes The missing target node initiates a deterministic retransmission request, which physically avoids the random backoff waiting and channel contention required by traditional carrier sense multiple access / collision avoidance mechanisms during the retransmission phase. This achieves a closed loop of network-wide data integrity with zero additional bandwidth cost, generating a complete network-wide service dataset. .

[0098] In the specific implementation process, the smart gateway executes zero-conflict directional micro-recording logic based on temporal void resource identification. This logic includes four core steps with temporal causal relationships: void resource identification, recording time slot mapping, silent directional retransmission, and dataset merging. The specific execution logic is as follows: First, the smart gateway performs a time-domain void resource identification operation. The smart gateway traverses the node time-slot mapping set stored locally. The system identifies idle time-slice indices that are not mapped to any registered physical node by the deterministic time-slot mapping function within the current polling cycle. These idle time-slice indices objectively correspond to the physical channel silence windows caused by uneven branch depth in the tree topology of LoRa-MESH ad hoc networks. The smart gateway performs time-series aggregation of all identified idle time-slice indices to construct a time-domain void resource pool. The time-domain void resource pool The specific construction logic is as follows: Set up a time-domain void resource pool The set of all time slices that satisfy a specific condition, which is: the total number of registered physical nodes. For any registered physical node, the deterministic time slot mapping function returns an empty set, meaning that the time slice is in an unused state at the physical layer. The deterministic time slot mapping function is based on the node time slot mapping set. The defined binary mapping relationship is used to represent the total number of registered physical nodes. Whether any registered physical node in the current time slice has scheduling authority to transmit or receive.

[0099] Secondly, the smart gateway performs the time slot mapping supplementation operation. The smart gateway reads the list of missing nodes. Identifier of each node to be supplemented in That is, the network interface identifier of the node to be scheduled that was determined in step S4100 to have failed to match data updates. And according to the first-in-first-out principle, from the temporal void resource pool Extracting independent idle time slot units sequentially The intelligent gateway establishes a connection from the identifier of the node to be supplemented. to idle time slot unit A one-to-one mapping relationship is used to generate a temporary supplementary scheduling table. This process is achieved through a data entry mapping function, which assigns an identifier to each node to be entered. Deterministically bound to the corresponding idle time slot unit Above, ensure the temporal orthogonality of the supplementary recording process. The idle time slot unit... In the time domain, it is divided into a dedicated downlink time slot for supplementary recording. and supplementary uplink time slots The dedicated downlink time slot for supplementary recording refers to idle time slot units The first half of the microsecond-level time window is used for the smart gateway to send wake-up preambles and control commands to the nodes; the supplementary recording dedicated uplink time slot refers to idle time slot units The latter half of the millisecond-level time window is used for nodes to upload retransmitted service data packets to the smart gateway. Dedicated downlink time slots are supplemented. and supplementary uplink time slots There is usually a transmit / receive switching protection interval between the two to ensure the time domain orthogonality of the supplementary recording process.

[0100] Next, the smart gateway performs a silent directional retransmission operation. After entering the supplementary recording stage, the nodes to be scheduled that were determined in step S4100 to have complete and non-missing data, i.e., valid response nodes, have automatically entered a deep sleep state according to the preset communication protocol, ceasing all radio frequency transmission activities. The collective sleep of non-target missing nodes leads to a significant reduction in the wireless channel noise floor, thereby forming a full-band electromagnetic silence window at the physical level. During this window period, the smart gateway will adjust its scheduling based on the temporary supplementary schedule. In the allocated downlink time slot for supplementary recording To the identifier of the node to be supplemented The corresponding physical node, i.e. the target missing node, sends a directed unicast supplementation request containing specific supplementation instructions. The directional unicast supplementary recording request. It is a Media Access Control (MAC) layer control frame whose frame structure contains the network interface identifier of the target missing node. And an instruction code to immediately retransmit the data from the previous cycle. Subsequently, the target missing node receives this directional unicast supplementation request. Subsequently, in the dedicated uplink time slot for supplementary recording Immediately wake up the radio frequency front end and retransmit its local data payload. This refers to supplementing the business data packets. Since all other valid response nodes in the entire network, except for the target missing node and the smart gateway, are in a dormant state at this time, there are no co-channel interference sources, so this transmission enjoys an extremely high signal-to-noise ratio.

[0101] Finally, the smart gateway performs a dataset merging operation. After receiving the supplementary data packet, the smart gateway repeats the reverse unpacking and state mapping logic in step S4100, extracts the supplementary data, and writes it into the network-wide device shadow table. In the middle and the identifier of the node to be supplemented The corresponding register locations that were originally empty or marked as missing. (When the list of missing nodes...) All node identifiers to be added When all data has been processed or the maximum number of retransmissions has been reached, the smart gateway locks the shadow table of all devices on the network. Encapsulate it into a complete business dataset covering the entire network. .

[0102] Step S4300: parse the complete business dataset of the entire network. It also performs heterogeneous protocol format mapping to generate application layer payloads. Establish a secure tunnel at the transport layer for the application layer payload. Encryption and encapsulation are performed to generate secure data packets for transmission. Securely transmit data packets Mapped to cloud business data stream And push it to the cloud-based IoT platform.

[0103] Specifically, this step is designed to respond to the complete business dataset from step S4200 across the entire network. By leveraging the edge computing protocol conversion capabilities of smart gateways, the complete business dataset of the entire network can be... The compact binary physical layer data based on the underlying LoRa-MESH self-organizing network protocol stack is used as the protocol adaptation object, mapped and converted into a standardized interconnection protocol format that can be recognized by the cloud application layer, generating cloud-based business data streams. At the data egress end, the physical access layer and business application layer protocols are decoupled, so that the cloud management platform does not need to be aware of the complex underlying MESH topology and timing scheduling logic. It only needs to process the full network status snapshot after edge cleaning and structuring, thereby reducing the cloud's parsing computing power overhead and saving the traffic cost of the WAN backhaul link.

[0104] In the specific implementation process, the smart gateway executes the heterogeneous protocol conversion and security reporting logic based on edge computing. This logic includes four time-sequential steps: dataset reading and parsing, heterogeneous protocol format mapping, secure tunnel encapsulation, and WAN burst reporting. The specific execution logic is as follows: First, the smart gateway performs dataset reading and parsing operations. The smart gateway's main control processor locks and reads the complete network-wide business dataset from local non-volatile memory through a direct memory access mechanism. This is the complete business dataset for the entire network. In memory space, it is represented as a weighted topological matrix with global weights from step S1300. A binary state bitmap with a strict address index mapping relationship. Each bit segment offset in the binary state bitmap physically corresponds to a globally weighted topology matrix. Network interface identifier of a specific node to be scheduled The intelligent gateway can directly locate and extract the business data of any node to be scheduled based solely on its physical memory address without requiring a traversal query.

[0105] Secondly, the smart gateway performs heterogeneous protocol format mapping. The smart gateway invokes a pre-built protocol adaptation engine to map the complete business dataset from the entire network. The compact binary physical layer data is converted into a standardized interconnect protocol format supported by the cloud-based IoT platform, i.e., the application layer payload. This process is implemented through data structure mapping functions. The specific execution logic is as follows: The smart gateway uses protocol conversion operators to process the complete network-wide business dataset as input. The data is parsed and analyzed based on a pre-defined tag mapping table. The bit field semantic rules defined in the code transcode and reassemble the binary data stream to generate the application layer payload. The protocol conversion operator is a transcoding algorithm logic running in the edge computing unit of the smart gateway. Its function is to perform format conversion operations from compact binary encoding to structured text encoding, achieving semantic alignment between heterogeneous network protocol stacks; the label mapping table It is a configuration file stored locally on the gateway. Its contents contain the lookup mapping relationship between physical layer binary bit fields and application layer semantic tags, which is used to guide the protocol conversion operator to interpret semantically meaningless bit data into semantic business metrics.

[0106] Next, the smart gateway performs a secure tunnel encapsulation operation. This is to ensure the application layer payload... To ensure confidentiality and integrity during transmission in untrusted environments of public wide area networks (WANs), the smart gateway invokes a hardware encryption coprocessor to establish a secure transport layer tunnel with the cloud-based IoT platform. Based on pre-configured encryption policies, the smart gateway transmits the application layer payload... Encapsulated into secure transmission data packets The secure transmission of data packets. The specific encapsulation logic is as follows: The smart gateway calls an encryption algorithm based on a transport layer security protocol, with the application layer payload... As plaintext input parameters to be processed, and with the session key As an encryption credential, it performs symmetric encryption and integrity verification operations to construct a secure data packet. The encryption algorithm based on the transport layer security protocol is a cryptographic operator pre-installed in the smart gateway firmware. Its value is an algorithm identifier conforming to the transport layer security protocol specification, used to process plaintext application layer payloads. Perform a mathematical transformation to convert it into a ciphertext state that cannot be decrypted by a third party; the session key It is a temporary credential generated by asymmetric encryption algorithms, such as RSA or ECC, during the handshake phase of establishing a connection between the smart gateway and the cloud-based IoT platform. Its value is a pseudo-random binary string with a high entropy, used as the sole decryption parameter for the symmetric encryption algorithm; the secure data transmission packet... It is a ciphertext data frame encapsulated by the transport layer security protocol stack. Its value is a binary sequence containing an encrypted header, encrypted payload and message authentication code. It is used to establish a communication carrier with anti-eavesdropping and anti-tampering capabilities in the public network environment to ensure the secure flow of business data from the edge to the cloud.

[0107] Finally, the smart gateway performs a WAN burst reporting operation. The smart gateway securely transmits data packets through a WAN communication module integrated within its hardware architecture. Mapped to cloud business data stream The data is then pushed to the cloud-based IoT platform using the TCP / IP protocol stack. This completes the secure, cross-layer delivery of data, from discrete sensing data at the underlying physical nodes to structured business data on the cloud management platform.

[0108] Example 2: This embodiment, based on Embodiment 1, provides a deterministic data polling system based on LoRa-MESH self-organizing networks, such as... Figure 4 As shown, the system includes a topology management module, a resource allocation module, a data polling module, and a data delivery module; The topology management module is used to analyze the analog radio frequency carrier signals fed back by nodes to be added to the network in order to aggregate and generate a set of physical features of the entire network. Based on the entire network's physical feature set The node impedance mapping table is obtained by solving. And according to the node impedance mapping table Perform hierarchical secondary sorting operations to construct a globally weighted topology matrix. .

[0109] The resource allocation module is used to allocate data based on preset service data frame structure parameters. And radio frequency communication parameters determine the atomic transmission time slot reference Combined with the global weighted topology matrix Perform spatial topology-temporal resource mapping to generate node time slot mapping sets. And through downlink channel broadcast, it drives the nodes to be scheduled to generate a network-wide synchronization state. .

[0110] The data polling module is used to respond to the polling cycle trigger signal. Synchronized with the entire network Send physical layer trigger signal Trigger signal through physical layer Alternatively, a hardware timer scheduler may match an interrupt to trigger the node to be scheduled to generate a local data payload. and local data payload Perform differentiated data transmission and aggregation logic to generate cascaded aggregated data packets. .

[0111] The data delivery module is based on the shadow table of all network devices. Cascaded aggregation data packets Execute reverse unpacking and state mapping logic to generate a list of missing nodes. Based on the list of missing nodes Node slot mapping set Mid-time domain void resource pool Execute targeted micro-recording logic to generate a complete business dataset for the entire network. The cloud-based business data stream is generated through heterogeneous protocol conversion and security reporting logic. .

[0112] The parts of the technical solutions provided in the embodiments of this application that are consistent with the implementation principles of corresponding technical solutions in the prior art have not been described in detail to avoid excessive elaboration.

[0113] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the invention. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A deterministic data polling method based on LoRa-MESH self-organizing networks, characterized in that, include: The analog radio frequency carrier signals fed back by the nodes to be connected to the network are analyzed to aggregate and generate a set of physical features of the entire network. Based on the set of physical features of the entire network, the node impedance mapping table is calculated, and a hierarchical secondary sorting operation is performed according to the node impedance mapping table to construct a global weighted topology matrix. Based on the preset service data frame structure parameters and radio frequency communication parameters, the atomic transmission time slot reference is determined. The spatial topology-temporal resource mapping is performed in combination with the global weighted topology matrix to generate the node time slot mapping set. The nodes to be scheduled are then driven to generate the network-wide synchronization state via downlink channel broadcast. In response to the polling cycle trigger signal and the network-wide synchronization status, a physical layer trigger signal is sent. The node to be scheduled is triggered to generate a local data payload by matching the interrupt through the physical layer trigger signal or the hardware timer task scheduler. Differentiated data transmission and aggregation logic is then performed on the local data payload to generate cascaded aggregated data packets. Based on the shadow table of all network devices, reverse unpacking and state mapping logic is performed on the cascaded aggregated data packets to generate a list of missing nodes. Based on the list of missing nodes, targeted micro-supplementation logic is performed on the time-domain hole resource pool of the node time slot mapping to generate a complete business dataset for the entire network. Finally, cloud business data stream is generated through heterogeneous protocol conversion and security reporting logic.

2. The deterministic data polling method based on LoRa-MESH ad hoc networks according to claim 1, characterized in that, The method for constructing the global weighted topological matrix includes: In response to the initial network discovery command, a network broadcast beacon is sent to the radio frequency channel, the analog radio frequency carrier signal fed back by the node to be joined is parsed, physical feature signals and power management unit status are extracted to generate a single-point physical feature set, and an aggregation operation is performed on all single-point physical feature sets to generate a full network physical feature set. Based on the physical feature set of the entire network, the topology impedance component, attenuation impedance component and interference impedance component are extracted respectively. The topology impedance component, attenuation impedance component and interference impedance component are fused to generate a single-point communication transmission impedance. The mapping relationship between the single-point communication transmission impedance and the node to be connected to the network is established to generate a node impedance mapping table. Extract the logical network depth of each node to be added to the network from the node impedance mapping table as the first sorting key value to perform first-level grouping sorting, extract the single-point communication transmission impedance of each node to be added to the network as the second sorting key value to perform second-level priority sorting, and generate a global weighted topology matrix.

3. The deterministic data polling method based on LoRa-MESH ad hoc networks according to claim 2, characterized in that, The method for extracting the topology impedance component includes: reading the time delay factor weight in the preset scenario adaptation weighting coefficient, obtaining the logical network depth of the node to be connected to the network, calculating the product of the time delay factor weight and the logical network depth, and generating the topology impedance component. The method for extracting the attenuation impedance component includes: reading the attenuation factor weight in the scenario adaptation weighting coefficient, obtaining the absolute value of the received signal strength indication of the node to be connected to the network, calculating the first ratio of the absolute value to the preset reference signal strength reference constant, and generating the attenuation impedance component based on the product of the attenuation factor weight and the first ratio. The method for extracting the interference impedance component includes: reading the interference factor weight in the scene adaptation weighting coefficient, obtaining the signal-to-noise ratio of the node to be connected to the network, calculating a second ratio with a preset noise reference constant as the numerator and the sum of the signal-to-noise ratio and the noise reference constant as the denominator, and performing a product operation between the interference factor weight and the second ratio to generate the interference impedance component. The scenario adaptation weighting coefficient consists of a delay factor weight, an attenuation factor weight, and an interference factor weight, and the sum of the three weights is 1. The single-point physical feature set includes a received signal strength indicator, which is the absolute value of the power level of the analog radio frequency carrier signal received by the radio frequency front-end of the node to be connected to the network.

4. The deterministic data polling method based on LoRa-MESH ad hoc networks according to claim 1, characterized in that, The method for generating the network-wide synchronization status includes: Based on the preset service data frame structure parameters and radio frequency communication parameters, the sensor acquisition delay, main control processing delay, physical layer radiation time consumption and radio frequency propagation margin of the node to be scheduled are measured respectively, and linear time domain superposition is performed to generate an atomic transmission time slot reference. Based on the global weighted topology matrix and the atomic transmission time slot reference, the spatial topology-temporal resource mapping logic is executed, and the absolute start transmission time is calculated and generated in combination with the preset hierarchical protection interval, thus generating the node time slot mapping set; Differential compression coding is performed on the node time slot mapping set to construct the timing configuration broadcast frame. The timing configuration broadcast frame is then broadcast and distributed through the downlink physical channel to trigger the nodes to be scheduled to perform non-volatile write and sleep-wake mode switching, and generate and output the network-wide synchronization status.

5. The deterministic data polling method based on LoRa-MESH ad hoc networks according to claim 4, characterized in that, The execution method of the spatial topology-temporal resource mapping logic includes: Based on the logical network depth in the global weighted topology matrix, the entire network data polling cycle is divided into multiple non-overlapping hierarchical transmission cycles. For each hierarchical transmission cycle, the product of the extracted intra-layer impedance sorting index and the atomic transmission time slot reference is calculated to obtain the discrete time slot offset of each node to be scheduled within the hierarchical transmission cycle. The intra-layer impedance sorting index is a sequence number generated based on the magnitude of the single-point communication transmission impedance. A preset hierarchical protection interval is configured between two adjacent hierarchical transmission cycles, and a timing superposition calculation is performed to generate an absolute start transmission time. The logic of the timing superposition calculation includes: obtaining the reference clock source value of the current polling cycle, the total duration consumed by all hierarchical transmission cycles, the discrete time slot offset, and the cumulative hierarchical protection interval duration, and performing a summation operation to obtain the absolute start transmission time. The total duration consumed by all hierarchical transmission cycles is the product of the logical network depth of the node to be scheduled minus one and the preset maximum hierarchical transmission cycle. The cumulative hierarchical protection interval duration is the product of the logical network depth and the hierarchical protection interval. The absolute start time of emission is calculated and the atomic transmission time slot reference is used to define the end time of emission. The absolute start time of emission and the end time of emission are associated and stored to construct a node time slot mapping set.

6. The deterministic data polling method based on LoRa-MESH ad hoc networks according to claim 4, characterized in that, The non-volatile write step includes: receiving a timing configuration broadcast frame, using the network interface identifier of the node to be scheduled as the index key value, extracting the absolute start transmission time and transmission end time belonging to the network interface identifier from the timing configuration broadcast frame; and writing the absolute start transmission time and transmission end time into the non-volatile storage unit of the node to be scheduled. The steps for switching between sleep and wake-up modes include: in response to the completion of a non-volatile write operation, controlling the node to be scheduled to enter a deep sleep state with only real-time clock counting; executing logic to compare the real-time clock count value with the absolute start transmission time stored in the non-volatile memory unit; in response to the matching result that the real-time clock count value and the absolute start transmission time value are consistent, driving the node to be scheduled to switch to the data transmission state; and controlling the node to be scheduled to re-enter the deep sleep state when the real-time clock count value reaches the transmission end time.

7. The deterministic data polling method based on LoRa-MESH ad hoc networks according to claim 1, characterized in that, The method for generating the cascaded aggregated data packets includes: In response to the polling cycle trigger signal generated by the hardware timer task scheduler when the accumulated pulse count value reaches the preset polling cycle value and the start of data polling in the network synchronization state, the absolute time base reference value and service instruction type of the entire network are read, an aggregated trigger frame is constructed using a bit-by-bit concatenation splicing algorithm, and the aggregated trigger frame is mapped to a physical layer trigger signal; In response to physical layer trigger signals or hardware timer task scheduler matching interrupts, dual-mode wake-up activation and parallel acquisition operations are started to obtain node local sensor interface signals. The node local sensor interface signals are mapped to local sensing vectors and local data payloads are generated through payload encapsulation operations. The global weighted topology matrix is ​​parsed to identify whether the node to be scheduled is a leaf node or a relay node. Differentiated data transmission and aggregation logic is performed on the local data payload to construct cascaded aggregated data packets. In response to the absolute start time of transmission, the cascaded aggregated data packets are transmitted to the next higher logical parent node.

8. The deterministic data polling method based on LoRa-MESH ad hoc networks according to claim 7, characterized in that, The method for executing the differentiated data transmission and aggregation logic includes: For the scheduled nodes identified as leaf nodes, a standard physical layer frame header is constructed and a cyclic redundancy check code is calculated. The standard physical layer frame header, local data payload, and cyclic redundancy check code are concatenated to generate a concatenated aggregated data packet. The concatenated aggregated data packet is configured as the uplink data packet of the next higher logical parent node. For nodes identified as relay nodes to be scheduled, the receive listening window is activated to buffer uplink data packets. The uplink data packets are unpacked and stripped to extract the valid data payload. Based on the node time slot mapping set, the local data payload and the valid data payload are aggregated and spliced ​​to construct a composite payload sequence. The composite payload sequence is then encapsulated with a standard physical layer frame header and a cyclic redundancy check code to generate a concatenated aggregated data packet.

9. The deterministic data polling method based on LoRa-MESH ad hoc networks according to claim 1, characterized in that, The method for generating the cloud service data stream includes: The node time slot mapping set is used to perform reverse unpacking operation on the concatenated aggregated data packets to extract the composite payload sequence. The composite payload sequence is processed by binary stream fixed-length slicing to separate node data blocks. The node data blocks are mapped to the network device shadow table according to the global weighted topology matrix. The network device shadow table is traversed to calculate the network data integrity index to generate a list of missing nodes. Traverse the node time slot mapping set to extract the time domain hole resource pool, map the missing node list to the time domain hole resource pool to build a temporary supplementary recording scheduling table, perform silent directional retransmission operation based on the temporary supplementary recording scheduling table to generate supplementary recording service data packets, and update the supplementary recording service data packets to the network-wide device shadow table to generate a complete network-wide service dataset. The system parses the complete business dataset across the entire network and performs heterogeneous protocol format mapping to generate application layer payloads. It then establishes a transport layer secure tunnel to encrypt and encapsulate the application layer payloads to generate secure transmission data packets. Finally, it maps the secure transmission data packets to cloud-based business data streams and pushes them to the cloud-based IoT platform.

10. A deterministic data polling system based on LoRa-MESH ad hoc networks, used to implement the deterministic data polling method based on LoRa-MESH ad hoc networks as described in any one of claims 1-9, characterized in that, The system includes a topology management module, a resource allocation module, a data polling module, and a data delivery module; The topology management module is used to parse the analog radio frequency carrier signals fed back by the nodes to be added to the network to aggregate and generate a set of physical features of the entire network, calculate the node impedance mapping table based on the set of physical features of the entire network, and perform a hierarchical secondary sorting operation based on the node impedance mapping table to construct a global weighted topology matrix. The resource allocation module is used to determine the atomic transmission time slot reference based on the preset service data frame structure parameters and radio frequency communication parameters, perform spatial topology-temporal resource mapping in combination with the global weighted topology matrix to generate a node time slot mapping set, and drive the nodes to be scheduled to generate a network-wide synchronization state via downlink channel broadcast. The data polling module is used to respond to the polling cycle trigger signal and the physical layer trigger signal sent by the network-wide synchronization status. It triggers the node to be scheduled to generate a local data payload by matching the interrupt through the physical layer trigger signal or the hardware timer task scheduler, and performs differentiated data transmission and aggregation logic on the local data payload to generate cascaded aggregated data packets. The data delivery module generates a list of missing nodes by performing reverse unpacking and state mapping logic on cascaded aggregated data packets based on the shadow table of all network devices. Based on the list of missing nodes, it performs targeted micro-supplementation logic on the time-domain void resource pool of the node time slot mapping set to generate a complete business dataset for the entire network. Finally, it generates a cloud-bound business data stream through heterogeneous protocol conversion and security reporting logic.

Images