An edge-computing-based internet-of-things device abnormality positioning method and system

Abstract

An edge computing-based IoT device anomaly positioning method and system, in which when the system detects a large number of street lamp terminals offline, the edge gateway controls the online terminals to enter an interference detection mode and collects background noise data. The terminals reporting noise data are marked as sensing nodes, and those failing to report are marked as communication blind area nodes. By comparing the noise intensity values of adjacent sensing nodes, the enhancement direction of the interference signal is determined and an interference gradient vector is generated. The system traces back along the gradient vector to find a topological convergence area to which all the vectors point. If the area exists communication blind area nodes, the blind area node located at the geometric center is determined as the abnormal source; if not, the sensing node with the highest noise intensity in the area is determined as the abnormal source. The application is used to improve the efficiency and accuracy of determining the noise source device in the scene where strong noise interference causes large-area communication blockage, thereby reducing the operation and maintenance of noise and time cost.

Description

Technical Field

[0001] This application belongs to the field of fault detection and location of Internet of Things (IoT) devices, and particularly relates to a method and system for anomaly location of IoT devices based on edge computing. Background Technology

[0002] With the deepening of smart city infrastructure construction, smart street light systems based on IoT technology have become an important carrier for urban data collection. In order to achieve remote centralized management and control of massive street light devices, a regional centralized management method based on power line communication (PLC) technology is usually adopted. This method typically deploys a smart concentrator as an edge gateway on the transformer substation side, using the existing street light power supply lines as the communication medium. The device polls and collects the status data of all individual lamp controllers under the same substation, performs protocol parsing and preliminary data cleaning locally, and only needs to back up the processed core data to the cloud. This method uses power lines as communication links, requires no additional wiring, and the signal coverage extends with the power lines, solving the problems of high communication costs and excessive cloud load caused by large-scale street light access.

[0003] However, power line carrier communication is not only used for data transmission, but its communication quality also depends on the physical environment of the power line. In actual operation, when a ballast or power supply at a certain point in time triggers a high-frequency noise interference fault, this high-frequency noise will couple to the power line as an interference source and radiate indiscriminately along the power line. Since power line carrier communication is essentially a high-frequency load signal on the power line, the noise generated by the faulty equipment often overlaps with the normal carrier communication intermittents. This results in the phenomenon that a large number of streetlights in the same concentrator area go offline or have communication timeouts at the same time, without any abnormality of the faulty node. This makes it difficult for maintenance personnel to quickly identify the overall line fault or the noise interference of the overall equipment when faced with the phenomenon of communication paralysis in the entire area, thus increasing the noise and time cost of troubleshooting the source of the fault. Summary of the Invention

[0004] This application provides an IoT device anomaly location method and system based on edge computing, which can improve the efficiency and accuracy of identifying the noise source device in scenarios where strong noise interference causes large-scale communication blockage, thereby reducing the noise and time costs of operation and maintenance troubleshooting.

[0005] In the first aspect, this application provides an abnormal location method for IoT devices based on edge computing. When the edge gateway detects that the proportion of street light terminals in the offline state exceeds a preset ratio, it broadcasts an environmental perception command to the power line network to control the street light terminals in the online state to suspend the transmission of business data and switch to interference listening mode. When the edge gateway determines that it has received the background noise intensity value reported by the street light terminal, it marks the street light terminal that reported the background noise intensity value as a sensing node, and marks other street light terminals other than the sensing node as communication blind zone nodes. The edge gateway maps the background noise intensity value of the sensing node to the corresponding node position in the electrical topology diagram, and determines the pairs of sensing nodes that are adjacent in electrical connection in the electrical topology diagram; The edge gateway compares the background noise intensity values ​​of each sensing node pair in turn to determine the direction of the interference signal amplification on the power line; The edge gateway generates an interference gradient vector based on the enhancement direction; In the electrical topology diagram, the edge gateway traces back along the interference gradient vector until it locates the topology convergence area that all interference gradient vectors point to. The edge gateway determines whether there are communication blind spots within the topology aggregation area; If present, the edge gateway will identify the communication blind spot node located at the geometric center of the topology convergence area as the abnormal source device; If it does not exist, the edge gateway will identify the sensing node with the highest background noise intensity value in the topology aggregation area as the abnormal source device.

[0006] By employing the above technical solution, the edge gateway controls the switching of the line light terminal to interference detection mode and reports the background noise intensity value. Combined with the electrical topology diagram, the background noise intensity value of the sensing node is mapped to the corresponding location, thus obtaining the spatial distribution characteristics of the interference signal in the power line network. Based on the comparison of background noise intensity values ​​between adjacent sensing node pairs, the direction of interference signal enhancement is determined, and an interference gradient vector is generated for path backtracking, enabling rapid local location of the interference source in complex tree-like topologies. By determining whether there are communication blind spots within the topology convergence area and combining the geometric center location or the maximum background noise intensity value to identify the abnormal source device, the positioning accuracy of abnormal devices in the power line communication network is improved, and the manual inspection workload required for abnormal device positioning is reduced. This method fully utilizes the local computing power of the edge gateway and the topological characteristics of the power line communication network, reducing the dependence of abnormal device positioning on cloud computing resources and improving the real-time performance of abnormal positioning.

[0007] In conjunction with some implementations of the first aspect, in some implementations, the edge gateway sequentially compares the background noise intensity values ​​of each pair of sensing nodes to determine the direction of signal enhancement on the power line, specifically including: The edge gateway calculates the absolute value of the difference between the background noise intensity values ​​of the first and second sensing nodes in the sensing node pair. When the edge gateway determines that the absolute value of the difference is greater than the preset effective fluctuation threshold, it identifies the node with the larger background noise intensity value between the first sensing node and the second sensing node as the high-potential node, and the node with the smaller background noise intensity value as the low-potential node. The edge gateway determines the direction of interference signal amplification on the power line as the direction from the low-potential node to the high-potential node.

[0008] By employing the above technical solution, and calculating the absolute value of the difference in background noise intensity between adjacent sensing node pairs, and comparing it with a preset effective fluctuation threshold, invalid data caused by measurement errors or environmental noise fluctuations can be effectively filtered out. Among node pairs with larger differences, the node with the larger background noise intensity value is identified as a high-potential node, and the node with the smaller difference is identified as a low-potential node. This is used to determine the direction of interference signal enhancement, improving the accuracy of interference propagation direction determination. This method, by introducing a judgment mechanism based on the absolute value of the difference and the effective fluctuation threshold, reduces the influence of environmental noise on interference direction determination, enhancing the anti-interference capability of abnormal device location. The method of determining the direction of interference signal enhancement based on high and low potential nodes improves the tracking efficiency of interference propagation paths in power line networks.

[0009] In conjunction with some implementations of the first aspect, in some implementations, the edge gateway generates an interference gradient vector based on the enhancement direction, specifically including: The edge gateway determines the enhancement direction as the direction of the interference gradient vector; The edge gateway calculates the difference in background noise intensity between pairs of sensing nodes forming the enhancement direction and determines the difference as the vector magnitude of the interference gradient vector; The edge gateway combines the vector pointer and vector magnitude to generate an interference gradient vector that describes the rate of change of noise space between sensing node pairs.

[0010] By employing the above technical solution, and determining the direction of interference signal enhancement as the vector direction of the interference gradient vector, and determining the difference in background noise intensity between adjacent sensing node pairs as the vector magnitude, an interference gradient vector describing the spatial rate of change of noise is generated, achieving a quantitative characterization of interference propagation characteristics. The interference gradient vector simultaneously contains both the direction and intensity information of interference propagation, improving the accuracy of abnormal device location. This method, by transforming qualitative interference propagation characteristics into quantitative vector expressions, enhances the mathematical foundation of the abnormal device location algorithm and improves the reliability of the location results. The quantitative analysis method based on interference gradient vectors improves the efficiency of identifying abnormal behavior in power line networks.

[0011] In some implementations of the first aspect, the edge gateway sends a spectrum sampling command to the sensing node located upstream of the electrical connection of the anomaly source device and receives background noise spectrum data fed back by the sensing node. The edge gateway identifies frequency bands in the background noise spectrum data whose energy amplitude exceeds a preset interference threshold as interference fingerprint features; The edge gateway identifies clean carrier frequency bands that do not overlap with interference fingerprint features in the frequency domain among the multiple communication carrier frequency bands supported by the current system. The edge gateway broadcasts a frequency band migration command containing configuration parameters for the clean carrier frequency band to the power line network to control street light terminals, except for abnormal source devices, to switch to the clean carrier frequency band for communication.

[0012] By employing the above technical solution, background noise spectrum data from upstream sensing nodes of the anomaly source device is collected, and interference fingerprint features are extracted to obtain the frequency domain distribution information of the interference signal. Among the multiple communication carrier frequency bands supported by the system, clean carrier frequency bands that do not overlap with the interference fingerprint features are identified, and other devices are controlled to switch to these frequency bands for communication, thus improving the spectrum utilization efficiency of the power line communication network. This method reduces the impact of abnormal devices on the overall communication quality of the network by performing frequency domain analysis on the interference signal and implementing communication frequency band migration. The interference avoidance mechanism based on spectrum analysis and frequency band migration enhances the anti-interference capability of the power line communication network and strengthens the stability of the communication system.

[0013] In conjunction with some implementations of the first aspect, in some implementations, the edge gateway identifies clean carrier frequency bands that do not overlap with interference fingerprint features in the frequency domain among multiple communication carrier frequency bands supported by the current system, specifically including: The edge gateway calculates the overlap bandwidth ratio of each communication carrier frequency band and interference fingerprint features in the frequency domain. If there are communication carrier frequency bands with an overlap bandwidth ratio of zero, the edge gateway will identify the communication carrier frequency bands with an overlap bandwidth ratio of zero as clean carrier frequency bands. If there are no communication carrier frequency bands with zero overlap bandwidth, the edge gateway calculates the average noise energy density within each communication carrier frequency band; The edge gateway identifies the communication carrier frequency band with the lowest average noise energy density as the clean carrier frequency band.

[0014] By employing the above technical solution, and calculating the frequency domain overlap bandwidth ratio between each communication carrier frequency band and the interference fingerprint characteristics, the impact of interference on different frequency bands can be quantitatively assessed, thereby selecting the optimal communication frequency band. When a completely non-overlapping frequency band exists, selecting this band can significantly reduce the impact of interference on communication quality. When a completely non-overlapping frequency band cannot be found, calculating the average noise energy density of each frequency band and selecting the band with the lowest energy density as the clean carrier frequency band can reduce background noise interference during communication and improve the channel's signal-to-noise ratio. This frequency band selection method based on quantitative indicators improves the accuracy of communication frequency band selection, reduces the trial-and-error costs during frequency band switching, and enables the system to adapt to changes in the interference environment more quickly. By selecting the optimal communication frequency band, the system's communication reliability is improved, the communication bit error rate is reduced, and the overall network throughput is increased.

[0015] In some embodiments, in conjunction with some implementations of the first aspect, after controlling the street light terminals other than the abnormal source devices to switch to the clean carrier frequency band for communication, the method further includes: The edge gateway controls the first neighboring node located upstream of the electrical connection of the abnormal source device to send test data packets to the second neighboring node located downstream of the electrical connection of the abnormal source device; The edge gateway receives link quality parameters for the test data packets fed back by the second neighboring node. The link quality parameters include the signal-to-noise ratio and the received signal strength. When the edge gateway determines that the signal-to-noise ratio is greater than the preset net channel threshold and the received signal strength is less than the preset attenuation threshold, it determines that the abnormal source device has formed a signal attenuation impedance on the clean carrier frequency band, and sends a bridging configuration command to the first and second neighboring nodes. The bridging configuration command is used to control the first and second neighboring nodes to establish a direct communication link that bypasses the abnormal source device, and adjust the signal transmission power of the direct communication link to the maximum power level supported by the device.

[0016] By adopting the above technical solution, and establishing test links between adjacent nodes on both sides of the anomalous source device, and evaluating link quality based on objective parameters such as signal-to-noise ratio and received signal strength, the attenuation characteristics of the anomalous source device on the communication signal can be accurately identified. When it is confirmed that the anomalous source device forms signal attenuation impedance on a clean carrier frequency band, establishing a direct communication link bypassing the anomalous source device and increasing the signal transmission power to the maximum power level can reduce signal attenuation loss during transmission. This adaptive link reconstruction scheme improves the network's anti-interference capability, reduces signal attenuation in the communication link, and reduces the risk of network segmentation. By using maximum transmission power for compensation at the physical layer, the signal penetration capability is improved, the stability of the communication link is enhanced, and the network has stronger fault adaptability.

[0017] In conjunction with some implementations of the first aspect, in some implementations, the bridging configuration command further includes configuration information for the physical layer modulation parameters of the directly connected communication link, specifically including: The edge gateway locks the data transmission rate of the direct communication link to the preset minimum transmission rate level; The edge gateway adjusts the forward error correction coding redundancy of the direct communication link to the preset highest redundancy level; The edge gateway prohibits the first neighboring node from making dynamic rate adjustments during data interaction between the second neighboring node and the first neighboring node in the direct communication link.

[0018] By adopting the above technical solution, a conservative and stable transmission strategy is formed by locking the transmission rate of the direct communication link at the lowest level, adjusting the forward error correction coding redundancy to the highest level, and prohibiting dynamic rate adjustment. The lower transmission rate reduces the impact of inter-symbol interference and improves the signal's time-domain resolution. The higher coding redundancy enhances the error correction capability during data transmission and improves the noise immunity of the communication link. Prohibiting dynamic rate adjustment avoids frequent changes in transmission parameters caused by channel state fluctuations, improving link stability. This fixed and conservative parameter configuration strategy reduces the bit error rate during transmission, improves data transmission reliability, and enables the communication link to maintain basic data interaction capabilities even under adverse transmission environments.

[0019] In a second aspect, embodiments of this application provide an edge computing-based IoT device anomaly location system, which includes: one or more processors and a memory; the memory is coupled to one or more processors, and the memory is used to store computer program code, the computer program code including computer instructions, and the one or more processors call the computer instructions to cause the system to perform the method described in the first aspect and any possible implementation thereof.

[0020] Thirdly, embodiments of this application provide a computer-readable storage medium including instructions that, when executed on a system, cause the system to perform the method described in the first aspect and any possible implementation thereof.

[0021] Fourthly, embodiments of this application provide a computer program product that, when run on a system, causes the system to execute the method described in any possible implementation of the first aspect.

[0022] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: 1. This application provides an edge computing-based method for anomaly localization of IoT devices. By controlling the switching of the street light terminal to interference detection mode and reporting background noise intensity values ​​through an edge gateway, and combining this with the electrical topology diagram, the background noise intensity values ​​of the sensing nodes are mapped to their corresponding locations, thus obtaining the spatial distribution characteristics of interference signals in the power line network. Based on the comparison of background noise intensity values ​​between adjacent sensing node pairs, the direction of interference signal enhancement is determined, and an interference gradient vector is generated for path backtracking, enabling rapid localization of the interference source in complex tree-like topologies. By determining whether there are communication blind spots within the topology convergence area, and combining this with the geometric center location or the maximum background noise intensity value, the source device of the anomaly is identified, improving the localization accuracy of anomaly devices in the power line communication network and reducing the manual inspection workload required for anomaly device localization. This method fully utilizes the local computing power of the edge gateway and the topological characteristics of the power line communication network, reducing the dependence of anomaly localization on cloud computing resources and improving the real-time performance of anomaly localization.

[0023] 2. This application provides a method for anomaly localization of IoT devices based on edge computing. By collecting background noise spectrum data from upstream sensing nodes of the anomaly source device and extracting interference fingerprint features, the frequency domain distribution information of the interference signal can be obtained. Among the multiple communication carrier frequency bands supported by the system, clean carrier frequency bands that do not overlap with the interference fingerprint features are identified, and other devices are controlled to switch to these frequency bands for communication, improving the spectrum utilization efficiency of the power line communication network. This method reduces the impact of abnormal devices on the overall communication quality of the network by performing frequency domain analysis on the interference signal and implementing communication frequency band migration. The interference avoidance mechanism based on spectrum analysis and frequency band migration enhances the anti-interference capability of the power line communication network and strengthens the stability of the communication system.

[0024] 3. This application provides an edge computing-based method for anomaly localization of IoT devices. By establishing test links between neighboring nodes on both sides of the anomaly source device and evaluating link quality based on objective parameters such as signal-to-noise ratio and received signal strength, the attenuation characteristics of the anomaly source device on the communication signal can be accurately identified. When it is confirmed that the anomaly source device forms signal attenuation impedance on a clean carrier frequency band, by establishing a direct communication link bypassing the anomaly source device and increasing the signal transmission power to the maximum power level, the attenuation loss of the signal during transmission can be reduced. This adaptive link reconstruction scheme improves the network's anti-interference capability, reduces signal attenuation in the communication link, and reduces the risk of network segmentation. By using maximum transmission power for compensation at the physical layer, the signal penetration capability is improved, the stability of the communication link is enhanced, and the network has stronger fault adaptability. Attached Figure Description

[0025] Figure 1This is a flowchart illustrating an abnormal location method for IoT devices based on edge computing, as described in an embodiment of this application.

[0026] Figure 2 This is another flowchart illustrating an abnormal location method for IoT devices based on edge computing, as described in this application.

[0027] Figure 3 This is another flowchart illustrating an abnormal location method for IoT devices based on edge computing, as described in this application.

[0028] Figure 4 This is a schematic diagram of the physical device structure of an IoT device anomaly location system based on edge computing provided in an embodiment of this application. Detailed Implementation

[0029] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification and appended claims of this application, the singular expressions “a,” “an,” “the,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to any or all possible combinations including one or more of the listed items.

[0030] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0031] The following example is used in conjunction with Figure 1 This application describes an abnormal location method for IoT devices based on edge computing, as described in the following embodiments: Please see Figure 1 This is a flowchart illustrating an abnormal location method for IoT devices based on edge computing, as described in this application.

[0032] S101. When the edge gateway detects that the proportion of street light terminals in the offline state exceeds a preset ratio, it broadcasts an environmental awareness command to the power line network to control the street light terminals in the online state to suspend the transmission of service data and switch to the interference listening mode. An edge gateway is a device deployed at the network edge, possessing computing, storage, and network communication capabilities. It acts as a bridge between the perception layer and the network layer, responsible for data aggregation, processing, and protocol conversion. Offline status refers to the state where streetlight terminals cannot establish a normal communication connection with the edge gateway or where heartbeat packets are lost. The preset ratio is a threshold parameter set by the system to trigger anomaly detection mechanisms. This ratio can be set according to the actual network scale and fault tolerance rate, for example, 10% or 20%, and is not limited to a specific value. An environmental awareness command is a specific control command used to instruct terminal devices to stop regular services and activate environmental noise monitoring. Interference detection mode refers to a specific operating state where the device turns off the transmitter and only turns on the receiver to measure channel background noise. In this step, the edge gateway continuously monitors the online status of streetlight terminals within its jurisdiction. Once it detects that the proportion of streetlight terminals in an offline state exceeds the preset ratio, the system determines that the current power line communication network may be suffering from severe channel interference. At this time, the edge gateway immediately broadcasts an environmental awareness command to the power line network. Upon receiving this instruction, the online street light terminal will immediately suspend its current business data transmission task, such as stopping the transmission of electricity metering data or status monitoring data, and quickly switch to interference detection mode to prepare for sampling and measuring the background noise on the power line.

[0033] One method for edge gateways to detect offline status and broadcast commands is through heartbeat-based status monitoring. The edge gateway periodically sends heartbeat probe messages to all registered streetlight terminals. If no response is received from a terminal within a preset time window, that terminal is marked as offline. The number of offline terminals is counted, and their proportion of the total number of registered terminals is calculated. When this proportion exceeds a preset threshold, a broadcast mechanism is triggered. Broadcasting environmental awareness commands can be achieved using broadcast frames in the Power Line Carrier Communication (PLC) protocol, ensuring that the commands cover all online nodes. Another implementation method is to utilize polling-based monitoring technology. The edge gateway polls each streetlight terminal sequentially according to the routing table, recording nodes that fail to respond. After one round of polling, the proportion of failed nodes is counted. If the proportion exceeds the threshold, the edge gateway uses a special control code defined by the application layer protocol as an environmental awareness command, which is sent via multicast or broadcast. After parsing the control code, the online terminal calls the underlying driver interface, disables the transmission interrupt, and enables the noise power spectral density scanning function of the receiving channel.

[0034] S102. When the edge gateway determines that it has received the background noise intensity value reported by the street light terminal, it marks the street light terminal that reported the background noise intensity value as a sensing node, and marks other street light terminals other than the sensing node as communication blind zone nodes. Background noise intensity refers to the power level of unwanted signals present on a communication channel when there is no effective signal transmission, usually expressed in decibels per milliwatt (dBm) or signal-to-noise ratio (SNR). A sensing node is a street light terminal that successfully receives environmental sensing commands, completes background noise measurement, and successfully reports the data to the edge gateway. A communication dead zone node is a street light terminal that fails to report background noise intensity under the current network conditions. These terminals may be offline, or although online, their reported data may be lost due to extremely poor channel quality. In this step, the edge gateway opens a receiving window and waits for data reports from street light terminals. When the edge gateway receives the background noise intensity value reported by a street light terminal, it parses and verifies the data source. After successful verification, the edge gateway marks these successfully reporting street light terminals as sensing nodes in its internal system logic. Correspondingly, for those street light terminals that fail to report background noise intensity within a specified time, the edge gateway uniformly marks them as communication dead zone nodes. This classification process is to distinguish which nodes can act as sensors to detect interference sources and which nodes may be the most severely affected victims or potential fault points.

[0035] The specific methods for implementing node labeling and classification can be based on hash tables or bitmap indexing techniques. The edge gateway maintains a status table containing unique identifiers (such as MAC addresses or logical IDs) for all streetlight terminals. When a background noise data packet reported by a terminal is received, its ID is parsed, and the status field corresponding to that ID is updated to "perceived node" in the status table, while the reported noise value is stored. After the receive window closes, the status table is traversed, and all IDs whose status fields have not been updated are marked as "communication blind zone nodes." Another implementation method utilizes database transaction processing technology. When the edge gateway receives data, it triggers a database update operation, inserting the terminal ID that reported the data into the perception node table. After the data collection phase is completed, a set difference operation is performed—that is, the records in the perception node table are subtracted from the total terminal table—to obtain the result set of "communication blind zone nodes," and these nodes are then inserted in batches into the "communication blind zone node table," completing the classification and labeling.

[0036] S103. The edge gateway maps the background noise intensity value of the sensing node to the corresponding node position in the electrical topology diagram, and determines the pair of sensing nodes that are adjacent in electrical connection in the electrical topology diagram. An electrical topology diagram is a graphical model that describes the physical connections and hierarchical structure between various devices such as streetlights, branch boxes, and transformers in a power line network. It typically includes nodes (devices) and edges (power lines). Mapping refers to the process of associating logical data (background noise intensity values) with specific locations in the physical or logical model (electrical topology diagram). Electrically adjacent nodes refer to two nodes in the power line network structure that have a direct physical line connection without passing through other devices marked as sensing nodes, or that are parent-child or sibling relationships in the logical topology tree and are physically the closest. A sensing node pair is a set of data units consisting of two sensing nodes that are electrically adjacent, used for subsequent differential analysis. In this step, the edge gateway retrieves the electrical topology diagram data stored in its local database. Then, based on the unique identifier of each sensing node, the edge gateway fills in the corresponding node positions in the topology diagram with the reported background noise intensity values. After mapping is complete, the edge gateway traverses the topology graph and, based on an adjacency search algorithm in graph theory, identifies all sensing nodes that are directly adjacent in electrical connection and pairs them up to form sensing node pairs. This step is a crucial step in transforming one-dimensional discrete measurement data into two-dimensional network spatial distribution data.

[0037] The specific method for implementing background noise mapping and neighbor node pair determination can be to use graph data structures (such as adjacency matrices or adjacency lists) to represent the electrical topology. The edge gateway first loads the electrical topology into an adjacency list in memory. Then, it traverses the received noise data list, finds the corresponding vertex in the adjacency list based on the node ID, and assigns the noise value as the weight attribute of that vertex. Next, it traverses each vertex in the adjacency list and checks all its adjacent vertices. If a vertex and its adjacent vertices both have valid noise value weights (i.e., both are sensing nodes), then these two vertices are stored as a "sensing node pair" in a list. Another implementation method is based on the combination of Geographic Information System (GIS) and topology data. If the topology map contains geographic coordinate information, the edge gateway can use spatial indexing algorithms (such as R-trees) to assist in positioning. First, the noise value is associated with point features on the GIS layer based on the ID. Then, using a topology tracing algorithm, starting from each sensing node, the nearest neighbor node is searched along the power line path. If the searched neighbor is also a sensing node, a pairing relationship is established. This approach is more intuitive and accurate when dealing with complex real-world cabling, such as ring networks or irregular branches.

[0038] S104. The edge gateway compares the background noise intensity values ​​of each sensing node pair in turn to determine the direction of the interference signal enhancement on the power line. The edge gateway sequentially compares the background noise intensity values ​​of each sensing node pair to determine the direction of interference signal enhancement on the power line. Specifically, this includes: the edge gateway calculating the absolute value of the difference between the background noise intensity values ​​of the first and second sensing nodes in the sensing node pair; if the edge gateway determines that the absolute value of the difference is greater than a preset effective fluctuation threshold, it identifies the node with the larger background noise intensity value as the high-potential node and the node with the smaller background noise intensity value as the low-potential node; the edge gateway determines the direction from the low-potential node to the high-potential node as the direction of interference signal enhancement on the power line.

[0039] Interference signals refer to unexpected electromagnetic signals transmitted on power lines that affect the quality of normal communication. The enhancement direction refers to the spatial trend of the interference signal power gradually increasing in the power line network, usually pointing towards the location of the interference source. The preset effective fluctuation threshold is a numerical standard used to filter measurement errors and minor environmental fluctuations; a significant interference gradient is considered to exist only when the noise difference between two nodes exceeds this threshold. High-potential nodes and low-potential nodes here borrow the concept of potential, referring to nodes with higher and lower background noise intensity values, respectively. In this step, the edge gateway analyzes each pair of sensing nodes determined in S103. Specifically, the edge gateway calculates the absolute value of the difference in background noise intensity values ​​between the first and second sensing nodes in each pair. Then, it compares this absolute value with the preset effective fluctuation threshold. If the absolute value of the difference is less than or equal to the threshold, the gradient information of the node pair is ignored; if it is greater than the threshold, it is determined that there is a significant change in interference between the two points. At this time, the edge gateway compares the noise values ​​of the two nodes, marking the node with the larger value as a high-potential node and the node with the smaller value as a low-potential node. Finally, based on the inverse logic of signal attenuation in physics (the closer to the interference source, the stronger the noise), it was determined that the direction of the amplification of the interference signal on the electric power line is from the low-potential node to the high-potential node.

[0040] One specific method for determining the direction of interference enhancement is to use a numerical comparison and logical judgment algorithm. The edge gateway traverses the list of sensing node pairs. For each pair of nodes (A, B), it reads their noise values ​​Noise(A) and Noise(B). It calculates Delta = |Noise(A) - Noise(B)|. If Delta > Threshold, it further determines: if Noise(A) > Noise(B), the direction is B->A; if Noise(B) > Noise(A), the direction is A->B. The determined direction information is stored as a directed edge attribute in the topology data structure. Another implementation method is to utilize fuzzy logic reasoning. Considering the potential fluctuations in noise measurements, the noise difference can be fuzzified into linguistic variables such as "no significant difference," "slight difference," and "significant difference." A fuzzy rule base is defined, for example, "if the difference is significant and the noise of node A is much greater than that of node B, then the direction strongly points to A." The fuzzy inference engine calculates the confidence level of the direction, and the direction with high confidence is selected as the final enhancement direction. This method has better robustness to measurement noise.

[0041] S105. The edge gateway generates an interference gradient vector based on the enhancement direction; The edge gateway generates an interference gradient vector based on the enhancement direction. Specifically, the edge gateway determines the enhancement direction as the vector direction of the interference gradient vector; the edge gateway calculates the difference in background noise intensity values ​​between the sensing node pairs that form the enhancement direction, and determines the difference as the vector magnitude of the interference gradient vector; the edge gateway combines the vector direction and the vector magnitude to generate an interference gradient vector that describes the rate of change of noise space between the sensing node pairs.

[0042] The interference gradient vector is a physical quantity that includes both direction and magnitude, used to describe the rate of change of the interference signal in the power line network space. The vector direction is the direction of interference signal enhancement determined in the previous step. The vector magnitude refers to the size of the interference gradient vector, which is defined here as the difference in background noise intensity values ​​between the sensing node pairs forming the enhancement direction. The rate of change of noise space reflects how quickly the interference signal decays or enhances with distance. In this step, the edge gateway directly uses the enhancement direction determined in S104 as the directional component of the interference gradient vector. At the same time, the edge gateway extracts the difference in background noise intensity values ​​between the sensing node pairs (i.e., the noise value of the high-potential node minus the noise value of the low-potential node), and uses this difference as the magnitude of the vector. Finally, the edge gateway combines the direction and magnitude to generate a complete interference gradient vector. This vector not only tells the system the direction of the interference source, but also implies the distance from the interference source or the intensity of the interference source through the magnitude—generally, the closer to the interference source, or the greater the power of the interference source, the more significant the noise difference (gradient) between adjacent nodes may be.

[0043] One method to generate the interference gradient vector is through vector algebra. In the coordinate system of the electrical topology graph (if virtual coordinates exist), let node A have coordinates of (x1, y1) and node B have coordinates of (x2, y2), with the direction determined as A->B. The unit direction vector d can be calculated by normalizing the coordinate difference. The magnitude m = Noise(B) - Noise(A). Then the interference gradient vector V = m*d. This vector is stored in the edge attributes of the topology graph. Another implementation method is to use a weighted directed graph model. The edge gateway constructs a directed graph corresponding to the electrical topology. For each determined enhancement direction, a directed edge is added to the graph. The calculated noise difference is used as the weight of the directed edge. In this way, the entire network is abstracted as a weighted directed graph, where the direction of the edge represents the interference search path, and the weight of the edge represents the degree of interference enhancement on that path. This representation facilitates the direct application of path search algorithms in graph theory.

[0044] S106. In the electrical topology diagram, the edge gateway traces back along the interference gradient vector until it locates the topology convergence area that all interference gradient vectors point to. Path backtracking refers to the process of traversing a directed graph or tree structure upstream along a specific direction (in this case, the direction indicated by the interference gradient vector) to find the starting point of a path. A topology convergence region refers to a local area in the electrical topology graph where multiple interference gradient vector paths eventually converge or point to each other. This region typically contains one or more nodes and is considered the most likely location of the interference source. In this step, the edge gateway operates on the electrical topology graph, which has already been labeled with interference gradient vectors. The algorithm starts from a low-noise node at the network edge, or from any node with a gradient vector, and strictly hops along the direction indicated by the interference gradient vector (i.e., the direction of noise amplification). Since the interference signal attenuates outward from the interference source in the network, the reverse gradient vectors will inevitably converge from all directions to the same central point or region, like rivers flowing into the sea. The edge gateway repeatedly performs this backtracking process until it finds that the path no longer extends, or that multiple paths converge at a node or a group of closely connected nodes; this location is then identified as the topology convergence region.

[0045] Path backtracking and convergence region localization can be achieved using variations of Depth-First Search (DFS) or Breadth-First Search (BFS). The edge gateway traverses all nodes with out-degrees (i.e., gradient vectors pointing to other nodes) along directed edges. During traversal, the number of times each node is visited (in-degree count) is recorded. At the end of the traversal, the nodes with the highest in-degree and zero out-degree (or whose out-degree points to nodes with noise values ​​lower than their own, forming local maxima) are the convergence points. Another implementation is a particle swarm optimization algorithm based on fluid dynamics simulation. Each sensing node is considered a particle, and the disturbance gradient vector is considered a force field acting on the particle. The trajectory of the particle under the force field is simulated. All particles eventually stop at the point of highest potential energy (i.e., the point of strongest noise), and this point and its neighborhood are the topological convergence region. This method can intuitively identify convergence regions through particle aggregation density when dealing with complex mesh topologies.

[0046] S107. The edge gateway determines whether there are communication blind spot nodes within the topology aggregation area; This step is a logical judgment process designed to select different anomaly source determination strategies based on the node composition within the topology convergence area determined in S106. The topology convergence area may contain one node or several electrically connected nodes. As previously mentioned, communication blind spot nodes are nodes that fail to report data. In this step, the edge gateway checks the list of all streetlight terminal nodes within the topology convergence area. The edge gateway queries the status markers of these nodes one by one (i.e., marked as "sensing node" or "communication blind spot node" in S102). The edge gateway determines whether there is at least one device marked as a "communication blind spot node" within the area. The logic behind this judgment is that if the interference source itself is a faulty streetlight terminal (e.g., a damaged power module causing strong interference), this terminal often cannot communicate due to its own malfunction or being located at the center of interference, thus manifesting as a communication blind spot node.

[0047] S108. The edge gateway identifies the communication blind zone node located at the geometric center of the topology convergence area as the abnormal source device. If present, the edge gateway identifies the communication blind spot node located at the geometric center of the topology convergence area as the source of the abnormality. This step corresponds to the case where the judgment result in S107 is "present". The geometric center refers to the location center of a group of nodes in the electrical topology or physical space. The source of the abnormality refers to the specific physical device that is ultimately determined to cause interference and lead to network abnormalities. In this step, since there are communication blind spot nodes in the topology convergence area, the system tends to assume that the source of interference is hidden in these nodes that cannot communicate. The edge gateway first filters out all communication blind spot nodes in the area. If there is only one blind spot node in the area, it is directly identified as the source of the abnormality. If there are multiple blind spot nodes, the edge gateway calculates the geometric center location of these blind spot nodes in the electrical topology. Usually, the device located at the center of the interference has the greatest impact and its own communication is most severely blocked. Therefore, the edge gateway identifies the communication blind spot node located at this geometric center location (or closest to the geometric center) as the source of the abnormality.

[0048] One method for determining the geometric center and locking the device is through topological distance matrix calculation. The edge gateway constructs a distance matrix for all blind zone nodes within the aggregation area and calculates the sum of the topological distances (i.e., the sum of hop counts) from each node to all other blind zone nodes. The node with the smallest sum of distances is the topological geometric center (the node with the highest ClosenessCentrality) and is identified as the source of the anomaly. Another implementation method is based on weighted centroid calculation using physical coordinates. If the physical latitude and longitude of the nodes are available, the edge gateway calculates the average of the coordinates of all blind zone nodes to obtain the centroid coordinates. Then, it finds the blind zone node with the closest Euclidean distance to these centroid coordinates and locks it as the source device of the anomaly.

[0049] S109. The edge gateway identifies the sensing node with the highest background noise intensity value in the topology aggregation area as the abnormal source device.

[0050] If not, the edge gateway identifies the sensing node with the highest background noise intensity value within the topology convergence area as the abnormal source device. This step corresponds to the case in S107 where the judgment result is "not present," meaning all nodes within the topology convergence area can communicate normally and report noise values. In this case, the interference source may be a device with strong radiated interference that has not completely cut off its own communication, or the interference source may be located on the power line in the area rather than the device itself (but is usually attributed to the closest device). In this step, the edge gateway directly analyzes the background noise intensity value of each sensing node within the topology convergence area. Since this area is the endpoint of gradient backtracking, it theoretically contains the noisiest node in the entire network. The edge gateway sorts the background noise intensity values ​​of all sensing nodes in the area and finds the one with the highest value. Based on the physical law that "the noise is strongest at the interference source," the system identifies the sensing node with the highest background noise intensity value as the abnormal source device.

[0051] One method for identifying the noisiest node is a simple sorting algorithm. The edge gateway extracts the noise values ​​of all nodes within the topology aggregation area and stores them in an array or list. The array is then sorted in descending order using quicksort or bubble sort. The node ID corresponding to the first element of the array is taken as the source device of the anomaly. Another approach is outlier detection based on statistical outliers. Although the area contains many high-noise nodes, the noise value of the source of the anomaly is usually significantly higher than that of neighboring nodes. The edge gateway calculates the mean and standard deviation of the noise values ​​within the area and checks if any node's noise value exceeds "mean + k times the standard deviation". If it does, the outlier node is identified; if no significant outlier exists (i.e., multiple nodes have similar noise levels), the common connection point (such as a branch box) in the area may be identified as the source of the anomaly, or it may be reported as a regional anomaly awaiting manual investigation.

[0052] In the above embodiments, by controlling the switch of the line light terminal to interference detection mode and reporting the background noise intensity value through the edge gateway, and mapping the background noise intensity value of the sensing node to the corresponding location in combination with the electrical topology diagram, the spatial distribution characteristics of the interference signal in the power line network can be obtained. Based on the comparison of the background noise intensity values ​​of adjacent sensing node pairs, the direction of interference signal enhancement is determined, and an interference gradient vector is generated for path backtracking, enabling rapid local location of the interference source in a complex tree-like topology. By determining whether there are communication blind spots within the topology convergence area, and combining the geometric center location or the maximum background noise intensity value to determine the abnormal source device, the positioning accuracy of abnormal devices in the power line communication network is improved, and the manual inspection workload required for abnormal device positioning is reduced. This method fully utilizes the local computing power of the edge gateway and the topological characteristics of the power line communication network, reduces the dependence of abnormal device positioning on cloud computing resources, and improves the real-time performance of abnormal positioning.

[0053] After locating the source of the anomaly, appropriate interference avoidance measures are needed to reduce its impact on the entire communication network. Frequency domain characteristic analysis of the interference signals generated by the source device can provide a basis for optimizing and adjusting communication frequency bands. The following section combines... Figure 2 Another method for anomaly location of IoT devices based on edge computing is described in the embodiments of this application: Please see Figure 2 This is another flowchart illustrating an abnormal location method for IoT devices based on edge computing, as described in this application.

[0054] S201. The edge gateway sends a spectrum sampling command to the sensing node located upstream of the electrical connection of the abnormal source device, and receives the background noise spectrum data fed back by the sensing node. The electrical upstream of the anomaly source device refers to one or more nodes in the power line communication network topology that are located between the anomaly source device and the edge gateway, and are the closest to the anomaly source device in terms of physical line or logical route. Sensing nodes here specifically refer to street light terminals that have spectrum analysis capabilities, are in normal communication status, and can respond to gateway commands. A spectrum sampling command is a higher-layer control message used to instruct the terminal to initiate lower-level analog-to-digital conversion and fast Fourier transform functions to capture the analog signal on the current channel and convert it into frequency domain data. Background noise spectrum data refers to the distribution curve data describing the noise power of the power line channel as a function of frequency during periods without effective service data transmission; it typically includes frequency point indices and corresponding amplitude values. In this step, after locating the anomaly source device, the edge gateway needs to acquire channel environment data near it to further analyze the interference characteristics generated by the device. Since the anomaly source device itself may be faulty or in an extremely unstable state and unable to reliably report data, the edge gateway chooses to send a command to its "upstream" sensing node. This upstream node is very close to the anomaly source in electrical distance, and the line noise it receives can highly reproduce the characteristics of the interference signal generated by the anomaly source. After receiving the instruction, the sensing node suspends its own transmission task, listens to and samples the channel, generates background noise spectrum data, and feeds it back to the edge gateway.

[0055] The specific methods for enabling edge gateways to send commands and receive spectrum data can employ a request-response mechanism based on the Simple Network Management Protocol (SNMP) or a custom application layer protocol. The edge gateway constructs a Get-Request message containing a specific object identifier (OID), which corresponds to the spectrum analysis module of the terminal device. The edge gateway searches its routing table for the optimal path directly to the upstream sensing node and unicasts the message. After parsing the message, the sensing node calls the underlying hardware driver interface to activate the analog-to-digital converter (ADC) to acquire a power line analog signal for a certain duration at a high sampling rate. Subsequently, it uses a digital signal processing unit (DSP) to execute a Fast Fourier Transform (FFT) algorithm to convert the time-domain waveform into power spectral density (PSD) data in the frequency domain. The sensing node encapsulates the calculated PSD data in a Get-Response message and sends it back to the edge gateway via the power line network. Another implementation method utilizes the channel estimation function built into the Orthogonal Frequency Division Multiplexing (OFDM) communication system. The edge gateway sends a channel sounding frame of a specific format to the sensing node. During the reception of the frame, the sensing node calculates the signal-to-noise ratio (SNR) spectrum or noise power spectrum of the current channel using the pilot subcarriers. Instead of directly performing raw sampling, the sensing node extracts the most recent channel estimation result stored in the physical layer (PHY) register and packages it as background noise spectrum data for uploading. This approach utilizes the existing functionality of the communication chip, reducing additional computational overhead.

[0056] S202. The edge gateway identifies frequency bands in the background noise spectrum data whose energy amplitude exceeds a preset interference threshold as interference fingerprint features. Energy amplitude refers to the signal strength value corresponding to each discrete frequency point or subcarrier in the background noise spectrum data, usually measured in decibels (dB) or microvolts (µV). The preset interference threshold is a criterion value used to distinguish normal background thermal noise from abnormal interference signals. This threshold can be set based on the statistical average of historical environmental noise plus a protection margin, and is not limited to a fixed value. Interference fingerprint characteristics refer to a set of continuous or discrete frequency intervals. These intervals represent the frequency range where abnormal interference signals mainly exist, constituting the unique frequency domain "fingerprint" of the interference source. In this step, the edge gateway analyzes the received background noise spectrum data point-by-point or segment-by-segment. The edge gateway traverses each frequency point in the spectrum data, reads its energy amplitude, and compares this amplitude with the preset interference threshold. If the energy amplitude of a frequency point exceeds the threshold, the frequency point is considered to be interfered with. The edge gateway merges all continuous frequency points exceeding the threshold into a single frequency band interval, or records all discrete frequency points exceeding the threshold. Ultimately, these identified high-energy frequency bands, when combined, are determined to be the interference fingerprint characteristics of the anomalous source device. This step is crucial in transforming the raw data into high-level features that can be used for decision-making.

[0057] One specific method for determining interference fingerprint features is to employ a sliding window-based smoothing and threshold decision algorithm. Since the original spectrum data may contain random spikes, the edge gateway first smooths the spectrum curve using a moving average filter or a Gaussian filter to eliminate the influence of high-frequency random noise. After processing, the edge gateway sets a constant power threshold (e.g., -80dBm). The algorithm scans from low to high frequencies. Once the amplitude curve crosses the threshold upwards, it is recorded as the "interference start frequency"; when the amplitude curve crosses the threshold downwards, it is recorded as the "interference end frequency." This pair of start and end frequencies constitutes an interference frequency band. This process is repeated until the entire spectrum is scanned, and the set of all recorded frequency bands constitutes the interference fingerprint feature. Another implementation method is to use a constant false alarm rate (CFAR) detection algorithm. Considering that the background noise level may vary in different frequency bands, using a single fixed threshold may be inaccurate. The edge gateway uses the CFAR algorithm to dynamically calculate an adaptive threshold suitable for the current local frequency band based on the average noise power of the reference cells surrounding the unit to be detected. The amplitude of each frequency point is compared with the adaptive threshold corresponding to that point; if it exceeds the threshold, it is determined to be an interference point. This method can more accurately extract significant interference signal features from fluctuating noise backgrounds.

[0058] S203. The edge gateway identifies clean carrier frequency bands that do not overlap with interference fingerprint features in the frequency domain among the multiple communication carrier frequency bands supported by the current system. Among the multiple communication carrier frequency bands supported by the current system, the edge gateway identifies clean carrier frequency bands that do not overlap with interference fingerprint features in the frequency domain. Specifically, this includes: the edge gateway calculating the overlap bandwidth ratio between each communication carrier frequency band and the interference fingerprint features in the frequency domain; if there is a communication carrier frequency band with an overlap bandwidth ratio of zero, the edge gateway identifies the communication carrier frequency band with the zero overlap bandwidth ratio as a clean carrier frequency band; if there is no communication carrier frequency band with an overlap bandwidth ratio of zero, the edge gateway calculates the average noise energy density within each communication carrier frequency band; and the edge gateway identifies the communication carrier frequency band with the lowest average noise energy density as a clean carrier frequency band.

[0059] A communication carrier frequency band refers to the set of frequency ranges defined by the system protocol that allow devices to transmit data, such as the CENELEC A-band and B-band defined in power line communication, or different sub-bands in OFDM technology. A clean carrier frequency band refers to the candidate frequency band that is least affected by interference and has the best expected communication quality under the current interference environment. The overlap bandwidth ratio refers to the percentage of the bandwidth length of the communication carrier frequency band that overlaps with the interference fingerprint features on the frequency axis, relative to the total bandwidth length of the communication carrier frequency band. The average noise energy density refers to the integral or sum of the noise energy amplitudes at all frequency points within a specific communication carrier frequency band, divided by the bandwidth of that frequency band, reflecting the average noise floor level within that frequency band. In this step, the edge gateway needs to find a new refuge for the network. First, the edge gateway obtains a list of all candidate communication carrier frequency bands supported by the system. Then, the frequency range of each candidate frequency band is compared with the interference fingerprint features (i.e., the set of interference frequency bands) extracted in S202. The edge gateway calculates the degree of overlap between each candidate frequency band and the interference frequency band. The logical judgment is divided into two levels: The first level prioritizes finding frequency bands that are "perfectly avoided," i.e., frequency bands with a zero overlap bandwidth. If such a frequency band exists, it is directly selected as the clean carrier band. The second level, if all candidate frequency bands are more or less affected by interference (i.e., there is no case with a zero overlap ratio), then as a second-best option, the average noise energy density within each frequency band is calculated. The band with the lowest overall noise level despite interference is then identified as the clean carrier band.

[0060] The specific method for calculating the overlap bandwidth ratio and selecting clean frequency bands can be achieved using an interval intersection algorithm. The edge gateway represents each communication carrier frequency band as an interval [Start_C, End_C], and each interference frequency band in the interference fingerprint features as an interval [Start_I, End_I]. For a given carrier frequency band, all interference frequency bands are traversed, and the length of the interval intersection is calculated. All intersection lengths are summed to obtain the total overlap bandwidth. The overlap ratio = total overlap bandwidth (End_C - Start_C). If the ratio is 0, it is marked as a candidate clean frequency band. If the ratios of all frequency bands are greater than 0, the process proceeds to the second stage. The second stage uses a numerical integration method. The edge gateway calls the original spectrum data, and for each carrier frequency band interval, it sums the power amplitude of all sampling points within that interval and divides it by the number of sampling points to obtain the average noise energy density. Finally, a sorting algorithm is used to select the frequency band corresponding to the minimum value. Another implementation method is based on fast bitmap computation. The entire communication spectrum is discretized into several small frequency particles (Bin), each Bin corresponding to one bit. Construct an "interference mask bitmap," setting the Bin of the area covered by the interference fingerprint to 1 and the rest to 0. Construct a "carrier band bitmap," setting the Bin of the area covered by that band to 1. Quickly determine if there is overlap using a bitwise AND operation (if the result is all 0, there is no overlap). If there is overlap, quickly calculate the total noise energy within any frequency band using a pre-calculated spectrum integral table (PrefixSum Array), then obtain the density and compare it.

[0061] S204. The edge gateway broadcasts a frequency band migration command containing configuration parameters of the clean carrier frequency band to the power line network to control street light terminals other than abnormal source devices to switch to the clean carrier frequency band for communication.

[0062] Configuration parameters refer to a series of settings required for a device to switch to a new frequency band, including center frequency, bandwidth, modulation scheme, and tone mask. A frequency band migration command is a network management command used to trigger the reconfiguration of physical layer parameters for all devices in the network or a specified group. Streetlight terminals other than the source of the abnormality refer to all normal nodes in the network that need to maintain communication capabilities, excluding the faulty device identified as an interference source. This is to prevent the faulty device from continuing to pollute the new frequency band during the switch, or to prevent the faulty device itself from becoming uncontrollable and unable to receive commands. In this step, the edge gateway encapsulates the detailed parameters of the clean carrier frequency band selected in S203 into the frequency band migration command. Then, the edge gateway broadcasts this command through the power line network. To ensure reliable transmission of the command, a highly redundant robust modulation mode may be used. Normal streetlight terminals that receive the command will parse the new frequency band parameters and perform a switchover at the predetermined effective time or immediately, adjusting their communication modules to operate on the new clean carrier frequency band. In this way, the entire network (except for the point of failure) is collectively "relocated" to a relatively quiet spectrum space, thereby avoiding strong interference from the source of the anomaly.

[0063] One method for implementing frequency band migration command broadcasting and handover is a synchronous handover mechanism based on beacon frames. The edge gateway modifies the content of periodically sent beacon frames, adding "frequency band migration countdown" and "new frequency band configuration information" fields. All online street light terminals, upon detecting the beacon frame, are aware that a frequency band migration is imminent. When the countdown reaches zero (e.g., after the Nth superframe), all terminals and the gateway synchronously adjust the parameters of the physical layer frequency synthesizer, switching to the clean carrier frequency band. This method utilizes the broadcast characteristics of beacon frames and the network-wide synchronization mechanism to ensure a smooth handover process. Another implementation method is a two-phase commit protocol. The edge gateway first broadcasts a "frequency band migration preparation" command, requiring all terminals to acknowledge and respond with an ACK. After receiving acknowledgments from the vast majority of terminals, the edge gateway then broadcasts a "execute handover" command. Upon receiving the execution command, the terminal performs a frequency band switch. If a large number of terminals do not respond during the preparation phase, the gateway can cancel the handover or adopt a batch handover strategy. For devices that are the source of the anomaly, the gateway can explicitly specify their ID in the command through a blacklist mechanism, or take advantage of the fact that the device is in a blind spot and cannot receive commands, thus leaving it in the old frequency band.

[0064] In the above embodiments, by collecting background noise spectrum data from upstream sensing nodes of the anomaly source device and extracting interference fingerprint features, the frequency domain distribution information of the interference signal can be obtained. Among the multiple communication carrier frequency bands supported by the system, clean carrier frequency bands that do not overlap with the interference fingerprint features are identified, and other devices are controlled to switch to these frequency bands for communication, improving the spectrum utilization efficiency of the power line communication network. This method reduces the impact of abnormal devices on the overall communication quality of the network by performing frequency domain analysis on the interference signal and implementing communication frequency band migration. The interference avoidance mechanism based on spectrum analysis and frequency band migration enhances the anti-interference capability of the power line communication network and strengthens the stability of the communication system.

[0065] After migrating the communication frequency band to a clean carrier band, it is still necessary to assess the impact of the anomalous source device on the communication link and take targeted link optimization measures. Since anomalous devices may exhibit signal attenuation characteristics on clean carrier bands, this application also provides another method for anomaly localization of IoT devices based on edge computing. The following section combines... Figure 3 This application describes yet another method for anomaly location of IoT devices based on edge computing: Please see Figure 3 This is another flowchart illustrating an abnormal location method for IoT devices based on edge computing, as described in this application.

[0066] S301, The edge gateway controls the first neighboring node located upstream of the electrical connection of the abnormal source device to send a test data packet to the second neighboring node located downstream of the electrical connection of the abnormal source device; The first neighboring node refers to the communication terminal located closer to the edge gateway in the electrical connection direction of the anomalous source device within the physical topology of the power line communication network. This is typically the closest working street light controller or relay node to the anomalous device. The second neighboring node refers to the communication terminal located further away from the edge gateway in the electrical connection direction of the anomalous source device, i.e., the downstream node of the anomalous device. The test data packet is a data frame with a specific bit sequence and length, generated by the edge gateway, used to probe the physical state of the communication link. This data packet does not carry actual service control information and is only used for link quality assessment. Upstream and downstream in the electrical connection are relative concepts defined based on the logical or physical location of the signal transmission path relative to the central control point (edge ​​gateway). In this step, the edge gateway has completed the initial location of the anomalous source device and has migrated the network to a clean carrier frequency band. To further confirm whether the anomalous source device is causing physical signal blockage or severe attenuation on the line, the edge gateway needs to initiate an active probe. The edge gateway sends a control command to the first neighboring node, instructing it to act as a signal transmitter. The first neighboring node responds to the command, constructs a test data packet, and sends it via the power line medium to the second neighboring node located on the other side of the anomalous device.

[0067] One method for sending test data packets is a directional diagnostic frame mechanism based on the Media Access Control (MAC) layer. The edge gateway sends a diagnostic request command containing the MAC address of the second neighboring node to the first neighboring node. Upon receiving the request, the communication chip of the first neighboring node calls the underlying physical layer interface to generate a channel estimation frame or sounding frame conforming to power line communication protocol standards (such as IEEE 1901.1 or G3-PLC). This frame contains a known pilot sequence and pseudo-random binary sequence (PRBS) payload. The first neighboring node uses a preset transmit power and a robust modulation scheme (such as BPSK) to unicast the diagnostic frame to the second neighboring node. This method utilizes the link testing function built into the communication protocol stack, requiring no upper-layer application encapsulation, resulting in high execution efficiency. Another implementation method is to use an application layer echo test mechanism. The edge gateway sends an application layer command to the first neighboring node, requesting it to send a specific number (e.g., 100) of custom application layer data packets to the second neighboring node. The first neighboring node constructs a data packet containing a sequence number and a timestamp at the application layer and routes it to the second neighboring node through the network layer.

[0068] S302. The edge gateway receives link quality parameters for the test data packet fed back by the second neighboring node. The link quality parameters include the signal-to-noise ratio and the received signal strength. Link quality parameters refer to a set of physical quantities used to quantitatively evaluate the transmission performance of power line communication channels. Signal-to-noise ratio (SNR) is the ratio of received useful signal power to background noise power, usually measured in decibels (dB), reflecting the clarity of the signal in a noisy environment. Received signal strength (RSSI) is the absolute power level of the radio wave or carrier signal received by the receiving node at its antenna or coupling interface, usually measured in dBm or dBuV, directly reflecting the remaining energy of the signal after passing through the transmission medium. In this step, the second neighboring node, acting as the receiver, captures the test data packet sent by the first neighboring node while continuously listening to the channel. The physical layer demodulation module of the second neighboring node calculates the current signal strength and SNR in real time when processing the preamble and pilot symbols of this data packet. The second neighboring node extracts these raw measurement data and encapsulates them in a feedback message. This feedback message is sent back to the edge gateway through the power line network. The edge gateway receives and parses the message to obtain accurate quality data for the section of the line crossing the source device of the anomaly.

[0069] One specific method for implementing link quality parameter feedback is a piggyback transmission mechanism based on acknowledgment frames (ACK). After the second neighboring node successfully demodulates the test data packet, it needs to reply with an ACK frame to the first neighboring node according to the communication protocol. The second neighboring node directly fills the calculated SNR and RSSI values ​​into the reserved fields or specific link state extension fields of the ACK frame. Upon receiving the ACK frame with link parameters, the first neighboring node parses and summarizes it, then reports it uniformly to the edge gateway. This method utilizes the existing communication handshake process, eliminating the need to establish a separate feedback connection and saving channel resources. Another implementation method is to use an independent Management Information Base (MIB) query mechanism. When receiving test data packets, the second neighboring node stores the measured SNR and RSSI values ​​in its local MIB register and updates the statistical values ​​(such as average and minimum values). After a predetermined delay following the sending of the test command, the edge gateway sends an SNMP Get request or a proprietary read register command to the second neighboring node to directly read the link quality data stored in the second neighboring node. This method allows the edge gateway to obtain more detailed historical statistical information, not just instantaneous values ​​from a single measurement.

[0070] S303. When the edge gateway determines that the signal-to-noise ratio is greater than the preset net channel threshold and the received signal strength is less than the preset attenuation threshold, it determines that the abnormal source device has formed a signal attenuation impedance on the clean carrier frequency band and sends a bridging configuration command to the first neighboring node and the second neighboring node.

[0071] When the edge gateway determines that the signal-to-noise ratio is greater than a preset net channel threshold and the received signal strength is less than a preset attenuation threshold, it determines that the abnormal source device has formed a signal attenuation impedance on the clean carrier frequency band. The edge gateway then sends a bridging configuration command to the first and second neighboring nodes. This bridging configuration command controls the first and second neighboring nodes to establish a direct communication link bypassing the abnormal source device and adjusts the signal transmission power of the direct communication link to the maximum power level supported by the device. The bridging configuration command also includes configuration information for the physical layer modulation parameters of the direct communication link, specifically: the edge gateway locks the data transmission rate of the direct communication link to a preset minimum transmission rate level; the edge gateway adjusts the forward error correction coding redundancy of the direct communication link to a preset maximum redundancy level; and the edge gateway prohibits the first and second neighboring nodes from dynamically adjusting the rate during data interaction on the direct communication link.

[0072] The preset net channel threshold is a reference value used to judge the background noise level of the channel. When the signal-to-noise ratio is higher than this threshold, it means that the background noise is low and the channel is relatively clean. The preset attenuation threshold is a reference value used to judge signal transmission loss. When the received signal strength is lower than this threshold, it means that the signal has experienced severe attenuation during transmission. Signal attenuation impedance refers to the high-frequency signal bypassing or absorption characteristics exhibited on the power line due to internal component failures (such as capacitor breakdown, coil short circuits, etc.) of the abnormal source device. It does not generate active noise, but it acts like a huge resistor or filter, swallowing the passing carrier signal. Bridging configuration commands are a complex set of control commands issued by the edge gateway to establish a special, highly reliable communication channel. A direct communication link refers to a point-to-point connection established directly between the first and second nearest nodes, logically ignoring intermediate nodes. Physical layer modulation parameters include modulation scheme (such as BPSK, QPSK, 16QAM), coding rate, carrier mode, etc. The preset minimum transmission rate level usually refers to the most robust modulation and coding scheme (MCS), such as the ROBO mode. The preset maximum redundancy level refers to adding the maximum proportion of parity bits to the forward error correction (FEC) coding in exchange for the highest error correction capability.

[0073] In this step, the edge gateway performs logical judgments on the parameters acquired by S302. If the signal-to-noise ratio is high (indicating no noise) but the signal strength is low (indicating signal loss), this not only rules out the possibility of noise interference but also directly points to the physical phenomenon of "signal being swallowed up," meaning that the abnormal device has created signal attenuation impedance. After confirming this fault mode, the edge gateway decides to take extreme measures to restore the link. The edge gateway sends instructions to the neighboring nodes at both ends, forcing them to establish a direct connection. To combat the huge attenuation, the instructions require the nodes to maximize their transmit power. At the same time, to ensure that weak signals can be demodulated, the edge gateway locks the transmission rate at the lowest level, which means that the energy of each bit is maximized; it sets the error correction redundancy to the highest level, meaning that even if there are many errors in the received bits, they can be corrected; and it prohibits dynamic rate adjustment to prevent the device from trying to increase speed during signal fluctuations, which could lead to connection interruption. This is a combination of measures designed to sacrifice bandwidth in exchange for connectivity under extremely severe attenuation conditions.

[0074] One way to implement refined technical solutions (bridging configuration and physical layer parameter adjustment) is through configuration file-based mode switching. The firmware of communication nodes pre-loads various communication profiles, including one for "disaster recovery mode" or "strong attenuation resistance mode." This profile hard-coded the maximum transmit power, BPSK modulation, 1 / 2 or 1 / 4 rate Turbo / LDPC codes, and the setting to disable the automatic rate adaptation algorithm. The edge gateway only needs to send a command containing the configuration file ID, and the first and second neighboring nodes can load the configuration, instantly completing the switching of all physical layer parameters. Another implementation method is item-by-item configuration based on physical layer registers. The edge gateway sends a series of write register operation commands to the first and second neighboring nodes via a remote management protocol. Specifically, this includes: writing to the power control register to set the gain value to maximum (e.g., 0xFF); writing to the modulation control register to lock the modulation mode to BPSK; writing to the FEC control register to set the repetition code to 4 times; and writing to the link adaptation control register to set the "Disable Auto-Rate" flag. After receiving and executing all write operations, the node restarts the physical layer state machine to take effect. This approach offers greater flexibility, allowing the gateway to fine-tune various parameters based on the actual attenuation level.

[0075] In the above embodiments, by establishing test links between adjacent nodes on both sides of the anomalous source device and evaluating link quality based on objective parameters such as signal-to-noise ratio and received signal strength, the attenuation characteristics of the anomalous source device on the communication signal can be accurately identified. When it is confirmed that the anomalous source device forms signal attenuation impedance on a clean carrier frequency band, by establishing a direct communication link bypassing the anomalous source device and increasing the signal transmission power to the maximum power level, the attenuation loss of the signal during transmission can be reduced. This adaptive link reconstruction scheme improves the network's anti-interference capability, reduces signal attenuation in the communication link, and reduces the risk of network segmentation. By using maximum transmission power for compensation at the physical layer, the signal penetration capability is improved, the stability of the communication link is enhanced, and the network has stronger fault adaptability.

[0076] The system in the embodiments of this invention is described below from the perspective of hardware processing. Please refer to [link / reference needed]. Figure 4 This is a schematic diagram of the physical device structure of an IoT device anomaly location system based on edge computing, provided in an embodiment of this application.

[0077] It should be noted that, Figure 4 The structure of the system shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0078] like Figure 4 As shown, the system includes a Central Processing Unit (CPU) 401, which can perform various appropriate actions and processes based on a program stored in Read-Only Memory (ROM) 402 or a program loaded from storage portion 408 into Random Access Memory (RAM) 403, such as executing the methods described in the above embodiments. The RAM 403 also stores various programs and data required for system operation. The CPU 401, ROM 402, and RAM 403 are interconnected via a bus 404. An Input / Output (I / O) interface 405 is also connected to the bus 404.

[0079] The following components are connected to I / O interface 405: input section 406 including a camera, infrared sensor, etc.; output section 407 including a liquid crystal display (LCD) and speakers, etc.; storage section 408 including a hard disk, etc.; and communication section 409 including a network interface card such as a LAN (Local Area Network) card and a modem, etc. Communication section 409 performs communication processing via a network such as the Internet. Drive 410 is also connected to I / O interface 405 as needed. Removable media 411, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., are installed on drive 410 as needed so that computer programs read from it can be installed into storage section 408 as needed.

[0080] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing computer programs for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 409, and / or installed from removable medium 411. When the computer program is executed by central processing unit (CPU) 401, it performs the various functions defined in the present invention.

[0081] It should be noted that the computer-readable medium shown in the embodiments of the present invention can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable compact disc read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In the present invention, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, wherein a computer-readable computer program is carried. The transmitted data signal can take many forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof.

[0082] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. Each block in a flowchart or block diagram may represent a module, segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0083] In another aspect, the present invention also provides a computer-readable storage medium, which may be included in the system described in the above embodiments; or it may exist independently and not assembled into the system. The storage medium carries one or more computer programs that, when executed by a processor of a system, cause the system to implement the methods provided in the above embodiments.

[0084] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0085] As used in the above embodiments, depending on the context, the term "when..." can be interpreted as "if...", "after...", "in response to determining...", or "in response to detecting...". Similarly, depending on the context, the phrase "when determining..." or "if (the stated condition or event) is interpreted as "if determining...", "in response to determining...", "when (the stated condition or event) is detected", or "in response to detecting (the stated condition or event)".

[0086] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive), etc.

[0087] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM or random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A method for anomaly localization of IoT devices based on edge computing, characterized in that, A method applicable to a system comprising an edge gateway and multiple street light terminals distributed in a tree topology along power lines, wherein the edge gateway stores an electrical topology diagram representing the physical connection order of the street light terminals, the method comprising: When the edge gateway detects that the proportion of street light terminals in an offline state exceeds a preset ratio, it broadcasts an environmental awareness command to the power line network to control the street light terminals in an online state to suspend service data transmission and switch to interference monitoring mode. When the edge gateway determines that it has received a background noise intensity value reported by a street light terminal, it marks the street light terminal that reported the background noise intensity value as a sensing node, and marks other street light terminals other than the sensing node as communication blind zone nodes. The edge gateway maps the background noise intensity value of the sensing node to the corresponding node position in the electrical topology diagram, and determines the pairs of sensing nodes that are adjacent in electrical connection in the electrical topology diagram; The edge gateway sequentially compares the background noise intensity values ​​of each pair of sensing nodes to determine the direction of signal amplification on the power line. The edge gateway generates an interference gradient vector based on the enhancement direction; The edge gateway traces back along the interference gradient vector in the electrical topology diagram until it locates the topology convergence area that all the interference gradient vectors point to. The edge gateway determines whether the communication blind spot node exists within the topology convergence area; If present, the edge gateway will identify the communication blind spot node located at the geometric center of the topology convergence area as the abnormal source device; If it does not exist, the edge gateway will identify the sensing node with the highest background noise intensity value in the topology aggregation area as the abnormal source device.

2. The method according to claim 1, characterized in that, The edge gateway sequentially compares the background noise intensity values ​​of each pair of sensing nodes to determine the direction of interference signal enhancement on the power line, specifically including: The edge gateway calculates the absolute value of the difference between the background noise intensity values ​​of the first sensing node and the second sensing node in the sensing node pair; When the edge gateway determines that the absolute value of the difference is greater than a preset effective fluctuation threshold, it identifies the node with the larger background noise intensity value between the first sensing node and the second sensing node as a high-potential node, and the node with the smaller background noise intensity value as a low-potential node. The edge gateway determines the direction of interference signal amplification on the power line by the direction from the low-potential node to the high-potential node.

3. The method according to claim 1, characterized in that, The edge gateway generates an interference gradient vector based on the enhancement direction, specifically including: The edge gateway determines the enhancement direction as the vector direction of the interference gradient vector; The edge gateway calculates the difference in background noise intensity values ​​between the pairs of sensing nodes forming the enhancement direction, and determines the difference as the vector magnitude of the interference gradient vector; The edge gateway combines the vector pointer and the vector magnitude to generate an interference gradient vector describing the rate of change of noise space between the sensing node pairs.

4. The method according to claim 1, characterized in that, The method further includes: The edge gateway sends a spectrum sampling command to the sensing node located upstream of the electrical connection of the anomaly source device, and receives background noise spectrum data fed back by the sensing node; The edge gateway identifies frequency bands in the background noise spectrum data whose energy amplitude exceeds a preset interference threshold as interference fingerprint features; The edge gateway identifies clean carrier frequency bands that do not overlap with the interference fingerprint features in the frequency domain among the multiple communication carrier frequency bands supported by the current system. The edge gateway broadcasts a frequency band migration command containing configuration parameters of the clean carrier frequency band to the power line network to control street light terminals other than the abnormal source device to switch to the clean carrier frequency band for communication.

5. The method according to claim 4, characterized in that, The edge gateway identifies clean carrier frequency bands that do not overlap with the interference fingerprint features in the frequency domain among the multiple communication carrier frequency bands supported by the current system, specifically including: The edge gateway calculates the overlap bandwidth ratio between each of the communication carrier frequency bands and the interference fingerprint features in the frequency domain; If there exists a communication carrier frequency band with an overlap bandwidth ratio of zero, the edge gateway will identify the communication carrier frequency band with an overlap bandwidth ratio of zero as a clean carrier frequency band. If there is no communication carrier frequency band with an overlap bandwidth ratio of zero, the edge gateway calculates the average noise energy density within each of the communication carrier frequency bands. The edge gateway identifies the communication carrier frequency band with the lowest average noise energy density as the clean carrier frequency band.

6. The method according to claim 4, characterized in that, After controlling the street light terminals other than the abnormal source device to switch to the clean carrier frequency band for communication, the method further includes: The edge gateway controls a first neighboring node located upstream of the electrical connection of the abnormal source device to send a test data packet to a second neighboring node located downstream of the electrical connection of the abnormal source device; The edge gateway receives link quality parameters for the test data packet fed back by the second neighboring node, the link quality parameters including signal-to-noise ratio and received signal strength. When the edge gateway determines that the signal-to-noise ratio is greater than a preset net channel threshold and the received signal strength is less than a preset attenuation threshold, it determines that the abnormal source device forms a signal attenuation impedance on the clean carrier frequency band and sends a bridging configuration command to the first neighboring node and the second neighboring node. The bridging configuration command is used to control the first neighboring node and the second neighboring node to establish a direct communication link bypassing the abnormal source device and adjust the signal transmission power of the direct communication link to the maximum power level supported by the device.

7. The method according to claim 6, characterized in that, The bridging configuration command also includes configuration information for the physical layer modulation parameters of the direct communication link, specifically including: The edge gateway locks the data transmission rate of the direct communication link to a preset minimum transmission rate level. The edge gateway adjusts the forward error correction coding redundancy of the direct communication link to a preset maximum redundancy level. The edge gateway prohibits the first neighboring node and the second neighboring node from making dynamic rate adjustments during data interaction on the direct communication link.

8. An IoT device anomaly location system based on edge computing, characterized in that, The system includes: One or more processors and a memory; the memory is coupled to the one or more processors, the memory being used to store computer program code, the computer program code including computer instructions, the one or more processors invoking the computer instructions to cause the system to perform the method as described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that, When the instructions are executed on the system, the system performs the method as described in any one of claims 1-7.

10. A computer program product, characterized in that, When the computer program product is run on the system, the system performs the method as described in any one of claims 1-7.

Images