Internet of Things communication optimization method and system based on edge AI

Abstract

The invention discloses an Internet of Things communication optimization method and system based on edge AI, and relates to the technical field of data communication optimization, and the method comprises the steps: edge equipment carries out the data preprocessing of collected data, and removes noise and irrelevant information; by setting a communication optimization feature extraction algorithm, performing real-time analysis on the data, and extracting features; selecting a communication path according to network conditions and communication requirements; adaptively adjusting the communication frequency and power according to the electric quantity of the equipment and the communication requirement; a communication fault prediction algorithm is set to predict communication faults, and measures are taken in advance. According to the invention, adaptive adjustment of communication frequency and power is realized, unnecessary energy consumption is avoided, and the service life of equipment is prolonged.

Description

Technical Field

[0001] The present invention relates to the technical field of data communication optimization, and specifically to an Internet of Things communication optimization method and system based on edge AI. Background Art

[0002] With the rapid development of information technology, the Internet of Things (IoT) has gradually become an important force in promoting social progress and economic development. IoT connects various devices and objects through the Internet through wireless sensor networks, radio frequency identification technology, cloud computing and other technologies, realizing the intelligent collection, transmission, processing and application of information. However, as the scale of IoT systems continues to expand, the number of devices has increased dramatically, and data traffic has exploded, the performance requirements for communication networks such as bandwidth, latency and energy consumption have become increasingly higher.

[0003] Traditional IoT communication architecture usually adopts a centralized data processing method, that is, all data needs to be transmitted to the cloud server through the network for processing before returning the results to the terminal device. This architecture has problems such as high latency, large network load, and high energy consumption when processing large-scale data, and it is difficult to meet the needs of application scenarios with high real-time requirements and fast response speed. Therefore, how to optimize the IoT communication architecture and improve system performance and energy efficiency has become one of the hot topics in the current IoT research field.

[0004] As an emerging computing paradigm, the core idea of ​​edge computing is to move computing resources from traditional data centers to the edge of the network, that is, to process data directly where it is generated. This distributed computing model can effectively reduce data transmission delays and bandwidth usage, and improve system response speed and resource utilization. At the same time, with the rapid development of artificial intelligence (AI) technology, especially breakthroughs in deep learning, reinforcement learning and other technologies, edge computing has been provided with powerful intelligent analysis and decision-making capabilities. Combining edge computing with AI technology to form edge AI can further enhance the intelligence level of the IoT system.

[0005] For example, the existing Chinese patent with publication number CN108848514A discloses a data communication optimization method and a data communication optimizer, including: obtaining each data communication optimization matrix, the data communication optimization matrix including current delay information and current bandwidth information; determining the optimal data communication optimization matrix according to the current delay information and current bandwidth information of each data communication optimization matrix; and outputting the optimal data communication optimization matrix.

[0006] However, this design optimization matrix is ​​based on the current state and does not take into account the dynamic changes in network conditions. In an actual network environment, the current delay information and the current bandwidth information may fluctuate frequently, requiring a more dynamic optimization strategy. To this end, the present invention provides an IoT communication optimization method and system based on edge AI. Summary of the invention

[0007] The purpose of the present invention is to provide an IoT communication optimization method and system based on edge AI to solve the existing problems raised in the above background technology.

[0008] To achieve the above object, the present invention provides the following technical solution: an IoT communication optimization method based on edge AI, comprising the following steps:

[0009] S1. The edge device preprocesses the collected data to remove noise and irrelevant information;

[0010] S2. Analyze the data in real time and extract features by setting a communication optimization feature extraction algorithm; select a communication path based on network conditions and communication requirements;

[0011] S3, adaptively adjust the communication frequency and power according to the device power and communication requirements;

[0012] S4. Set a communication failure prediction algorithm to predict communication failures and take measures in advance.

[0013] A further improvement of the present invention is that the specific creation steps of the communication optimization feature extraction algorithm include:

[0014] S21. First, the original communication message data is cut into pieces, and the continuous byte sequence is converted into a discrete Token sequence, where each Token contains a message data fragment with uniform data characteristics;

[0015] S22, performing communication path pattern matching on the message data fragments, and calculating the similarity of Token sequences in different communication data;

[0016] S23, setting a sequence similarity threshold, and including the communication data with a similarity greater than the similarity threshold in step S22 into the communication feature data set;

[0017] S24, returning the communication feature data set to step S22;

[0018] S25. Setting a similarity termination threshold, when the Token sequence similarity of all data in the communication feature data set is greater than the similarity termination threshold, terminating the iteration.

[0019] A further improvement of the present invention is that the communication path pattern matching obtains a communication path matching integral matrix by performing binary matching on two random sets of data in the discrete Token sequence. Assume that the expression form of the two random sets of data is A i =(A1, A2, ..., A anum ) and B j =(B1, B2, ..., Bbnum ), anum represents the data set A i The number of packets, bnum represents the data set B j The number of messages is calculated to calculate the scores of all positions in the communication path matching score matrix. INM(i, j) represents the score of the communication path matching integral matrix at coordinate (i, j). According to the score of INM(i, j), the communication path matching integral matrix is ​​completely filled, and INM(anum, bnum) is calculated to be the score of the final selected communication path. Then, starting from the position (anum, bnum), a matching path search is performed. When the node position (i, j) meets the judgment function F1={(INM(i, j)=INM(i-1, j-1)+1)∧(A i =B j )}, (i, j) is the final node position, and the next node position is (i-1, j-1); when the node position (i, j) meets the judgment function F2 = {(INM(i, j) = INM(i-1, j)) ∧ (INM(i, j) ! = INM(i, j-1))}, (i, j+1) is the final node position, and the next node position is (i-1, j); when the node position (i, j) meets the judgment function F3 = {(INM(i, j) = INM(i, j-1)) ∧ (INM(i, j)) ! = INM(i, j-1))}, =INM(i-1, j))}, (i+1, j) is the final selected node position, and the next node position is (i, j-1); when the node position (i, j) meets the judgment function F4 = {(INM(i, j) = INM(i-1, j))∧(INM(i, j) = INM(i, j-1))}, (i, j+1) or (i+1, j) is the final selected node position, and the next node position is (i-1, j) or (i, j-1); until backtracking to the node position (1, 1), the final communication path is obtained.

[0020] The present invention is further improved in that the specific step S3 comprises:

[0021] S31, set the initial power P initial and a minimum power threshold TP, monitoring the remaining power of the device in real time, and triggering a communication frequency and power adjustment strategy when the remaining power of the device is less than the minimum power threshold;

[0022] S32. Calculate the communication demand adaptation value of the device power according to the communication demand, including communication frequency, power, power, data rate and delay

[0023] , where R actual(f, P) represents the actual data rate achieved by the device at the current communication frequency f and power P, D actual (f, P) represents the actual delay achieved by the device at the current communication frequency f and power P, E current represents the remaining power of the device, R represents the current required data rate, D represents the current delay, and R req Indicates the communication frequency requirement, D req represents the delay requirement, α1 represents the weight of balancing data rate, α2 represents the weight of delay, and α3 represents the weight of power consumption;

[0024] S33, randomly generate a set of frequencies and powers, and obtain the final frequency and power as output through the device power communication demand adaptation value and genetic algorithm.

[0025] The present invention is further improved in that the communication fault prediction algorithm includes collecting power, data rate, delay and environmental parameters and historical fault records to establish, the environmental parameters include air humidity and equipment temperature, and calculating the communication fault impact value. E cun Indicates the remaining power of the device, R cun The characterization of data rate, D cun Indicates the delay, represents the mean value of air humidity, Indicates the average temperature of the device.

[0026] A further improvement of the present invention is that the communication fault prediction algorithm also includes establishing a fault type database, which includes hardware faults, software faults, network faults, safety faults and no faults, and collecting fault types in the fault type database corresponding to historical communication fault impact values ​​as training data for the communication fault prediction algorithm, based on a support vector machine model, and then calculating the current communication fault impact value as input to the communication fault prediction algorithm, outputting the predicted fault type, and when the output fault type is one of hardware faults, software faults, network faults and safety faults, extracting the current power, setting a power threshold and a communication fault impact standard value, and calculating the ratio of the current communication fault impact value to the communication fault impact standard value as an adjustment coefficient, and when the current power is greater than or equal to the power threshold, multiplying the communication frequency and power by the adjustment coefficient to obtain new communication frequency and power, and sending them to the management personnel together with the fault type via SMS or email.

[0027] A further improvement of the present invention is that the data preprocessing includes data cleaning and format conversion, and the specific steps include: using the mean method to interpolate missing values ​​on the collected data set, and after deleting duplicate data in the data set, converting data with different character encodings into a unified encoding.

[0028] On the other hand, the present invention provides an IoT communication optimization system based on edge AI, comprising:

[0029] The IoT device layer, including sensors, actuators and IoT devices, is responsible for data collection and preliminary processing by actuators; the edge device layer, including edge gateways and edge servers, is deployed close to the data source and is responsible for local data processing and intelligent analysis; the cloud platform layer provides global management, complex data analysis, long-term storage and advanced application services; the communication network layer connects the IoT device layer, edge device layer and cloud platform layer.

[0030] A further improvement of the present invention is that the platform layer includes an edge AI processing module, a communication optimization module, an adaptive communication strategy module and a fault prediction module; the edge AI processing module is used to perform data preprocessing on the collected data through the edge device to remove noise and irrelevant information; the communication optimization module is used to analyze the data in real time and extract features by setting a communication optimization feature extraction algorithm; select a communication path according to network conditions and communication requirements; the adaptive communication strategy module is used to adaptively adjust the communication frequency and power according to the device power and communication requirements; the fault prediction module is used to set a communication fault prediction algorithm to predict communication failures and take measures in advance.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] The present invention firstly avoids unnecessary energy consumption and prolongs the service life of the equipment by adaptively adjusting the communication frequency and power according to the power level of the equipment and the communication demand; the adaptive adjustment strategy can ensure the provision of high-quality communication services when the power level is sufficient, and can reduce the communication demand when the power level is low, thereby ensuring the continuous operation of the equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 This is a flow chart of the IoT communication optimization method based on edge AI of the present invention;

[0034] Figure 2 This is a flow chart of the communication optimization feature extraction algorithm of the Internet of Things communication optimization method based on edge AI of the present invention;

[0035] Figure 3 A flow chart of the invention's edge AI-based IoT communication optimization method for adaptively adjusting communication frequency and power;

[0036] Figure 4 This is a framework diagram of the IoT communication optimization system based on edge AI in the present invention. DETAILED DESCRIPTION

[0037] The technical solution of the present invention is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations on the technical solution of the present invention. The embodiments of the present invention and the technical features in the embodiments may be combined with each other unless there is a conflict.

[0038] The term "and / or" is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " generally indicates that the associated objects before and after are in an "or" relationship.

[0039] Example 1

[0040] Figure 1 The flowchart of the edge AI-based IoT communication optimization method disclosed in this embodiment is shown, and the steps are as follows:

[0041] S1. The edge device performs data preprocessing on the collected data to remove noise and irrelevant information; the data preprocessing includes data cleaning and format conversion, and the specific steps include: using the mean method to interpolate missing values ​​on the collected data set, and deleting duplicate data in the data set before converting data with different character encodings into a unified encoding.

[0042] S2. Analyze the data in real time and extract features by setting a communication optimization feature extraction algorithm; select a communication path based on network conditions and communication requirements; Figure 2 The flow chart of the communication optimization feature extraction algorithm of the IoT communication optimization method based on edge AI of the present invention is shown. The specific creation steps of the communication optimization feature extraction algorithm include:

[0043] S21. First, the original communication message data is cut into pieces, and the continuous byte sequence is converted into a discrete Token sequence, where each Token contains a message data fragment with uniform data characteristics;

[0044] S22, S22, perform communication path pattern matching on the message data fragments, and calculate the similarity of Token sequences in different communication data; the communication path pattern matching obtains a communication path matching integral matrix by performing binary matching on two random sets of data in the discrete Token sequence, and assumes that the expression form of the two random sets of data is A i =(A1, A2, ..., A anum ) and B j =(B1, B2, ..., B bnum ), anum represents the data set A i The number of packets, bnum represents the data set B jThe number of messages is calculated to calculate the scores of all positions in the communication path matching score matrix. INM(i, j) represents the score of the communication path matching integral matrix at coordinate (i, j). According to the score of INM(i, j), the communication path matching integral matrix is ​​completely filled, and INM(anum, bnum) is calculated to be the score of the final selected communication path. Then, starting from the position (anum, bnum), a matching path search is performed. When the node position (i, j) meets the judgment function F1={(INM(i, j)=INM(i-1, j-1)+1)∧(A i =B j )}, (i, j) is the final node position, and the next node position is (i-1, j-1); when the node position (i, j) meets the judgment function F2 = {(INM(i, j) = INM(i-1, j)) ∧ (INM(i, j) ! = INM(i, j-1))}, (i, j+1) is the final node position, and the next node position is (i-1, j); when the node position (i, j) meets the judgment function F3 = {(INM(i, j) = INM(i, j-1)) ∧ (INM(i, j)) ! = INM(i, j-1))}, =INM(i-1, j))}, (i+1, j) is the final selected node position, and the next node position is (i, j-1); when the node position (i, j) meets the judgment function F4 = {(INM(i, j) = INM(i-1, j))∧(INM(i, j) = INM(i, j-1))}, (i, j+1) or (i+1, j) is the final selected node position, and the next node position is (i-1, j) or (i, j-1); until backtracking to the node position (1, 1), the final communication path is obtained.

[0045] S23, setting a sequence similarity threshold, and including the communication data with a similarity greater than the similarity threshold in step S22 into the communication feature data set;

[0046] S24, returning the communication feature data set to step S22;

[0047] S25. Setting a similarity termination threshold, when the Token sequence similarity of all data in the communication feature data set is greater than the similarity termination threshold, terminating the iteration.

[0048] S3, adaptively adjust the communication frequency and power according to the device power and communication requirements; Figure 3 The flowchart of the edge AI-based IoT communication optimization method of the present invention for adaptively adjusting the communication frequency and power is shown, and the specific steps include:

[0049] S31, set the initial power P initial and a minimum power threshold TP, monitoring the remaining power of the device in real time, and triggering a communication frequency and power adjustment strategy when the remaining power of the device is less than the minimum power threshold;

[0050] S32. Calculate the communication demand adaptation value of the device power according to the communication demand, including communication frequency, power, power, data rate and delay Where R actual (f, P) represents the actual data rate achieved by the device at the current communication frequency f and power P, D actual (f, P) represents the actual delay achieved by the device at the current communication frequency f and power P, E current represents the remaining power of the device, R represents the current required data rate, D represents the current delay, and R req Indicates the communication frequency requirement, D req represents the delay requirement, α1 represents the weight of balancing data rate, α2 represents the weight of delay, and α3 represents the weight of power consumption;

[0051] S33, randomly generate a set of frequencies and powers, and obtain the final frequency and power as output through the device power communication demand adaptation value and genetic algorithm.

[0052] S4. Set a communication fault prediction algorithm to predict communication faults and take measures in advance. The communication fault prediction algorithm is established by collecting power, data rate, delay and environmental parameters and historical fault records. The environmental parameters include air humidity and device temperature, and calculate the communication fault impact value. E cun Indicates the remaining power of the device, R cun The characterization of data rate, D cun Indicates the delay, represents the mean value of air humidity, Represents the mean value of the device temperature; establishes a fault type database, the fault type database includes hardware faults, software faults, network faults, safety faults and no faults, collects the fault types in the fault type database corresponding to the historical communication fault impact values ​​as the training data of the communication fault prediction algorithm, based on the support vector machine model, and then calculates the current communication fault impact value as the input of the communication fault prediction algorithm, outputs the predicted fault type, and when the output fault type is one of the hardware fault, software fault, network fault and safety fault, extracts the current power, sets the power threshold and the communication fault impact standard value, calculates the ratio of the current communication fault impact value to the communication fault impact standard value as the adjustment coefficient, and when the current power is greater than or equal to the power threshold, multiplies the communication frequency and power by the adjustment coefficient to obtain new communication frequency and power, and sends them to the management personnel together with the fault type through SMS or email.

[0053] The threshold and weight may be set by default according to the present invention, or may be set by an operator.

[0054] Example 2

[0055] Figure 4 The framework diagram of the IoT communication optimization system based on edge AI of the present invention is shown. Based on the same inventive concept as Example 1, the present invention provides an IoT communication optimization system based on edge AI, including:

[0056] The IoT device layer, including sensors, actuators, and IoT devices, is responsible for data collection and preliminary actuator processing; the edge device layer, including edge gateways and edge servers, is deployed close to the data source and is responsible for local data processing and intelligent analysis; the cloud platform layer provides global management, complex data analysis, long-term storage, and advanced application services; the communication network layer connects the IoT device layer, edge device layer, and cloud platform layer;

[0057] The platform layer includes an edge AI processing module, a communication optimization module, an adaptive communication strategy module and a fault prediction module; the edge AI processing module is used to perform data preprocessing on the collected data through edge devices to remove noise and irrelevant information; the communication optimization module is used to analyze the data in real time and extract features by setting a communication optimization feature extraction algorithm; select a communication path according to network conditions and communication requirements; the adaptive communication strategy module is used to adaptively adjust the communication frequency and power according to the device power and communication requirements; the fault prediction module is used to set a communication fault prediction algorithm to predict communication failures and take measures in advance.

[0058] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0059] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0060] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0061] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0062] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the purpose of the present invention and the claims, which all fall within the protection of the present invention.

Claims

1. An IoT communication optimization method based on edge AI, characterized by: The following steps are involved: S1. The edge device preprocesses the collected data to remove noise and irrelevant information; S2. Analyze the data in real time and extract features by setting a communication optimization feature extraction algorithm; select a communication path based on network conditions and communication requirements; S3, adaptively adjust the communication frequency and power according to the device power and communication requirements; S4. Set a communication failure prediction algorithm to predict communication failures and take measures in advance.

2. The IoT communication optimization method based on edge AI according to claim 1 is characterized in that: The specific creation steps of the communication optimization feature extraction algorithm include: S21. First, the original communication message data is cut into pieces, and the continuous byte sequence is converted into a discrete Token sequence, where each Token contains a message data fragment with uniform data characteristics; S22, performing communication path pattern matching on the message data fragments, and calculating the similarity of Token sequences in different communication data; S23, setting a sequence similarity threshold, and including the communication data with a similarity greater than the similarity threshold in step S22 into the communication feature data set; S24, returning the communication feature data set to step S22; S25. Setting a similarity termination threshold, when the Token sequence similarity of all data in the communication feature data set is greater than the similarity termination threshold, terminating the iteration.

3. The IoT communication optimization method based on edge AI according to claim 2 is characterized in that: The communication path pattern matching obtains the communication path matching integral matrix by performing binary matching on two random sets of data in the discrete Token sequence. Suppose the expression form of the two random sets of data is A i =(A1, A2, ..., A anum ) and B j =(B1, B2, ..., B bnum ), anum represents the data set A i The number of packets, bnum represents the data set B j The number of messages is calculated to calculate the scores of all positions in the communication path matching score matrix. INM(i, j) represents the score of the communication path matching integral matrix at coordinate (i, j), and the communication path matching integral matrix is ​​completely filled according to the score of INM(i, j); Calculate INM (anum, bnum) to get the final selected communication path score; then start from the position (anum, bnum) to find the matching path. When the node position (i, j) meets the judgment function F1 = {(INM(i, j) = INM(i-1, j-1) + 1) ∧ (A i =B j )}, (i, j) is the final node position, and the next node position is (i-1, j-1); when the node position (i, j) meets the judgment function F2 = {(INM(i, j) = INM(i-1, j)) ∧ (INM(i, j) ! = INM(i, j-1))}, (i, j+1) is the final node position, and the next node position is (i-1, j); when the node position (i, j) meets the judgment function F3 = {(INM(i, j) = INM(i, j-1)) ∧ (INM(i, j)) ! = INM(i, j-1))}, =INM(i-1, j))}, (i+1, j) is the final selected node position, and the next node position is (i, j-1); when the node position (i, j) meets the judgment function F4 = {(INM(i, j) = INM(i-1, j))∧(INM(i, j) = INM(i, j-1))}, (i, j+1) or (i+1, j) is the final selected node position, and the next node position is (i-1, j) or (i, j-1); until backtracking to the node position (1, 1), the final communication path is obtained.

4. The IoT communication optimization method based on edge AI according to claim 3 is characterized in that: The specific steps of S3 include: S31, set the initial power P initial and a minimum power threshold TP, monitoring the remaining power of the device in real time, and triggering a communication frequency and power adjustment strategy when the remaining power of the device is less than the minimum power threshold; S32. Calculate the communication demand adaptation value of the device power according to the communication demand, including communication frequency, power, power, data rate and delay Where R actual (f, P) represents the actual data rate achieved by the device at the current communication frequency f and power P, D actual (f, P) represents the actual delay achieved by the device at the current communication frequency f and power P, E current represents the remaining power of the device, R represents the current required data rate, D represents the current delay, and R req Indicates the communication frequency requirement, D req represents the delay requirement, α1 represents the weight of balancing data rate, α2 represents the weight of delay, and α3 represents the weight of power consumption; S33, randomly generate a set of frequencies and powers, and obtain the final frequency and power as output through the device power communication demand adaptation value and genetic algorithm.

5. The IoT communication optimization method based on edge AI according to claim 4 is characterized in that: The communication fault prediction algorithm includes collecting power, data rate, delay and environmental parameters and historical fault records to calculate the communication fault impact value. E cun Indicates the remaining power of the device, R cun The characterization of data rate, D cun Indicates the delay, represents the mean value of air humidity, Indicates the average temperature of the device.

6. The IoT communication optimization method based on edge AI according to claim 4 is characterized in that: The communication fault prediction algorithm also includes establishing a fault type database, which includes hardware faults, software faults, network faults, safety faults and no faults, collecting fault types in the fault type database corresponding to historical communication fault impact values ​​as training data for the communication fault prediction algorithm, based on a support vector machine model, and then calculating the current communication fault impact value as input to the communication fault prediction algorithm, outputting the predicted fault type, and when the output fault type is one of hardware faults, software faults, network faults and safety faults, extracting the current power, setting a power threshold and a communication fault impact standard value, and calculating the ratio of the current communication fault impact value to the communication fault impact standard value as an adjustment coefficient, and when the current power is greater than or equal to the power threshold, multiplying the communication frequency and power by the adjustment coefficient to obtain a new communication frequency and power, which are sent to the management personnel together with the fault type via SMS or email.

7. The IoT communication optimization method based on edge AI according to claim 6 is characterized in that: The data preprocessing includes data cleaning and format conversion, and the specific steps include: using the mean method to perform missing value interpolation processing on the collected data set, and deleting duplicate data in the data set and converting data with different character encodings into a unified encoding.

8. An edge AI-based IoT communication optimization system, used to execute the edge AI-based IoT communication optimization method according to any one of claims 1 to 7, characterized in that: include: The IoT device layer, including sensors, actuators, and IoT devices, is responsible for data collection and preliminary processing by actuators; The edge device layer, including edge gateways and edge servers, is deployed close to the data source and is responsible for local data processing and intelligent analysis; The cloud platform layer provides global management, complex data analysis, long-term storage, and advanced application services; The communication network layer connects the IoT device layer, edge device layer, and cloud platform layer.

9. The IoT communication optimization system based on edge AI according to claim 8, characterized in that: The platform layer includes an edge AI processing module, a communication optimization module, an adaptive communication strategy module, and a fault prediction module; The edge AI processing module is used to pre-process the collected data through the edge device to remove noise and irrelevant information; The communication optimization module is used to analyze the data in real time and extract features by setting a communication optimization feature extraction algorithm; select a communication path according to network conditions and communication requirements; The adaptive communication strategy module is used to adaptively adjust the communication frequency and power according to the device power and communication requirements; The fault prediction module is used to set a communication fault prediction algorithm to predict communication faults and take measures in advance.

Images