CN117155845B - Internet of things data interaction method and system - Google Patents

Internet of things data interaction method and system Download PDF

Info

Publication number
CN117155845B
CN117155845B CN202311421852.3A CN202311421852A CN117155845B CN 117155845 B CN117155845 B CN 117155845B CN 202311421852 A CN202311421852 A CN 202311421852A CN 117155845 B CN117155845 B CN 117155845B
Authority
CN
China
Prior art keywords
data
module
adopting
sub
data transmission
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202311421852.3A
Other languages
Chinese (zh)
Other versions
CN117155845A (en
Inventor
张博伦
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing University of Posts and Telecommunications
Original Assignee
Beijing University of Posts and Telecommunications
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing University of Posts and Telecommunications filed Critical Beijing University of Posts and Telecommunications
Priority to CN202311421852.3A priority Critical patent/CN117155845B/en
Publication of CN117155845A publication Critical patent/CN117155845A/en
Application granted granted Critical
Publication of CN117155845B publication Critical patent/CN117155845B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/0464—Convolutional networks [CNN, ConvNet]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/14—Fourier, Walsh or analogous domain transformations, e.g. Laplace, Hilbert, Karhunen-Loeve, transforms
    • G06F17/141—Discrete Fourier transforms
    • G06F17/142—Fast Fourier transforms, e.g. using a Cooley-Tukey type algorithm
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/25—Fusion techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/092—Reinforcement learning
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16Y—INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]
    • G16Y10/00—Economic sectors
    • G16Y10/75—Information technology; Communication
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/02—Topology update or discovery
    • H04L45/08—Learning-based routing, e.g. using neural networks or artificial intelligence
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/12—Shortest path evaluation
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/12—Shortest path evaluation
    • H04L45/121—Shortest path evaluation by minimising delays
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/14—Routing performance; Theoretical aspects
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/12—Protocols specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Abstract

The invention relates to the technical field of the Internet of things, in particular to a data interaction method and a data interaction system of the Internet of things, which comprise the following steps: based on the equipment data of the Internet of things, evaluating the performance of a data transmission path by adopting a shortest path algorithm, and generating a preliminary routing table; based on the preliminary routing table, a Bayesian decision theory is utilized to dynamically select a proper communication protocol, and an optimized routing and protocol configuration table is obtained. In the invention, the transmission path is evaluated by the shortest path algorithm, so that high-efficiency low-delay data transmission is ensured. Bayesian decision theory provides flexible communication protocol selection, enhancing communication stability and economy. And the edge calculation is used for unloading cloud computing tasks through fast Fourier transform, so that data redundancy is reduced, and transmission efficiency is improved. The integrity and accuracy are enhanced by the convolutional neural network data fusion, and high-quality data are ensured. And the intelligent transmission strategy is strengthened and learned, and the data loss rate caused by network fluctuation is reduced. The association rule mining depth analyzes time-space data, and optimizes a data transmission strategy.

Description

Internet of things data interaction method and system
Technical Field
The invention relates to the technical field of the Internet of things, in particular to an Internet of things data interaction method and system.
Background
The internet of things is an emerging technical field, and relates to the connection of various physical devices, sensors, automobiles, household appliances and the like to the internet so as to realize real-time interaction, monitoring, control and analysis of data. The field covers a plurality of aspects such as hardware, software, communication protocols, data analysis and the like, and aims to realize interconnection and interworking among devices so as to improve efficiency, safety, convenience and automation degree.
The data interaction method of the Internet of things refers to data exchange and communication modes between equipment and a sensor and between the equipment and a cloud server in the Internet of things. The method comprises the steps of data acquisition, transmission, storage, processing, analysis and the like, so that the equipment can cooperate with each other, and real-time data sharing and remote control are realized. The method and the device mainly aim at ensuring that the devices are interconnected and intercommunicated so as to improve efficiency, safety, convenience and automation degree. The method comprises real-time connection between the device and the sensor, and data transmission between the device and the cloud server. The interaction mode enables the equipment to monitor the change of the physical world in real time, perform data analysis and insight, remotely manage and maintain, and optimize resources. Typically, these methods enable connection between devices, data collection and transmission through various communication technologies (e.g., wi-Fi, bluetooth, zigbee, loRa, etc.), and then data storage and processing in a cloud server or data center, and enable users to access data, monitor, control, and analyze through application programs or network interfaces. The core objective of the Internet of things data interaction method is to realize an intelligent, remote and data-driven Internet of things ecosystem.
The existing data interaction method of the internet of things often depends on a fixed transmission path and a communication protocol too much, and lacks of dynamics and adaptability. This may not only lead to inefficiency and high delay in data transmission, but may also increase the cost of data transmission. In addition, effective edge calculation and data fusion are not performed, so that a large amount of redundant and incomplete data are transmitted to the cloud, calculation and storage pressure is increased, and data misunderstanding can be caused. The consideration of the data transmission strategy in the existing method is relatively single, the dynamic change of the network environment is not considered, and the data loss and the transmission failure are easy to cause. Correlation analysis of time-space data is also typically ignored, and the value and potential of the data is not fully exploited.
Disclosure of Invention
The invention aims to solve the defects in the prior art, and provides a data interaction method and system of the Internet of things.
In order to achieve the above purpose, the present invention adopts the following technical scheme: the data interaction method of the Internet of things comprises the following steps:
s1: based on the equipment data of the Internet of things, evaluating the performance of a data transmission path by adopting a shortest path algorithm, and generating a preliminary routing table;
S2: based on the preliminary routing table, dynamically selecting a proper communication protocol by using a Bayesian decision theory to obtain an optimized routing and protocol configuration table;
s3: performing edge calculation processing on the data by adopting fast Fourier transform according to the optimized route and protocol configuration table to generate an edge calculation result set and a cloud data packet to be transmitted;
s4: performing multi-mode data fusion on the edge calculation result set by using a convolutional neural network to obtain fused edge data;
s5: based on the fused edge data, adopting a reinforcement learning algorithm to perform collaborative learning optimization data transmission to obtain a collaborative optimization data transmission strategy;
s6: and carrying out time-space data correlation analysis by adopting a correlation rule mining algorithm in combination with the collaborative optimized data transmission strategy to generate a final optimized data transmission plan.
As a further scheme of the invention, based on the equipment data of the Internet of things, the performance of a data transmission path is evaluated by adopting a shortest path algorithm, and the step of generating a preliminary routing table specifically comprises the following steps:
s101: collecting geographic position and network topology information of the Internet of things equipment by adopting a node detection method to obtain a network topology graph;
S102: based on the network topology diagram, a shortest path algorithm is adopted to evaluate paths from a source node to a destination node, and a shortest path table among the nodes is obtained;
s103: based on the shortest path table among the nodes, adopting a delay calculation method to evaluate the communication delay and cost of the paths to obtain a communication path performance table;
s104: and selecting a path with optimal performance by adopting a path selection algorithm based on the communication path performance table to generate a preliminary routing table.
As a further scheme of the present invention, based on the preliminary routing table, a bayesian decision theory is utilized to dynamically select a suitable communication protocol, and the steps for obtaining the optimized routing and protocol configuration table are specifically as follows:
s201: based on the preliminary routing table, adopting a data sampling method to collect past communication data and error rate to obtain a historical communication data table;
s202: based on the historical communication data table, evaluating the applicability of the communication protocol by adopting a Bayesian statistical model to obtain a communication protocol applicability table;
s203: based on the communication protocol applicability table, adopting a decision tree algorithm to dynamically select an optimal communication protocol to obtain an optimized communication protocol;
S204: and generating an optimized route and protocol configuration table by adopting a merging strategy method based on the optimized communication protocol and the preliminary route table.
As a further scheme of the present invention, according to the optimized routing and protocol configuration table, performing edge calculation processing on data by using fast fourier transform, and generating an edge calculation result set and a cloud data packet to be transmitted specifically includes:
s301: based on the optimized route and protocol configuration table, acquiring the data of the to-be-processed Internet of things equipment by adopting a data flow interception method to obtain an original data set;
s302: based on the original data set, preprocessing data by adopting a wavelet transformation method, including filtering and denoising, so as to obtain a preprocessed data set;
s303: based on the preprocessing data set, adopting a fast Fourier transform method to realize edge calculation, and generating an edge calculation result set and a cloud data packet to be transmitted;
s304: and based on the edge calculation result set and the cloud data packet to be transmitted, distributing according to the optimized routing and protocol configuration table by adopting a data packaging method to finish edge calculation data distribution.
As a further scheme of the invention, the convolution neural network is utilized to perform multi-mode data fusion on the edge calculation result set, and the step of obtaining the fused edge data specifically comprises the following steps:
S401: preprocessing the multi-mode data by adopting a data standardization method based on the edge calculation result set to generate a standardized multi-mode data set;
s402: extracting key information features by adopting a feature extraction algorithm based on the standardized multi-mode data set to obtain a key feature data set;
s403: based on the key characteristic data set, designing a convolutional neural network structure to perform multi-mode data fusion to obtain a CNN model structure;
s404: and training and fusing the key characteristic data set based on the CNN model structure to obtain fused edge data.
As a further scheme of the present invention, based on the fused edge data, collaborative learning is performed to optimize data transmission by adopting a reinforcement learning algorithm, and the step of obtaining a collaborative optimized data transmission policy specifically includes:
s501: defining a reinforcement learning state, action and rewarding mechanism based on the fused edge data to obtain reinforcement learning environment definition;
s502: based on the reinforcement learning environment definition, performing preliminary learning by adopting a Q-learning algorithm to generate a preliminary strategy network;
s503: performing strategy iteration based on the preliminary strategy network, and continuously optimizing data transmission to obtain an optimized strategy network;
S504: and based on the optimized strategy network, evaluating the efficiency and accuracy of data transmission to obtain a cooperatively optimized data transmission strategy.
As a further scheme of the present invention, in combination with the cooperatively optimized data transmission policy, a correlation rule mining algorithm is adopted to perform a time-space data correlation analysis, and the step of generating a final optimized data transmission plan specifically includes:
s601: based on the cooperatively optimized data transmission strategy, carrying out data preprocessing to obtain a preprocessed data set;
s602: based on the preprocessed data set, generating a preliminary association rule by adopting an Apriori algorithm to obtain a preliminary association rule set;
s603: based on the preliminary association rule set, a confidence and support filtering method is adopted to screen high association rules, and a high association rule set is obtained;
s604: and based on the high association rule set, combining the time-space data characteristics, carrying out deep association analysis, and generating a final optimized data transmission plan.
The Internet of things data interaction system is used for executing the Internet of things data interaction method, and comprises a route optimization module, a protocol selection module, an edge calculation module, a data fusion module and a data transmission optimization module.
As a further scheme of the invention, the route optimization module collects network topology by adopting a node detection method based on the equipment information of the internet of things, and uses a shortest path algorithm to perform path evaluation to generate a preliminary routing table;
the protocol selection module collects communication data by adopting a data sampling method based on the preliminary routing table, and adopts a Bayesian statistical model to evaluate the applicability of the communication protocol to generate an optimized routing and protocol configuration table;
the edge calculation module collects equipment data by adopting a data flow interception method based on the optimized routing and protocol configuration table, performs edge calculation by using a fast Fourier transform method, and generates an edge calculation result set and a cloud data packet to be transmitted;
the data fusion module is used for preprocessing by adopting a data standardization method based on the edge calculation result set, and carrying out data fusion by utilizing a convolutional neural network structure to obtain fused edge data;
the data transmission optimization module defines a reinforcement learning environment based on the fused edge data, optimizes data transmission through a Q-learning algorithm, and generates a collaborative optimized data transmission strategy.
As a further scheme of the invention, the route optimization module comprises a node detection sub-module, a path evaluation sub-module, a delay calculation sub-module and a path selection sub-module;
The protocol selection module comprises a data sampling sub-module, a Bayesian evaluation sub-module, a decision tree selection sub-module and a merging strategy sub-module;
the edge computing module comprises a data interception sub-module, a data preprocessing sub-module, an edge computing sub-module and a data distribution sub-module;
the data fusion module comprises a data standardization sub-module, a feature extraction sub-module, a network design sub-module and a data fusion sub-module;
the data transmission optimization module comprises a reinforcement learning environment definition sub-module, a Q-learning sub-module, a strategy iteration sub-module and a data transmission evaluation sub-module.
Compared with the prior art, the invention has the advantages and positive effects that:
in the invention, the data transmission path is more efficiently evaluated by the shortest path algorithm, and the high efficiency and low delay of data transmission are ensured. Bayesian decision theory provides a flexible communication protocol selection mechanism, so that communication is more stable and economical. The edge calculation is converted into cloud calculation through fast Fourier transform, so that part of calculation tasks are unloaded, the redundancy of data is reduced, and the transmission efficiency is improved. And the convolutional neural network is utilized for data fusion, so that the integrity and accuracy of the data are enhanced, and the high quality of the data is ensured. The application of the reinforcement learning algorithm ensures that the data transmission strategy is more intelligent and self-adaptive, and reduces the data loss rate caused by network fluctuation. Through the association rule mining algorithm, the deep association analysis of time-space data is realized, and the strategy and the plan of data transmission are further perfected.
Drawings
FIG. 1 is a schematic workflow diagram of the present invention;
FIG. 2 is a S1 refinement flowchart of the present invention;
FIG. 3 is a S2 refinement flowchart of the present invention;
FIG. 4 is a S3 refinement flowchart of the present invention;
FIG. 5 is a S4 refinement flowchart of the present invention;
FIG. 6 is a S5 refinement flowchart of the present invention;
FIG. 7 is a S6 refinement flowchart of the present invention;
FIG. 8 is a system flow diagram of the present invention;
FIG. 9 is a schematic diagram of a system framework of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
In the description of the present invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," and the like indicate orientations or positional relationships based on the orientation or positional relationships shown in the drawings, merely to facilitate describing the present invention and simplify the description, and do not indicate or imply that the devices or elements referred to must have a specific orientation, be configured and operated in a specific orientation, and therefore should not be construed as limiting the present invention. Furthermore, in the description of the present invention, the meaning of "a plurality" is two or more, unless explicitly defined otherwise.
Embodiment one: referring to fig. 1, the present invention provides a technical solution: the data interaction method of the Internet of things comprises the following steps:
s1: based on the equipment data of the Internet of things, evaluating the performance of a data transmission path by adopting a shortest path algorithm, and generating a preliminary routing table;
s2: based on the preliminary routing table, a Bayesian decision theory is utilized to dynamically select a proper communication protocol, so as to obtain an optimized routing and protocol configuration table;
s3: performing edge calculation processing on the data by adopting fast Fourier transform according to the optimized route and protocol configuration table to generate an edge calculation result set and a cloud data packet to be transmitted;
s4: carrying out multi-mode data fusion on the edge calculation result set by using a convolutional neural network to obtain fused edge data;
s5: based on the fused edge data, adopting a reinforcement learning algorithm to perform collaborative learning optimization data transmission to obtain a collaborative optimization data transmission strategy;
s6: and (3) combining the cooperatively optimized data transmission strategy, adopting a correlation rule mining algorithm to perform time-space data correlation analysis, and generating a final optimized data transmission plan.
Firstly, in the step S1, the shortest path algorithm is used for evaluating the data transmission path, so that the time delay and the energy consumption of data transmission can be reduced, and the efficiency of data transmission is improved. This helps to ensure timeliness and reliability of the data.
Then, in step S2, a bayesian decision theory is adopted to dynamically select a communication protocol, so that the communication mode can be optimized according to the actual network condition and the equipment performance. This makes the data transmission more adaptable to changing environments, improving the adaptability and stability of the network.
The fast Fourier transform and the edge calculation in the step S3 can efficiently process data at the equipment end, and reduce the dependence on a central server, thereby reducing the pressure and the cost of data transmission. This contributes to an increase in data processing speed and a reduction in energy consumption.
In the step S4, the convolutional neural network is used for multi-mode data fusion, so that the data of different sensors can be effectively combined, richer and comprehensive information is provided, and the quality of data analysis and decision making is improved.
In step S5, a reinforcement learning algorithm is adopted to perform collaborative learning to optimize data transmission, and the devices can jointly learn and optimize a data transmission strategy according to the behaviors and performances of each other. This allows the network to adaptively provide the best transmission mode, improving the intelligence and efficiency of the network.
Finally, in step S6, a correlation rule mining algorithm is used to perform a time-space data correlation analysis, so that a correlation rule between data can be found, and a data transmission plan can be planned better. This helps to improve the correlation and utility value of the data.
Referring to fig. 2, based on the data of the internet of things device, the performance of the data transmission path is evaluated by adopting a shortest path algorithm, and the step of generating the preliminary routing table specifically includes:
s101: collecting geographic position and network topology information of the Internet of things equipment by adopting a node detection method to obtain a network topology graph;
s102: based on the network topology diagram, a shortest path algorithm is adopted to evaluate paths from a source node to a destination node, and a shortest path table among the nodes is obtained;
s103: based on the shortest path table among the nodes, adopting a delay calculation method to evaluate the communication delay and cost of the paths to obtain a communication path performance table;
s104: based on the communication path performance table, a path selection algorithm is adopted to select a path with optimal performance, and a preliminary routing table is generated.
First, in step S101, a node detection method is used to collect geographic location and network topology information of an internet of things device. This is beneficial in constructing accurate network topology graphs so that subsequent route calculations can be based on actual device locations and connectivity relationships, rather than on theoretical or static information. This helps to improve the accuracy and practicality of the routing table.
Next, in step S102, a shortest path algorithm is employed to evaluate a path from a source node to a destination node based on the network topology. The method has the advantages that the shortest path can be calculated quickly, and the time delay and the energy consumption of data transmission are reduced. This helps to ensure the efficiency of data transmission and a fast response.
In step S103, the communication delay and cost of the path are evaluated by using a delay calculation method. This is beneficial in comprehensively considering a number of factors of path performance, not only path length, but also communication quality and cost. By taking delay and cost into account, routing can be better weighed, ensuring that data transmission is both fast and economical.
Finally, in step S104, a path selection algorithm is adopted based on the communication path performance table to select a path with optimal performance, and a preliminary routing table is generated. The method has the beneficial effects of ensuring high performance of data transmission, enabling data to flow in a network in an optimal mode, and improving efficiency of the whole Internet of things system.
Referring to fig. 3, based on the preliminary routing table, using bayesian decision theory, a suitable communication protocol is dynamically selected, and the steps for obtaining the optimized routing and protocol configuration table are specifically as follows:
s201: based on the preliminary routing table, adopting a data sampling method to collect past communication data and error rate to obtain a historical communication data table;
s202: based on the historical communication data table, evaluating the applicability of the communication protocol by adopting a Bayesian statistical model to obtain a communication protocol applicability table;
S203: based on the communication protocol applicability table, adopting a decision tree algorithm to dynamically select the most suitable communication protocol to obtain an optimized communication protocol;
s204: based on the optimized communication protocol and the preliminary routing table, a merging strategy method is adopted to generate an optimized routing and protocol configuration table.
First, in step S201, past communication data and error rates are collected by a data sampling method, and a history communication data table is obtained. This is beneficial in establishing an accurate knowledge of past communication performance, allowing decisions to be made based on actual communication history. This helps to better accommodate changes in the actual communication environment.
Next, in step S202, the applicability of the communication protocol is evaluated by using a bayesian statistical model, and a communication protocol applicability table is obtained. The advantage of this step is that the performance of the different protocols can be estimated from historical data and statistical models to more accurately select the applicable protocol. This contributes to improvement in communication quality and stability.
In step S203, a decision tree algorithm is used to dynamically select the most appropriate communication protocol. This is beneficial in making decisions based on real-time communication requirements and protocol suitability, thereby ensuring that the best protocol is selected in each communication scenario. This contributes to improvement in communication efficiency and adaptability.
Finally, in step S204, an optimized route and protocol configuration table is generated by adopting a merging strategy method based on the optimized communication protocol and the preliminary route table. The advantage of this step is that it integrates protocol selection and route optimization, ensuring that data transmission takes place in an optimal way throughout the network. This helps to improve network performance and response speed.
Referring to fig. 4, according to the optimized routing and protocol configuration table, the edge calculation processing is performed on the data by adopting the fast fourier transform, and the steps of generating the edge calculation result set and the cloud data packet to be transmitted are specifically as follows:
s301: based on the optimized route and protocol configuration table, acquiring the to-be-processed equipment data of the Internet of things by adopting a data flow interception method to obtain an original data set;
s302: based on an original data set, preprocessing data by adopting a wavelet transformation method, including filtering and denoising, so as to obtain a preprocessed data set;
s303: based on the preprocessing data set, adopting a fast Fourier transform method to realize edge calculation, and generating an edge calculation result set and a cloud data packet to be transmitted;
s304: based on the edge calculation result set and the cloud data packet to be transmitted, a data packaging method is adopted, and distribution is carried out according to the optimized route and protocol configuration table, so that edge calculation data distribution is completed.
Firstly, in step S301, data of the internet of things device to be processed is collected by a data stream interception method, so as to obtain an original data set. This is beneficial to ensure that the system is able to capture and process real-time data from the device in a timely manner. This is a key step in ensuring real-time and data accuracy.
Next, in step S302, the raw data is preprocessed using wavelet transform, including filtering and denoising. This is beneficial to improving the quality and stability of the data, reducing uncertainty and errors in the edge computation phase. This helps to improve the accuracy of data analysis and computation.
In step S303, based on the preprocessed data set, edge calculation is performed by using a fast fourier transform method, and an edge calculation result set and a cloud data packet to be transmitted are generated. The method has the advantages that the calculation is pushed to the equipment end, the dependence on a central server is reduced, and the pressure and the cost of data transmission are reduced. Meanwhile, the edge calculation can quickly generate a useful calculation result, and the response speed is improved.
Finally, in step S304, based on the edge calculation result set and the cloud data packet to be transmitted, a data packaging method is adopted, and distribution is performed according to the optimized routing and protocol configuration table. This is beneficial for transmitting data in an optimal manner, depending on network configuration and performance requirements. This helps to improve data transmission efficiency and reliability.
Referring to fig. 5, the steps of performing multi-mode data fusion on the edge calculation result set by using a convolutional neural network to obtain fused edge data are specifically as follows:
s401: preprocessing the multi-mode data by adopting a data standardization method based on the edge calculation result set to generate a standardized multi-mode data set;
s402: extracting key information features by adopting a feature extraction algorithm based on a standardized multi-mode data set to obtain a key feature data set;
s403: based on the key characteristic data set, designing a convolutional neural network structure to perform multi-mode data fusion to obtain a CNN model structure;
s404: and training and fusing the key characteristic data set based on the CNN model structure to obtain fused edge data.
First, in step S401, the multimodal data is preprocessed by a data normalization method to generate a normalized multimodal data set. This is beneficial to ensure that the data of the different modalities has a consistent data format and range for efficient feature extraction and fusion. Data normalization helps to improve data consistency and comparability.
Next, in step S402, a feature extraction algorithm is used to extract key information features from the normalized multi-modal dataset, resulting in a key feature dataset. This is beneficial in reducing redundancy of the data, extracting the most relevant information, and providing more useful data for subsequent multi-modal data fusion.
In step S403, a convolutional neural network structure is designed to perform multi-modal data fusion, so as to obtain a CNN model structure. The method has the advantages that the deep learning technology can be fully utilized to process the complex relation of the multi-mode data, and the efficiency and the accuracy of data fusion are improved.
Finally, in step S404, training and fusing the key feature data set based on the CNN model structure, to obtain fused edge data. This is beneficial to combining information of different modalities, thereby achieving a more comprehensive and rich data expression. This helps to improve the information density and quality of the data.
Referring to fig. 6, based on the fused edge data, collaborative learning is performed to optimize data transmission by using a reinforcement learning algorithm, and the steps of obtaining a collaborative optimized data transmission policy are specifically as follows:
s501: defining a reinforcement learning state, action and rewarding mechanism based on the fused edge data to obtain reinforcement learning environment definition;
s502: based on the reinforcement learning environment definition, performing preliminary learning by adopting a Q-learning algorithm to generate a preliminary strategy network;
s503: based on the preliminary policy network, performing policy iteration, and continuously optimizing data transmission to obtain an optimized policy network;
S504: based on the optimized strategy network, the efficiency and accuracy of data transmission are evaluated, and a cooperatively optimized data transmission strategy is obtained.
First, in step S501, based on the fused edge data, the state, action and rewarding mechanism of reinforcement learning are defined, and the definition of reinforcement learning environment is obtained. This is beneficial in creating a clear problem description and learning framework so that reinforcement learning algorithms can understand the goals and limitations of data transmission. This helps to ensure that the learning process is performed efficiently.
Next, in step S502, a Q-learning algorithm is used to perform preliminary learning, and a preliminary policy network is generated. The advantage of this step is that the possibility of data transmission strategies can be explored initially and decisions optimized according to the rewarding mechanism. This is the initial stage of reinforcement learning, which is used to build the learning basis.
In step S503, policy iteration is performed based on the preliminary policy network, and continuous optimization is performed on data transmission, so as to obtain an optimized policy network. The advantage of this step is that the efficiency and performance of the data transmission strategy can be increased by continuous learning and improvement. The iteration of the strategy helps to continuously adapt to different communication environments and data requirements.
Finally, in step S504, based on the optimized policy network, the efficiency and accuracy of data transmission are evaluated, and a cooperatively optimized data transmission policy is obtained. This is beneficial to generating an optimal data transmission strategy to meet the performance requirements of the internet of things system and to optimize the speed and quality of data transmission.
Referring to fig. 7, in combination with a co-optimized data transmission policy, a correlation rule mining algorithm is adopted to perform a time-space data correlation analysis, and the steps of generating a final optimized data transmission plan are specifically as follows:
s601: based on a data transmission strategy of collaborative optimization, carrying out data preprocessing to obtain a preprocessed data set;
s602: based on the preprocessed data set, generating a preliminary association rule by adopting an Apriori algorithm to obtain a preliminary association rule set;
s603: based on the preliminary association rule set, a confidence and support filtering method is adopted to screen high association rules, and a high association rule set is obtained;
s604: and based on the high association rule set, combining the characteristics of the time-space data, carrying out deep association analysis, and generating a final optimized data transmission plan.
First, in step S601, data preprocessing is performed based on a co-optimized data transmission policy, to obtain a preprocessed data set. This is beneficial to ensure that the data sets are consistent and available for association rule mining. Data preprocessing may include operations such as data cleansing, deduplication, and formatting to improve the quality and usability of the data.
Next, in step S602, an Apriori algorithm is used to generate a preliminary association rule, so as to obtain a preliminary association rule set. This step helps to discover potential associations between data, knowing frequent combinations between data items. Preliminary association rules may be used as a basis for further screening and optimization.
In step S603, based on the preliminary association rule set, a confidence and support filtering method is adopted to screen out association rules with high confidence and high support, and a high association rule set is obtained. This is beneficial to eliminate less relevant or less important rules, improving the accuracy and efficiency of the analysis.
Finally, in step S604, based on the high association rule set, and in combination with the time-space data characteristics, deep association analysis is performed to generate a final optimized data transmission plan. This step helps to combine the data transfer plan with time and space factors to better meet the actual needs. By analyzing the relevance among the data items, the time and the position of data transmission can be better arranged, and the transmission efficiency and the resource utilization rate are improved.
Referring to fig. 8, the data interaction system of the internet of things is configured to execute the data interaction method of the internet of things, where the data interaction system of the internet of things includes a route optimization module, a protocol selection module, an edge calculation module, a data fusion module, and a data transmission optimization module.
The route optimization module collects network topology by adopting a node detection method based on the equipment information of the Internet of things, and performs path evaluation by using a shortest path algorithm to generate a preliminary routing table;
the protocol selection module collects communication data by adopting a data sampling method based on the preliminary routing table, and adopts a Bayesian statistical model to evaluate the applicability of the communication protocol to generate an optimized routing and protocol configuration table;
the edge calculation module collects equipment data by adopting a data flow interception method based on the optimized routing and protocol configuration table, performs edge calculation by using a fast Fourier transform method, and generates an edge calculation result set and a cloud data packet to be transmitted;
the data fusion module is used for preprocessing by adopting a data standardization method based on the edge calculation result set, and carrying out data fusion by utilizing a convolutional neural network structure to obtain fused edge data;
the data transmission optimization module defines a reinforcement learning environment based on the fused edge data, optimizes data transmission through a Q-learning algorithm, and generates a collaborative optimized data transmission strategy.
Firstly, the implementation of the route optimization module can greatly improve the efficiency of data transmission by adopting a node detection method, the collection of network topology and the application of a shortest path algorithm. By generating the preliminary routing table, the system can more effectively determine the transmission path of the data, reduce the delay of data transmission and network congestion, and further improve the stability and reliability of data transmission.
Second, implementation of the protocol selection module helps the system dynamically select the most appropriate communication protocol. According to the evaluation method based on the Bayesian statistical model, the system can select the most suitable communication protocol according to the current network environment and equipment requirements, so that communication overhead is reduced, and communication quality and efficiency are improved.
The application of the edge computing module allows the system to process data near the Internet of things equipment, and the requirement of data transmission to the cloud is reduced. The edge calculation is performed by using methods such as fast Fourier transform, so that a calculation result can be generated more quickly, the delay of data transmission is reduced, the bandwidth occupation is reduced, and the instantaneity is improved.
Next, the implementation of the data fusion module fuses data from different devices into one body through the application of data normalization and convolutional neural networks. This helps to eliminate data inconsistencies, improving data consistency and availability, and enabling more accurate and efficient subsequent data processing.
Finally, implementation of the data transmission optimization module may continually optimize the data transmission strategy by defining a reinforcement learning environment and using a Q-learning algorithm. This means that the system can automatically adjust parameters and policies of data transmission according to actual conditions to adapt to changing network environments and demands, thereby maximizing efficiency and resource utilization of data transmission.
Referring to fig. 9, the route optimization module includes a node detection sub-module, a path evaluation sub-module, a delay calculation sub-module, and a path selection sub-module;
the protocol selection module comprises a data sampling sub-module, a Bayesian evaluation sub-module, a decision tree selection sub-module and a merging strategy sub-module;
the edge computing module comprises a data interception sub-module, a data preprocessing sub-module, an edge computing sub-module and a data distribution sub-module;
the data fusion module comprises a data standardization sub-module, a feature extraction sub-module, a network design sub-module and a data fusion sub-module;
the data transmission optimization module comprises a reinforcement learning environment definition sub-module, a Q-learning sub-module, a strategy iteration sub-module and a data transmission evaluation sub-module.
Route optimization module:
and the node detection submodule: the network topology is collected by the node detection method, so that the position and connection information of the equipment of the Internet of things can be accurately mastered, and the manageability and maintainability of the network are improved.
Path evaluation sub-module: and the shortest path algorithm is used for path evaluation, so that delay of data transmission is reduced, the data can reach a destination rapidly, and the efficiency and stability of the data transmission are improved.
A delay computation sub-module: the delay calculation can help the system to better know the data transmission delay in the network and optimize the route selection, thereby improving the real-time performance and quality of data transmission.
A path selection sub-module: the application of the path selection sub-module is beneficial to selecting the optimal data transmission path according to the path evaluation result, so that the risk of network congestion is reduced, and the reliability of data transmission is improved.
A protocol selection module:
data sampling submodule: communication data in the network can be known in real time through data sampling, actual data support is provided for protocol selection, and the system is helped to dynamically select a communication protocol according to data quantity and requirements.
Bayesian evaluation submodule: the communication protocol applicability is evaluated by using the Bayesian statistical model, the most suitable protocol can be deduced according to actual conditions, and the communication intelligence and adaptability are improved.
Decision tree selection submodule: the application of the decision tree selection sub-module facilitates the system in making decisions based on a number of factors (e.g., bandwidth, delay, etc.) to select an optimal communication protocol.
Combining policy sub-modules: the implementation of the merging strategy sub-module can merge different protocol selection strategies together, so that the flexibility and decision performance of the system are improved.
And an edge calculation module:
a data interception sub-module: through data interception, the system can capture data generated by the Internet of things equipment, unnecessary data transmission is reduced, and network load is reduced.
And a data preprocessing sub-module: the data preprocessing is beneficial to cleaning and normalizing the data, improves the quality and usability of the data, and reduces the error of subsequent calculation.
An edge calculation sub-module: the edge calculation can process real-time data nearby the Internet of things equipment, so that the requirement of data transmission to the cloud is reduced, and the instantaneity and efficiency are improved.
A data distribution sub-module: the data distribution sub-module is beneficial to transmitting the processed data to a correct destination, ensures the data to be transmitted as required, and reduces the risk of network congestion.
And a data fusion module:
data normalization sub-module: data normalization helps to ensure that data from different sources has consistent formats and units, providing a consistent data basis for subsequent data fusion.
And a feature extraction sub-module: the feature extraction is beneficial to extracting key information from the data, reduces the dimension of the data and improves the efficiency and accuracy of subsequent processing.
Network design submodule: the network design sub-module can help the system to construct convolutional neural network and other structures so as to better perform multi-mode data fusion.
And a data fusion sub-module: the implementation of the data fusion sub-module enables the system to integrate data from different sources into a whole, and provides richer and comprehensive data expression.
And a data transmission optimization module:
reinforcement learning environment definition sub-module: defining a reinforcement learning environment helps the system establish a learning framework for data transmission, so that the data transmission strategy can be automatically optimized according to actual conditions.
Q-learning submodule: the application of the Q-learning algorithm enables the system to continuously learn and optimize the data transmission strategy according to the rewarding mechanism, and improves the self-adaptability of the system.
Strategy iteration submodule: the strategy iteration sub-module is helpful for continuously improving and adjusting the data transmission strategy, and ensures that the strategy is suitable for the continuously changing network environment and requirements.
A data transmission evaluation sub-module: the implementation of the data transmission evaluation sub-module can help the system evaluate the efficiency and accuracy of data transmission in real time, so that the data transmission is fed back to the reinforcement learning algorithm for optimization.
The present invention is not limited to the above embodiments, and any equivalent embodiments which can be changed or modified by the technical disclosure described above can be applied to other fields, but any simple modification, equivalent changes and modification made to the above embodiments according to the technical matter of the present invention will still fall within the scope of the technical disclosure.

Claims (5)

1. The data interaction method of the Internet of things is characterized by comprising the following steps of:
based on the equipment data of the Internet of things, evaluating the performance of a data transmission path by adopting a shortest path algorithm, and generating a preliminary routing table;
based on the preliminary routing table, dynamically selecting a proper communication protocol by using a Bayesian decision theory to obtain an optimized routing and protocol configuration table;
performing edge calculation processing on the data by adopting fast Fourier transform according to the optimized route and protocol configuration table to generate an edge calculation result set and a cloud data packet to be transmitted;
performing multi-mode data fusion on the edge calculation result set by using a convolutional neural network to obtain fused edge data;
based on the fused edge data, adopting a reinforcement learning algorithm to perform collaborative learning optimization data transmission to obtain a collaborative optimization data transmission strategy;
combining the cooperatively optimized data transmission strategy, adopting a correlation rule mining algorithm to perform time-space data correlation analysis, and generating a final optimized data transmission plan;
based on the preliminary routing table, a Bayesian decision theory is utilized to dynamically select a proper communication protocol, and the steps of obtaining an optimized routing and protocol configuration table are specifically as follows:
Based on the preliminary routing table, adopting a data sampling method to collect past communication data and error rate to obtain a historical communication data table;
based on the historical communication data table, evaluating the applicability of the communication protocol by adopting a Bayesian statistical model to obtain a communication protocol applicability table;
based on the communication protocol applicability table, adopting a decision tree algorithm to dynamically select an optimal communication protocol to obtain an optimized communication protocol;
generating an optimized route and protocol configuration table by adopting a merging strategy method based on the optimized communication protocol and the preliminary route table;
according to the optimized route and protocol configuration table, performing edge calculation processing on data by adopting fast Fourier transform, and generating an edge calculation result set and a cloud data packet to be transmitted specifically comprises the following steps:
based on the optimized route and protocol configuration table, acquiring the data of the to-be-processed Internet of things equipment by adopting a data flow interception method to obtain an original data set;
based on the original data set, preprocessing data by adopting a wavelet transformation method, including filtering and denoising, so as to obtain a preprocessed data set;
based on the preprocessing data set, adopting a fast Fourier transform method to realize edge calculation, and generating an edge calculation result set and a cloud data packet to be transmitted;
Based on the edge calculation result set and the cloud data packet to be transmitted, distributing according to the optimized route and protocol configuration table by adopting a data packaging method to finish edge calculation data distribution;
the convolution neural network is utilized to conduct multi-mode data fusion on the edge calculation result set, and the step of obtaining the fused edge data specifically comprises the following steps:
preprocessing the multi-mode data by adopting a data standardization method based on the edge calculation result set to generate a standardized multi-mode data set;
extracting key information features by adopting a feature extraction algorithm based on the standardized multi-mode data set to obtain a key feature data set;
based on the key characteristic data set, designing a convolutional neural network structure to perform multi-mode data fusion to obtain a CNN model structure;
training and fusing the key feature data set based on the CNN model structure to obtain fused edge data;
and combining the cooperatively optimized data transmission strategy, and carrying out time-space data association analysis by adopting an association rule mining algorithm, wherein the step of generating a final optimized data transmission plan comprises the following specific steps:
based on the cooperatively optimized data transmission strategy, carrying out data preprocessing to obtain a preprocessed data set;
Based on the preprocessed data set, generating a preliminary association rule by adopting an Apriori algorithm to obtain a preliminary association rule set;
based on the preliminary association rule set, a confidence and support filtering method is adopted to screen high association rules, and a high association rule set is obtained;
and based on the high association rule set, combining the time-space data characteristics, carrying out deep association analysis, and generating a final optimized data transmission plan.
2. The internet of things data interaction method according to claim 1, wherein based on internet of things equipment data, a shortest path algorithm is adopted to evaluate performance of a data transmission path, and the step of generating a preliminary routing table is specifically:
collecting geographic position and network topology information of the Internet of things equipment by adopting a node detection method to obtain a network topology graph;
based on the network topology diagram, a shortest path algorithm is adopted to evaluate paths from a source node to a destination node, and a shortest path table among the nodes is obtained;
based on the shortest path table among the nodes, adopting a delay calculation method to evaluate the communication delay and cost of the paths to obtain a communication path performance table;
and selecting a path with optimal performance by adopting a path selection algorithm based on the communication path performance table to generate a preliminary routing table.
3. The internet of things data interaction method according to claim 1, wherein based on the fused edge data, collaborative learning optimization data transmission is performed by adopting a reinforcement learning algorithm, and the step of obtaining a collaborative optimization data transmission policy specifically comprises the following steps:
defining a reinforcement learning state, action and rewarding mechanism based on the fused edge data to obtain reinforcement learning environment definition;
based on the reinforcement learning environment definition, performing preliminary learning by adopting a Q-learning algorithm to generate a preliminary strategy network;
performing strategy iteration based on the preliminary strategy network, and continuously optimizing data transmission to obtain an optimized strategy network;
and based on the optimized strategy network, evaluating the efficiency and accuracy of data transmission to obtain a cooperatively optimized data transmission strategy.
4. The data interaction system of the Internet of things is characterized by comprising a route optimization module, a protocol selection module, an edge calculation module, a data fusion module and a data transmission optimization module;
the route optimization module collects network topology by adopting a node detection method based on the equipment information of the Internet of things, and performs path evaluation by using a shortest path algorithm to generate a preliminary routing table;
The protocol selection module collects communication data by adopting a data sampling method based on the preliminary routing table, and adopts a Bayesian statistical model to evaluate the applicability of the communication protocol to generate an optimized routing and protocol configuration table;
the edge calculation module collects equipment data by adopting a data flow interception method based on the optimized routing and protocol configuration table, performs edge calculation by using a fast Fourier transform method, and generates an edge calculation result set and a cloud data packet to be transmitted;
the data fusion module is used for preprocessing by adopting a data standardization method based on the edge calculation result set, and carrying out data fusion by utilizing a convolutional neural network structure to obtain fused edge data;
the data transmission optimization module defines a reinforcement learning environment based on the fused edge data, optimizes data transmission through a Q-learning algorithm, and generates a collaborative optimized data transmission strategy.
5. The internet of things data interaction system according to claim 4, wherein the route optimization module comprises a node detection sub-module, a path evaluation sub-module, a delay calculation sub-module and a path selection sub-module;
the protocol selection module comprises a data sampling sub-module, a Bayesian evaluation sub-module, a decision tree selection sub-module and a merging strategy sub-module;
The edge computing module comprises a data interception sub-module, a data preprocessing sub-module, an edge computing sub-module and a data distribution sub-module;
the data fusion module comprises a data standardization sub-module, a feature extraction sub-module, a network design sub-module and a data fusion sub-module;
the data transmission optimization module comprises a reinforcement learning environment definition sub-module, a Q-learning sub-module, a strategy iteration sub-module and a data transmission evaluation sub-module.
CN202311421852.3A 2023-10-31 2023-10-31 Internet of things data interaction method and system Active CN117155845B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311421852.3A CN117155845B (en) 2023-10-31 2023-10-31 Internet of things data interaction method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311421852.3A CN117155845B (en) 2023-10-31 2023-10-31 Internet of things data interaction method and system

Publications (2)

Publication Number Publication Date
CN117155845A CN117155845A (en) 2023-12-01
CN117155845B true CN117155845B (en) 2024-01-23

Family

ID=88903057

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311421852.3A Active CN117155845B (en) 2023-10-31 2023-10-31 Internet of things data interaction method and system

Country Status (1)

Country Link
CN (1) CN117155845B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117395166B (en) * 2023-12-11 2024-02-13 四海良田(天津)智能科技有限公司 Intelligent agricultural management platform based on Internet of things

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113315806A (en) * 2021-04-14 2021-08-27 深圳大学 Multi-access edge computing architecture for cloud network fusion
CN116739389A (en) * 2023-08-14 2023-09-12 广东创能科技股份有限公司 Smart city management method and system based on cloud computing

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10223169B2 (en) * 2016-12-28 2019-03-05 Intel Corporation Technologies for adaptive collaborative optimization of internet-of-things systems
US11429474B2 (en) * 2020-08-03 2022-08-30 Bank Of America Corporation Enterprise IOT system for onboarding and maintaining peripheral devices

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113315806A (en) * 2021-04-14 2021-08-27 深圳大学 Multi-access edge computing architecture for cloud network fusion
CN116739389A (en) * 2023-08-14 2023-09-12 广东创能科技股份有限公司 Smart city management method and system based on cloud computing

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
面向边缘计算的分布式深度神经网络研究;邹颖;中国优秀硕士学位论文全文数据库;全文 *

Also Published As

Publication number Publication date
CN117155845A (en) 2023-12-01

Similar Documents

Publication Publication Date Title
CN107871164B (en) Fog computing environment personalized deep learning method
Peralta et al. Fog computing based efficient IoT scheme for the Industry 4.0
CN117155845B (en) Internet of things data interaction method and system
Lu The current status and developing trends of Industry 4.0: A Review
CN104469833B (en) A kind of heterogeneous network operation management method based on user's perception
CN108924198A (en) A kind of data dispatching method based on edge calculations, apparatus and system
CN113435472A (en) Vehicle-mounted computing power network user demand prediction method, system, device and medium
CN103177298A (en) Smart energy network control method
Thantharate et al. ADAPTIVE6G: Adaptive resource management for network slicing architectures in current 5G and future 6G systems
Yu et al. A multi-objective metaheuristics study on solving constrained relay node deployment problem in WSNS
Huang Quality of service optimization in wireless transmission of industrial Internet of Things for intelligent manufacturing
Tedesco et al. A preliminary discussion of measurement and networking issues in cyber physical systems for industrial manufacturing
Feng et al. Bidirectional green promotion of 6G and AI: architecture, solutions, and platform
AU2020103373A4 (en) Machine learning based network intelligentization for automatically- configurable cellular communication systems
CN111077851A (en) Chemical process fault diagnosis system based on gas chromatography fog calculation framework
CN112215326B (en) Distributed AI system
Kurniawan et al. Mobile computing and communications-driven fog-assisted disaster evacuation techniques for context-aware guidance support: A survey
Jiang et al. Human-Cyber-Physical Ubiquitous Intelligent Communication: System Architecture, Key Technologies, and Challenges
Lu et al. Framework of industrial networking sensing system based on edge computing and artificial intelligence
CN111445125B (en) Agricultural robot computing task cooperation method, system, medium and equipment
Xiao et al. Mobile-edge-platooning cloud: a lightweight cloud in vehicular networks
CN113204915A (en) PHM design method based on CPS
Lin et al. Hypergraph-Based Autonomous Networks: Adaptive Resource Management and Dynamic Resource Scheduling
Wang et al. Self-maintained Network Digital Twin for Human-centric Wireless Metaverse
Gu et al. Consensus control and feedback graph co-design for MIMO discrete-time multi-agent systems

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant