Data transmission method and device of Internet of Things equipment, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a data transmission method and device of Internet of Things equipment, electronic equipment and a storage medium, and relates to the technical field of Internet of Things, and the method comprises the steps: obtaining business data collected by the Internet of Things equipment, and determining the target business priority of the business data; acquiring network index data of communication equipment connected with the Internet of Things equipment; performing data volume prediction according to the service data to obtain a corresponding predicted data volume; performing transmission mode selection according to the predicted data volume and the network index data to obtain a first transmission strategy for the service data; and issuing the first transmission strategy to the Internet of Things equipment, so that the Internet of Things equipment transmits the service data according to the first transmission strategy, thereby optimizing the data transmission mode, realizing dynamic adjustment of the transmission strategy, improving the network adaptability, improving the success rate of data transmission, and reducing the probability of network congestion.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Things, and in particular to a data transmission method for Internet of Things devices, a data transmission device for Internet of Things devices, an electronic device, and a computer-readable storage medium. Background Art

[0002] Light commercial IoT devices typically refer to Internet of Things (IoT) devices designed for lightweight commercial scenarios. These devices are characterized by low cost, easy deployment, and low power consumption. These devices can upload data through data transmission units. However, this data upload process presents at least the following issues: fixed transmission frequency: regardless of the scenario, the device uploads data at fixed intervals, which can lead to data waste; poor network adaptability: in areas with weak signals, the device will still attempt high-frequency transmission, which can easily lead to data transmission failures and power consumption; and a single traffic processing method: unable to handle burst traffic, which can easily lead to network congestion. Summary of the Invention

[0003] The embodiments of the present invention provide a data transmission method, apparatus, electronic device, and computer-readable storage medium for IoT devices to solve or partially solve the problems of fixed transmission frequency, poor network adaptability, and single traffic processing method in data transmission of IoT devices.

[0004] An embodiment of the present invention discloses a data transmission method for an Internet of Things device, comprising:

[0005] Obtaining business data collected by IoT devices and determining target business priorities for the business data;

[0006] Obtain network indicator data of the communication device connected to the IoT device;

[0007] Performing data volume prediction based on the business data to obtain a corresponding predicted data volume;

[0008] Selecting a transmission mode according to the predicted data volume and the network indicator data to obtain a first transmission strategy for the business data;

[0009] The first transmission policy is issued to the IoT device, so that the IoT device transmits the service data according to the first transmission policy.

[0010] The embodiment of the present invention further discloses a data transmission device for an Internet of Things device, comprising:

[0011] A data acquisition module is used to acquire business data collected by IoT devices and determine the target business priority of the business data;

[0012] A network indicator acquisition module is used to obtain network indicator data of the communication device connected to the Internet of Things device;

[0013] A data volume prediction module is used to predict the data volume based on the business data and obtain the corresponding predicted data volume;

[0014] a transmission mode determination module, configured to select a transmission mode according to the predicted data volume and the network indicator data, and obtain a first transmission strategy for the service data;

[0015] The data transmission module is used to send the first transmission strategy to the IoT device, so that the IoT device transmits the business data according to the first transmission strategy.

[0016] An embodiment of the present invention further discloses an electronic device, comprising a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus;

[0017] The memory is used to store computer programs;

[0018] The processor is configured to implement the method described in the embodiment of the present invention when executing the program stored in the memory.

[0019] An embodiment of the present invention further discloses a computer-readable storage medium having instructions stored thereon. When executed by one or more processors, the processors are enabled to execute the method according to the embodiment of the present invention.

[0020] The embodiments of the present invention include the following advantages:

[0021] In an embodiment of the present invention, in an Internet of Things scenario, an edge device can obtain business data collected by the Internet of Things device, determine the target business priority of the business data, and simultaneously obtain network indicator data of the communication device to which the Internet of Things device is connected. Then, the data volume is predicted based on the business data to obtain the corresponding predicted data volume. Then, a transmission method is selected based on the predicted data volume and the network indicator data to obtain a first transmission strategy for the business data. The first transmission strategy is then sent to the Internet of Things device so that the Internet of Things device transmits the business data according to the first transmission strategy. Thus, in the process of data transmission by the Internet of Things device, by predicting the amount of data to be transmitted and determining the transmission strategy based on the prediction results and network indicator data, traffic demand prediction is achieved, which is conducive to selecting a suitable transmission method for transmitting business data according to different needs, optimizing the data transmission method, and at the same time, dynamically adjusting the transmission strategy, improving network adaptability, thereby improving the success rate of data transmission, and reducing the probability of network congestion. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 This is a flowchart of a data transmission method for an Internet of Things device provided in an embodiment of the present invention;

[0023] Figure 2 Schematic diagram of the device communication architecture provided in an embodiment of the present invention;

[0024] Figure 3 is a schematic diagram of a data transmission process provided in an embodiment of the present invention;

[0025] Figure 4 This is a structural block diagram of a data transmission device for an Internet of Things device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0026] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0027] As an example, the corresponding light business IoT devices can upload data through the data transmission unit. In the process of uploading data, there are at least the following problems: fixed transmission frequency: regardless of the scenario, the device uploads data at fixed time intervals, which can easily waste traffic; poor network adaptability: in areas with weak signals, the device will still try to transmit at a high frequency, which can easily lead to data transmission failure and power consumption; single traffic processing method: unable to process burst traffic, which can easily lead to network congestion.

[0028] In this regard, in the present invention, in the Internet of Things scenario, the edge device can obtain the business data collected by the Internet of Things device, and determine the target business priority of the business data, and at the same time obtain the network indicator data of the communication device connected to the Internet of Things device, and then predict the data volume based on the business data to obtain the corresponding predicted data volume, and then select the transmission method based on the predicted data volume and the network indicator data to obtain a first transmission strategy for the business data, and then send the first transmission strategy to the Internet of Things device, so that the Internet of Things device transmits the business data according to the first transmission strategy, so that in the process of transmitting data by the Internet of Things device, by predicting the amount of data to be transmitted and determining the transmission strategy based on the prediction results and network indicators and other data, the traffic demand is predicted, which is conducive to selecting a suitable transmission method for transmitting business data according to different needs, optimizing the data transmission method, and at the same time, dynamically adjusting the transmission strategy, improving network adaptability, thereby improving the success rate of data transmission and reducing the probability of network congestion.

[0029] Reference Figure 1 , shows a flowchart of a data transmission method for an Internet of Things device provided in an embodiment of the present invention, which may specifically include the following steps:

[0030] Step 101: Acquire service data collected by IoT devices and determine the target service priority of the service data;

[0031] In an IoT system, IoT devices, edge devices, and cloud servers form a layered, collaborative architecture. With clear divisions of labor and close collaboration among the three, they achieve a closed loop from data collection to intelligent decision-making. IoT devices can collect relevant business data in real time and transmit it to edge devices. Edge devices filter and process the business data sent by IoT devices, then upload the results to the cloud server, which stores it, coordinates across regions, and issues global policies. This is not a limitation of the present invention.

[0032] Among them, the Internet of Things device can be a device used to perform data collection or execute instructions, such as temperature and humidity sensors, smart door locks, monitoring equipment, etc.; the edge device can be an edge gateway, edge server, etc., which is not limited by the present invention.

[0033] In one example, referring to Figure 2 , which shows a schematic diagram of the device communication architecture provided in an embodiment of the present invention. The processing flow between IoT devices, edge gateways / nodes, and cloud servers can be as follows: 1. Upload raw data; 2. Receiving transmission strategies; 3. Aggregating gradients and regional models; 4. Receiving dynamic protocol instructions; 5. Distributing global model parameters; 6. Local model updates; 7. Regional model training; 8. Global model iteration. Specifically, a corresponding local model can be deployed on the IoT device, a corresponding regional model (such as a traffic prediction model or a network prediction model) can be deployed on the edge gateway / node, and a corresponding cloud-based large model can be deployed on the cloud server. The IoT device can use the local model to preprocess service data, the edge gateway / node can use the regional model to perform gradient aggregation, optimization, and traffic prediction, and the cloud server can use the cloud-based large model to perform global state prediction.

[0034] In the specific implementation, "raw data uploading" refers to the IoT device uploading the corresponding business data to the edge gateway; "receiving transmission strategy" refers to the IoT device dynamically receiving the transmission strategy returned by the edge gateway in the process of transmitting business data to the edge gateway, and transmitting the business data according to the transmission strategy to ensure the stability and success rate of data transmission; "aggregation gradient and regional model" refers to the edge gateway confirming the direction and amplitude of parameter adjustment in the model training process to determine the corresponding model gradient, and the regional model refers to the deployment of the corresponding network prediction model in the edge gateway, and the process of performing network prediction on the corresponding area through the network prediction model, where the area refers to the range of devices divided based on the same gateway or base station. Usually, groups of devices with similar functions can be divided into the same area to facilitate the management of multiple devices in the same area; "receiving dynamic protocol instructions" refers to the edge The process of edge devices dynamically adjusting the transmission protocol of IoT devices; "issuing global model parameters" refers to the process of cloud servers sending global model parameters to edge devices. The corresponding cloud-based large model can be deployed in the cloud server. Through the cloud-based large model, the resource allocation strategy can be dynamically adjusted, and the edge model (i.e., network prediction model, etc.) on the edge device can be coordinated and linked. In this way, the cloud-based large model can aggregate the edge model gradients corresponding to different regions, and update the global model parameters of the cloud-based large model through an adaptive weighted algorithm, and then send them back to the corresponding edge gateway; "local model update" refers to the process of edge devices updating local models; "regional model training" refers to the process of edge devices performing model training; "global model iteration" refers to the process of cloud servers performing global model iteration. Through the above process, the linkage between IoT devices, edge devices and cloud servers can be achieved.

[0035] In addition, a corresponding dynamic protocol controller can be deployed on the IoT device, through which multi-protocol stack management can be realized, so that the transmission protocol in the data transmission process can be adjusted according to the transmission strategy issued by the edge gateway / node, and dynamic adjustment of the protocol can be realized; accordingly, a semantic communication module can also be deployed on the edge gateway / node, through which feature extraction and compression can be realized, and then based on the compression result, the compressed data can be reported to the cloud server, reducing the amount of data transmission and reducing traffic consumption; and a corresponding feedback optimization module can be deployed on the cloud server, through which real-time monitoring and tuning can be realized, so as to dynamically adjust the corresponding model and improve the rationality and intelligence of traffic management, etc. The present invention does not impose any restrictions on this.

[0036] In a specific implementation, IoT devices can collect corresponding business data and upload it to edge devices, which then perform processing operations on the business data. The business data may include device status data, sensor data, and historical traffic records. Device status data can be used to characterize the state of the IoT device itself, sensor data can be data related to the environment collected by the IoT device, and historical traffic records can be the traffic corresponding to business data previously transmitted by the IoT device to the edge device. The present invention does not impose any restrictions on this.

[0037] After receiving business data, the edge device can first determine the target business priority corresponding to the business data, so as to determine which transmission method to use for the business data based on the target business priority. It should be noted that the process of IoT devices transmitting business data to edge devices is a continuous process. Based on this, when the IoT device is triggered to transmit business data to the edge device, the edge device can dynamically adjust the transmission strategy based on the business data after receiving the business data, so as to transmit the business data according to the adjusted transmission strategy, thereby better ensuring the stability and efficiency of data transmission. Among them, the business priority can be used to represent the priority of business data transmission.

[0038] Optionally, the target business priority of business data can be achieved by configuring specific rules. For example, alarm information can be set to the highest priority. Alarm information represents potential security and operational risks. Heartbeat packets and key parameter monitoring need to be transmitted in real time (higher priority) to ensure system operation, while other historical logs, periodic environmental data (such as daily average humidity), and other data can be processed at a lower priority. For example, different types of business data can be configured through a priority mapping table, and the priority can be dynamically adjusted based on data content analysis (such as keyword matching, etc.) and time sensitivity. The present invention does not limit this.

[0039] Step 102: Obtain network indicator data of the communication device to which the IoT device is connected;

[0040] While obtaining business data, edge devices can also obtain network indicator data of the communication devices connected to the IoT devices. The network indicator data can be used to characterize the current network status of the communication devices.

[0041] The communication devices may be gateways, base stations, etc., and the network indicator data may represent the current network status of the gateways, base stations, etc. Optionally, the network indicator data may include real-time reference signal received power, base station load, channel congestion probability, etc. By obtaining network indicator data, the current network status of IoT devices can be effectively analyzed, allowing edge devices to select appropriate transmission methods for IoT devices based on the current network status, thereby improving the stability and efficiency of data transmission and increasing regional network utilization.

[0042] Step 103: predicting the data volume based on the business data to obtain the corresponding predicted data volume;

[0043] When the edge device receives the business data sent by the IoT device, it can first predict the data volume based on the business data to obtain the corresponding predicted data volume. The predicted data volume can be used to predict the amount of data transmitted by the IoT device within a certain period of time in the future (such as 1 hour, etc.). The edge device can analyze the traffic demand of the IoT device based on the predicted data volume to ensure the stability and efficiency of data transmission.

[0044] In some feasible implementations, business data includes at least device status information and traffic data. The device status information can be used to characterize the device status of the IoT device, which may include device parameters of the IoT device, such as battery power, working mode, transmission rate, etc.; and the traffic data can be data that the IoT device needs to transmit and is related to the device function, such as temperature and humidity data, etc. In the process of predicting the data volume, the edge device can first obtain the traffic prediction model, then normalize the device status information and traffic data to obtain the corresponding feature data, and then input the feature data into the traffic prediction model to obtain the corresponding predicted data volume. The predicted data volume is used to characterize the amount of data expected to be generated by the IoT device within the target time length. The traffic prediction model is used to predict the amount of data that the IoT device needs to transmit, thereby realizing the prediction of traffic demand. Based on the prediction result, it is conducive to selecting a suitable transmission method for data transmission according to different traffic demands, so that the edge device can dynamically adjust the transmission method according to different demands, greatly improving the flexibility of data transmission.

[0045] For example, taking a smart agricultural greenhouse as an example, the device status information may include:

[0046] Battery level: The current remaining power is 60% (affects device battery life and transmission frequency).

[0047] Working mode: such as "energy saving mode" (reducing sensor sampling rate).

[0048] Transmission rate: such as LoRaWAN (Long Range Wide Area Network) (low speed) or 4G (high speed).

[0049] Traffic data can include:

[0050] Temperature and humidity data: collected every 5 minutes, with a single data volume of 0.5KB.

[0051] Light intensity: collected every 10 minutes, with a single data volume of 0.2KB.

[0052] Soil moisture: collected every 30 minutes, with a single data volume of 1KB.

[0053] For the above data, the edge device can uniformly scale parameters of different dimensions to the range of [0, 1]. For example, if the battery power is 60%, the normalization formula can be power / 100, that is, the normalized value can be 0.6; the transmission rate (LoRaWAN), low speed is 0.2, medium speed is 0.5, and high speed is 1. Assuming it is low speed, the normalized value can be 0.2; the temperature and humidity sampling frequency is 5 minutes, the normalization formula can be (1 / sampling interval), that is, the normalized value is 0.2, etc. Optionally, for other parameters, normalization can also be performed based on the corresponding normalization formula to obtain the corresponding normalized value. After obtaining the normalized values ​​corresponding to each parameter, the normalized device status and traffic data can be combined to generate corresponding feature data, such as feature vector = [0.6 (battery charge), 0.2 (transmission rate), 0.2 (temperature and humidity frequency), 0.1 (light frequency), 0.03 (soil frequency)]. This feature vector can then be input into a pre-trained LSTM (Long Short-Term Memory) model (suitable for time series data prediction) to obtain the predicted data volume for the next hour. For example, the model can output: predicted data volume = 12KB (for the next hour), which can be further broken down into: 6KB (0.5KB × 12 times) for temperature and humidity, 1.2KB for light, 2KB for soil, and 2.8KB for metadata. The traffic prediction model can then predict the amount of data that IoT devices need to transmit, thus enabling traffic demand prediction. This prediction facilitates the selection of appropriate transmission methods for different traffic demands, allowing edge devices to dynamically adjust transmission methods based on varying demands, greatly improving data transmission flexibility.

[0054] Step 104: selecting a transmission mode according to the predicted data volume and the network indicator data to obtain a first transmission strategy for the service data;

[0055] After obtaining the corresponding predicted data volume through the traffic prediction model, the edge device can further select the transmission method based on the predicted data volume and network indicator data, and determine the first transmission strategy for the IoT device, so that the IoT device can transmit business data based on the first transmission strategy, thereby determining the transmission strategy based on the prediction results and network indicators and other data, realizing the prediction of traffic demand, which is conducive to selecting the appropriate transmission method for business data transmission according to different needs, optimizing the data transmission method, and at the same time realizing dynamic adjustment of the transmission strategy, improving network adaptability, thereby improving the success rate of data transmission, and reducing the probability of network congestion.

[0056] In some feasible implementations, the network indicator data includes at least the real-time reference signal receiving power, and the first transmission strategy includes at least the first transmission protocol and the first target transmission frequency. The edge device can determine the first target transmission frequency corresponding to the predicted data volume, and select the protocol based on the real-time reference signal receiving power and the target service priority, determine the first transmission protocol for the service data, and then use the first target transmission frequency and the first transmission protocol as the first transmission strategy for the service data.

[0057] Among them, the real-time reference signal received power (RSRP) is used to measure the strength of the base station signal received by the IoT device. The strength of the real-time reference signal received power directly affects the network connection quality and rate of the IoT device. The transmission protocol can be the transmission protocol used by the IoT device when transmitting data, such as 4G, TCP (Transmission Control Protocol), NB-IoT (Narrow Band Internet of Things), LoRaWAN, etc. Selecting a suitable transmission protocol according to different network qualities can effectively ensure the stability and rate of data transmission; the transmission frequency can be the frequency used by the IoT device when transmitting data. Selecting a suitable transmission frequency according to different network qualities can reduce the data transmission volume and reduce traffic consumption while ensuring data transmission stability.

[0058] For transmission frequency, the edge device can obtain a transmission frequency mapping table, which includes at least the mapping relationship between traffic intervals and transmission frequencies. Then, the first target transmission frequency corresponding to the predicted data volume can be extracted from the transmission frequency mapping table. By constructing a corresponding mapping table, the transmission frequency that matches the predicted data volume can be obtained quickly and accurately, thereby dynamically adjusting the transmission strategy based on traffic demand, which can not only reduce traffic consumption, but also ensure the stability of data transmission and reduce the probability of network congestion.

[0059] Among them, the transmission frequency includes at least a first preset transmission frequency for transmission according to a preset time interval, a second preset transmission frequency for transmission using semantic compression, and a second preset transmission frequency for transmission according to data priority. The traffic interval includes at least a first traffic interval, a second traffic interval, and a third traffic interval. The traffic corresponding to the first traffic interval, the second traffic interval, and the third traffic interval increases in sequence. The edge device can compare the preset data volume with each traffic interval. If the predicted data volume is in the first traffic interval, the first preset transmission frequency is used as the first target transmission frequency for the business data; if the predicted data volume is in the second traffic interval, the second preset transmission frequency is used as the first target transmission frequency for the business data; if the predicted data volume is in the third traffic interval, the third preset transmission frequency is used as the first target transmission frequency for the business data. By constructing a corresponding mapping table, the transmission frequency that matches the predicted data volume can be obtained quickly and accurately, thereby dynamically adjusting the transmission strategy based on traffic demand, which can not only reduce traffic consumption, but also ensure the stability of data transmission and reduce the probability of network congestion.

[0060] For example, as shown in Table 1 below:

[0061] Predicted traffic range Network quality Adjustment strategy <10Kbps excellent Extend the interval to 15-30 minutes 10-50Kbps generally Enable semantic compression transmission >50Kbps congestion Prioritize transmission of critical data

[0062] Table 1

[0063] Among them, the first traffic interval can be <10Kbps. When the preset data volume is in this traffic interval, the corresponding transmission frequency can be to extend the transmission interval to 15-30 minutes; the second traffic interval can be 10-50Kbps. When the preset data volume is in this traffic interval, the corresponding transmission frequency can enable semantic compression transmission for transmission, that is, by semantically compressing the original data, reducing the amount of data transmission and reducing traffic consumption; the third traffic interval can be >50Kbps. When the preset data volume is in this traffic interval, the corresponding transmission frequency can adopt a method of prioritizing the transmission of critical data, that is, for business data that needs to be transmitted, it can be divided into critical data and non-critical data, and then the critical data is transmitted first, and after the critical data is transmitted, the non-critical data is transmitted to improve the success rate of critical data transmission, thereby dynamically adjusting the transmission strategy based on traffic demand, which can not only reduce traffic consumption, but also ensure the stability of data transmission and reduce the probability of network congestion.

[0064] For the transmission protocol, the edge device can obtain the protocol mapping table, which includes at least the mapping relationship between the reference signal receiving power, service priority and transmission protocol. Then, the first transmission protocol that matches the real-time reference signal receiving power and the target service priority is extracted from the protocol mapping table, thereby dynamically adjusting the transmission strategy based on traffic demand, which can not only reduce traffic consumption, but also ensure the stability of data transmission and reduce the probability of network congestion.

[0065] Furthermore, the protocol mapping table also includes the mapping relationship between the reference signal receiving power, service priority and redundant backup method. The first transmission strategy also includes a target redundant backup method. In the process of backing up data, the success rate of data backup can be guaranteed by selecting the corresponding redundant backup method. Optionally, the target redundant backup method corresponding to the target reference signal receiving power and the target service priority can be extracted from the protocol mapping table.

[0066] For example, as shown in Table 2 below:

[0067] RSRP (dBm) Business urgency Preferred Protocol Redundant backup >-90 high 4G / TCP NB-IoT -90～-100 middle NB-IoT LoRaWAN <-100 Low LoRaWAN Local Cache

[0068] Table 2

[0069] Among them, when the real-time reference signal received power is in the range of >-90dBm and the target service priority is high, 4G / TCP can be selected as the transmission protocol, and NB-IoT can be used as a redundant backup method; when the real-time reference signal received power is in the range of -90 to -100 and the target service priority is medium, NB-IoT can be selected as the transmission protocol, and LoRaWAN can be used as a redundant backup method; when the real-time reference signal received power is in the range of <-100 and the target service priority is low, LoRaWAN can be selected as the transmission protocol, and local cache can be used as a redundant backup method, so as to dynamically adjust the transmission strategy based on traffic demand, which can not only reduce traffic consumption, but also ensure the stability of data transmission and reduce the probability of network congestion.

[0070] It should be noted that 4G / TCP, etc., have high signal strength, can ensure high speed and low latency, and are suitable for high-priority services (such as video surveillance and remote control); for NB-IoT, low-power wide area network (LPWAN) is selected under medium signal conditions to balance energy consumption and reliability (such as environmental monitoring); for LoRaWAN, it relies on long-distance communication technology when the signal is extremely weak, sacrificing speed to ensure connectivity (such as agricultural sensors, etc.) to ensure the stability of data transmission, and thus adjusts the transmission strategy in multiple dimensions according to different signal strengths and business priorities, realizing intelligent disaster recovery of IoT transmission, which can not only reduce traffic consumption, but also ensure the stability of data transmission and reduce the probability of network congestion.

[0071] Step 105: Send the first transmission policy to the IoT device, so that the IoT device transmits the service data according to the first transmission policy.

[0072] After the first transmission strategy is determined, the edge device can send the first transmission strategy to the corresponding IoT device so that the IoT device can dynamically adjust the transmission mode based on the adjusted first transmission strategy and transmit the business data based on the new transmission strategy. In the process of data transmission by the IoT device, by predicting the amount of data to be transmitted and determining the transmission strategy based on the prediction results and network indicators and other data, the traffic demand is predicted, which is conducive to selecting the appropriate transmission mode for business data transmission according to different needs, optimizing the data transmission method, and realizing dynamic adjustment of the transmission strategy, thereby improving network adaptability, thereby improving the success rate of data transmission and reducing the probability of network congestion.

[0073] In a specific implementation, the first transmission strategy may include a first transmission frequency and a first transmission protocol, and the IoT device may transmit the service data based on the first transmission frequency and the first transmission protocol.

[0074] In some feasible implementations, according to the adjusted first transmission strategy, if the first target transmission frequency is the second preset transmission frequency, feature extraction is performed on the business data to obtain data features corresponding to the business data, and then the data features are transmitted according to the target transmission protocol.

[0075] Optionally, feature extraction methods may include numerical feature extraction, image feature extraction, and time series feature extraction, etc. For example, in the process of numerical feature extraction, the corresponding slope, mean, variance, peak value, etc. may be extracted as data features; in the process of image feature extraction, key image contours, etc. may be extracted as data features through neural networks; in the process of time series feature extraction, corresponding data features may be extracted through Fourier transform, etc. By performing feature extraction on business data, the transmission volume may be effectively reduced and traffic consumption may be reduced.

[0076] For example, assuming that the original data reports the temperature value [25, 26, 27, 28, 29]°C every minute and the slope is 1°C / min, the reporting slope can be set to 1 to reduce the transmission volume to achieve the purpose of reducing traffic consumption.

[0077] In other feasible implementations, according to the adjusted first transmission strategy, if the first target transmission frequency is the third preset transmission frequency, the business data is divided into critical data and non-critical data, and then the critical data is transmitted first, and after the critical data is transmitted, the non-critical data is transmitted.

[0078] Optionally, by grading the data, business data can be divided into critical data and non-critical data based on the grading results. For example, data such as equipment alarms, control instructions, real-time payment requests, etc. can be regarded as critical data, and the data characteristics of critical data are low latency, high reliability, etc. Historical logs, periodic status reports, etc. can be regarded as non-critical data, and the data characteristics of non-critical data can be tolerable delay, etc. By classifying business data by characteristics, the transmission delay of critical data can be reduced, and the probability of business interruption due to network congestion can be reduced.

[0079] In addition, for edge devices, regional network data and network prediction models can also be obtained. The regional network data includes at least one of time series network indicators, regional device density, and business traffic patterns. The time series network indicators, regional device density, and business traffic patterns are then input into the network prediction model to output corresponding network prediction information. Among them, the network prediction information includes at least one of the probability of network congestion within the future target time period, the available bandwidth resources corresponding to each region, and the optimal transmission path. The region is an area divided based on the same communication device. Therefore, the edge device can also analyze the network conditions corresponding to the area where the IoT device is located based on the regional network data, so as to aggregate the network prediction information corresponding to different regions, dynamically adjust the transmission strategy of the IoT device, and / or perform cross-regional resource adjustments, etc., thereby improving the resource utilization of the regional network.

[0080] For the network prediction model, historical network data corresponding to multiple areas can be obtained. The historical network data includes at least one of base station load, channel utilization, and device distribution density. Then, the base station load, channel utilization, and device distribution density are used to construct a graph to obtain the corresponding global network status graph, and the global network status graph is used as the network prediction model.

[0081] For example, edge devices can collect regional network data corresponding to each area, including at least the average hourly base station load, channel utilization and packet loss rate, and device density in each area, and then divide the data into time windows (hourly) and time grids (1km 2 ) slices and extract the following features:

[0082] Network congestion index: congestion probability = (packet loss rate * base station load) / maximum channel capacity;

[0083] Device activity during peak periods (e.g., traffic accounts for 30% between 6 and 8 p.m.).

[0084] Finally, the Neo4j graph database is used to construct nodes (base stations, device groups) and edges (transmission paths, load relationships), and the time node dependencies are mined through the graph neural network (GNN) to construct the corresponding global network status graph, so that the corresponding network prediction information can be predicted through the global network status graph.

[0085] For example, input features include: time series network indicators (RSRP, signal-to-noise ratio, packet loss rate), regional device density, and business traffic patterns (such as peak hour distribution); output predictions include: network congestion probability in the next hour, available bandwidth resources in each region, and optimal transmission path recommendations.

[0086] Furthermore, in addition to deploying corresponding network prediction models on edge devices, corresponding cloud-based large models can also be deployed on cloud servers. Cloud-edge collaboration can be achieved through the cloud-based large model of the cloud server and the edge model of the edge device (traffic prediction model, network prediction model, etc.), thereby realizing collaborative optimization of the device group and improving regional network utilization.

[0087] In some feasible implementations, the network indicator data includes at least the transmission success rate and delay information. The edge device can send the transmission success rate and delay information to the cloud server, and then receive the second transmission strategy for the transmission success rate and delay information returned by the cloud server. The second transmission strategy is that the cloud server inputs the transmission success rate and delay information into the cloud big model for dynamic adjustment, and obtains the transmission strategy. The second transmission strategy is then sent down to the IoT device, so that the IoT device transmits business data according to the second transmission strategy. The edge device uploads the corresponding network indicator data to the cloud server, and the cloud server dynamically adjusts the transmission strategy through the cloud big model, which can adapt to network changes in real time and ensure business continuity. At the same time, through the dynamic allocation of resources, it can reduce traffic consumption and can adapt to personalized transmission in different scenarios.

[0088] Among them, after receiving the transmission success rate and delay information, the cloud server can input the transmission success rate and the delay information into the cloud big model for congestion prediction, obtain the corresponding congestion probability, and determine the transmission strategy corresponding to the congestion probability. In this way, the cloud server can dynamically adjust the transmission strategy based on the prediction results by predicting the corresponding red probability, which can significantly improve the network reliability, resource utilization and business continuity of the Internet of Things system.

[0089] Optionally, for the dynamic adjustment process, the cloud server can continuously monitor real-time indicators such as transmission success rate and delay, and compare them with preset thresholds. When it is found that the actual congestion probability of the target area exceeds the preset threshold, the cloud server can take corresponding adjustment measures: such as adjusting the transmission priority of the IoT devices in the corresponding area (for example, postponing low-priority tasks) or protocol type (for example, switching from a high-speed but unstable method to a stable but slower method), etc., and sending the adjustment results to the edge device, which is further sent to the corresponding IoT device by the edge device, so that the IoT device can transmit data based on the adjusted transmission strategy to reduce the load in the area to which it belongs and ensure the continuity of data transmission.

[0090] Among them, the second transmission strategy includes at least a second transmission protocol, a second target transmission frequency, a target base station identifier and priority adjustment information, then the Internet of Things device can transmit business data according to the second transmission protocol and / or the second target transmission frequency; and / or, access the target base station corresponding to the target base station identifier; and / or, adjust the target business priority according to the priority adjustment information, etc., thereby realizing dynamic adjustment of the transmission strategy, improving network adaptability, and then improving the success rate of data transmission and reducing the probability of network congestion.

[0091] In addition, the cloud server can also implement a dynamic resource allocation strategy. The edge device uploads the corresponding regional network data to the cloud server. The cloud server can predict the regional network data based on the cloud large model and generate a cross-regional bandwidth scheduling strategy based on the prediction results. For example: priority mapping: routing high-priority services (such as fault alarms) to low-load base stations to ensure real-time performance; load balancing: in the base station overload area, automatically guide the device to switch to an adjacent base station or an alternative communication protocol (such as switching from 4G to LoRaWAN), etc. The present invention does not impose any restrictions on this.

[0092] It should be noted that, for the above-mentioned edge devices and cloud servers, a collaborative mechanism between the cloud-based large model and the edge model can be realized through a hierarchical federated learning framework. Optionally, the data volume, network quality and model gradient corresponding to the area to which the IoT device belongs can be obtained, and then the data volume, network quality and model gradient can be sent to the cloud server. The cloud server updates the weight of the area in the cloud-based large model based on the data volume, network quality and model gradient.

[0093] In the specific implementation, for edge devices, a local lightweight model can be deployed to be responsible for device-level traffic prediction, and information such as model gradients can be uploaded to the regional edge gateway regularly, and then uploaded to the cloud server by the edge gateway. The large cloud model deployed in the cloud server can aggregate the edge model gradients of different regions, and update the global model parameters through an adaptive algorithm (based on regional data volume and network quality, etc.), and send them back to the corresponding edge device, so that the data can be called through the federated learning engine for model training and prediction.

[0094] The model gradient can be the direction and magnitude of parameter adjustment during the training process, reflecting the impact of local data on the model. In an embodiment of the present invention, the model gradient can be the change in the model parameters of the edge device relative to the prediction error, which is used to guide the adjustment of the model parameters. The adaptive weighting algorithm can be based on the amount of data in each region and its network quality, giving a weight. The weight is used to determine the importance of information from each region to the global model parameter update. Regions with large data volumes and good network quality are generally given higher weights. The data from such regions can more effectively reflect the actual network status.

[0095] In some feasible implementations, the process of calculating the regional weight may be:

[0096] ω i = Regional data volume i * network quality i / ∑ (data volume j * network quality j);

[0097] Network quality: Scored based on transmission success rate and latency information;

[0098] Parameter aggregation: θ 全局 =∑(ω i *θ 边缘 , i).

[0099] For example, assume that there are three edge regions (A, B, and C) in an IoT system. The data volume, network quality score (based on transmission success rate and latency), and local model parameters of each region are shown in Table 3 below:

[0100] area Data volume (GB) Network quality score (0-1) Edge model parameters (θ_edge) A 10 0.9 (high) θ_A=1.2 B 5 0.6 (medium) θ_B=0.8 C 2 0.3 (low) θ_C=0.5

[0101] Table 3

[0102] The above formula can be used to calculate:

[0103] The regional weight of region A is 0.714 (rounded off);

[0104] The regional weight of region B is 0.238 (rounded off);

[0105] The area weight of area C is 0.048 (rounded off).

[0106] Then, the following results can be calculated based on the global parameter aggregation formula:

[0107] θ global =(0.714×1.2)+(0.238×0.8)+(0.048×0.5)=0.857+0.190+0.024=1.071

[0108] It can be seen from this that area A has a high data volume and good network quality, while area C has a low data volume and poor network quality. By adjusting the cloud-based large model with parameters that reflect high data volume and good network quality, we can effectively avoid low-quality data from contaminating the cloud-based large model, making the prediction results of the cloud-based large model more consistent with the actual network status, and enabling the cloud server to dynamically adapt to different network environments, thereby optimizing resource allocation.

[0109] In addition, the network quality scoring process can be calculated using the following formula:

[0110] Network quality = 0.7 × S + 0.3 × (1-L max / L)

[0111] Among them, S represents the transmission success rate, and L represents the delay information. The network quality corresponding to each area can be calculated using the above formula.

[0112] It should be noted that the embodiments of the present invention include but are not limited to the above examples. It is understandable that those skilled in the art can also make settings according to actual needs under the guidance of the ideas of the embodiments of the present invention, and the present invention does not limit this.

[0113] In an embodiment of the present invention, in an Internet of Things scenario, an edge device can obtain business data collected by the Internet of Things device, determine the target business priority of the business data, and simultaneously obtain network indicator data of the communication device to which the Internet of Things device is connected. Then, the data volume is predicted based on the business data to obtain the corresponding predicted data volume. Then, a transmission method is selected based on the predicted data volume and the network indicator data to obtain a first transmission strategy for the business data. The first transmission strategy is then sent to the Internet of Things device so that the Internet of Things device transmits the business data according to the first transmission strategy. Thus, in the process of data transmission by the Internet of Things device, by predicting the amount of data to be transmitted and determining the transmission strategy based on the prediction results and network indicator data, traffic demand prediction is achieved, which is conducive to selecting a suitable transmission method for transmitting business data according to different needs, optimizing the data transmission method, and at the same time, dynamically adjusting the transmission strategy, improving network adaptability, thereby improving the success rate of data transmission, and reducing the probability of network congestion.

[0114] In order to enable those skilled in the art to better understand the technical solutions in the embodiments of the present invention, the following examples are provided for illustrative purposes:

[0115] As an example, see Figure 3 , shows a flow chart of data transmission provided in an embodiment of the present invention, the Internet of Things device collects data through the corresponding data acquisition module and performs preprocessing, and then transmits the corresponding data to the edge device; the edge device can perform model training by learning the latest data distribution to generate a corresponding local model, and then make dynamic transmission decisions through the local model, adjust the transmission strategy according to the model prediction results and send the transmission strategy to the Internet of Things device, the Internet of Things device transmits data based on the transmission strategy, and at the same time feeds back corresponding parameter indicators to the edge device, the edge device performs feedback optimization based on the feedback parameter indicators, and further optimizes the local model, thereby improving the accuracy of the model prediction by continuously optimizing the local model, and in the process of data transmission by the Internet of Things device, by predicting the amount of data to be transmitted and determining the transmission strategy based on the prediction results and network indicators and other data, the traffic demand is predicted, which is conducive to selecting a suitable transmission method for transmitting business data according to different needs, optimizing the data transmission method, and at the same time, dynamically adjusting the transmission strategy, improving network adaptability, thereby improving the success rate of data transmission, and reducing the probability of network congestion.

[0116] As another example, a traffic management system can be built based on IoT devices, edge devices, and cloud servers. For edge devices, this can include a corresponding cross-layer perception module. This module includes a physical layer responsible for collecting operational data from light business IoT devices, including device status, sensor data, and historical traffic records. This data is collected in real time by sensors and data acquisition devices and stored in the system's database; a network layer for real-time monitoring of network metrics such as RSRP, base station load, and channel congestion probability; and a service layer for determining data priority and importance (such as alarm priority and log data timeliness). Simultaneously, a large-scale pre-trained Transformer-based model (with >1B parameters) can be deployed on edge devices to integrate historical network data from multiple regions (including base station load, channel utilization, and device density) to construct a global network status map. This model, by inputting features such as time-series network metrics (RSRP, signal-to-noise ratio, packet loss rate), regional device density, and service traffic patterns (such as peak hour distribution), can output predictions: network congestion probability within the next hour, available bandwidth resources in each region, and optimal transmission path recommendations.

[0117] In a specific implementation, in addition to deploying the corresponding edge model on the edge device, the corresponding cloud-based large model can also be deployed on the cloud server, so that the edge model and the cloud-based large model can be coordinated. The coordination mechanism between the cloud-based large model and the edge model can include:

[0118] (1) Hierarchical Federated Learning Framework

[0119] Edge model, used for local lightweight model (Tiny-Transformer) responsible for device-level traffic prediction, and regularly uploads model gradients to the regional edge gateway.

[0120] The large cloud model is used to aggregate the gradients of multi-region edge models, update the global model parameters through an adaptive weighted algorithm (based on regional data volume and network quality), and send them back to the edge nodes.

[0121] (2) Real-time feedback optimization

[0122] The cloud-based large model receives real-time metrics such as transmission success rate and latency from edge devices and dynamically adjusts resource allocation strategies. Optionally, it can implement congestion avoidance and elastic scaling. For congestion avoidance, if the actual congestion probability in the target area exceeds a preset threshold, the cloud will be triggered to proactively intervene and forcibly adjust the transmission priority or protocol of the devices in that area. For example, when the congestion probability is greater than or equal to 70%, only urgent data transmission is allowed, low-priority data transmission tasks are suspended, or the transmission protocol is adjusted based on RSRP. For elastic scaling, the cloud can release redundant bandwidth resources during low-load periods for high-throughput services (such as OTA firmware upgrades). For example, it prioritizes data transmission for services that require higher bandwidth and lower latency, such as OTA upgrades and sensor data that need to be reported in real time. Services that do not require high bandwidth and are less sensitive to latency, such as sensor data and simple log record reporting, can be transmitted after the high-priority services are completed.

[0123] In the above process, the cloud can generate a cross-regional bandwidth scheduling strategy based on the prediction results of the cloud-based large model: ① Priority mapping: route high-priority services (such as fault alarms) to low-load base stations to ensure real-time performance; ② Load balancing: in the base station overload area, automatically guide the device to switch to an adjacent base station or an alternative communication protocol (such as switching from 4G to LoRaWAN).

[0124] Optionally, for edge models and large cloud models, the corresponding model architecture can be constructed using Tiny-Transformer (4-layer attention head) + physical constraint module (embedded device hardware characteristics). At the same time, for IoT devices, the latest data distribution can be learned online, and a preset percentage of adjustable parameters can be retained; for edge devices, gradients can be aggregated regularly, such as recording water flow, battery voltage, and network signal strength every hour to generate region-specific models; for the cloud, global models can be integrated and benchmark parameters can be issued (the common part of all regional models is averaged), etc.

[0125] In addition, when adjusting the transmission strategy, if switching between different transmission protocols is involved, the system can also switch the transmission protocol through the corresponding dynamic protocol controller. For example, the system supports LoRaWAN / NB-IoT / 4G multi-protocol stacks to dynamically switch according to network quality; and supports adaptive selection of TCP / UDP / CoAP transmission protocols to match business needs.

[0126] Furthermore, IoT devices can also adjust the way they upload periodic detection data based on the transmission strategy issued by the edge device. For example, they can use the semantic communication module to extract corresponding feature values ​​(such as the temperature change slope, etc.) from the periodic detection data, and then upload the extracted feature values ​​instead of the original data, thereby reducing the amount of data transmission and reducing traffic consumption.

[0127] Through the above process, in the process of IoT devices transmitting data, the amount of data that needs to be transmitted is predicted, and the transmission strategy is determined based on the prediction results and network indicators and other data, thereby realizing the prediction of traffic demand. This is conducive to selecting appropriate transmission methods for business data transmission according to different needs, optimizing the data transmission method, and at the same time, dynamically adjusting the transmission strategy, improving network adaptability, thereby increasing the success rate of data transmission and reducing the probability of network congestion.

[0128] It should be noted that for the sake of simplicity, the method embodiments are described as a series of actions. However, those skilled in the art should be aware that the embodiments of the present invention are not limited by the order of the actions described, because according to the embodiments of the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in this specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of the present invention.

[0129] Reference Figure 4 , shows a structural block diagram of a data transmission device for an Internet of Things device provided in an embodiment of the present invention, which may specifically include the following modules:

[0130] The data acquisition module 401 is used to acquire the business data collected by the IoT device and determine the target business priority of the business data;

[0131] A network indicator acquisition module 402 is used to obtain network indicator data of a communication device connected to the IoT device;

[0132] The data volume prediction module 403 is used to perform data volume prediction based on the business data to obtain the corresponding predicted data volume;

[0133] a transmission mode determination module 404, configured to select a transmission mode based on the predicted data volume and the network indicator data, and obtain a first transmission strategy for the service data;

[0134] The data transmission module 405 is configured to send the first transmission policy to the IoT device, so that the IoT device transmits the service data according to the first transmission policy.

[0135] In some feasible implementations, the service data includes at least device status information and traffic data, and the data volume prediction module 403 is specifically configured to:

[0136] Obtain traffic prediction model;

[0137] Normalizing the device status information and the flow data to obtain corresponding feature data;

[0138] The characteristic data is input into the traffic prediction model to obtain a corresponding predicted data volume, where the predicted data volume is used to characterize the amount of data that the IoT device is expected to generate within a target time period.

[0139] In some feasible implementations, the network indicator data includes at least a real-time reference signal received power, the first transmission strategy includes at least a first transmission protocol and a first target transmission frequency, and the transmission mode determination module 404 is specifically configured to:

[0140] determining a first target transmission frequency corresponding to the predicted data amount;

[0141] Selecting a protocol based on the real-time reference signal received power and the target service priority to determine a first transmission protocol for the service data;

[0142] The first target transmission frequency and the first transmission protocol are used as a first transmission strategy for business data.

[0143] In some feasible implementations, the transmission mode determination module 404 is specifically configured to:

[0144] Acquire a transmission frequency mapping table, wherein the transmission frequency mapping table at least includes a mapping relationship between flow intervals and transmission frequencies;

[0145] A first target transmission frequency corresponding to the predicted data amount is extracted from the transmission frequency mapping table.

[0146] In some feasible implementations, the transmission frequency includes at least a first preset transmission frequency for transmission according to a preset time interval, a second preset transmission frequency for transmission using semantic compression, and a second preset transmission frequency for transmission according to data priority. The traffic interval includes at least a first traffic interval, a second traffic interval, and a third traffic interval. The traffic corresponding to the first traffic interval, the second traffic interval, and the third traffic interval increases in sequence. The transmission mode determination module 404 is specifically used to:

[0147] If the predicted data volume is within the first traffic interval, using the first preset transmission frequency as a first target transmission frequency for the service data;

[0148] If the predicted data volume is within the second traffic interval, using the second preset transmission frequency as the first target transmission frequency for the service data;

[0149] If the predicted data volume is within the third traffic interval, the third preset transmission frequency is used as the first target transmission frequency for the business data.

[0150] In some feasible implementations, the transmission mode determination module 404 is specifically configured to:

[0151] Acquire a protocol mapping table, where the protocol mapping table includes at least a mapping relationship between a reference signal received power, a service priority, and a transmission protocol;

[0152] A first transmission protocol matching the real-time reference signal received power and the target service priority is extracted from the protocol mapping table.

[0153] In some feasible implementations, the protocol mapping table further includes a mapping relationship between a reference signal received power, a service priority, and a redundancy backup mode, the first transmission strategy further includes a target redundancy backup mode, and the apparatus further includes:

[0154] A target redundancy backup mode corresponding to the target reference signal received power and the target service priority is extracted from the protocol mapping table.

[0155] In some feasible implementations, the data transmission module 405 is specifically configured to:

[0156] If the first target transmission frequency is the second preset transmission frequency, performing feature extraction on the service data to obtain data features corresponding to the service data;

[0157] The data characteristics are transmitted according to the target transmission protocol.

[0158] In some feasible implementations, the data transmission module 405 is specifically configured to:

[0159] If the first target transmission frequency is the third preset transmission frequency, dividing the service data into critical data and non-critical data;

[0160] The critical data is transmitted first, and the non-critical data is transmitted after the critical data has been transmitted.

[0161] Some possible implementations also include:

[0162] A model acquisition module, configured to acquire regional network data and a network prediction model, wherein the regional network data includes at least one of a time series network indicator, a regional device density, and a service traffic pattern;

[0163] A network prediction module, configured to input the time series network indicators, the regional device density, and the service traffic pattern into the network prediction model and output corresponding network prediction information;

[0164] The network prediction information includes at least one of the following: a probability of network congestion within a future target time period, available bandwidth resources corresponding to each area, and an optimal transmission path;

[0165] The areas are areas divided based on the same communication device.

[0166] Some possible implementations also include:

[0167] A network data acquisition module 401 is configured to acquire historical network data corresponding to a plurality of the areas, wherein the historical network data includes at least one of base station load, channel utilization, and device distribution density;

[0168] A graph construction module is used to construct a graph using the base station load, the channel utilization rate and the device distribution density to obtain a corresponding global network status graph, and use the global network status graph as a network prediction model.

[0169] In some feasible implementations, the network indicator data includes at least a transmission success rate and delay information, and the apparatus further includes:

[0170] An information sending module, configured to send the transmission success rate and the delay information to a cloud server;

[0171] an information receiving module, configured to receive a second transmission strategy for the transmission success rate and the delay information returned by the cloud server, wherein the second transmission strategy is obtained by the cloud server inputting the transmission success rate and the delay information into a cloud-based large model for dynamic adjustment;

[0172] A policy issuing module is used to issue the second transmission policy to the IoT device, so that the IoT device transmits the service data according to the second transmission policy.

[0173] In some feasible implementations, the information receiving module is specifically configured to:

[0174] The transmission success rate and the delay information are input into a cloud-based large model to perform congestion prediction, obtain a corresponding congestion probability, and determine a transmission strategy corresponding to the congestion probability.

[0175] In some feasible implementations, the second transmission strategy includes at least a second transmission protocol, a second target transmission frequency, a target base station identifier, and priority adjustment information, and the strategy issuing module is specifically configured to:

[0176] transmitting the service data according to the second transmission protocol and / or the second target transmission frequency;

[0177] and / or, accessing the target base station corresponding to the target base station identifier;

[0178] And / or, adjusting the target service priority according to the priority adjustment information.

[0179] Some possible implementations also include:

[0180] A gradient acquisition module is used to obtain the data volume, network quality and model gradient corresponding to the area to which the IoT device belongs;

[0181] A weight updating module is used to send the data volume, the network quality and the model gradient to the cloud server, and the cloud server updates the weight of the area in the cloud large model according to the data volume, the network quality and the model gradient.

[0182] As for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0183] In addition, an embodiment of the present invention further provides an electronic device, comprising: a processor, a memory, and a computer program stored in the memory and runnable on the processor. When the computer program is executed by the processor, the various processes of the data transmission method embodiment of the above-mentioned Internet of Things device are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0184] The embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the various processes of the above-mentioned embodiment of the data transmission method for an IoT device are implemented, and the same technical effects are achieved. To avoid repetition, the details are not described here. The computer-readable storage medium is, for example, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0185] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0186] It should be understood by those skilled in the art that the embodiments of the present invention may be provided as methods, apparatuses, or computer program products. Therefore, the embodiments of the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the embodiments of 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, EEPROM, Flash, and eMMC, etc.) containing computer-usable program code.

[0187] The embodiments of the present invention are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the 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 the 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 terminal device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the process 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.

[0188] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0189] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0190] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they become aware of the basic creative concepts. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.

[0191] Finally, it should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or terminal device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or terminal device that includes the element.

[0192] The above is a detailed introduction to a data transmission method for an IoT device and a data transmission device for an IoT device provided by the present invention. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the ideas of the present invention, there may be changes in the specific implementation methods and application scopes. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A data transmission method for an Internet of Things device, characterized in that: include: Obtaining business data collected by IoT devices and determining target business priorities for the business data; Obtain network indicator data of the communication device connected to the IoT device; Performing data volume prediction based on the business data to obtain a corresponding predicted data volume; Selecting a transmission mode according to the predicted data volume and the network indicator data to obtain a first transmission strategy for the business data; The first transmission policy is issued to the IoT device, so that the IoT device transmits the service data according to the first transmission policy.

2. The method according to claim 1, characterized in that The business data includes at least device status information and traffic data, and performing data volume prediction based on the business data to obtain a corresponding predicted data volume includes: Obtain traffic prediction model; Normalizing the device status information and the flow data to obtain corresponding feature data; The characteristic data is input into the traffic prediction model to obtain a corresponding predicted data volume, where the predicted data volume is used to characterize the amount of data that the IoT device is expected to generate within a target time period.

3. The method according to claim 1 or 2, characterized in that The network indicator data includes at least a real-time reference signal received power, the first transmission strategy includes at least a first transmission protocol and a first target transmission frequency, and selecting a transmission mode based on the predicted data volume and the network indicator data to obtain a first transmission strategy for the service data includes: determining a first target transmission frequency corresponding to the predicted data amount; Selecting a protocol based on the real-time reference signal received power and the target service priority to determine a first transmission protocol for the service data; The first target transmission frequency and the first transmission protocol are used as a first transmission strategy for business data.

4. The method according to claim 3, characterized in that The determining a first transmission frequency corresponding to the predicted data amount includes: Acquire a transmission frequency mapping table, wherein the transmission frequency mapping table at least includes a mapping relationship between flow intervals and transmission frequencies; A first target transmission frequency corresponding to the predicted data amount is extracted from the transmission frequency mapping table.

5. The method according to claim 4, characterized in that The transmission frequency includes at least a first preset transmission frequency for transmission according to a preset time interval, a second preset transmission frequency for transmission using semantic compression, and a second preset transmission frequency for transmission according to data priority. The traffic interval includes at least a first traffic interval, a second traffic interval, and a third traffic interval. The traffic corresponding to the first traffic interval, the second traffic interval, and the third traffic interval increases in sequence. Extracting the target transmission frequency corresponding to the predicted data amount from the transmission frequency mapping table includes: If the predicted data volume is within the first traffic interval, using the first preset transmission frequency as a first target transmission frequency for the service data; If the predicted data volume is within the second traffic interval, using the second preset transmission frequency as the first target transmission frequency for the service data; If the predicted data volume is within the third traffic interval, the third preset transmission frequency is used as the first target transmission frequency for the business data.

6. The method according to claim 3, characterized in that The selecting a protocol according to the real-time reference signal received power and the target service priority to determine a first transmission protocol for the service data includes: Acquire a protocol mapping table, where the protocol mapping table includes at least a mapping relationship between a reference signal received power, a service priority, and a transmission protocol; A first transmission protocol matching the real-time reference signal received power and the target service priority is extracted from the protocol mapping table.

7. The method according to claim 6, characterized in that The protocol mapping table further includes a mapping relationship between a reference signal received power, a service priority, and a redundancy backup mode. The first transmission strategy further includes a target redundancy backup mode. The method further includes: A target redundancy backup mode corresponding to the target reference signal received power and the target service priority is extracted from the protocol mapping table.

8. The method according to claim 5, characterized in that The transmitting the service data according to the first transmission strategy includes: If the first target transmission frequency is the second preset transmission frequency, performing feature extraction on the service data to obtain data features corresponding to the service data; The data characteristics are transmitted according to the target transmission protocol.

9. The method according to claim 5, characterized in that The transmitting the service data according to the first transmission strategy includes: If the first target transmission frequency is the third preset transmission frequency, dividing the service data into critical data and non-critical data; The critical data is transmitted first, and the non-critical data is transmitted after the critical data has been transmitted.

10. The method according to claim 1, characterized in that Also includes: Obtaining regional network data and a network prediction model, wherein the regional network data includes at least one of a time series network indicator, a regional device density, and a service traffic pattern; Inputting the time series network indicators, the regional device density, and the service traffic pattern into the network prediction model, and outputting corresponding network prediction information; The network prediction information includes at least one of the following: a probability of network congestion within a future target time period, available bandwidth resources corresponding to each area, and an optimal transmission path; The areas are areas divided based on the same communication device.

11. The method according to claim 10, characterized in that Also includes: Acquire historical network data corresponding to the plurality of areas, the historical network data including at least one of base station load, channel utilization, and device distribution density; The base station load, the channel utilization rate and the device distribution density are used to construct a graph to obtain a corresponding global network status graph, and the global network status graph is used as a network prediction model.

12. The method according to claim 1, characterized in that The network indicator data includes at least transmission success rate and delay information, and the method further includes: Sending the transmission success rate and the delay information to a cloud server; receiving a second transmission strategy for the transmission success rate and the delay information returned by the cloud server, where the second transmission strategy is obtained by the cloud server inputting the transmission success rate and the delay information into a cloud-based large model for dynamic adjustment; The second transmission policy is issued to the IoT device, so that the IoT device transmits the service data according to the second transmission policy.

13. The method according to claim 11, characterized in that The transmission success rate and the delay information are input into a cloud-based large model for dynamic adjustment to obtain a transmission strategy, including: The transmission success rate and the delay information are input into a cloud-based large model to perform congestion prediction, obtain a corresponding congestion probability, and determine a transmission strategy corresponding to the congestion probability.

14. The method according to claim 12, characterized in that The second transmission strategy includes at least a second transmission protocol, a second target transmission frequency, a target base station identifier, and priority adjustment information, and transmitting the service data according to the second transmission strategy includes: transmitting the service data according to the second transmission protocol and / or the second target transmission frequency; and / or, accessing the target base station corresponding to the target base station identifier; And / or, adjusting the target service priority according to the priority adjustment information.

15. The method according to claim 12, characterized in that Also includes: Obtaining the data volume, network quality, and model gradient corresponding to the area to which the IoT device belongs; The data volume, the network quality, and the model gradient are sent to the cloud server, and the cloud server updates the weight of the region in the cloud large model according to the data volume, the network quality, and the model gradient.

16. A data transmission device for an Internet of Things device, characterized in that: include: A data acquisition module is used to acquire business data collected by IoT devices and determine the target business priority of the business data; A network indicator acquisition module is used to obtain network indicator data of the communication device connected to the Internet of Things device; A data volume prediction module is used to predict the data volume based on the business data and obtain the corresponding predicted data volume; a transmission mode determination module, configured to select a transmission mode according to the predicted data volume and the network indicator data, and obtain a first transmission strategy for the service data; The data transmission module is used to send the first transmission strategy to the IoT device, so that the IoT device transmits the business data according to the first transmission strategy.

17. An electronic device, characterized in that: comprising a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; The memory is used to store computer programs; The processor is configured to implement the method according to any one of claims 1 to 15 when executing a program stored in the memory.

18. A computer-readable storage medium having instructions stored thereon, which, when executed by one or more processors, cause the processors to perform the method according to any one of claims 1 to 15.

Images