Low-power wide-area energy efficiency algorithm based on adaptive data rate and micro machine learning

Abstract

The invention provides an energy efficiency optimization algorithm for a low-power wide area network (LPWAN), and aims to reduce the energy consumption of LPWAN equipment in a data transmission process and prolong the service life through adaptive data rate adjustment and a micro machine learning (TinyML) technology. The algorithm core comprises a self-adaptive data rate adjustment mechanism, and the transmission rate is dynamically adjusted according to the real-time link quality index and the signal strength. A TinyML model is integrated for local data processing and intelligent decision making, and the data uploading amount is reduced; according to the intelligent dormancy awakening strategy, energy consumption use of equipment is predicted and optimized according to an equipment use mode; and a multi-parameter optimization scheduling algorithm is adopted, and signal quality, equipment electric quantity and environmental factors are integrated for scheduling. In addition, the security of data transmission and user privacy protection are also considered, and a lightweight encryption technology is adopted. The practical application test in various LPWAN technologies verifies the effectiveness of the algorithm, realizes the obvious reduction of energy consumption and the prolonging of the service life of equipment, and has high expandability and adaptability.

Description

Technical Field

[0001] This invention relates to the field of Internet of Things (IoT) communication technology, and in particular to an energy efficiency optimization algorithm for low-power wide-area networks (LPWAN). Background Technology

[0002] In the field of IoT communication technology, Low-Power Wide-Area Networks (LPWANs) have become a key technology due to their ability to achieve long-distance, low-power data transmission. LPWAN technology allows a large number of devices to connect to the Internet with lower power consumption and cost, making it ideal for applications such as smart cities, environmental monitoring, and smart agriculture. However, existing LPWAN technologies still suffer from high energy consumption during data transmission, especially when frequent communication or wide-area coverage is required. This limits device lifespan, increases maintenance costs, and impacts the economic efficiency and sustainability of IoT solutions.

[0003] Traditional LPWAN networks, such as LoRaWAN and NB-IoT, while designed with low power consumption in mind, still have significant room for improvement in energy efficiency optimization due to the complexity of network environments and the diversity of devices in practical applications. For example, fixed data rates may not be suitable for all communication scenarios, leading to excessive power consumption in some situations. Furthermore, the lack of intelligent sleep and wake-up strategies can cause devices to continue consuming power even when communication is not needed. In addition, with the proliferation of IoT devices, effectively managing and optimizing their energy consumption has become a pressing technical challenge.

[0004] To overcome the limitations of existing technologies, researchers have been exploring new algorithms and methods to further improve the energy efficiency of LPWANs. These include adaptive data rate adjustment, intelligent sleep / wake-up strategies, and the application of TinyML (Tiny Machine Learning) technology. The integration and optimization of these technologies are expected to significantly improve the overall energy efficiency of LPWAN networks, extend device lifespan, and reduce operating costs, thereby promoting the widespread application and development of IoT technology. Summary of the Invention

[0005] The technical solution proposed in this invention is an energy efficiency optimization algorithm for Low Power Wide Area Networks (LPWAN). It aims to significantly reduce energy consumption of LPWAN devices during data transmission, extend device lifespan, and improve overall network transmission efficiency through a series of innovative technical means. The following is a detailed description of the technical solution, corresponding one-to-one with the claims: 1. Adaptive Data Rate Adjustment Mechanism: The core of this algorithm includes an adaptive data rate adjustment module, which can monitor the Link Quality Indicator (LQI) and Signal Strength Index (RSSI) of the LPWAN network in real time. Based on these indicators, the module dynamically adjusts the data transmission rate to adapt to the current network conditions, thereby reducing energy consumption while ensuring data transmission reliability. The optimization of LPWAN network energy consumption is achieved in the following details: 101. Real-time Monitoring: First, the adaptive data rate adjustment mechanism continuously collects the Link Quality Indicator (LQI) and Signal Strength Index (RSSI) of the LPWAN network through a real-time monitoring module. These key indicators reflect the current network communication quality and signal stability, providing basic data for subsequent data rate adjustments.

[0006] 102. Data Analysis: The collected LQI and RSSI data are then transmitted to the data analysis unit. Here, the data is processed using advanced statistical analysis and machine learning algorithms to identify trends and patterns in network state changes.

[0007] 103. Dynamic Adjustment Algorithm: Based on data analysis, the dynamic rate adjustment engine uses an adaptive algorithm to calculate and determine the optimal data transmission rate. This algorithm can automatically adjust the transmission rate according to different network conditions and communication needs to achieve the best balance between energy consumption and transmission efficiency.

[0008] 104. Policy Library Application: The algorithm has a built-in rate adjustment policy library, which contains a variety of predefined rate adjustment policies suitable for different network environments and communication scenarios. These policies can be updated and optimized based on feedback from actual applications.

[0009] 105. Feedback Loop: To ensure continuous optimization of transmission rates, the adaptive data rate adjustment mechanism also includes a feedback loop mechanism. This mechanism monitors the adjusted network performance and further fine-tunes the rate settings based on actual transmission results.

[0010] 106. User Configuration Interface: To provide greater flexibility and adaptability, the algorithm provides a user configuration interface, allowing users to manually set parameters for rate adjustment according to the needs of specific applications, such as adjusting sensitivity and rate variation range.

[0011] 107. Energy Consumption Assessment: Finally, the algorithm integrates an energy consumption assessment component to predict and evaluate energy consumption under different rate settings. This provides users with important decision support, helping them understand energy consumption changes and make more informed adjustments.

[0012] Through this series of orderly steps, the adaptive data rate adjustment mechanism can accurately respond to changes in the LPWAN network status and realize intelligent adjustment of data transmission rate, thereby significantly reducing energy consumption and improving the energy efficiency of the entire network while ensuring data transmission reliability.

[0013] 2. TinyML Model Integration: The algorithm integrates an optimized TinyML model on the LPWAN device, enabling efficient data processing and intelligent decision-making on the device side. Localized processing reduces reliance on cloud data processing and lowers energy consumption due to data transmission. The following are the specific implementation details of this technical solution: 201. Model Selection and Customization: Select a TinyML model architecture suitable for the resource constraints (such as computing power, memory, and power consumption) of LPWAN devices. These models are customized and optimized to ensure they can run efficiently on resource-constrained devices.

[0014] 202. Data Preprocessing: The data collected at the device first goes through a preprocessing module to clean, normalize, and reduce dimensionality, thereby providing high-quality and simplified data input for the TinyML model.

[0015] 203. Feature Extraction: Through the feature extraction module, the algorithm automatically identifies and extracts the data features most valuable to the decision-making process, reducing the amount of data that needs to be transmitted while retaining key information.

[0016] 204. Localized processing reduces reliance on cloud or central server data processing, thereby reducing data transmission requirements and related energy consumption.

[0017] 205. Model Update Mechanism: To maintain the accuracy and adaptability of the model, the algorithm includes a lightweight model update mechanism that allows the model to self-optimize and update based on the latest data and feedback without frequent external intervention.

[0018] 206. Resource Management: The TinyML model takes into account the energy and computing resources of the device, and adopts energy-saving mode and interruption mechanism to ensure that the operation of the model will not affect the normal communication and other tasks of the device.

[0019] 207. Intelligent Decision Support: The model's output is used to support intelligent decision-making, such as predicting equipment usage patterns, identifying abnormal situations, or automating control commands, thereby further improving equipment operating efficiency.

[0020] 208. Security Considerations: Data security and privacy protection were considered during the design and integration of the TinyML model to ensure that sensitive information is processed and stored locally, thereby reducing the risk of data leakage.

[0021] 209. User configurability: Provides a user interface that allows users to configure the parameters of the TinyML model according to specific application scenarios, such as threshold settings and decision logic, to meet the needs of different applications.

[0022] 210. Integration Testing and Validation: Extensive testing and validation are conducted before the model is integrated into LPWAN devices to ensure the model's accuracy, reliability, and performance under various network conditions.

[0023] Through the detailed implementation steps described above, the TinyML model integration technology solution can significantly reduce energy consumption caused by data transmission while ensuring data processing quality, and improve the intelligence level and operating efficiency of LPWAN devices.

[0024] 3. Intelligent Sleep / Wake-up Strategy: This involves in-depth analysis of device usage patterns, collecting key data such as communication frequency, data transmission volume, and user interactions. Subsequently, an integrated TinyML model is used to learn from this data, predicting peak and off-peak periods of communication demand. Based on these predictions, a logical system is developed that intelligently determines when to put devices into sleep mode to reduce energy consumption and wakes them up promptly when communication demand increases. This strategy also includes a dynamic adjustment mechanism that responds in real-time to changes in network and device status, ensuring rapid device response by setting wake-up conditions such as predetermined times or specific events. Furthermore, the strategy provides user-configurable options, allowing adjustment of sleep / wake-up parameters according to specific needs, and continuously optimizes performance through periodic evaluation and feedback loops, ultimately achieving significant energy reduction while maintaining communication efficiency and improving the energy efficiency ratio of LPWAN devices.

[0025] 4. Multi-parameter Optimization Scheduling Algorithm: The algorithm further includes a multi-parameter optimization scheduling module. This module comprehensively considers multiple parameters such as signal quality, device power consumption, and environmental factors to perform real-time data transmission optimization scheduling. The optimization of LPWAN network data transmission is achieved through the following specific steps: Parameter collection: The algorithm first collects key parameters that affect data transmission efficiency and energy consumption, including but not limited to signal quality indicators (such as LQI and RSSI), current device power, environmental conditions (such as temperature and humidity), and expected network congestion.

[0026] Real-time analysis: Through the real-time analysis module, the algorithm performs in-depth analysis on the collected parameters, identifies their relationship with data transmission efficiency and energy consumption, and predicts network performance under different transmission conditions.

[0027] Optimization Model: Based on the results of real-time analysis, the algorithm uses an optimization model to calculate the optimal transmission time window and frequency. This model employs advanced optimization techniques, such as linear programming or dynamic programming, to ensure that energy consumption is minimized while meeting data transmission requirements.

[0028] Dynamic scheduling decision: Based on the output of the optimization model, the algorithm dynamically determines the priority and timing of data transmission. It can intelligently arrange the order of data packet transmission, avoiding transmission when the signal is weak or the battery is low, thereby reducing retransmissions and improving the transmission success rate.

[0029] Adaptive adjustment: The multi-parameter optimization scheduling algorithm has strong adaptability and can automatically adjust the transmission strategy according to real-time feedback and changes in network status to ensure that the optimal energy efficiency ratio can be achieved in various environments.

[0030] User-defined rules: The algorithm allows users to define transmission rules based on specific application scenarios, such as prioritizing the transmission of critical data when the battery is fully charged, or transmitting batch data during periods of network idleness.

[0031] Energy consumption and efficiency assessment: The algorithm integrates energy consumption and efficiency assessment tools to help users and system administrators understand the impact of different scheduling strategies on network energy efficiency and make adjustments accordingly.

[0032] Communication protocol integration: The multi-parameter optimization scheduling algorithm is seamlessly integrated with the existing LPWAN communication protocol to ensure that scheduling decisions can be correctly executed by the device and work in coordination with other elements in the network.

[0033] Security and reliability: While optimizing scheduling, the algorithm also considers the security and reliability of data transmission, ensuring that critical data can be delivered safely even under poor network conditions.

[0034] Through the detailed steps described above, the multi-parameter optimization scheduling algorithm can precisely control the time and frequency of data transmission, significantly improving the energy efficiency of LPWAN networks while ensuring high-efficiency data transmission and long-term stable operation of equipment.

[0035] 5. Security and Privacy Protection Mechanisms: An innovative lightweight encryption technology is integrated, specifically tailored for the low-power and resource-constrained environments of LPWAN devices. This encryption technology employs efficient algorithms and compact data structures to ensure fast encryption and decryption operations even on resource-constrained LPWAN devices. Through carefully designed key management and protocols, the algorithm achieves end-to-end encryption during data transmission, effectively preventing the risk of data interception or tampering during transmission. Furthermore, the lightweight encryption technology is designed to balance energy consumption and performance, ensuring that while providing security, it does not significantly impact the device's battery life or response speed. Through this mechanism, the algorithm not only protects the confidentiality and integrity of user data but also safeguards user privacy, meeting the growing security needs of the Internet of Things (IoT) while maintaining the energy efficiency advantages of LPWAN networks.

[0036] 6. LPWAN Device Hardware Modules: These modules collectively form the physical foundation of the device. First, there's the link quality monitoring module, which utilizes advanced signal processing technology to capture and analyze key indicators such as LQI and RSSI in real time, ensuring accurate awareness of network status. Next is the computing module, which possesses sufficient processing power to execute TinyML models, performing efficient data preprocessing, feature extraction, and local inference without relying on external computing resources. Furthermore, to enhance data security, the device integrates an encryption module responsible for encrypting data before transmission and decrypting it after reception. This module employs lightweight encryption algorithms, ensuring data security while avoiding unnecessary burdens on device performance and energy consumption. The close collaboration between these modules not only improves the performance of the LPWAN device but also provides robust hardware support for achieving highly optimized energy efficiency management and data security.

[0037] 7. Algorithm Deployment and Execution Method: This involves the algorithm initialization phase, during which all necessary hardware module configurations and system parameter settings are completed to ensure the device is in optimal startup condition. Subsequently, the real-time monitoring module begins operation, continuously collecting key data such as link quality and signal strength, providing input for subsequent steps. Next, the adaptive data rate adjustment mechanism dynamically adjusts the transmission rate based on the collected data to adapt to current network conditions and reduce energy consumption. Simultaneously, the intelligent sleep / wake-up strategy automatically manages the device's sleep and wake-up states based on the device's communication mode and the prediction results of the TinyML model, reducing energy consumption during non-communication periods. Furthermore, a multi-parameter optimization scheduling algorithm comprehensively considers multiple parameters such as signal quality, device power consumption, and environmental factors to perform real-time data transmission optimization scheduling, ensuring high data transmission efficiency and network energy efficiency. Throughout the execution process, the various modules and strategies coordinate with each other, forming an efficient and adaptive energy efficiency management system, enabling LPWAN devices to maximize energy utilization efficiency while ensuring communication performance.

[0038] 8. Computer-readable storage medium: Designed specifically for executing the aforementioned energy efficiency optimization algorithms. This program is a comprehensive software package integrating several key components, including an adaptive data rate adjustment module for dynamically adjusting data transmission rates to ensure optimized energy consumption based on real-time network conditions; a TinyML model integration module, allowing for efficient data processing and intelligent decision-making at the device end; an intelligent sleep / wake-up strategy module, automatically managing the device's energy consumption status based on predicted changes in communication demand; a multi-parameter optimization scheduling module, comprehensively considering various factors to intelligently schedule data transmission tasks to improve network efficiency; and a security and privacy protection module, protecting the security and user privacy during data transmission through lightweight encryption technology. These modules work together to maximize the energy efficiency of LPWAN devices through the program on this storage medium, while ensuring data security and network stability.

[0039] 9. System-level Implementation: The system comprises numerous LPWAN devices employing the aforementioned energy efficiency optimization algorithm and a centralized control unit. These devices utilize technologies such as adaptive data rate adjustment, TinyML model integration, intelligent sleep / wake-up strategies, multi-parameter optimized scheduling, and security and privacy protection to achieve individual energy consumption reduction and performance improvement. The central control unit is responsible for centralized management and coordination of these devices, monitoring, configuring, and optimizing the entire network by collecting status information and performance data from each device. The system design considers scalability, allowing for the deployment of a large number of devices across a wide geographical area, while achieving large-scale network energy efficiency optimization through intelligent scheduling and resource allocation at the network layer. Furthermore, the system includes intelligent management functions such as fault detection, self-repair, and automatic upgrades, ensuring long-term stable operation and ease of maintenance. Through this system-level implementation, this invention not only improves the energy efficiency of individual devices but also achieves efficient management and energy utilization of the entire LPWAN network, providing a solid technical foundation for the sustainable development of the Internet of Things.

[0040] By implementing the above technical solutions, this invention can not only significantly improve the energy efficiency of LPWAN networks, but also ensure the security of data transmission and the long-term stable operation of devices, providing strong technical support for the widespread application of Internet of Things (IoT) technology.

[0041] The following three technical effects are particularly significant when using the technical solution proposed in this invention: 1. Significantly Reduced Energy Consumption and Extended Device Lifespan: The adaptive data rate adjustment mechanism of this invention can adjust the data transmission rate in real time based on the Link Quality Index (LQI) and Signal Strength Index (RSSI), avoiding high-energy-consuming transmission attempts when the signal is poor, while improving transmission efficiency when the signal is good, thus maintaining low energy consumption under different network conditions. The intelligent sleep / wake-up strategy optimizes the device's sleep cycle by analyzing device communication patterns and predicting using the TinyML model, waking the device only when necessary for communication, significantly reducing energy consumption during non-communication periods and significantly extending battery life.

[0042] 2. Improved Data Transmission Efficiency and Network Throughput: The multi-parameter optimization scheduling algorithm comprehensively considers multiple dimensions such as signal quality, device power consumption, and environmental factors to achieve dynamic optimization of data transmission, avoiding network congestion and invalid transmissions, and improving network transmission efficiency. The innovation of this algorithm lies in its ability to respond to changes in network status in real time. Through predictive analysis and adaptive adjustment, it achieves reasonable allocation of transmission resources, thereby improving the throughput and service quality of the entire LPWAN network.

[0043] 3. Enhanced Data Security and System Reliability: The lightweight encryption technology employed in this invention is specifically designed for LPWAN devices, enabling end-to-end data encryption with extremely low power consumption. This ensures data transmission security while avoiding performance bottlenecks that traditional encryption methods may cause on low-power devices. The integration of the TinyML model provides the device with local intelligent processing capabilities, reducing reliance on the cloud, lowering latency, and improving response speed. In the event of network instability or partial device failure, local decision-making capabilities ensure continuous system operation and high reliability.

[0044] These three technical effects not only demonstrate the innovation of this invention in reducing energy consumption, improving efficiency, and enhancing security, but also showcase its broad application potential and long-term technical value in practical Internet of Things applications. Attached Figure Description

[0045] Figure 1 This is a flowchart of the adaptive data rate adjustment process of the present invention. Figure 2 The flowchart below illustrates the multi-parameter optimization scheduling algorithm of this invention. Figure 3 The algorithm deployment and execution method flow of this invention Detailed Implementation

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] The specific implementation details of the energy efficiency optimization algorithm of this invention on LPWAN devices are as follows: System initialization and parameter setting: Set the initial parameter set, including but not limited to data collection frequency (once every 5 seconds), initial transmission rate (50kbps), and sleep mode threshold (automatically enters when battery level is below 20%). Initialize the adaptive data rate adjustment module, TinyML model integration module, intelligent sleep / wake-up strategy module, multi-parameter optimization scheduling module, and security and privacy protection module.

[0048] Real-time monitoring and data collection: The monitoring module collects environmental parameters such as LQI, RSSI, equipment temperature, and humidity in real time, as well as equipment power consumption and storage space usage. Data collection frequency and monitoring cycle (10 minutes per cycle).

[0049] Adaptive data rate adjustment: Based on real-time LQI and RSSI data, the transmission rate is dynamically adjusted using a preset rate adjustment algorithm, such as an exponential backoff strategy based on LQI. The algorithm parameters include the rate adjustment step size (1 kbps) and the maximum / minimum rate thresholds (10 kbps-100 kbps). This achieves dynamic optimization of the transmission rate.

[0050] TinyML Model Integration and Local Data Processing: Integrates TinyML models for data preprocessing, feature extraction, and decision-making. The model training cycle is set to automatically update every 24 hours. Model accuracy threshold (95%) and resource consumption limit (not exceeding 30% of device CPU).

[0051] Intelligent sleep / wake-up strategy execution: Based on TinyML model predictions and device usage patterns, the sleep cycle and wake-up timing are intelligently set, taking into account device power consumption and communication needs. The execution cycle of the sleep / wake-up strategy is evaluated every 30 minutes, with a power drop rate threshold (no more than 1% per minute).

[0052] Multi-parameter optimized scheduling implementation: Taking into account factors such as signal quality, device power consumption, and environmental factors, the optimal time window and frequency for data transmission are calculated through an optimization algorithm. The optimization algorithm weights are set as follows: signal quality 0.5, device power consumption 0.3, and environmental factors 0.2, with the scheduling cycle updated every 15 minutes.

[0053] Security and privacy protection: Lightweight encryption technology is used to encrypt transmitted data, with a key update cycle of every 48 hours to ensure data security. Encryption algorithm selection (AES-128), key length (128 bits), and encryption processing time (not exceeding 10% of data transmission time).

[0054] Performance monitoring and feedback adjustment: Monitor data transmission success rate, network latency, device power consumption, etc., and set performance indicator thresholds. For example, automatically adjust the strategy when the transmission failure rate exceeds 5%. Performance monitoring indicators, adjustment thresholds, and feedback adjustment cycle (real-time monitoring, feedback every 5 minutes).

[0055] User configuration and interface interaction: Provides a user interface that allows users to adjust algorithm parameters according to specific needs, such as sleep mode and data transmission frequency. User configuration options and interface response time (no more than 2 seconds).

[0056] System-level implementation and network management: Device management and network coordination are implemented at the central control unit, setting network performance goals and optimization strategies. This includes network management strategies, device management cycles (synchronized hourly), and optimization strategy update frequency (reviewed quarterly).

[0057] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A low-power wide-area network (LPWAN) energy efficiency optimization algorithm, characterized in that, The process includes the following steps: dynamically monitoring and evaluating LPWAN network link quality indicators (LQI) and signal strength (RSSI); automatically adjusting data transmission rates based on real-time monitoring data to optimize energy consumption and transmission efficiency; and utilizing TinyML models to perform local processing and intelligent analysis of data collected at the device end, reducing unnecessary data reporting.

2. A low-power wide-area network (LPWAN) energy efficiency optimization algorithm, characterized in that, The adaptive data rate adjustment mechanism includes: setting a threshold parameter for rate adjustment and automatically adjusting the transmission rate according to real-time changes in link quality; achieving a smooth transition in rate adjustment and avoiding transmission instability caused by sudden rate changes.

3. A low-power wide-area network (LPWAN) energy efficiency optimization algorithm, characterized in that, The TinyML model includes: a TinyML model optimized for the computing power of LPWAN devices, used to perform data preprocessing, feature extraction, and pattern recognition; model training uses algorithms feasible on low-power devices to ensure real-time updates and adaptability of the model.

4. A low-power wide-area network (LPWAN) energy efficiency optimization algorithm, characterized in that, The intelligent sleep / wake-up strategy includes: intelligently setting the sleep cycle and wake-up time of the device based on historical data analysis of device usage patterns and TinyML model prediction; the sleep / wake-up strategy can be dynamically adjusted according to environmental changes and network status to achieve optimal energy consumption.

5. A low-power wide-area network (LPWAN) energy efficiency optimization algorithm, characterized in that, The multi-parameter optimization scheduling algorithm includes: real-time optimization scheduling of data transmission time and frequency by integrating multiple parameters such as signal quality, device power, and environmental factors; the scheduling algorithm can dynamically adjust the transmission strategy according to network load and device status to improve the overall energy efficiency of the network.

6. A low-power wide-area network (LPWAN) energy efficiency optimization algorithm, characterized in that, The security and privacy protection mechanisms include: using a lightweight encryption algorithm to encrypt the transmitted data end-to-end to ensure data security during transmission; and the encryption mechanism supports fast key exchange to adapt to the low power consumption and low computing power characteristics of LPWAN devices.

7. An LPWAN device for implementing the algorithm, characterized in that, include: The hardware module used to monitor link quality can collect LQI and RSSI data in real time; The computation module for executing TinyML models has data preprocessing and intelligent decision-making capabilities; the encryption module for data encryption ensures the security of data transmission and user privacy.

8. The LPWAN device is further characterized in that the computing module is capable of: flexibly selecting and adjusting the TinyML model according to different application scenarios and data types; and realizing rapid deployment and updating of the model to adapt to the ever-changing network environment and user needs.

9. A method for implementing the algorithm, characterized in that, Includes the following steps: Deploy and initialize the algorithm on the LPWAN device, including configuring adaptive data rate adjustment parameters and the TinyML model; collect network status data in real time through the device monitoring module to provide input for the algorithm; execute the adaptive data rate adjustment, intelligent sleep-wake strategy and multi-parameter optimization scheduling steps in the algorithm; adjust the data transmission strategy according to the algorithm output to achieve energy efficiency optimization of the device.

10. A computer-readable storage medium, characterized in that, The system stores a computer program for executing the algorithm, which includes: a software module for implementing adaptive data rate adjustment; a software module for implementing TinyML model integration and intelligent decision-making; a software module for implementing intelligent sleep-wake strategies; a software module for implementing multi-parameter optimized scheduling; and an encryption software module for implementing security and privacy protection.

Images