面向数据增强的物联网无线传输与智能服务关键技术研究

In order to meet the demand of diversified mass energy constrained devices in the Internet of Things (IoT) for low latency, high reliability, and high-precision intelligent services, this project focuses on low-energy efficient transmission, multidimensional data optimization modeling, and information mining, from the perspective of hierarchical IoT data enhancement technologies. This project specifically centers on: 1) For high-concurrency and privacy data communication needs, we propose highly reliable transmission protocols and interference coordination algorithms, and design safe transmission schemes for improving energy efficiency and spectral efficiency; 2) For low-power and large-connection business requirements, we design the edge node behavior model with caching and computing capabilities, and propose an intelligent resource allocation and transmission strategy based on the reinforcement learning; 3) For the demand of high-precision positioning in complex data acquisition environment, a position inference model based on perceptual information fusion is proposed, which uses deep learning to analyze the video image information and integrates the interest area trajectory to realize intelligent services; 4) Aiming at the problem of identifying individual heterogeneous terminals and accurate recognition of motion, we establish a motion fingerprint library for specific individuals, use image analysis to quantitatively model individual sensual actions from a cognitive perspective, and provide intelligent service adaptation based on different IoT business requirements. The implementation of this project will provide the theoretical and technical support for the green, intelligent, and sustainable development of the Internet of Things.

为了满足物联网中多元化海量能量受限设备对低时延、高可靠、高精度智能化服务的需求，本项目分别从低能耗高效传输、多维数据优化建模和信息挖掘等角度出发，围绕层次化物联网数据增强技术进行研究。包括：1）针对高并发隐私数据通信需求，提出高可靠传输协议及干扰协调算法，设计提升能效和谱效的安全传输方案；2）针对低功耗大连接业务需求，设计具有缓存及计算能力的边缘节点行为模型，提出基于增强学习的智能资源分配和传输策略；3）针对复杂数据采集环境高精度定位需求，提出基于感知信息融合位置推断模型，利用深度学习解析视频图像信息，结合兴趣区域轨迹实现智能化服务；4）针对异种终端个体识别与动作精确认知问题，建立动作指纹库识别特定个体，利用图像解析从认知角度将个体的感性动作量化建模，结合不同物联网业务需求提供智能化服务适配。该项目的实施将为物联网的绿色、智能、可持续发展提供理论和技术支持。