基于动态信道自学习的无线体域网高能效通信理论研究

Wireless Body Area Network (WBAN) is a human centered network, which featured by various kinds of application scenarios, different requirements of Quality of Service (QoS), dynamic channel environment and limited resource. High energy efficiency communication is the key to reduce the energy consumption of the nodes and ensure long-term and continuous work. As influenced by the physical geometry, the tissue composition and the movement of the human body, the channel quality of different parts of the human body is different and time variant. However, the traditional fixed mode data transmission scheme didn’t take the dynamic characteristics of the human body channel and the diversity of WBAN applications, so it couldn’t realize the adaptive optimization of energy efficiency. According to the characteristics of WBAN applications and the high energy efficiency requirements, this project proposes a human body/near human body dynamic channel self-learning based energy efficient communication theory for WBAN. First, the Finite Difference Time Domain (FDTD) method will be used to establish the mathematical path loss model of the heterogeneous and dynamic human body channel. Real-time link quality prediction will be achieved based on multi-information fusion technology. Then, according to the abstraction of the quality of service for the typical WBAN applications, the online WBAN applications classification model will be given and the real-time communication QoS requirements will be analyzed. On the basis of link quality prediction and QoS requirements analysis, the optimal parameters for the physical layer, the link layer and the application layer of the transmission system will be calculated, which not only meet the communication requirements but also has the optimal energy efficiency. The research results of this project will provide theoretical and technical support for the research of low power data transmission protocols of WBAN and the standardization of communication technology.

无线体域网（WBAN）是以人为中心的网络，具有业务多样性、通信需求差异性、环境动态性和资源受限性的特点。高能效通信是降低能耗，保证节点长期连续工作的关键。受人体几何结构、组织成分和运动姿态的影响，人体不同部位的链路质量存在差异性和时变性；而传统的固定模式的数据传输方案缺少对信道环境和业务特征的研究，无法实现能效的自适应优化。本项目拟结合WBAN的特点，提出基于人体/近人体动态信道自学习的高能效通信理论。首先利用时域有限差分方法建立人体运动状态下路径损耗数学模型，并通过多元信息融合实现链路质量实时预测；然后通过对WBAN典型业务的通信服务质量抽象化感知，建立业务在线分类模型，实现通信需求实时分析；在链路质量预测和通信需求分析的基础上，利用约束优化理论计算当前链路质量和通信需求下能效最优的物理层、链路层及应用层传输参数。本项目研究成果将为WBAN高能效传输方案设计及通信标准化提供理论和技术支持

无线体域网（WBAN）是以人为中心的网络，具有业务多样性、通信需求差异性、环境动态性和资源受限性的特点。高能效通信是降低能耗，保证节点长期连续工作的关键。受人体几何结构、组织成分和运动姿态的影响，人体不同部位的链路质量存在差异性和时变性；而传统的固定模式的数据传输方案缺少对信道环境和业务特征的研究，无法实现能效的自适应优化。本项目结合WBAN业务多样性和信道多变性的特点，提出了基于近人体动态信道自学习的高能效通信理论。首先，基于多源信息融合理论提出了基于可穿戴传感器的无线体域网业务场景分类方法,利用可穿戴传感采集的多源信息建立基于K-means的WBAN业务聚类和通信需求分类模型。进一步，利用人体不同部位佩戴的可穿戴设备采集到的人体的运动信息学习不同部位节点之间的信道状态；最后在链路质量预测和通信需求分析的基础上，利用约束优化理论计算当前链路质量和通信需求下能效最优的物理层、链路层及应用层传输参数，实现无线体域网通信带宽和传输功耗的双重优化。本项目相关的研究成果共发表SCI检索期刊论文5篇，国际会议论文3篇，申请发明专利9项（包含国外专利3项），培养研究生3名，项目成果作为“穿戴式信息连续监测与分析技术及系统”的部分内容，获广东省科学技术奖技术发明类二等奖。