在能量收集直通通信（device to device, D2D）中，物理层安全技术着眼于物理层，为能量收集终端的数据业务搭建了第一道安全屏障。传统的物理层安全算法把能量收集终端抽象成一个节点，利用电池能量和物理信道, 结合无线信道衰落、时变特征，来建立数据安全传输链路。然而，能量收集终端都是由用户携带，把能量收集终端抽象成一个节点，忽略用户行为对安全链路建立的影响，并不能真实反映实际能量收集D2D网络中，数据安全传输的情形, 因此传统的物理层安全算法具有一定的局限性。针对当前能量收集D2D网络物理层安全机制的不足，本项目在提取用户行为特征的基础上，对能量收集D2D网络安全机制和算法展开如下研究：首先，提取用户行为特征，并对行为特征建模；其次，研究用户行为特征、电池能量、物理信道之间的相互作用对建立安全链路的影响；最后，在用户行为特征导向下，基于电池能量、物理信道构建安全传输机制和算法模型。

In energy harvesting device to device (D2D) communications, physical layer security serves as the first safety barrier for the sensitive data. The proposed physical layer security algorithms set up secure links for the sensitive data by abstracting the energy harvesting device as a node, exploring the battery energy and physical channel and considering the fading and the time varying characteristics of wireless channels. However, the proposed physical layer security algorithms are imperfect, and even cannot reveal the real security situation of sensitive data transmission in energy harvesting D2D networks, since they neglect the influence of user behavior on security links setup. Aiming to cover the shortages of the proposed physical layer security algorithms, this project will focus on the research of user behaviors oriented security mechanisms and algorithms in energy harvesting D2D networks. First, the user behaviors patterns will be mined, and the user behaviors characteristics will be extracted and mathematically modeled. Then, the influence of the interaction among user behaviors characteristics, battery energy and physical channel on the security links setup will be researched. Finally, under the orientation of user behavior characteristics, the security mechanisms and algorithms will be formulated based on battery energy and physical channel.