中文摘要：

随着互联网技术的飞速发展，物联网技术也在不断发展进步，物理对象之间的网络互联性与交互性更加紧密也使得物理对象面临着越来越严峻的安全威胁，物理对象的感知工作和安全威胁问题是学术界和工业界急需解决的。物理对象感知方法研究为网络安全防护提供了工具和手段，在对比分析现有典型物理对象感知系统的基础上，针对目前物理对象感知系统中存在的问题，提出了基于BGP最近邻优先的感知方法，同时对网络中的通信协议和感知策略进行研究，提出基于设备网络特征分类的有效物理对象识别方法。对比实验表明，这些方法在IP地址生成、目标设备判别以及针对特定网络协议的感知上，在快速性和准确性上有了一定提高。这些物理对象感知方法弥补了现有工具的不足，在一定程度上为今后全球网络分析提供了研究可能。

英文摘要：

With the rapid development of Internet technology, IoT technology also develops quickly. Due to the more connectivity and in- teractivity of network between physical objects which are faced increasing security threats. The physical object discovering works and se- curity threats urgently need to be solved in academia and industry, and physical object discovery methods provide research condition for network security. Through the comparison and analysis of the existing typical physical object discovering system, a discovering method based on BGP Nearest Neighbor First has been proposed, which is intended to solve the problems in present system. At the same time, by studying the communication protocol and sensing strategy in the network, another effective physical object recognition method based on device network feature classification has been put forward. Compared with traditional method, these methods have achieved high accuracy and efficiency in IP generation, target device discrimination and specific network protocol recognition, which have made up for the defi- ciencies of existing tools. To a certain extent it provides the possibility of research for future global network analysis.