摘要:Internet was designed about thirty years ago, and used as an academic communication and research platform for researchers at universities and research agencies of the United States at that time. However, with the great development and improvement of science and technology and many surprising inventions like desktop, laptop, mobile phone, PDA, intelligent sensor and so forth, Internet has evolved from being an academic communication and research platform to a huge ubiquitous information platform. The number of users of the Internet and the number of applications running on the Internet are both getting larger and larger, which have brought many challenges to the Internet itself. These challenges include the super-linear growth of routing table size in the Internet core routers which is also called routing scalability, mobility support for end host, multi-homing technology that provides the capability for a network site to be connected to more than one Internet Service Providers, and traffic engineering for a multi-homed site. These challenges are assumed to be caused from the primitive design principle’s contradiction with more and more newly developed applications over Internet. Many proposals have been published in the past few years to address these challenges, and these proposals can be divided into two categories, “evolutionary” and “revolutionary”, which are completely two different ways to consider the future of ubiquitous Internet. In this paper we reconsider the challenges faced by current Internet, and then propose a “clean-slate” and ID/Locator separation based architecture for future ubiquitous Internet (AFUI).