Research On Key Technologies Of Mobility Management For Future Internet Architecture

The mobility management technology is an emerging communication technology. Its main function is to guarantee the continuous communication services while the mobile terminals are moving. As the number and types of mobile terminals have increased dramatically, the mobility problem has become more and more complex. It brings a lot of challenges to the mobility management techniques over the traditional IP network. What’s more. it also challenges the architecture of traditional IP network. Therefore, a lot of researchers from academics and industry are turning their attentions to support of mobility over the next generation networks.Among the next generation network architectures, locator and identifier separation architecture (e.g., LISP) and control plane and data plane separation architecture (e.g., SDN) have attracted much attention in solving mobility problem. However, the researches of mobility management over these two network architectures are just starting up. In this paper, based on these two network architectures, we design a mobility management to solve mobility of both terminals and VMs. The main contributions are as follows:(1) Since network mobility supported LISP network architecture brings serious signaling loads, we propose a new mechanism NEMO-LISP (NEtwork MObility management in LISP). In our solution, a lightweight entity has been introduced to help a group of mobile nodes to access and handover to networks. Finally, compared to traditional mobility solution with LISP, NEMO-LISP brings fewer signaling load and higher handover efficiency.(2) In order to solve the problem that SDN network architecture brings serious triangle routing while solving VM migration, we design a redirection VM migration scheme based on target switch. This scheme takes avoiding triangle routing problem as constraint condition to seek a redirection VM migration with the lowest network influence. Besides, in SDN data center network, this article proposes a DARD VM location selection algorithm (user Distribution Aware Re-Deployment) from the view of online user distribution to solve the problem of unreachable user service level which is caused by unreasonable VM initial deployment. Furthermore, we proposed a TRM (Traffic Redirection VM Migration) scheme to maintain the active flows without being interrupted.