Based On The IPv4Network Network Through The Technology Research And Realization

At a ceremony held on3February,2011, the Internet Assigned Number Authority(IANA) allocated the last five remaining IPv4addresses to the five big RegionalInternet Registries (RIRs). With this action, the pool of available IPv4addressesmanaged by IANA is now fully depleted and the world will usher in the era of IPaddress depletion. The fundamental solution to this grim status quo of the internet is the128bit IPv6address. However, due to high cost, commercialization of IPv6has reacheda bottleneck, despite some small-scale adoption within the education network. Onecompelling alternative to IPv6is usage of Network Address Translation (NAT).NAT technology temporarily solves the IP address shortage problem, butintroduces another problem: although NAT can protect the internal network, it preventspublic network users from accessing the network services of a host in the internalnetwork which does not have a public service port mapping; it also limits a largenumber of applications that utilize P2P technology. This issue has spurred theemergence of a variety of traversal technology, each with pros and cons.Commercializing a traversal technology necessitates that the technology requiresminimal network changes and minimum investment. In this light, the STUNTtechnology was born. It can bypass the private network to access external networks,while requiring minimal investment and no change to the existing network topology.However, STUNT has a shortcoming in its application: it is not able to bypass asymmetric network. This research project aims to utilize port forecasting techniques,along with port traversal technology, to predict the external IP address and port usedwhen a connection is first established, thus enabling STUNT to bypass a symmetricNAT network. In addition, I hope to improve the efficiency of a STUNT connectionwhile maintaining the success rate, which is affected by several factors, at a certainacceptance level. This research project will focus on analyzing the main factors andfinding a solution to ensure a high success rate of prediction. Some possible solutionsinclude shortening the length of the critical time, failed retries, and multi-portprediction. This research project is organized as follows: the first part focuses on literaturereview and experimental data collection; in the second part, the project is divided intoseveral different functions of the network application modules and a flowchart is createdfor each module; and in the last part, I use VC++to write simple software to test thesolution and present the numeric results.