Abstract

This thesis deals with the study of Network Address Translation-Protocol Translation (NAT-PT), its limitations, and the way of avoiding the drawbacks of the protocol. NAT-PT is a transition mechanism for establishing communication between an IPv6 network and legacy systems. RFC 2766 describes the semantics of this mechanism. However, the proposed mechanism as described by RFC 2766 has a number of serious drawbacks that are of primary concern to its users. Due to these limitations, this mechanism is not widely accepted by the Internet community. Some of the most critical limitations of the proposed NAT-PT have been identified as end-to-end security, scalability, DoS attacks, etc. NAT-PT does not allow network layer and, in some cases application layer end-to-end security. As a result, the use of NAT-PT increases the threats to the existing vulnerable network security. The current study addresses the security related drawbacks of the existing NAT-PT model, and proposes a modified NAT-PT model. The modified model is able to establish secured communication between IPv6 and IPv4 as well as to correct other problems that may arise from the use of the existing NAT-PT. In addition, the current study also outlines a formal validation of the NAT-PT model with a model checker tool SPIN, which is a very powerful validation tool for distributed systems