3.1 INTRODUCTION

IPv4 protocol uses 32-bit IP address. With 32-bit, we can make approximately four billions of numbers, but we can not fully utilize the 4 billion address space mainly because IP address space has been assigned by the class. Class is identified by the leftmost 3 bits of IPv4 address. Several novel mechanisms such as CIDR and Network Address Translator (NAT) are devised to better utilize IPv4 address space. ^[1] However, significant portions of assigned addresses are currently unused. NAT has shortcomings disrupting end-to-end communications.

There are also high demands for IP address assignment from many countries, companies and individuals. As the evolution of mobile communications, there is high chance that we need tremendous amount of IP address. For example, 4G mobile communications assume that all mobile terminals including current cellular terminals are IP enabled. Evolution of p-to-p communications also increases the demand for globally unique IP addresses. Thus, 32-bit address space may not enough to satisfy these demands.

IPv6 uses 128-bit length identifier to distinguish a host from others in Internet. Numerically, the length of IPv6 identifier is four times longer than the length of IPv4 identifier. Thus, there might be worries about the size of routing table. However, IPv6 address space is designed to be hierarchically managed.

In IPv4, there are three address types depending on the packet transmission scope or the number of receivers, such as unicast, multicast and broadcast addresses. Besides, special addresses, such as network address, ^[9] direct broadcast address, limited broadcast...