Abstract
|
In order to connect to the Internet, each device needs an IP address. However, the number of IPv4 addresses is limited and insufficient to respond to the explosion of demand for new addresses for connected devices, including connected objects and smartphones. On February 2011, the IANA (Internet Assigned Numbers Authority) reported having exhausted the /8 blocks of IPv4 addresses intended for regional Internet registry (RIR). Gradually, the RIRs have exhausted their stock in turn. Consequently, it is necessary to deploy the new version of the Internet Protocol (IPv6), which considerably extends the address space. As IPv4 and IPv6 are incompatible (the headers are different from each other), the transition from the current version (IPv4) to the new version (IPv6) can't be done in a period of time, and this deployment has to be done gradually. In order to counter this issue, three solutions are possible: a) put a double stack on each device, b) translation or c) tunneling. Tunneling is the best solution possible. However, as each technology, tunneling is influenced by scalability. In this paper, we are proposing a solution for the IPv4/IPv6 transition, based on the LISP protocol (Locator/Identifier Separation Protocol), and we are studying its impact in relation to the IPv6 manual tunnel and the native IPv4/IPv6 networks. For this reason, we will perform an experimental study of the scalability under GNS3 by increasing the number of customers and varying the different technologies in order to deduce the best solution. For the performance measurements, we used VoIP traffic generated by IP SLA (Service Level Agreement). The evaluation criteria are the delay, jitter, MOS score, and loss rate. The results of this research will be important for the network administrators and various Internet service providers (ISP) for planning IPv6 migration networks.
|
Keywords
|
IPv4, IPv6, manual tunnel, LISP, GNS3, IP SLA, VoIP, Scability.
|