IPv6 delay and loss performance evolution

Wed, 03/19/2008 - 10:30 by Damien Saucez • Categories:

The paper titled "IPv6 delay and loss performance evolution" written by Zhou et al. in the International Journal of Communication Systems proposes to analyze the performances of IPv6 paths compared to IPv4 paths between the same nodes.

The dataset is the RIPE database between the October 1st, 2003 and October 31th 2005. The principle is simple: process the RIPE database for the TTM boxes running both IPv4 and IPv6 and compare the performances in term of delay and loss ratio for the two protocols.

The results show that globally IPv6 works worst than IPv4 in term of delay and loss. However, the authors propose to make a distinction between native IPv6 and IPv6 tunnels. With this distinction, we observe that the performances of tunnels are worst than native IPv6.

Nevertheless, it is important to notice that delay is higher with IPv6 and presents more variation with the time for both IPv6 techniques. Fortunately, we can observe that delays are more affected by tunnels than the native solution. On the contrary, there seems to have no strong correlation between the loss ratio and the IPv6 solution.

Authors argue that performances are worst mostly because of the "experimental" status of IPv6 in most of the networks (e.g., IPv6 traffic is less prior than IPv4 one, software routing of IPv6...).

From the paper, we must remember that native IPv6 must be preferred in term of delay but remains outperformed by IPv4. For the loss ratio, the IPv6 solution has no major impact. In the next future, performances of IPv6 should reach IPv4 as the problems comes from (i) the software implementation vs hardware, (ii) the use of tunnels and (iii) the lack of good management tools and policies for IPv6.

Note: a good point for the paper is that it can be understood by only looking at the plots and their legend!