Performance Analysis of Ipv4 To Ipv6 Transition Methods
Performance Analysis of Ipv4 To Ipv6 Transition Methods
Indian Journal of Science and Technology, Vol 9(20), DOI: 10.17485/ijst/2016/v9i20/90005, May 2016 ISSN (Online) : 0974-5645
Abstract
It is an arduous and indolent process to switch to IPv6 technology from the existing IPv4. This paper aims at providing
a lucid performance analysis of key techniques used in IPv4 to IPv6 transition. Sustainable and real time topologies
are built for each of the three robust techniques, namely, dual stack, tunneling, and network address translation. These
implementations are done in an open source Network simulator GNS3 (1.3.13), a wide compatible and realistic simulator.
All the topologies are analyzed for latency, efficiency and throughput using wire shark packet analyzer. Networks are built
using different commercially used cisco 7200, 3600 and 3700 series routers and used serial and fast ethernet cables for
connecting the nodes. All topologies are configured as private, to analyze each technique performance at their maximum
potential. This analysis can be found useful in employing the right transition technique depending on the network scenario
used as it weigh the advantage and limitation for each technique. The analysis depicts the competence of tunneling for its
highest latency comparatively. Among the three methods, Dual Stack displays 100% efficiency in communicating within
the network. Network address translation show 94% efficiency as it plays an important role when IPv4 only needs to
communicate with IPv6 nodes. In most cases, IPv6 show better performance than IPv4, which lucidly explains the potential
of IPv6. The analysis can further be extended to hardware implementation by constructing large topologies and with
various other sophisticated routers produced by different vendors.
Keywords: GNS3, IPV4, IPv6, Performance Analysis (Throughput; Latency; Efficiency), Transition Techniques (Dual Stack;
Tunneling; Translation), Wireshark
2 Vol 9 (20) | May 2016 | www.indjst.org Indian Journal of Science and Technology
C. V. Ravi Kumar, Kakumanilakshmi Venkatesh, Marri Vinay Sagar and Kala Praveen Bagadi
Vol 9 (20) | May 2016 | www.indjst.org Indian Journal of Science and Technology 3
Performance Analysis of IPv4 to IPv6 Transition Methods
R1 is 1945msec (avg.). Whereas, the lower part of the that link for the analysis. For better understanding, only
Figure 2 displays the trace route details of R1 to R8. ICMP v4 and v6 are individually filtered and analyzed on
It is clearly evident that IPv6 is having less latency and the parameters throughput and efficiency.
it is about 962msec (avg.), which is nearly half of that In networking, throughput is defined as the amount
of IPv4’s time. One inference that can be drawn from of data transferred from one node to another in a given
the above results is that with the distance increased, the time interval. Whereas, the efficiency is coined as the
performance of IPv6 is better in terms of latency in this number of packets successfully reached the destination.
scenario. Another important parameters to analyze are It is observed from the Figure 3(a) that the throughput of
throughput and efficiency of the data. the ICMPv4 packets is 0.003Mbits/s with 100% efficiency
As mentioned in the beginning of the section, all in this network scenario. Results of ICMPv6 in Figure
the analysis is carried in wireshark packet analyzer. For 3(b), shows that throughput is almost similar to that of
the analysis in this topology, we captured the data pass- ICMPv4 traffic. Figure 3(c) provides the comparative
ing through the interface f0/0 of R10. 3001:1:3:2/64 or analysis of ICMPv4 and ICMPv6 packets’ throughput in
10.1.3.0/24 link. Considerable traffic is generated across log scale over time.
Figure 3. (a) IPv4 ICMP packet throughput. (b) IPv6 ICMP packet throughput. (c) IPv6 and IPv4 ICMP packet throughput
comparative analysis.
4 Vol 9 (20) | May 2016 | www.indjst.org Indian Journal of Science and Technology
C. V. Ravi Kumar, Kakumanilakshmi Venkatesh, Marri Vinay Sagar and Kala Praveen Bagadi
3.2 Tunneling Implmentation carried with 100% efficiency and with 0.003 Mbits/sec.
Similar to that, Figure 7(b) gives a insight on the v4 tunnel
The Figure 4 shows the tunneling topology with ten routers
where the v4 traffic is carried in v6 packets. Data through-
similar to that of the dual stack topology. Here R1, R4, R7
put and efficiency are same as that of the above.
are IPv4 configured and have IPv4 loopbacks configured
to it. R2, R5, R8 areIPv6 configured and they are assigned
with their respective loopbacks. R3, R6, R9, R10 are
tunnel initiators and each of these have IPv4 loopbacks.
This topology contains two IPv6 tunnels for allowing v6
traffic over v4 network and one IPv4 tunnel for allow-
ing v4 traffic over v6 traffic. First IPv6 tunnel is present
between R3 and R10, second in between R9 and R10.
The IPv4 tunnel is located between the R6 and R10. The
traffic is generated throughout the network and the data is
captured for interfaces s0/0 and s0/1 of R10 in order to get Figure 5. Trace Route between R1 and R4 lpb.
the analysis of both v4 and v6 tunnels.
In the Figure 5 the third hop in the list i.e., 10.1.6.1 is
v4 over v6 tunnel. The total latency of the IPv4 packet
over v6 network is 575 msec (avg). The second hop in
the Figure 6, i.e., 3001:1:A::2 represents the IPv6 tunnel
over v4 network. The latency of the v6 packet over v4
network is being 648 msec (avg) which is nearly same
as that of v4 packet latency.
The Figure 7(a) lucidly depicts the v6 traffic is carried
as v4 traffic as the v6 tunnel is present there and data is Figure 6. Trace Route between R2 and R5 lpb.
Vol 9 (20) | May 2016 | www.indjst.org Indian Journal of Science and Technology 5
Performance Analysis of IPv4 to IPv6 Transition Methods
Figure 7. (a) ICMP packet throughput analysis over v6 tunnel. (b) ICMP packet throughput analysis over v4 tunnel.
6 Vol 9 (20) | May 2016 | www.indjst.org Indian Journal of Science and Technology
C. V. Ravi Kumar, Kakumanilakshmi Venkatesh, Marri Vinay Sagar and Kala Praveen Bagadi
both cases which nearly 1050 msec(avg). In this case Trace drawback found during translation is IPv4 configured R4
Route is able to show only the intermediate nodes till NAT was not able to ping IPv6 configured R5 since both the
server because of the address translation. In the above sce- routers are connected to the same interface of NAT con-
nario IPv6 to IPv4 address translation took place at router figured R10, which translates and redirects the traffic only
R10. So, an IPv4 configured R1 is able to ping IPv6 con- to other interfaces.
figured R8. Similarly any version IP address can ping any
version IP address with the help of NAT-64 and NAT-PT.
The Figures 11(a) and 11(b), shows the analysis of data
captured at interfaces s0/1 and s0/0 of R10 router. Figure
11(a) shows the traffic when data from IPv4 R1 is sent to
IPv6 R8. 94.7% of the v4 traffic is converted to IPv6 and
sent with the throughput of 0.03 Mbits/sec and similarly
there is only 92% efficiency found in the conversion of v6
traffic to IPv4 with throughput of 0.03 Mbits/sec. Another
Vol 9 (20) | May 2016 | www.indjst.org Indian Journal of Science and Technology 7
Performance Analysis of IPv4 to IPv6 Transition Methods
8 Vol 9 (20) | May 2016 | www.indjst.org Indian Journal of Science and Technology