Anda di halaman 1dari 6

2016 International Conference on Informatics and Computing (ICIC)

Performance Evaluation of Routing Protocols RIPng, OSPFv3, and EIGRP in an


IPv6 Network

Siti Ummi Masruroh Fadly Robby Nashrul Hakiem


Department of Informatics, FST Department of Informatics, FST Department of Informatics, FST
UIN Syarif Hidayatullah UIN Syarif Hidayatullah UIN Syarif Hidayatullah
Jakarta, Indonesia Jakarta, Indonesia Jakarta, Indonesia
ummi.masruroh@uinjkt.ac.id fadly.robby11@mhs.uinjkt.ac.id hakiem@uinjkt.ac.id

Abstract—A Routing Protocol is a rule that determines how Interior Gateway Routing Protocol (EIGRP). Each routing
routers can communicate with each other. The development protocol has its own advantages and disadvantages. The
of modern computer networks such as the Internet makes determination and selection of routing protocols depends
routing protocols increasingly required to find the best and on several parameters that affect the quality of a network
efficient route. This study examined the performance of [3].
routing protocols RIPng, OSPFv3, and EIGRP in an IPv6 The growth of the Internet is very high which also
network by using GNS3 based on parameters of throughput, results in the depletion of the allocation of resources in
jitter and packet loss. Throughput is the actual speed of a Internet protocol version 4 (IPv4). Due to the demand for
network, the higher the throughput then the faster the data
content and web-based applications that create a large
transfer can be completed. Jitter is the difference in the
arrival time interval between packets at the destination
inventory, fewer and fewer IPv4 addresses are available.
node. To get a better value of QoS (Quality of Service), the IPv6 was introduced in 1994 as a solution to anticipate the
value of jitter must be kept to a minimum. Packet loss is the problems that existed in the address limitations of IPv4 [4].
number of lost data packets during the data transmission Changes to IPv6 also encouraged the development of the
process. To get a better QoS value, the value of packet loss existing routing protocols to support IPv6. Some routing
must also be kept to a minimum. The results of this research protocols that support IPv6 are such as RIPng, OSPFv3,
indicate that the RIPng routing protocol has the highest and EIGRP.
throughput value, whilst a combination of OSPFv3-EIGRP As each routing protocol has advantages in certain
routing protocols has the smallest jitter value. Lastly, the circumstances, then this research will conduct simulation
RIPng routing protocol has the smallest value for packet loss experiments of network performance by applying the
incorporation of each routing protocol to obtain a protocol
Keywords- Routing Protocol; QoS; throughput; Jitter; combination that has the maximum performance on the
packet loss; Routing Redistribution; RIPng; OSPFv3; network. It is said that the use of three or more routing
EIGRP protocols can improve network efficiency [5].
This research aims to evaluate the performance of the
I. INTRODUCTION network using a combination of three types of dynamic
routing protocol, namely RIPng, OSPFv3, and EIGRP that
In large-scale networks a complex routing protocol already support IPv6 with the parameters of throughput,
plays a very important role. The task of a routing protocol jitter, and packet loss. The purpose of this evaluation is to
is to determine the communication channels that can pass determine the best combination of routing protocols that
messages from one node to another node in order to get to can be implemented in a computer network.
the destination node in a network. The objective of an
efficient routing protocol is to determine the best path II. DISCUSSION
between the source node and the destination node, and to
maintain those lines [1]. With a routing protocol,
In this research we use simulation methodology that
communication between nodes can be performed on
consist of the following stages:
several different networks.
Along with the increasingly rapid development of A. Problem Formulation
technology, the growth of communication networks is also Increasingly complex network conditions cause a
getting larger. In modern communication networks such as Network Administrator to look to performing a network
the Internet network, a dynamic routing protocol is used design with the goal of good performance. Application of a
more often than a static routing protocol. Along with the proper routing protocol on a network can help improve
development of the network, it requires a dynamic routing throughput and efficiency.
design that can accommodate these changes without the The performance of a network can be measured by
intervention of the network administrator when the several variables, namely throughput, jitter and packet loss.
network develops or changes [2]. The application of each routing protocol has an impact on
Some dynamic routing protocols can be used in a the performance of a network. The right combination of
network, such as the Open Short Path First (OSPF), routing protocols can be expected to provide optimum
Routing Information Protocol (RIP), the Intermediate performance in a network.
System to Intermediate System (IS-IS) and the Enhanced In [5] has been discussed a performance comparison

978-1-5090-1648-8/16/$31.00 ©2016 IEEE


2016 International Conference on Informatics and Computing (ICIC)

of mixed protocols based on EIGRP, IS-IS and OSPF for E. Simulation


real-time applications, however this research uses an IPv4. The simulation used Graphical Network Simulator 3
The proposed research uses an IPv6 to analyze (GNS3) version 1.3.10, and VirtualBox version 5.0.4 that
performance of RIPng, OSPFv3, and EIGRP. runs on the Windows 10 operating system. The Router
B. Conceptual Model 3725 Series with Cisco IOS operating system that runs in
GNS3, and the operating system Ubuntu Server 14:04
Simulations will be carried out with a maximum of Trusty Tahr running in VirtualBox as a PC. To determine
three combinations of different routing protocols. The the performance of the network, the authors used version
design of the network model will be based on real life 3.0.11 Iperf applications running on the PC.
topology using seven routers. The design of the topology
has been split into three different routing protocol areas, F. Verification and Validation
namely, Area 0, Area 1, and Area 2. The topology will be The verification and validation stage performed
designed by using GNS3 which is also used to conduct experiments on each scenario to determine whether the
simulations. The tools which are used in the design are: simulated network designed in the previous stages had
1) 7 Cisco 3725 Series Router Units been run in accordance with the provisions of the Model
2) 2 PC Units Conceptual phase, Input Output Data and Modelling.
3) 9 Connection Serial DTE / DCE’s
4) 2 Copper Straight-Through Connections G. Experimentation
There are seven scenarios to be simulated with
different routing protocols and also combinations of
routing protocols. Experiments were conducted by sending
a TCP packet under each scenario as much as five times
the size of the different windows, and UDP packets five
times at different times. So the total experiment was
performed 10 times. For the TCP packets, the window size
used was 2, 4, 8, 16, and 32 Kbytes, while the time spent
for the UDP packet was 20, 30, 40, 50, and 60 seconds.
The value which was output at the end of the experiment
was the throughput of the TCP packets, jitter and packet
Figure 1. Draft Area Topology loss in the UDP packets.

C. Input Output Data III. RESULTS


1) Input A. First Scenario
There are three essential input attributes required,
which are Node, Bandwidth, and Window size. A node is a
network junction or connection point. Each terminal,
computer, router and other member is adjusted for the
number of routers used. The simulation used seven router
units and two PC units. The bandwidth is a measure that
indicates how much data can be passed through a network
connection. The amount of bandwidth used in this
simulation was 1 Mbits/s. The window size is the value or
the maximum size of the data that can be sent without
packet acknowledgment (confirmation). The window sizes
to be used were 2, 4, 6, 8, and 32 Kbytes. Figure 2. RIPng throughput graph

2) Output Figure 2 shows the change in the value of the


The output in this simulation is based on the major throughput for each RIPng trial. The smallest throughput
issues in a joint performance analysis of routing protocols values shown in the first experiment used 2 Kbyte window
RIPng, OSPFv3, and EIGRP on IPv6 networks. These size. The throughput value then increased in the second
issues are Throughput, Jitter and Packet Loss. Throughput experiment using a 4 Kbyte window size. In the
is the maximum amount of traffic that can be sent in a subsequent experiments, the throughput value did not
second. Jitter represents the time it takes a packet of data to change significantly.
be sent from the sending node and received at the Figure 3 shows the change in the value of the RIPng
destination node. Packet loss is the percentage of the jitter as each trial showed a fluctuating value. The jitter
amount of data sent compared to the data actually received. value shown in the third trial at a time of 40 seconds was
the highest while the smallest jitter value was indicated in
D. Modelling
the fifth trial at a time of 60 seconds.
Simulation will be conducted within the seven
scenarios. Scenario 3 consists of a routing protocol
combination, namely RIPng, OSPFv3, EIGRP, RIPng-
OSPFv3, RIPng-EIGRP, OSPFv3-EIGRP, and RIPng-
OSPFv3-EIGRP.
2016 International Conference on Informatics and Computing (ICIC)

Figure 6 shows the change in the value of the OSPFv3


jitter as each trial showed a fluctuating value. The jitter
value shown in the fourth experiment was the highest at a
time of 50 seconds while the smallest jitter value was
indicated in the fifth trial at a time of 60 seconds.

Figure 3. RIPng jitter graph

Figure 7. OSPFv3 packet loss graph

Figure 7 shows the change in the value of the OSPFv3


packet loss for each trial. The highest value for packet loss
was shown in the fifth trial at a time of 60 seconds and the
value of the smallest packet loss was shown in the first
trial at a time of 20 seconds.

Figure 4. RIPng packet loss graph


C. Third Scenario

Figure 4 shows the change in the value of the RIPng


packet loss for each trial. The highest value of the packet
loss was shown in the fourth and fifth trial at a time of 50
seconds and 60 seconds respectively. The value of the
smallest packet loss was shown in the first, second and
third trial with times of 20, 30 and 40 seconds respectively.
B. Second Scenario

Figure 8. EIGRP throughput graph

Figure 8 shows the change in the value of the EIGRP


throughput for each trial. The smallest throughput value
was shown in the first experiment using a 2 Kbyte window
size. The value of the throughput increased in the second
experiment using a 4 Kbyte window size. In the
subsequent experiments, the throughput value did not
change significantly.
Figure 5. OSPFv3 throughput graph

Figure 5 shows the change in the value of the OSPFv3


throughput for each trial. The smallest throughput value
was shown in the first experiment using a 2 Kbyte window
size. The value of the throughput increased in the second
experiment using a 4 Kbyte window size. Subsequently,
the throughput value did not change significantly.

Figure 9. EIGRP jitter graph

Figure 9 shows the change in the value of the EIGRP


jitter as each trial showed a fluctuating value. The jitter
value shown in the fourth experiment at a time of 50
seconds was the highest while the smallest jitter value was
indicated in the fifth trial at a time of 60 seconds.

Figure 6. OSPFv3 jitter graph


2016 International Conference on Informatics and Computing (ICIC)

Figure 10. EIGRP packet loss graph Figure 13. RIPng-OSPFv3 packet loss graph

Figure 10 shows the change in the value of the EIGRP Figure 13 shows the change in the value of the RIPng-
packet loss for each trial. The highest value of the packet OSPFv3 packet loss for each trial. The highest value of the
loss was shown in the third trial at a time of 40 seconds. packet loss was shown in the fifth trial at a time of 60
The value of the smallest packet loss was shown in the first seconds, while the value of the smallest packet loss was
and second trial at a time of 20 and 30 seconds shown in the first trial at a time of 20 seconds.
respectively.
E. Fifth Scenario
D. Fourth Scenario

Figure 14. RIPng-EIGRP throughput graph


Figure 11. RIPng-OSPFv3 throughput graph
Figure 14 shows the change in the value of the RIPng-
Figure 11 shows the change in the value of the RIPng-
EIGRP throughput for each trial. The smallest throughput
OSPFv3 throughput for each trial. The smallest throughput
value was shown in the first experiment using a 2 Kbyte
value was shown in the first experiment using a 2 Kbyte
window size. The value of the throughput increased in the
window size. The value of the throughput increased in the
second experiment using a 4 Kbyte window size. In
second experiment using a 4 Kbyte window size. In the
subsequent experiments, the throughput value did not
subsequent experiments, the throughput value did not
change significantly.
change significantly.

Figure 12. RIPng-OSPFv3 jitter graph


Figure 15. RIPng-EIGRP jitter graph
Figure 12 shows the change in the value of the RIPng-
OSPFv3 jitter as each trial showed a fluctuating value. The Figure 15 shows the change in the value of the RIPng-
highest jitter value was shown in the first experiment at a EIGRP jitter as each trial showed a fluctuating value. The
time of 20 seconds while the smallest jitter value was jitter value shown in the fourth experiment was the highest
indicated in the fifth trial at a time of 60 seconds. at a time of 50 seconds, while the smallest jitter value was
indicated in the fifth trial with a time of 60 seconds.
2016 International Conference on Informatics and Computing (ICIC)

Figure 16. RIPng-EIGRP packet loss graph Figure 19. OSPFv3-EIGRP packet loss graph

Figure 16 shows the change in the value of the RIPng- Figure 19 shows the change in the value of the
EIGRP packet loss for each trial. The highest value of OSPFv3-EIGRP packet loss for each trial. The highest
packet loss was shown in the fifth trial at a time of 60 value of the packet loss was shown in the fifth trial at a
seconds. The value of the smallest packet loss was shown time of 60 seconds. The value of the smallest packet loss
in the first and second trial at a time of 20 and 30 seconds was shown in the first trial at a time of 20 seconds.
respectively.
G. Seventh Scenario
F. Sixth OSPFv3-EIGRP

Figure 17. OSPFv3-EIGRP throughput graph


Figure 20. RIPng-OSPFv3-EIGRP throughput graph
Figure 17 shows the change in the value of the
Figure 20 shows the change in the value of the RIPng-
OSPFv3-EIGRP throughput for each trial. The smallest
OSPFv3-EIGRP throughput for each trial. The smallest
throughput value was shown in the first experiment using a
throughput value was shown in the first experiment using a
2 Kbyte size window. The throughput value increased in
2 Kbyte window size. The value of the throughput
the second experiment using a 4 Kbyte window size. In the
increased in the second experiment using a 4 Kbyte
subsequent experiments, the throughput value did not
window size. In the subsequent experiments, the
change significantly.
throughput value did not change significantly.

Figure 21. RIPng-OSPFv3-EIGRP jitter graph


Figure 18. OSPFv3-EIGRP jitter graph
Figure 18 shows the change in the value of the Figure 21 shows the change in the value of the RIPng-
OSPFv3-EIGRP jitter as each trial showed a fluctuating OSPFv3-EIGRP jitter as each trial showed a fluctuating
value. The jitter value shown in the third trial was the value. The jitter value shown in the second trial had the
highest at a time of 40 seconds, while the smallest jitter highest value at a time of 30 seconds while the smallest
value was indicated in the fifth trial at a time of 60 seconds. jitter value was indicated in the fifth trial with a time of 60
seconds.
2016 International Conference on Informatics and Computing (ICIC)

value of jitter, while the RIPng OSPFv3 routing protocol


indicated the highest jitter value.

Figure 22. RIPng-OSPFv3-EIGRP packet loss graph

Figure 22 shows the change in the value of the RIPng-


OSPFv3-EIGRP packet loss for each trial. The value of
packet loss shown in the fifth trial was the highest at a time Figure 25. Packet loss value comparison chart
of 60 seconds. The value of the smallest packet loss was
shown in the first trial at a time of 20 seconds. Figure 25 shows a comparison of the average value of
packet loss for each combination of routing protocol. The
IV. EVALUATION RIPng-EIGRP routing protocol indicated the best average
value of packet loss, while the packet loss of the OSPFv3
routing protocol had the greatest value.
V. CONCLUSION
Based on the simulation results obtained, there were
mixed results in the experiments for each routing protocol
and routing protocol combinations. Comparison of the
performance of the routing protocols and the combined
routing protocols of RIPng, OSPF and EIGRP in terms of
the TCP packet throughput, the best value was obtained by
RIPng with an average throughput value of 935 Kbits/s
and the least was from OSPFv3 with an average
throughput value of 808.8 Mbits/s. For the UDP packet
Figure 23. Throughput value comparison chart testing, the best jitter value was achieved by OSPFv3
EIGRP with an average jitter value of 41 ms and the
Figure 23 shows a comparison of the average value of highest jitter came from the RIPng-OSPFv3 protocol with
throughput for each combination of routing protocols. The an average value of jitter of 60 ms. For the best value of
RIPng routing protocol had an average value that showed packet loss, this was achieved by the RIPng protocol with
the best throughput. However the difference in the average an average value of packet loss of 4.4 % and the highest
throughput values for the RIPng routing protocol with the was for the OSPFv3 protocol with an average value of
EIGRP routing protocol, RIPng-EIGRP, EIGRP and packet loss of 14.4 %.
OSPFv3-RIPng-OSPFv3-EIGRP was not too great. The
lowest average throughput was shown by the OSPFv3
routing protocol, followed by the RIPng-OSPFv3 protocol. REFERENCES
[1] M. Tarique, M. S. Islam, and M. A. Riaz, “Performance Analysis
of the Routing Protocols for Video Streaming Over Mobile Ad
Hoc,” vol. 4, no. 3, pp. 133–150, 2012.
[2] V. Bahl, “Performance Issues and Evaluation considerations of
web traffic for RIP & OSPF Dynamic Routing Protocols for
Hybrid Networks Using OPNET,” vol. 2, no. 9, pp. 1–9, 2012.
[3] D. Xu and L. Trajkovi, “Performance Analysis of RIP , EIGRP ,
and OSPF using OPNET,” 2013.
[4] F. Heriyanto, “Perbandingan Internet Protocol Versi 4 Dan Versi
6,” Fak. Ilmu Komputer, Univ. Sriwij., vol. 6, 2010.
[5] S. Farhangi, a Rostami, and S. Golmohammadi, “Performance
Comparison of Mixed Protocols Based on EIGRP , IS-IS and
OSPF for Real-time Applications,” vol. 12, no. 11, pp. 1502–1508,
2012.

Figure 24. Jitter value comparison chart

Figure 24 shows a comparison of the average value of


the jitter for each combination of routing protocol. The
OSPFv3-EIGRP routing protocol had the best average

Anda mungkin juga menyukai