s the technology advances, and considering the needs of the growing users of Internet each day, Internet Protocol is one of the major concerns. IPv6 is simply the upgraded version of IPv4, and makes all the attempts to overcome the drawback of the previous 4 version of Internet Protocol. Today an end to end pervasive connectivity is the need of hour. At one end revolution of Internet enabled connected devices are required because all devices have to be always connected for proper communication. Keeping this in mind, there are two networks designed, one for each-IPv4 and IPv6. Attempt has been made to bring the easy to understand comparison between both the protocols on the basis of recital analysis of IPv4 and IPv6 in OMNeT++ simulation environment.
The IPv4 and IPv6 network have been designed in the Network Editor of the OMNeT++ simulation tool. A group of parameters have been taken which illustrate the working features, performance differences between both the protocols. After preparing the respective designs which show the various wired hosts, routers, network configurator, channel controller, channel installer and servers, and various type of connections between them; the relevant .INI file is made and Author ? : Associate Professor, Computer Science Engineering Department, Suresh GyanVihar University, Jaipur, Rajasthan, India. E-mail : [email protected] Author ? : Dean, Apaji Institute of Mathematics & Applied Computer Technology (AIM & ACT) Banasthali Vidyapeeth, 304022 Rajasthan. E-mail : [email protected] necessary coding is done in the C++ file which have .cc extension. Modules are then made to run and hence the respective simulation is performed. All the results and comparable issues are enclosed. In order to bring out the basic and foremost differences between the network protocols, both the IPv4 and IPv6, networks are designed. The different aspects in terms of parameters, featured attributes of these protocols are stated.
The designed network contains at least 90 wired hosts which play an important role in bringing out the basic performance of the protocol network. Along with these hosts there is a router which has the responsibility to transfer the different packets to the different hosts aligned in the network. All the management of the protocol is done from the "channel controller" which is also laid in to the network in course of designing. This channel controller needs no connection to be established with any of the devices and it automatically governs its working. Apart from the wired hosts, as mentioned earlier, there is IPv4 Network Configurator. The basic task of this device is to configure the different devices used in the IPv4 network like the v4 wired hosts etc. The parameters of these devices like, data rate, In this network, IPv4 protocol has been used for all routers, server and host by using flat Network Configurator, which assigns IP addresses to the network devices. There is total 30 numbers of hosts comprising a LAN and connected to a server via routers. Data rate channel has been setup between host and router with parameters as delay=0.1ms and data rate=100Mbps and between router and server with the following parameters like delay is set to 10ms and data rate=100Mbps. A ThruputMeter is also connected to routers as it provides through put measurement utility with parameters set as delay=10ms and data rate=100Mbps.
Complete Network description file for IPv4 based lan network is shown in Fig. 3, which describe the whole information about the network and the connections which have been made in the network:
All the wired hosts are entitled to receive the message packet by the communication mechanism. As specified that there are 90 wired hosts assumed in the NED. The module hierarchy of each wired hosts are further described. The programmer is allowed to create a new channel type which is capable to encapsulate all the data rate settings. In order to avoid the litter in global namespace, this type of channel can be defined inside the network itself. So some kind of mechanism is required to control and manage the activities of the channel created within the network. Hence Channel Installer is also placed in the designed network. The IPv6 network is also designed using 90 wired hosts. Channel Installer and Network Configurator are similar to the IPv4 network with slight difference in the working and major difference in the performance output. Similarly in this network, IPv6 protocol has been used instead of IPv4 for all routers, server and host by using flatNetworkConfigurator6, which assigns IPv6 addresses to the network devices. There is total 30 numbers of hosts comprising a LAN and connected to a server via routers. Data rate channel has been setup between host and router with parameters as delay=0.1ms and data rate=100Mbps and between router and server with the following parameters like delay is set to 10ms and data rate=100Mbps. A ThruputMeter is also connected to routers as it provides through put measurement utility with parameters set as delay=10ms and data rate=100Mbps.
Complete Network description file for IPv6 based lan network is shown below, which describe the whole information about the network and the connections which have been made in the network: The Fig. 8 illustrates the designed network of the IPv6 Protocol. This design may look a little similar to that of IPv4. But still there are several differences in the structure and the execution flow of both the protocols.
Under OMNeT++
Both the IPv4 and IPv6 networks are loaded with FTP traffic beginning at 50 bytes up to 100 MB with an inter-request time of 2000 seconds. The performance metrics for both IP networks are then measured and analyzed.
The first step to see the results of the simulation of the network is to build the entire network. Because every time any changes are made in the design or the code then the network is required to be reconfigured every time. As the constructions of the project will register all the functions built into the system tool and necessary updating of the INI files is done so that results obtained are according to the changes. The simulation time can be from few seconds to many hours. More is the simulation time, better are the obtained results, and the simulation time chosen was 6 hours.
Server 18 and 19 shows number of packets dropped by queue in IPv4 and IPv6 based LAN respectively. It is observed that there is no packet loss in case of IPv6 but in IPv4 some packets were dropped by queue represented by purple and highlighted using yellow color, as it's very small in number that's why we can't visualize clearly in graph. At about 50 Mbps there is no packet loss but if load on network exceed, packet loss increases, it is clearly visualize from table given below where highlights shows the no. of packets dropped. First thing is the bit rate, the bit rate, which is the total number of bits transmitted in some unit time (second). The receive bit rate for IPv4 was 13450.5094235678 bps and send bit rate for the same IPv4 was 15868.578533435bps. When it is compared with the IPv6 bit rate, it was less. The bit rate observed in IPv6 case was 36291.2904392990 bps.
Another very important key point in the wired transmission of the packets considered is the time in which the data packets are being delivered. Total messages created in case of IPv4 are 3219, and the total number of messages created in IPv6 protocol was 11073. The time at which these critical values were observed was 2.0159 minutes. It clearly explains the better output results in case of IPv6 protocol.
These were some of the major things which were observed during the simulation of both the networks.
Altogether it contributed to the better performance of the Ipv6 protocol over IPv4.
IV.
In this research work various performance parameters like throughput, packet loss, latency, etc. for both the protocols IPv4 and IPv6 based on wired networks were evaluated. Baseline IPv4 network, baseline IPv6 network have been simulated. The simulation has been done by using OMNeT++, which is a disceret event simulator. A comparative study of parameters was carried out in two different networks based on IPv4 and IPv6 respectively. This thesis analyses the performance of IPv4 and IPv6 Networks in OMNeT++. The network consists of 100Mbp links. The networks are loaded with FTP traffic to analyze their throughput, packet throughput, Delay, and response time. When network is loaded with FTP traffic the throughput is low for IPv4 compared to IPv6 during the low load and the difference is very small. When the FTP traffic increases throughput of both IPv4 and IPv6 increases, But Ipv6 shows a better result. The throughput for IPv4 and IPv6 is constant when the FTP traffic reaches the link bandwidth. Packet throughput is initially low for IPv4 than for IPv6, due to low FTP traffic. As the volume of data increases the number of packets in the IPv4 network is more than the IPv6 network. When the volume of FTP traffic is increased the delay in the IPv4 network is more than that of IPv6 because the IPv4 network has a higher number of packets to be processed than the IPv6 network.
In case of packet loss it was found that it is more in IPv4 as compared with IPv6. It was also found that IPv4 and IPv6 versions of IP protocol behave roughly the same in terms of Latency, with difference in overhead due to large header format of IPv6 may be because IPv6 is still in developing phase.
Thus, the analysis of IPv4 and IPv6 networks presents us with their performance characteristics through statistical analysis. The statistics obtained from simulation tells us that the performance of IPv6 is much better than IPv4. IPv6 performs better under specific circumstances.
So far the performance is concerned; the IPv6 protocol has better transmission efficiency despite the larger size of the header and the packet frame. Another key aspect is the jitter. Jitter is basically a slight irregular directional flow of the electrical signals, which are actually the data packets. When the simulation was in a running state, then more or less there was no major difference observed in the jitter values of both the protocols. Although in a comparison, IPv6 showed less jitter than IPv4 protocol.
With the extinct of the address spaces in IPv4, there is an immediate need to adopt IPv6 protocol as early as possible, so as to avoid future impediments in the Internet network.
V.
Future work can be done on satellite and wireless IPv4 and IPv6 networks. In future more research can be done on various aspects like study of IPsec as to observe the increase in overhead due to use of encryption and decryption concept using OMNeT++.
Comparison | IPv4 | IPv6 |
Parameter | ||
Send Bit Rate | 15868.578 | 33977.568 bps |
bps | ||
Receive Bit | 13450.509 | 33425.290 bps |
Rate | bps | |
and IPv6 simulation | ||
So the above description clearly states the | ||
working concept and technical aspect of both Internet | ||
Protocols. Basically five vital comparisons were | ||
considered and traced down while running this | ||
simulation on the OMNeT++ tool. |
Network Migration And Performance Analysis Of Ipv4 And Ipv6. ISSN 1857-7431. European Scientific Journal March edition 1857 - 7881. 2012. 8 (5) . M GC. University Faisalabad (Print)
IPv4/IPv6 Performance Analysis: Transport Layer Protocol Impact to Transmission Time. Internet Technology and Applications (iTAP), International Conference, (Nitra, Slovakia
Performance analysis of networks with IPv4 and IPv6. Computational Intelligence and Computing Research (ICCIC), IEEE International Conference, 2010.
Understanding current IPv6 performance: a measurement study. 10.1109/ISCC.151. Proceedings of 10th IEEE Symposium on Computers and Communications (ISCC'05), (10th IEEE Symposium on Computers and Communications (ISCC'05)Cartagena, Murcia, Spain