The computer networking
The computer network is defined as two or more computers that are connected together to share resources such as hardware, data, and software. Most common are the local area network (LAN) and the wide area network (WAN). Routers play a very important in the computer networking. Routing protocol is the set of rules defining the way router machines find the way that packets containing information have to follow to reach the intended destination.
The purpose of this project is to research about the behaviour of the two major routing protocols in the present day world. Open shortest path first (OSPF) and intermediate system to intermediate system (IS-IS). Both these protocols belong to the link state routing and both do interior gateway routing. In the process of deciding, which routing protocol is to be implemented in any network certain parameters are considered and checked as to which protocol suits the best, there are a few parameters which have not been tested for these two protocols. In this project we are comparing the OSPF and IS-IS routing protocols by considering few important parameters such as delay, resource usage, convergence time and network throughput. This provides additional help in deciding which protocol is best suited to be implemented in the real network.
The comparison is carried under simulated environment; the simulator used is the optimized network engineering tools (OPNET).
Computer networks form the major connectivity scheme in the present world. Computer networks are the most widely implemented criterion and used by the majority of the people they represent the sense of being attached and close by in large occasions. All thanks to the internet which created this boom in the modern world by bringing everyone together by large margins.
Anybody anywhere in the world is able to stay in touch with and communicate by the means of internetwork, large business transactions and defence activities are attributed to the internet. From personal mailing to the banking is done in a matter of seconds using the internet in the present day world. The computer networks are mainly divided into wired or wireless categories by the nature of implementation. The primitive level of networking consists of cables, servers, routers and switches.
Routers are one of the primitive elements in computer networking; they are purposely customized, configured devices which are used to forward data among computer networks beyond directly connected devices. Routing protocols is the implementation of the routing algorithm in the form of the software. Routing protocols uses the metrics to determine which path to utilize for the transmission of the packets; some of the metrics used are hop count, bandwidth, delay load etc. Here in this article we are concerned about two main routing protocols which are widely used across the globe by the large network providers.
Link state protocols requires each router to maintain partial information map at least, when there is any change in the network links a notification called as link state advertisements (LSA) is flooded across the network informing about the change, all the routers update the changes and recompute their routing tables accordingly. Hence, this method is more reliable, efficient, easier to debug and uses less bandwidth than the distance-vector protocols. OSPF and IS-IS are the two main link state routing protocols used by the network providers. OSPF is the open shortest path first routing protocol and IS-IS is the intermediate system to intermediate system routing protocol. They both fall under the interior gate way routing protocols.
In order to implement the best routing protocol in our network, we should consider a number of aspects such as the metric used, convergence time, resource utilization, bandwidth utilization, system delay, network throughput etc. In this case we are interested to know which protocol between the two proves to be better when they are tested against the delay, convergence time, resource usage and network throughput parameters. By which we can conclude by saying that which routing protocol is yields best results when applied in the best possible circumstances.
What is the significance of these parameters?
Delay: Network delayis an important design and performance characteristic of acomputer networkortelecommunications network. The delay of a network specifies how long it takes for a bit of data to travel across the network from one node or endpoint to another. It is typically measured in multiples or fractions of seconds. Delay may differ slightly, depending on the location of the specific pair of communicating nodes.
Resource usage: It is the amount of CPU and the memory used by the routing protocol to complete the task. Less the CPU and memory used by the routing protocol, better it is considered to be.
Convergence time: two routers are said to be converged only if it agrees on what the network topology looks like. When there is any change in the network, the time taken to notify all the routers about the change and time taken to update their routing tables is known as convergence time. This is one of the most important parameter to be considered in deployment of the routing protocol.
Network throughput: the average amount of time taken to transmit a packet from one node to another in the network is called as network throughput. It is usually calculated in bits per second. It is the sum of the data rates that are delivered to all the terminals in the network.
By considering the above parameters it seems to be very interesting to know about which protocol is better than the other. The biggest motivation in doing this project is because; the comparison of these protocols has not been carried out before using the above parameters.
All the protocols are entitled to have set backs in some parameters and do better in few others, but yet we get to choose from the best present at the moment. There may be huge difference in the performance of the protocol when compared with few parameters and vice versa with few other parameters. Hence we can conclude by saying that most appropriate parameters are considered while applying the routing protocol in the large broadcast network.
This paper presents the simulation model which is created for the analysis of both the routing protocols OSPF and IS-IS. In the first simulation model the OSPF protocol is applied and implemented to the network and the set of defined parameters are under investigation. Similarly in the second scenario the IS-IS protocol is implemented and analysed for the defined parameters under study. Accordingly the obtained results are compared with the first model i.e. the results of the OSPF model.
OSPF and IS-IS falls under the same set of routing protocols such as link state routing protocols they fall under the domain of interior gateway routing protocols. There are many similarities between the two protocols but there are some important aspects and parameters to be considered between the both which needs to be recognised and also analyse the pros and cons of the difference. And probably merge the differences and improve upon building a better protocol.
Basically, the first link state protocol was ARPANET, with major ideas being flooding of link state information to everyone, use of Dijkstra's algorithm to compute the paths, and incremental updates as and when there is a change in the network. The introduction of the IS-IS brought upon few changes and improvements such as making the link state distribution robust and incorporation of LAN's, exchange of parameter information such that the parameters can be set separately.
OSPF was designed for the IP specific networks and had a very loose relation with the IS-IS, IS-IS was bit complex to apply and overhead between them was minimal. There are few considerable differences between the two such as theories in electing the designated router (DR), passing over the upper layer information to other areas and parameter synchronization.
IDENTIFICATION OF THE PROBLEM
"Perhaps if the idea of using IS-IS for IP was thought of before Work on OSPF had started, there wouldn't be two protocols. But once A group starts on something it's hard to stop. So there wouldn't Be two protocols." Radia .
There are few major differences in both the protocols but the base is same, the differences such as:-
Designated Router election: it's "deterministic" in IS-IS, which means, given the same set of routers, the same router will be elected. It's "sticky" in OSPF, meaning that once you get to be DR, you stay DR. This makes things more stable...if the highest priority router is flaky in IS-IS, every time it goes up it takes over, only to crash again. But I was told when designing IS-IS that determinism was important, which requires the behaviour in IS-IS.
LSP distribution on a LAN: IS-IS does it with CSNP's. OSPF with explicit acks. I believe both ways are just fine.
Originally IS-IS passed no upper layer information into areas, and you just exited via the nearest level 2 router, whereas OSPF always fed all information into the area so you could choose the optimal exit point. Both have now been modified so you can choose any point on the tradeoffs between extra routing info and optimal routing.
Parameter synchronization: IS-IS allow neighbours to have different values for things like Hello Timer, and still interwork
The base journal which is followed here is "the Dissemination of Routing Information in Broadcast Networks: OSPF versus IS-IS" the journal compares the two routing protocols against a set of few parameters which helps to choose the appropriate protocol according to the situation. The set of parameters under review are the longest arrival time of a routing update packet to all the routers, the average arrival time of the routing update packets to all the routers, the total required bandwidth and number of memory access a router performs, which is the evidence of the internal work it performs.
The simulated results suggested that the scheme suggested in IS-IS is more efficient than that of OSPF in terms of the arrival times of routing update packets. In particular, the average arrival time of routing update packets in OSPF is 2-10 times longer than in IS-IS. In terms of the bandwidth each scheme consumes, there are scenarios where OSPF outperforms IS-IS and vice versa. In terms of the number of memory accesses routers perform in each scheme, IS-IS outperforms OSPF.
Hence, the critical evaluation of the parameters which I have noted upon can yield us good results as in terms of most appropriate routing protocol to choose from.
Current status and developments of research or technology
LITERATURE TECHNOLOGY SURVEY AND ANALYSIS
The difference between OSPF and IS-IS is noted down in one of the drafts, the notable differences can be summed up as:-
IS-IS does not support NBMA (non-broadcast multiple access) and point-to-multipoint links but OSPF does.
IS-IS rides directly above layer two, versus on IP like OSPF, which may offer a security advantage (IS-IS attacks cannot be routed).
IS-IS cannot support multiple links but OSPF can, because it rides on layer2 directly.
On broadcast networks, OSPF elects a DR and BDR which cannot be pre-empted, whereas IS-IS elects only a single DIS which may be pre-empted.
OSPF designates a backbone area (area 0) for inter-area advertisements; IS-IS organizes the domain into two layers.
IS-IS routers belong to exactly one area, OSPF routes can belong to multiple areas.
OSPF has more strict requirements for forming neighbour adjacencies. The hello and dead intervals must match, and the subnet mask must match (except on point-to-point links)
The above are some notable differences which can be taken into consideration while deciding the deployment conflict between OSPF and IS-IS. But there should be some more of notable differences which could well be helpful during the process, which definitely gives more leverage to the administrator. Hence, the parameters considered in this paper can help for the cause suitably.
One of the users of OSPF and IS-IS has noted down few security parameters differences between the 2 protocols which are:-
IS-IS runs directly over layer 2 and hence Cannot support virtual links unless some explicit tunnelling is implemented.
Packets are intentionally kept small so that they don't require hop-by-hop fragmentation.
Uses ATM/SNAP encapsulation on ATM but there is hacks to make it use VcMux encapsulation.
Some OSs that support IP networking has been implemented to differentiate Layer 3 packets in kernel. Such OSs require a lot of kernel modifications to support IS-IS for IP routing. Ditto goes for any HW that's clever and tries to identify L3 control packets in the ASICs.
Can never be routed beyond the immediate next hop and hence shielded from IP spoofing and similar Denial of Service attacks.
Need to provide code points of access for each data link protocol types (Frame Relay, Ethernet, ATM, PPP, etc).
Doesn't need to rely on network layer protocols (like ARP) to communicate with the neighbouring systems. Some implementations however, do rely on ARP or static routing to communicate with the neighbours on LAN.
OSPF runs over IP and hence can support virtual links.
Can use IP fragmentation services.
Can use VcMux encapsulation on ATM.
If an OS already supports IP, no it requires no changes to support OSPF.
Can be routed to a destination multiple hops away and thus vulnerable to DoS attacks and IP spoofing
Transmitted with additional IP header information, thereby increasing some packet overhead. I wouldn't fret over this because there aren't many control packets running helter-skelter in a network anyways!
The above theory states the encapsulation technique in OSPF and IS-IS. This gives an admin a clear indication about how to go about the security constraints in the network. But yet this plays a part of the role in the protocol selection and may be one of the most important of all aspects if the network is large and belongs to a very important organisation.
"Dissemination of routing information in broadcast networks: OSPF versus IS-IS" by ORAN SHARON, Haifa University is the journal posted in the IEEE website. This journal is used by me as the base copy for my dissertation, here the comparison of OSPF and IS-IS protocol is carried out by using
- The longest arrival time of a routing update packet at all the routers,
- The average arrival time of routing update packets at all the routers,
- The total required bandwidth and,
The number of memory accesses a router performs, which is evidence of the amount of internal work it performs.
In this article the he has only considered multi-access broadcast networks, both IS-IS and OSPF are link state routing protocols, as described earlier. IS-IS is the ISO standard routing protocol to support Connectionless Network Protocol (CLNP), which is an ISO standard protocol in the network layer . OSPF was designed by the IETF and is very similar to IS-IS, and many of the ideas of IS-IS are embedded in OSPF.
The model is created as per the broadcast network model assumed by each scheme. A broadcast network is a network where a single copy of a packet transmitted by a router/host is received by all other routers/hosts attached to it. In we show the broadcast network models in IS-IS and OSPF where the transmission infrastructure is shown as a cloud. We denote the broadcast network in each model by BN.
The network design consists of 7 routers connected across each other and FDDI broadcast network made in the centre and 2 routers made as DR (designated router) and BDR (border designated router) and the simulation process is carried out, by implementing OSPF first in the network and analysing the behaviour of the packet transmission and noting down all the simulated results, later IS-IS is implemented to the same network and tested against all the defined parameters. Finally comparing the results of each obtained.
The obtained results show that in the model of broadcast networks the scheme suggested in IS-IS is more efficient than that of OSPF in terms of the arrival times of routing update packets. In particular, the average arrival time of routing update packets in OSPF is 2-10 times longer than in IS-IS. In terms of the bandwidth each scheme consumes, there are scenarios where OSPF outperforms IS-IS and vice versa. In terms of the number of memory accesses routers perform in each scheme, IS-IS outperforms OSPF. Thus, giving a clear idea about the packet processing inside the two routing protocols. This document mainly concentrates on packet transmission and behaviour of the packet attributes and entities. It has nothing to do with calculation of the network delay and network throughput. The convergence time is the most important aspect to consider in the network behaviour and needs to be addressed. The amount of system utilization and CPU cycles required is one more critical parameter to be considered and needs to be addressed.
CHALLENGES AND EXPECTED CONTRIBUTIONS
Creating a network with designated router (DR) and border designated router (BDR), by making the topology with addition of number of routers to route between each other and making it look like a real corporate network as we are in ways to simulate a corporate network here seems to be quite a challenge, the important task is in the implementation of these two protocols in the simulated network and make it work. The careful consideration of the traffic that is generated and parameters to be jotted down which no one else has done before motivates me and is quite challenging to me.
The contribution towards the networking field by this dissertation could be very useful as it helps a great deal when it comes to know the real behavior of these protocols. Delay, resource utilization, convergence time and network throughput are the most important parameters to be considered by the administrators while the deployment of the protocol in the real scenario, these parameters have not been addressed before hence it can be a productive contribution towards the field of networking.
The comparison of these two protocols has been done previously by considering other set of parameters and concepts; I have come up with a set of critical parameters which needs to be assessed before the deployment of the protocols. Hence, this task seems quite daunting and challenging by the nature of its structure.
The main objective of this project is to find the critical performance evaluation comparison between the two link state routing protocols OSPF and IS-IS. By considering the delay, resource utilization, convergence time and network throughput parameters in Optimized Network Engineering Tools (OPNET). By creating a suitable network topology and analyzing the results by using the graphical representations.
- To create a network domain in OPNET.
- To analyze and understand the link state routing protocols.
- To critically study the behavior of the OSPF and IS-IS routing protocols.
- To design a new network topology to implement the OSPF and IS-IS.
- To implement the two routing protocols to the network separately.
- To capture the parameters under study separately.
- To analyze the protocols performance by using the captured parameters.
- To measure the graph by comparing between the two obtained results.
The output solely depends on the results obtained and no one can be sure of what the results will yield until the process is carried out. Though the strong feeling is that IS-IS may stand strong and believed to come good in comparison with the OSPF. In certain parameters OSPF can out beat the IS-IS as per previous researches. We can demonstrate the results with the use of graphs and tables.
METHOD OF THE PROJECT
This project is an extension to the existing projects on the performance evaluation comparison between the OSPF and IS-IS. Previously the process has been carried out using the real time systems in the test bed and also has been simulated using OPNET and NS2 simulators. In this project I will be using OPNET as the simulator to create a new network topology to address the issue.
THE METHOD TO BE EMPLOYED BY THE PROJECT
- Initially the network topology is created on paper to support both the protocols.
- The developed network topology is implemented using the OPNET simulator.
- The network which has been produced replicates a real corporate network.
- The OSPF protocol is implemented first to the network to be tested on the parameters.
- Traffic is generated to make the network active.
- The parameters are captured and saved separately.
- The IS-IS protocol is implemented to the network and traffic is generated.
- The parameters are captured and the comparison is made against the OSPF parameters and IS-IS parameters like delay, network utilization, convergence time and network throughput.
- The performance analysis is carried out with different values from which the average value will be considered.
- Finally the measurement is carried out using the UDP measurement.
HOW THE METHOD OF THE PROJECT TO BE ILLUSTRATED, VALIDATED OR EVALUATED
This project will be carried out by simulating the real corporate system by the help of Optimized Network Engineering Tools (OPNET). The simulated network is made active by generating traffic and the routing protocols such as OSPF and IS-IS is implemented upon the network. The parameters in which we are interested in is captured and saved from both the protocols, for the later comparison of the same. The different values are generated from which we consider the average value and measure the difference between the values such that we can draw upon which protocol is more efficient when compared delay, network utilization, convergence time and network throughput parameters.
- Desktop or Laptop PC with internet access
- Windows XP OS
- OPNET simulator software
- OSPF and IS-IS source codes
- MS OFFICE
- Access to IEEE
RISK MANAGEMENT STRATEGY
This project depends on OPNET simulation software. The OSPF and IS-IS source codes are needed for developing a new protocol. In case of any errors in the module then it needs to be resolved by getting help from the OPNET developer's team through e-mail and sort out any software bugs if present.
The parameters are available. It needs to be developed and integrated in OPNET and tested. Any doubts if aroused need to be clarified with the author via e-mail.
The CONVERGENCE TIME parameter is the most vulnerable parameter and depends on various aspects and scenarios in which the alteration takes place. Such that the recovery of the routes by the routing protocols depends upon which route in the network failed. That particular node is to shut down while comparing the protocols which may not be quite possible. Hence, the default or the plain testing of this parameter may be considered. Though, my best efforts will be put into this to generate most accurate and efficient results.
J. Moy. (July 1991).RFC1245 - OSPF Protocol Analysis. Available: http://www.faqs.org/rfcs/rfc1245.html. Last accessed 23 April 2010.
Radia Perlman. (Sat, 31 Aug 2002).Re: what is the fundamental difference between OSPF and IS-IS?.Available: http://www.ietf.org/mail-archive/web/ospf/current/msg00620.html. Last accessed 24 April 2010.
Stretch. (January 23, 2009).Internet draft: IS-IS and OSPF differences.Available: http://packetlife.net/blog/2009/jan/23/internet-draft-and-ospf-differences/. Last accessed 24 April 2010.
U. (January 23, 2009).Show me everything on WAN protocols.Available: http://searchenterprisewan.techtarget.com/sDefinition/0,,sid200_gci212728,00.html. Last accessed 24 April 2010.
U. (March 3, 2007).Differences in IS-IS and OSPF encapsulation.Available: http://routingfreak.wordpress.com/2007/03/03/differences-in-is-is-and-ospf-encapsulation/. Last accessed 25 April 2010.
U. (March 4, 2008).Shortest Path First (SPF) Calculation in OSPF and IS-IS.Available: http://routingfreak.wordpress.com/2008/03/04/shortest-path-first-calculation-in-ospf-and-is-is/. Last accessed 25 April 2010.
Wayne E. Bouchard. (Jun 21 13:36:20 2005).OSPF -vs. - ISIS. Available: http://www.merit.edu/mail.archives/nanog/2005-06/msg00420.html. Last accessed 25 April 2010.
Oran Sharon, Haifa University. (2001 IEEE). Dissemination of Routing Information in Broadcast Networks: OSPF versus IS-IS.Dissemination of Routing Information in Broadcast Networks: OSPF versus IS-IS. 01 (1), 70.
Radia Perlman. (1991 IEEE). a comparison between two routing protocols : ospf and is-is..OSPF vs IS-IS. 01 (1), 24.
Timothy G. Griffin, F. Bruce Shepherd, and Gordon Wilfong. (2002 IEEE). The Stable Paths Problem and Interdomain Routing.OSPF vs IS-IS. 01 (1), 19.
The project does not involve any humans or animal for experimentation. No act is done to affect human emotions and feelings. All the project work is carried within computer lab. Software's used for this project come under GPL. Therefore no ethical issues involved in the project.