Network Performance While Running EIGRP And OSPF Information Technology Essay

ABSTRACT

This MSc project proposal is aimed to scrutinize(analyse) the network performance while running EIGRP and OSPF routing protocols on the network with best effort and differentiated services model and to compare their performance on data, voice and video traffic during congestion and link failure.

The proposed proposal explains (illustrate) the aim, objectives, plan and contribution to accomplish the proposed project. The comparison analysis will be based on the parameters such as routing protocol traffic, convergence time, throughput and utilization for data traffic. The voice and video performance will be analyze using parameters such as delay, jitter, traffic and packet loss. Optimized network engineering tool (OPNET) will be used as a simulation software for network design, configuration and analysis.

INTRODUCTION

In an IP network routers must provides routes so that the traffic can travel from source node to destination node. Dynamic routing protocols allow routers to automatically learn of available route to the destination network. Data, voice and video traffic are forwarded by the routers based on the information in the routing table [1].

Different algorithms are used by different dynamic routing protocols to determine the shortest path from the source to the destination. Every router has the basic information about the directly connected networks, they share this information among immediate neighbours and this information is then propagated throughout the network. In this manner, all routers are updated and gain knowledge about the network topology. The dynamic routing protocols should ensure fast convergence, best and loop free routes and data should be transported securely to a particular network. The metric used by the dynamic routing protocols depends on hop count, bandwidth, delay, load, reliability and cost. There are two major classes of routing protocols

1. Interior gateway routing protocols (RIP, EIGRP, OSPF)

2. Exterior gateway routing protocol (BGP)

All the routing protocols provide the same functionality so it should be identified which routing protocol is better for a particular environment because companies and organizations are always looking forward towards to improve network performance.

Link failures, routers failure and changes in the network topology effect the routing protocol performance. The convergence time of the routing protocols also depends on the configured timers of the routing protocols. In the event of the link failure and changes in the topology can increase the amount of delay and packet loss. The protocol traffic increases during convergence and thus the utilization of the link increases.

In VoIP network, data and voice traffic are integrated. In best effort network, all type of traffic is serviced in similar manner and no specific priority is given to real time traffic (voice and video). The reconvergence of the network can be detrimental for real time applications as these types of applications are delay sensitive [2].

Quality of service mechanism should apply to real time applications so that the packets which are delay sensitive should be serviced first. QoS model such as differentiated service model can be used to provide low delay services to the critical traffic while providing the best effort service to non critical data traffic [3].

This project provides a simulation based study on the performance comparison of EIGRP and OSPF on the proposed network and to investigate their effect on voice and video performance with link failure and congestion. OPNET simulator will be used to carry out the simulation experiments.

PROBLEM DESCRIPTION

The problem description consist of the following points given below

Evaluation and performance comparison of EIGRP and OSPF routing protocols on the proposed network topology.

Effect of link failure and congestion on the convergence of EIGRP and OSPF routing protocols.

VoIP performance comparison between EIGRP and OSPF routing protocols.

Video performance comparison between EIGRP and OSPF routing protocols.

EIGRP AND OSPF routing protocol performance over best effort and diffserv model.

EIGRP and OSPF routing protocols are mainly used as interior gateway protocols in an enterprise interior network. OSPF is more widely used as it is more scalable than EIGRP but it is complicated to set up on the network.

On the other hand, EIGRP is proprietary protocol but is easy to set up or configure on the network topology. EIGRP provides faster convergence in medium scale network [4].

Congestion and link failure causes routing instabilities which can degrade the network performance [5]. Few milliseconds of disruption can degrade the voice and video quality [5]. Routing instability can cause data, voice and video traffic to be delayed, dropped and reach out of order at a particular destination, thus degrading the network performance.

The routing protocols must be analyzed to determine which protocol is better for a particular environment, thus utilizing the network resources efficiently and speeding up the network performance. The delay for real time applications must be less than 150 ms and the jitter should be less than 30 ms [2]. Packet loss should be less than one percent and the mean opinion score must be maintained above level 3 for real time applications [2].

QoS models are used to provide a different level of service to different types of network traffic and can be applied at any point in the network. QoS improve network services by providing special services to voice and video calls without losing quality and satisfying the service level demand.

Diffserv is a scalable QoS model and is based on per hop behaviour. In diffserv model different types of traffic are first classified and then marked.

The objective is to scrutinize the performance of the network using EIGRP and OSPF routing protocols on the proposed network topology and then measured the different network performance parameters.

The project research questions are

What will be the effect of EIGRP and OSPF routing protocol on the network performance and to compare which one gives better performance?

What will be the effect of link failure and congestion on EIGRP and OSPF routing protocols?

Investigate the effect of EIGRP and OSPF routing protocols on voice and video performance?

Investigate the improvement in voice and video performance using diffserv compare to best effort model?

RATIONALE

The project outcomes will be beneficial for companies in the designing phase of the network and to decide which one is effective for a particular environment by not restricting to one protocol. Another important outcome of this research is that the network administrator can understand the effect of routing protocols on real time applications during network outages. This will help to choose the most suitable dynamic routing protocol for voice and video applications.

LITERATURE SURVEY

Large enterprise network and service providers depend on routing protocols to maintain the routing table up to date. The router should be aware of where to send the traffic to reach the final destination. For efficiently and effectively distributing data, the choice of routing protocol becomes very important. There are several factors such as routing protocol traffic, convergence, selection of best path if more than one path exists, differentiate one routing protocol from other, CPU and memory consumption, selecting proprietary or standard protocols and administration ease are also other factor in the routing protocol selection. Three routing protocols OSPF, EIGRP and RIP are compared in [6]. The performance metrics used for comparison are routing protocol traffic, convergence activity, convergence duration, memory and bandwidth utilization, queuing delay and throughput. The network used for simulation contains different locations across various states within United States of America. Each of the site location contains servers, routers and terminal devices. Hundred Mbps links are used to connect routers to end device and OC links have been used between routers. Three routing protocols OSPF, EIGRP and RIP are configured on the same network for comparison. The result shows that EIGRP provides better performance compared to the other two protocols. The study shows that EIGRP provides improve memory utilization, CPU utilization and intelligent bandwidth control. The convergence of EIGRP was faster as compared to the other routing protocols and is easy to maintain the network and has better control on timers.

VoIP performance is studied over different routing protocol in [2]. The protocols which are compared for voice are EIGRP, OSPF and RIP. The voice performance is compared using metrics such as MOS value, end to end delay, jitter and packet loss. RIP provides poor performance, utilize more network bandwidth and poor MOS value. In contrast with RIP, EIGRP and OSPF provide acceptable MOS value and higher throughput. In OSPF the MOS value and jitter remain at acceptable level during link failure and congestion. EIGRP was showing little inefficient behaviour during the link failure. For efficient real time data delivery network reconvergence is an important factor. The results conclude that OSPF protocol is efficient to support VoIP in the enterprise network.

Diffserv model normally focus on aggregate flows, because it provides QoS guarantee only for the aggregate flows [7]. In diffserv model the traffic is classified on per class basis. Instead of ToS field DS field is used in diffserv model. QoS can improve delay, jitter and packet loss for real time application. It allows efficient use of available bandwidth for multiple applications. QoS in IP networks offer benefit in term of network capacity for new applications and services, it can solve congestion problem with the help of congestion avoidance mechanism.

Link failures are major cause of quality degradation in voice and video applications. Service availability is an important factor in voice and video services [5]. These failure does not occur regularly. To validate the performance and behaviour of routing protocols and equipment failure ever have been introduced. The result shows that the impact was due to reliability of routing equipment not due to the routing protocols. More reliable hardware and software are required as the number of nodes and links increases.

CHALENGES AND EXPECTED CONTRIBUTION

For network connectivity routing protocol is an important factor. Various comparisons have been done between routing protocols studying voice performance with less focus on video performance. The main point is to evaluate the network performance using voice and video traffic running in the same network.

The main contribution of the project is to investigate the network performance by running EIGRP and OSPF routing protocols on a network and to conclude which routing protocol provide better performance in terms of voice and video. The other point is to check the network performance improvement by applying diffserv model on the network.

PROJECT AIM

The purpose of this project is to analyze and compare EIGRP and OSPF, their effect on Network, Voice and Video performance on best effort and Differentiated Services model with link failures and congestion.

OBJECTIVES OF PROJECT

Investigate the impact of OSPF and EIGRP routing protocols on network performance by observing convergence, utilization and queuing delay with link failures.

Investigate the impact of OSPF and EIGRP routing protocols on Voice performance by observing delay, jitter, voice traffic and packet loss with link failure.

Observe the effect of OSPF and EIGRP routing protocols on Video performance by monitoring delay, jitter, video traffic and packet loss with link failure.

METHODOLOGY

OPNET is a powerful network simulator and can be used to design and study networking devices, protocols and networks varying from small to large global internetworks.

The most obvious advantage of using simulation is the time and money. The design can be built in a shorter time and can easily be tested. Building topologies and making alterations to the design in a real experiment can be expensive and difficult. It is easy to design, test and then redesign and re-test the topologies in opnet which is not only cheaper but also faster. Lots of details about the network behavior can easily be taken from the simulation. Such details are sometimes not measureable in a real experiment. The disadvantage of simulation can be the simulation errors.

OPNET modeler will be used as a simulation tool for network modeling, evaluation and comparison purposes.

Network Model and Design:

Design the network topologies that will be used in the simulations.

Design the routing protocol scheme in the network domain.

Traffic (data, voice and video)

Choose parameters for the simulation.

How the Method of the Project to be Illustrated, Validated or Evaluated

To investigate the network performance, simulation based analysis will be used by running different protocols on the network and then analysing voice and video performance. A medium sized network topology will be design to achieve all the objectives. Different scenarios will be designed depending on the protocols.

Network topology that will be used in the simulation is shown below:

Figure 1

Implementation in OPNET

In this topology we have Branch office, Head Quarter and Backbone network. The Branch office router 6 creates traffic congestion on Router 4. Similarly, on the other side Head quarter router 5 creates traffic congestion on Router 2. In this proposed project topology we mark the different voice, video and other traffic at edge routers and after marking we give priority to voice and video traffic. I have selected this topology from different research papers.

VoIP Support on Differentiated Service using Expedited Forwarding [1].

VoIP performance on Differentiated Service Enabled Network [2].

In this topology Branch1, Branch 2 and Head Brach LAN is using workstations, IP Phones, Fast Ethernet switches, routers and 10mbps links. Branch1 and Branch2 are connects with Head Branch via R1, R2 and R3 routers. The R1, R2 and R3 are connect with each other via using serial link and serial link speed is 1.544mbps.Congestion occurs at R1, R2 and R3 because LAN sends high speed data on slow serial link. The use of Quality of Service will improve the network performance and minimize the network congestion. The network diagram is implemented in Ms-Viso and OPNET.

Routing protocols that will be selected for the network topology are:

EIGRP

OSPF

Network performance comparison and evaluation will be done using simulation. Voice and video performance will be analyzed through jitter, delay, voice and video traffic and packet loss.

Three types of traffic will be generated on the network:

Project Plan

This proposed project will start in Jan 2011 and will be completed in three monthâ€™s time. In order to complete the objectives we have to follow the time management schedule.

Project schedule

Tasks No

Task Description

Duration (Days)

Start Date

Finish Date

1. First Phase

31 Days

15/01/11

14/02/11

1.1

First meeting with supervisor to discuss about project schedule.

15/01/11

15/01/1

1.2

Relevant background reading and other research literature review on Ipv4, Routing protocols, Voice, and QoS.

16/01/11

30/01/11

1.3

Studying and learning of Opnet modeller tutorials, scenarios, and network design.

31/02/11

14/02/11

2. Implementation Phase

46 Days

16/02/11

02/04/11

2.1

Meeting with supervisor to get approval for the design of project model.

16/02/11

2.2

Designing the proposed project network model in OPNET modeller tools.

17/02/11

22/02/11

2.3

Validation of network model in OPNET modeller.

23/02/11

28/02/11

2.4

Implementation of background traffic (Databases, ftp, http) on Best Effort network model in OPNET modeller.

01/02/11

06/03/11

2.5

Implementation of voice and video with above stated traffic on Best Effort network model in OPNET modeller.

07/03/11

12/03/11

2.6

Implementation of background traffic (Databases, ftp, http) on Diff Serv. Network model in OPNET modeller.

13/03/11

18/03/11

2.7

Implementation of voice and video with above stated traffic on Diff. Serv. Network model in OPNET modeller.

19/03/11

24/03/11

2.8

Collection of results and analysis.

25/03/11

30/03/11

2.9

Meeting with supervisor regarding obtained results and objectives.

31/03/11

2.10

Changes and troubleshooting in network model if needed according to discussion with supervisor about project results and conclusion.

01/03/11

02/04/11

3. Final Phase

15 Days

03/04/11

17/04/11

3.1

Meeting with supervisor to start final project documentation.

03/04/11

3.2

Start project report documentation.

04/04/11

09/04/11

3.3

Meeting with supervisor with draft project report in order to get feedback.

10/04/10

10/04/11

3.4

Prepare presentation in power point.

11/04/11

13/04/11

3.5

Preparing final submission report

14/04/11

17/04/11

Submission.

GANTT CHART

H:11.jpg

Resource Requirement

The resources required for the completion of this project are given below

University library:

To search and read books, journals and other reading material for project research.

Access to university computer Lab:

To search online journals and digital research papers (IEEE, SPRINGER, ACM and etc).

Access to OPNET modeller software:

Access to the lab computers having OPNET modeller software (12.1, 14.5 AND 15) licenced to run IP version 4 .

Access to Microsoft Office:

Access to Microsoft office tools ( Excel, Word, Power point and Ms project ) for the preparation of project report and presentation.

Risk Management Strategy

There are number of risks involved during the completion of my proposed project which are given below :

Short Time Period (High Risk)

Three month time is very short period of time for the completion of this project. Although i will strictly act upon to complete each and every task according to the proposed project schedule and Gantt chart to overcome this problem. But there is still no assurance to complete this project with 100% confidence in 3 monthâ€™s time. Therefore this risk is considered as high value risk because the time is considered as important constraint in research projects.

Overambitious Aims and objectives (High Risk)

This project objectives are very ambitious. This type of risk generally occurs when we have not enough knowledge and experience in the research area of the proposed project. To overcome this kind of risk we should have to do hard work and spend as much time as possible in reading the research papers and project literature review according to the time management schedule (Gantt chart and project schedule).

Less Knowledge in Simulation Environment (Medium Risk)

The lack of knowledge and experience in using OPNET modeler as a simulation tool is another risk involved in this proposed project. To overcome this medium level risk, given time schedule is planned to spend more time in studying tutorials, built in topologies and scenarios in OPNET modeler.

Crash of Software/Hardware (Low Risk)

The research of this proposed project will be carry out in the university labs. There is also a possibility of low level risk in case of failure of critical softwareâ€™s, hardware and operating system in labs computers, due to which we can loss project research literature materials and saved simulation results. To decrease this kind of low level risk we have to make backup copies of our work on personnel hard disk, two different computers or on other type of media.

Order Now