In the past there is a problem for the researchers to select the better network simulator for their research. Our project is based on the comparison of different network simulators to evaluate the performance of each simulator and we have to conclude that which one is the better network simulator. For comparison of different network simulators we use MANET protocol (AODV) and simulate this protocol on each simulator and evaluate the performance on different parameters. We take different parameters for comparison of different network simulators compares them on the basis of memory utilization, CPU utilization, computational time and scalability.
As there are number of network simulators like NS-2, OMNeT++, Opnet,J-Sim, GloMoSim etc selecting a simulator from the number of network simulators is a very crucial task. The focus of our thesis is to compare the different network simulators to evaluate the performance of network simulators.
This research based project is helpful for the researchers to select the better network simulator for their research. This is a guideline for the researchers to select the efficient network simulator for the research purpose.
Get your grade
or your money back
using our Essay Writing Service!
This project involves the detail study of MANET protocols AODV and DSDV. This report includes there simulation results by comparing different network simulators on different parameters by using these protocols and there graphs are studied in detail.It covers all aspect of AODV and DSDV routing protocols. The report presents the detailed study, comparison and analysis different network simulators. This research has gone through a series of sequential steps to its final form, which is explained by the chapters included in this report.
Provides Introduction to the project and also the scope of the project.
Briefly describes Mobile Ad hoc Networks and types of mobile adhoc networks.
Explains Adhoc On Demand Distance Vector (AODV) routing protocol
Explains network simulators in detail
Explains our work that we had done on different network simulators and also explains the problems that we suffered during working on different network simulators
Shows the detail of performance parameters
Shows Results in detail and conclusion is also added in this chapter.
TABLE OF CONTENTS
Chapter 1 Introductionâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦1
1.2 Scope of the projectâ€¦..................................................................1
1.3 Statement of Problemâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦1
Chapter 2 Mobile Adhoc Networks............................................3
2.2 Categories of MANET protocolsâ€¦â€¦â€¦â€¦â€¦â€¦â€¦.â€¦..â€¦â€¦4
2.2.1 Reactive routing protocolsâ€¦....................................4
2.2.2 Proactive routing protocolsâ€¦â€¦â€¦â€¦â€¦.â€¦â€¦â€¦.â€¦.5
2.2.3 Hybrid routing protocolsâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦.â€¦......6
2.3 Types of MANETSâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦.....6
Chapter 3 Adhoc On Demand Distance Vector Routing protocol â€¦â€¦â€¦â€¦â€¦â€¦....7
3.2) Advantages and disadvantagesâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦.
Chapter 4 Network Simulatorsâ€¦â€¦â€¦...12
Chapter 5 Our work on network Simulatorsâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦..
5.1) Simulators and workingâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦.
Chapter 6 Performance parametersâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦
6.1) CPU Utilizationâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦
6.2) Memory Usageâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦.
6.3) Computational Timeâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦
Chapter 7 Results and conclusionâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦.
7.1) CPU Utilization
7.2) Memory Utilization
7.3) Computational Time
Comparison of different network simulators by using some well known routing protocols for performance evaluation in order to draw a conclusion that which simulator is more suitable for simulation.
In this project we will implement some of the well known protocols by using different well known network simulators and then evaluate the performance of each simulator. The selected network simulators will be extensively studied and a comparative study of these simulators will also be given. We will evaluate the performance of different network simulators by simulating well known routing protocols on the basis of different parameters.
1.3) STATEMENT OF PROBLEM
There are a lot of network simulators available like NS-2, NS-3, OMNET++, Opnet, and JSim etc. Selecting a simulator for performance evaluation is very crucial task. With a variety of network simulators readily available it is important to select the best simulator that best fit for your simulation needs. The objective of this research is to simulate some of the well known protocols on different network simulators and then on the basis of observations and results, do a comparative analysis of these simulators.
MOBILE ADHOC NETWORK (MANET)
Always on Time
Marked to Standard
Mobile Adhoc Network (MANET) is an adhoc (temporary), rapidly deployable and self configuring network of mobile devices connected by wireless links. As nodes in MANET are wireless therefore one can use mobiles, laptops, PDAs and PC etc.MANET operates without the support of any infrastructure therefore MANET is also called as infrastructureless networks. In MANET every device can move independently and can change its links to the other devices frequently and every node act as a communicating device can transmit and receive data at the same time and also act as a wireless router.MANET can be standalone network or it can be connected to external networks (internet).MANET has broad applications like military applications, rescue operations, disaster management etc. Due to the broad applications of MANET a lot of research work has been done on MANET and its protocols. There are two types of MANET protocols reactive and proactive depends on the user requirements and network needs.
2.1) CATEGORIES OF MANET PROTOCOLS:
MANET protocols are categories in three categories a proactive, reactive and hybrid routing protocols.
2.1.1) PROACTIVE (TABLE DRIVEN) ROUTING PROTOCOL:
Proactive routing protocols are also called as table driven routing protocols. These protocols maintain and up-to date the routing information on each node connected to the network. In proactive routing protocols all nodes in the network are needed to maintain the consistent network topology view. If the change has been occurring in the network topology than the respective updates must be propagated throughout the network and notify the change in the network. In proactive the routing table is updated periodically and the periodic updates start with the hope count of metric 1.In the proactive routing protocols packets are flooded all over the networks to all nodes even to those nodes that are not transmitting the packets. The advantage of proactive routing protocol is that the connection times are fast as the routing information is already available when the first packet is sent. The disadvantage of this routing protocol is that it uses resources all the time to communicate information even then when there is no traffic on the network.
2.1.2) REACTIVE ROUTING PROTOCOLS:
On-demand routing protocols were designed to reduce the overheads in proactive protocols by maintaining information for active routes only. This means that routes are determined and maintained for nodes that require sending data to a particular destination. Route discovery usually occurs by flooding a route request packets through the network. When a node with a route to the destination (or the destination itself) is reached a
route reply is sent back to the source node using link reversal if the route request has traveled through bidirectional links or by piggy-backing the route in a route reply packet via flooding. Reactive protocols can be classified into two categories: source routing and
Hop-by-hop routing. In source routed on-demand protocols, each data packets carry the complete source to destination address. Therefore, each intermediate node forwards these packets according to the information kept in the header of each packet. This means that the intermediate nodes do not need to maintain up-to-date routing information for each active route in order to forward the packet towards the destination. Furthermore, nodes do not need to maintain neighbor connectivity through periodic beaconing messages. The major drawback with source routing protocols is that in large networks they do not perform well.
2.1.3) HYBRID ROUTING PROTOCOLS:
Hybrid routing protocols are another type of MANET protocols that has the features of both proactive routing protocols and reactive routing protocols. These protocols are both on demand as well as periodically update their routing tables. Examples of such type of protocols are TORA, OORP (OrderOne Routing Protocol), SSR (Scalable Source Routing), and ZRP (Zone Routing Protocol) etc.
2.2 TYPES OF MANET
Mainly there are three types of MANETS
1) VANETS (Vehicular Adhoc networks)
2) IVANETS (Intelligent Vehicular Adhoc Networks)
3) INVANETS (Internet Based Vehicular Adhoc Networks)
ADHOC ON DEMAND DISTANCE VECTOR (AODV)
Adhoc on demand distance vector (AODV) routing algorithm is a routing protocol designed for mobile adhoc network (MANET).It has the capability of both unicast and multicast routing.AODV is a reactive routing protocol meaning that it establishes the route on demand, at the time when the rite is desired by the source node. It maintains the route as long as the sources needed that routes.AODV avoids the count to infinity problem by using the sequence number on route updates. An important feature of AODV is the maintenance of time-based states in each node: a routing entry not recently used is expired. AODV is relatively fast to the topological changes in the network and updates only the nodes aï¬€ected by these changes.
This Essay is
a Student's Work
This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.Examples of our work
In AODV the network remain silent unless the connection is needed by any node in the network. At that time the node needs the connection broadcasts the request for the connection. Other nodes in the network forward this request and record the node from where they heard this request, creating the temporary route to the needy node. When a node receives such a message and already has a route to the desired node, it sends a message backwards through a temporary route to the requesting node. The needy node then begins using the route that has the least number of hops through other nodes. The entries that are not used in theÂ routing tablesÂ are recycled after a time. If a link fails, the routing error is send back to the transmitting node and the process repeats.
The advantage of AODV is that it generates no extra traffic for communication along existing links. Also, distance vector routing is simple, and doesn't require much memory or calculation. However AODV requires more time to establish a connection, and the initial communication to establish a route is heavier than some other approaches.
3.2) ADVANTAGES AND DISADVANTAGES
The main advantage of this protocol is that routes are established on demand and destination sequence numbers are used to find the latest route to the destination. The connection setup delay is lower. One of the disadvantages of this protocol is that intermediate nodes can lead to inconsistent routes if the source sequence number is very old and the intermediate nodes have a higher but not the latest destination sequence number, thereby having stale entries. Also multiple RouteReply packets in response to a single RouteRequest packet can lead to heavy control overhead. Another disadvantage of AODV is that the periodic beaconing leads to unnecessary bandwidth consumption.
In this chapter we discuss about network simulators as we know that there are a number of network simulators currently available, selecting the best simulator is a very crucial task. After studying different network simulators we have choose four network simulators for our project. The selected simulators are:
The network simulator (NS) it is a discrete event simulator. Its an open source network simulator. Â NS is used for the simulation of local area and wide area networks and also for the simulation of different routingÂ protocols, and is heavily used inÂ ad-hoc networkingÂ research. It is used for the simulation wired as well as wireless networks. NS was built inÂ C++Â and provides a simulation interface throughÂ OTcl (Object Tool C command Language), anÂ object-orientedÂ dialect ofÂ Tcl. The user describes a network topology by writing OTcl scripts, and then the main NS program simulates that topology with specified parameters. In NS2 simulation can be displayed in graphical format called NAM (Network Animator). NAM interface provide the ability that one can (forward, pause, stop and play) the speed.
Network simulator-3 it is an open source simulator, NS-3 is a discrete-event network simulator, helpful for research and educational uses it is a free software and still under development. We say that it is an replacement of ns-2 but not an extension so NS-3, like NS-2 it does not have an OTcl API. Is is written is C++ language and python. Latest version of NS-3 is NS-3.10.it improves scalability of the simulator and also improve performance in the sense of memory and CPU utilization, it also support parallel simulation as its developers says. NS-3 actually improve performance, his feature set of the simulator is also about being extended.
Likens-2 and NS-3 OMNeT++ is not a network simulator by its definition but It is a general discrete event and component-based modular and open architecture Simulator framework. The most common use of OMNeT++ is the simulation of computer networks, but it is also used forÂ queuing networkÂ simulations, and other areas as well. Actually OMNeT ++ shows a framework approach also containing hardwired support for networks or we say other areas. It provides us a structure for writing such a simulation. Provides application of various simulation models frameworks most of them is open source.
A network simulation in OMNeT++ is implementing it as a compound module which comprehends other compound modules, like the ones which model host nodes.
Global Mobile Information System Simulator (GloMoSim) is a scalable simulation environment for large wireless and wireless communication networks. GloMoSim uses parallel discrete-event simulation capability provided by Parsec. GloMoSim simulates network with up to thousand nodes linked by a heterogeneous communications capability that includes multicast, asymmetric communications using direct satellite broadcasts, multi-hop wireless communications using ad-hoc networking, and traditional Internet protocols. The following table lists the GloMoSim models currently available at each of the major layers. The node aggregation technique is introduced into GloMoSim to give signs can't tenets to the simulation performance. Initializing each node as a separate entity inherently limits the scalability because the memory requirements increase dramatically for a model with large numbers of nodes. With nodes aggregation, a single entity can simulate several network nodes in the system. Node aggregation technique implies that the number of nodes in the system can be increased while maintaining the same number of entities in the simulation. In GloMoSim, each entity represents a geographical area of the simulation. Hence the network nodes which a particular entity represents are determined by the physical position of the nodes. GLOMOSIM has a Visualization Tool that is platform independent because it is coded in Java.
OUR WORK ON NETWORK SIMULATORS
5.1) SIMULATORS AND WORKING
There are many network simulators currently working but we select four simulators:
Our project is comparison of network simulators, for comparison of network simulators we simulate AODV on each network simulator. The main reason to choose AODV is the availability of AODV in each simulator. For measuring CPU utilization, memory usage and the computation time we use the Linux command as given below:
Top | grep process name
e.g top | grep omnetpp
With use of this command we can calculate the CPU utilization, memory usage and computation time of the specific process.
As we haven't work in any of the given network simulators that is why we face a lot of problems but due to the guideness of our teachers we learned a lot about these simulators and about their working. While working on NS-2 we face some of the problems in the initial state during the installation phase, but after exploring about NS-2 we have learn a lot and learn how it works. The protocol that we have simulated i.e. AODV is built-in in NS-2 but we have created connections between the nodes deployed nodes randomly.
Some of the peoples think that NS-3 is the new version of NS-2 but it is not correct, it is not new version of NS-2 but it is itself another network simulator. While working in NS-3 we have faced a lot of problems because it is a new simulator and not used as much as NS-2 and also there is not as much documentation about it. As it has different versions like ns-3.7, ns-3.9.ns-3.9Dev and ns-3.10, first of all we have installed ns-3.7 and ns-3.9 but AODV is not available in these versions after that we have add the patch file of AODV in ns-3.9 but AODV was still not working. After exploring about NS-3 on the internet we have found that there is another version i.e. ns-3.9Dev in which AODV is built-in. we have installed that version and found AODV in the working condition.
5.1.3) OMNET ++
Another network simulator OMNET++ is totally different form NS-2 and NS-3. While using OMNET++ we have found that there is no MANET protocol, for MANET protocols we have to add INETMANET framework in which all the MANET protocols are available, so after installation of OMNET++ we had add the INETMANET framework in OMNET++.When we had ran the AODV on OMNET++ we found the it take more time in simulation as compared to NS-2 and NS-3.
While working in GloMoSiM we had faced some of the problems in the initial state as in the other network simulators but during the passive of time we had sort out that problems. The protocols that we had simulated i.e. AODV is also built-in in GloMoSiM.
We are going to compare different network simulators on the basis of following Parameters
5.1) CPU UTILIZATION
The CPU Utilization is measured by varying the number of nodes
Number of nodes=200, 400, 600, 800, 1000
Network Area=1000 x 1000
5.2) MEMORY UTILIZATION
The memory utilization is also measured by varying the number of nodes
Number of nodes=200, 400, 600, 800, 1000
Network Area=1000 x 1000
5.3) COMPUTATIONAL TIME
It is the measure of the time taking by each network simulator to simulate specific number of nodes. The computational time varies by increasing number of nodes.
Number of nodes=200, 400, 600, 800, 1000
Network Area=1000 x 1000
This is measured by increasing the number of nodes and in the response what is the effect on the network and CPU utilization.
Number of nodes=200, 400, 600, 800, 1000
Network Area=1000 x 1000
RESULTS & CONCLUSION
7.1) CPU UTILIZATION
For the evaluation of this parameter we had ran the AODV for 500sec and keeping the area 1000, 1000
Figure 1 No of nodes vs CPU utilization
Figure 1 shows the effect on the CPU by increasing the number of nodes.The above figure shows that as the number of nodes are increased then the CPU utilization also increased.First of all we had ran the AODV on NS-2 for 200 nodes and record the CPU utilization at 200 nodes similarly we had followed these steps 5 times by varying the number of nodes.We observed that the CPU utilization in NS-2 is between 94% to 99% and if we open another process during the simulation then NS-2 shares its utilized CPU with that process and half of its CPU is allocated to that process which is recently open.
7.1.2 ) NS-3
For NS-3 the simulation parameters remain same as in NS-2 like simulation time is set as 500sec and keeping the X,Y dimensions 1000,1000.
Figure 2 No of nodes vs CPU utilization
Figure 2 shows the CPU utilization of NS-3 in against the number of nodes.As the number of nodes increased the CPU utilization also increased.The values are taken by increasing the number of nodes from 200 to 1000 nodes.In NS-3 the CPU utilization is observed between 89% to 98%.And also there is the same situation as in NS-2 about the running of another process that if we run another process during the simulation then NS-3 like NS-2 shares its half of the utilized CPU with other process.
For OMNET++ the simulation parameters also remain same as in NS-2 and NS-3 like simulation time is set as 500sec and keeping the X,Y dimensions 1000,1000.
Figure 3 No of nodes vs CPU utilization
The above figure shows the CPU utilization in against of number of nodes. In OMNET++ the CPU utilization is not too much as compared to NS-2 and NS-3 i.e these both simulators has 90 plus utilization of CPU. The CPU utilization increases as increasing the number of nodes.In OMNET++ the CPU utilization is between 1% to 3% from 200 nodes to 1000 nodes.
For GloMoSiM the simulation parameters also remain same as in NS-2, NS-3 and OMNET++ like simulation time is set as 500sec and keeping the X,Y dimensions 1000,1000.
Figure 4 No of nodes vs CPU utilization
The above figure shows the CPU utilization in against the number of nodes.In GloMoSiM the CPU utilization is between 6% to 7%,more than OMNET++ but much less than NS-2 and NS-3.The CPU utilization increases as the number of nodes increases. We had noted the values on 200 nodes to 100 nodes to check the CPU utilization of GloMoSiM.
7.2) MEMORY USAGE
The below given figure shows that how much memory is used by NS-2 against varying the number of nodes.
Figure 5 No of nodes vs Memory usage
Figure 5 show the memory usage of NS-2 in against of the number of nodes. As the number of nodes increases the memory also increases. Initially we has noted the values for 200 nodes by simulating AODV for 500sec and keeping X, Y dimension 1000, 1000 and similarly followed these steps 5 times up to 1000 nodes. The memory used by NS-2 from node 200 to 1000 is between 4MB to 50 MB.
Figure 6 No of nodes vs Memory usage
Figure 6 shows the memory utilization of the NS-3 by varying the number of nodes.The memory usage of NS-3 is less than NS-2.The memory usage of NS-3 is between 1MB to 3MB from 200 nodes to 1000 nodes.As the number of nodes are increased the memory usage also increased.
Figure 7 No of nodes vs Memory usage
The above figure shows the memory usage of the OMNET++ by varying number of nodes. After simulating the AODV for 500sec for 200 nodes to 1000 nodes by keeping the area 1000, 1000 we get these points. This figure shows that as increasing the number of nodes the memory also increases. It uses more memory as compared to NS-3 but less as compared to NS-2.
Figure 8 No of nodes vs Memory usage
The above figure shows the memory usage of GloMoSiM by varying number of nodes. By increasing the number of nodes the memory also increases.GloMoSiM use less memory as compared to the other three simulators. For this parameter we had simulate AODV for 500sec and keeping the area 1000,1000 from 200 nodes to 1000 nodes.
7.3) COMPUTATIONAL TIME
The below figure shows the computation time in against of the number of nodes.The computational time is the time taken by the network simulator for the simulation of specific number of nodes.The graph shows the straight line because as the number of nodes increases obviously it takes more time to compute them.
Figure 9 No of nodes vs Computational Time
Figure 10 No of nodes vs Computational Time
The above figure shows the computational time of NS-3 by varying number of nodes.The simulation parameters are same as in all other simulators, keeping area 1000,1000 and simulation time 500 sec.The computational time is little bit more as compared to NS-2.
Figure 11 No of nodes vs Computational Time
Figure 11 show the computation time of the OMNET++ in against the number of nodes. It take much less time to compute the specific number of nodes as compared to NS-2 and NS-3.For the computation time we had ran AODV for 500sec from 200 nodes to 1000 nodes by keeping the X,Y dimensions 1000,1000.
The below figure shows the computational time of GloMoSiM by varying the number of nodes.The computational time of GloMoSiM is little bit more than OMNET++ but much less as compared to NS-2 and NS-3.For the computation time of GloMoSiM we use the same simulation parameters as in all other network simulators.
Figure 12 No of nodes vs Computational Time