Category: CSE Project Reports

Research methodology 

As this a pure development Performance Comparison Of TCP Against CBR Using Reactive And Proactive Protocols project, quantitative research methodology is used across this project and two important scenarios are created and the results are compared using the graphs. In general the quantitative methodology includes the empirical investigation and analysis of the comparison results and in this project scenarios are compared in order to test the performance of TCP against CBR across MANETs using the AODV routing protocol and the actual network setup’s used are given below

 Scenario 1 

Number of nodes	200
Network Scale	Campus
Network size	100X100 meters
Model Family	MANET
Routing protocol	AODV
Application	File Transfer Protocol(FTP)
Traffic type	TCP
TCP Parameter	Tahoe

 Scenario 2 

Number of nodes	200
Network Scale	Campus
Network size	100X100 meters
Model Family	MANET
Routing protocol	AODV (AODV)
Application	Voice
Traffic type	CBR
TCP Parameter	Default

 Scenario3 

Number of nodes	200
Network Scale	Campus
Network size	100X100 meters
Model Family	MANET
Routing protocol	DSR
Application	File Transfer Protocol(FTP)
Traffic type	TCP
TCP Parameter	Tahoe

 Scenario4 

Number of nodes	200
Network Scale	Campus
Network size	100X100 meters
Model Family	MANET
Routing protocol	DSR
Application	Voice
Traffic type	CBR
TCP Parameter	Default

 Scenario5

Number of nodes	200
Network Scale	Campus
Network size	100X100 meters
Model Family	MANET
Routing protocol	DSDV
Application	File Transfer Protocol(FTP)
Traffic type	TCP
TCP Parameter	Tahoe

 Scenario6

Number of nodes	200
Network Scale	Campus
Network size	100X100 meters
Model Family	MANET
Routing protocol	DSDV
Application	Voice
Traffic type	CBR
TCP Parameter	Default

 Scenario7

Number of nodes	200
Network Scale	Campus
Network size	100X100 meters
Model Family	MANET
Routing protocol	OLSR
Application	File Transfer Protocol(FTP)
Traffic type	TCP
TCP Parameter	Tahoe

 Scenrio8 

Number of nodes	200
Network Scale	Campus
Network size	100X100 meters
Model Family	MANET
Routing protocol	OLSR
Application	Voice
Traffic type	CBR
TCP Parameter	Default

 OPNET modeler 

OPNET modeler is used as the simulation tool in this project, as the scope of coding is less in this tool and generating the scenarios with respect to the proactive and reactive routing protocols is very easy in OPNET when compared to the simulations tools like NS2. OPNET can be installed on windows operating system, where as the other simulation tools kike NS2 should be installed on Linux. 

Basic simulation model is as shown in the below diagram 

Project Plan 

Task	Start date	End date	Duration
Proposal			One week
Literature review report			11 days
Design of the application			Two weeks
Research Analysis			8 days
Simulation and testing			Three weeks
Dissertation report			Three weeks

 Conclusion                  

The main aim of this project is to evaluate the performance of TCP against CBR across MANETs using two proactive routing protocols like DSDV and OLSR and reactive routing protocols like DSR and AODV. A total of 8 scenarios are created in this project, where four scenarios are created to generate the TCP traffic using a simple file transfer application and four scenarios are used to generate the CBR traffic using a voice application. All the scenarios are compared against the simulation results using the OPNET modeler and the best performing traffic against the best routing protocol in comparison to the AODV routing protocol is done in this project.

This paper is written and submitted by Sujana Priya V

Behavior of TCP over CBR 

In general User Datagram protocol (UDP) is the widely accepted protocol to handle the high Constant Bit Rate transmission that was required to transmit data like video or audio. But there are few limitations in the usage of UDP to support the CBR due to the firewalls and thus a limited connectivity is achieved. There has been a wide research done on the alternative protocols to handle the CBR and Skype the famous VOIP application could mark its success in using the TCP across CBR.

These days TCP has improved a lot in terms of loss recovery mechanisms like SACK and thus transport based on TCP can be used to handle the real-time CBR workloads. Due to the increase of mobile and multimedia applications, always a new set of Quality of Service Standards are set and most of the  existing wireless routing protocols are supporting these QoS services and they are proved to failure in case of mobile environments like MANETS. To meet the basic requirements of mobile multimedia requirements a separate set of QoS standards are always required and the constant bit rate of transmission stands on first among the requirements.

Energy consumption across the MANETs to support the multimedia services would be high and in these conditions TCP is proved to be a victim of error correlation. There are many attempts proposed to improve the energy efficiency of TCP over achieving CBR across the MANETs and most of them concentrate on the error control mechanisms .

Block diagram of MANET is as shown below 

 

AODV can be considered as the hop-by-hop routing protocol as they maintain the route information for some predetermined lifetime and it would be around 3 seconds across AODV routing protocol. In general a route entry is maintained for a constant time period even the usage of that particular entry is out of scope at a given point of time. It is always advantageous to maintain a short life time for a route entry across the dynamic routing protocols like AODV for MANETs.

In most of the cases the life time is determined by period of Constant Bit Rate (CBR). If the period of CBR is sufficiently long, then there is no proper requirement to maintain the route and node information and it is simply wastage of computation and as a result the overall overhead of the network will be increased. It is always advisable to update the routing table in prior to the generation of CBR data. In static environments, the life time of each node is set to the CBR period and this might not work for the dynamic environments like MANETS, where the mobility of nodes decides the actual performance of the network.

In general the CBR period of each and every node should be set dynamically across the AODV routing protocol. Whenever the AODV protocol sends a RREQ message to the other nodes a separate CBR period field is added to the message and broadcasted across the network. Once the message reaches the desired destination, the destination nodes fetches the CBR period field from the RREQ message and adds a new CBR period filed to the RREP message and sends back to the actual source.

Thus each node created a new life time based on the CBR period of the message and based on this life time the node or route entry is expired and thus the overhead to the system is reduced a lot. Control overhead messages can be reduced a lot with the implementation of dynamic CBR period field in the RREQ and RREP messages that were broadcasted by the AODV routing protocol across MANETs .

This paper is written and submitted by Sujana Priya V

There are many factors that were responsible for the performance degradation of TCP over MANETS and few of them are as listed below 

Bit Error Rates: High Bit Error rates are always responsible for the performance evaluation of TCP across MANETs. In general wireless routes are heavily porn to high bit error rates due to some internal reasons like signal degradation and multipath fading and thus finally leads to packet losses and TCP data segment losses. Now the TCP will unnecessarily invoke the congestion control mechanism and thus the overall system performance is affected a lot. 

Route Failures: There could many chances of route failures across MANETS due to its mobility and thus it simply affects the performance of TCP. In general the route reestablishment time taken by the MANETs depends on the routing protocol used across them and in some cases it would take more time and now the TCP would in a congestion state and initiates the congestion control mechanism, as a result the overall system performance is affected. 

Multipath routing: Multipath routing is the basic characteristic of MANETs and also one of the important factors responsible for performance degradation of TCP over mobile Adhoc networks. In general multiple routes are maintained across the sender and receiver and thus results in out-of-sequence packets at the receiver. Due to this duplicate ACKs are generated across the network and this scenarios is similar to a congestion situation across TCP and thus again the TCP would invoke the congestion control mechanism to reduce the duplicates.

This paper is written and submitted by Sujana Priya V

Performance of TCP over MANETS 

Transmission Control protocol (TCP) is the widely used transmission protocol across the data services like file transfer and internet and thus its usage is spread towards MANETs as well. In general there could be multiple TCP connections over mobile adhoc networks and thus evaluating the performance of TCP is really a tedious job in case of multiple connections. In general performance of MANET depends on few attributes like scalability, traffic load and mobility. Extensive research has been done towards the performance of MANETs across the Constant Bit Rate (CBR) against file transfer protocol (FTP), TCP and User datagram protocol (UDP).

Scalability of the routing protocols across MANETs purely depends on the attributes as mentioned above. In most of the case TCP in the current form can’t support the MANETs in case of route failures and in terms of congestion control methodologies. There are many existing techniques which discuss the performance of TCP over MANETs, but still it is proved to be hard nut to crack and there is lot of research being done towards the performance of TCP across the mobile adhoc networks.

Performance of TCP can be improved in case of route failures by a feedback mechanism known as TCP-Feedback or TCP-F, where the intermediate node detects the link or route failures and sends a notification message known as Route Failure Notification (RFN) to the actual TCP source. Once the TCP senders receive this RFN notification, it quits sending the required packets and blocks the complete communication process. Now if the intermediate node is able to discover the path to destination, it immediately sends the Route Re-establishment notification (RRN) to the TCP senders and thus now the source will reactivate the state and continue with the data transmission process. Thus from this discussion it is clear that, the actual TCP throughput is completely related to the average repair time taken towards the link or route failures across the MANETs.

One more notification methodology was introduced later known as Explicit Link Failure Notification (ELFN) to improve the performance of TCP over mobile adhoc networks and is as discussed. Whenever an intermediate node intimates the TCP sender regarding the link failure, TCP sender will freeze its state along with the transmission window and the timer set. In this case the use of route re-establishment is not done; instead the TCP sender will constantly sends the packets and checks for the availability of route across the MANETs. Across many simulations using NS2 and OPNET, ELFN is proved to be the best notification procedure than RFN and can be considered as the better choice to improve the performance of TCP over MANETS. 

In general TCP provides end-to-end data delivery across the wired networks and proved to be an efficient means of data transmission but it abruptly fails in case of wireless networks like MANETs and this is mainly due to the reason that TCP can’t work in few conditions like packet loss or delay, where in a general MANET packet delivery delay and packet losses are simply encouraged.

This paper is written and submitted by Sujana Priya V.

Mobile Adhoc networks are widely used across the networking world these days and there are many routing protocols in place to manage the routing across these MANETS. The Adhoc on demand distance vector routing protocol (AODV) is specially designed for the purpose of supporting both the Unicast and multicast routing across MANETs .

The basic nature of this algorithm runs on the principle like on demand, where the actual routing and the corresponding routes are built as per the decision of the source node dynamically. Source passes the instructions to add or modify the predefined routes based on the demand of networking parameters like bandwidth and congestion. A typical tree based structure is maintained by the AODV routing protocols, where this tree is composed of group members of multicast routing and the nodes connects these members. AODV is a reactive protocol even the basic characteristics of the protocol are similar to proactive protocols .

Technical implementation of AODV is done based on the Dynamic Source Routing Protocol (DSR), where it inherits the route discovery from the DSR and Destination Sequence Distance Vector (DSDV), where it inherits the concept of maintaining sequence of the hello messages being transmitted. Time based states for each node is the important and useful feature available with the AODV routing protocol, where a node which is not used recently is automatically expired. AODV updates the neighboring nodes if there is any change across the route and even in case of route failures. AODV always use a most recent route to reach the destination nodes .

Table of Contents 

1.INTRODUCTION AND PROBLEM DEFINITION

1.1 INTRODUCTION

     1.2PROBLEM DEFINITION

2.AIMS AND OBJECTIVES

2.1 AIM

     2.2 OBJECTIVES

3.LITERATURE REVIEW 

    3.1 ROUTE DISCOVERY IN AODV

    3.2 ABOUT DSR

    3.3 ABOUT DSDV

    3.4 PERFORMANCE OF TCP OVER MANETs

    3.5 FACTORS RESPONSIBLE FOR LOW PERFORMANCE OF TCP OVER MANETs

    3.6 BEHAVOIUR OF TCP OVER CBR

4.RESOURCE REQUIREMENTS

   4.1 SOFTWARE REQUIREMENTS

     4.2 HARDWARE REQUIREMENTs

5. RESEARCH  METHODOLOGY

6.DEVELOPMENT SCHEDULE

7.CONCLUSION

8.ETHICAL AND LEGAL ISSUES

8.REFERENCES

This paper is written and submitted by Sujana Priya V

Problem definition:      

                       Comparison of the performances of various traffic sources across the Mobile Adhoc Networks using the respective routing protocol is not clearly explained in the existing literature which is already discussed in the project back ground section. In this project, the performance of TCP traffic is compared against the CBR traffic across the MANETs designed using 200 nodes across a capacity area of 100X100 meters by using the Adhoc Distance vector routing protocol is done using two scenarios.

                          The proactive routing protocol like DSDV and the reactive routing protocols like DSR and AODV are considered across the comparison process. Performance of TCP against CBR is compared against these three routing protocols and the performance of AODV is evaluated with respect to the other routing protocols. In this project 6 scenarios are used, where three scenarios are used to generate the TCP traffic across the proactive routing protocols and the other scenarios are used to generate the CBR traffic across the reactive routing protocols. Performance of TCP and CBR is compared against all the routing protocols and the overall assessments like the best source of traffic and best routing protocol to handle this traffic are evaluated. TCP traffic is generated using a simple file transfer application and CBR traffic is generated using a voice application. 

Aims and Objectives:

Aim: To simulate the performance of TCP against the CBR using the parameters 10²*10² and 500*500 areas across 200 nodes against the proactive routing protocol like DSDV and the reactive routing protocols like DSR and AODV. 

Objectives:

Following are the project objectives 

• To understand the importance of TCP across the real time  CBR   transmission across the real time applications
• To compare the performance of TCP against CBR across MANETs using proactive routing protocol like DSDV and reactive routing protocols like DSR and AODV and evaluate the respective performance metrics like throughput, delay and packet delivery ratio.
• In this project, the research considered is the selection of traffic type across the MANET and the impact of this traffic on the performance of the network.
• To simulate the proactive routing protocols like DSDV and reactive routing protocols like DSR and AODV using OPNET.
• To check the performance of TCP against CBR using an area of 10²*10² and 500*500 across 200 nodes using OPNET
• To document the simulation results and observations

This paper is written and submitted by Sujana Priya V

Mobile communication is the main purpose of using the Mobile Adhoc Networks and these networks are called as MANETs which don’t have a particular infrastructure setup and can also be operated in any conditions and so these are called as infrastructure less networks.

                  Across the MANETs the routing process is depended on the routing protocols that are developed in a large extent and the performance of the routing using these routing protocols is based on the routing mechanism developed and internal technology used in particular routing protocol.

                  In general MANET routing protocols are classified into three types and they are the proactive routing protocols which are also called as table driven routing protocols, reactive routing protocols which are also called as on-demand routing protocols and the third one id hybrid routing protocol. This classification routing protocols is based on the extent of support for the infrastructure of the routing and mode of operation that are applied throughout the network. Generally, a MANET is formed by connecting several nodes using some links with an aim of sending the information packet using a route between the source and destination. The routing protocol used in the network will decide the routing between the nodes.

                    An arbitrary graph is used generally to represent the MANET and sometimes it is also represented using the dynamic graphs and the nodes are used to represent the mobile terminals, and the links that connect these nodes will represent all the routing paths that are available between the nodes. Generally in MANETs multi-hop routing is utilized because of the reason that the range of radio frequency is high, radio frequency is one of the wireless communication formats and its range is more when compared to the other formats and so high end implementation is needed for enabling the routing among the routing protocols.

                      In MANETs, several factors influences the performance of the routing protocols that are selected to use across the MANETs, and these factors include security level employed across the network, maintenance of the route, configuration of router, various types of applications supported by MANETs and different kinds of traffic that are sent throughout the network.

                      MANETs supports different types of traffics and the most important and frequently used traffics are TCP, VBR and CBR traffics here VBR means Variable bit rate and CBR means Constant bit rate. The traffic type selected across the routing procedure will influence the routing protocol performance. The performance of the routing protocol is also based on the nodes selected in the MANETs generally two types of nodes can be used in MANETs and they are mobile nodes and fixed nodes.

                          MANETs are basically dynamic in nature and so it supports a large variety of applications and the most important and most commonly used applications of MANETs are FTP, video conferencing, VOIP, Email, voice and web applications. The characteristic of the traffic sent across the MANET is decided by the selected type of application. The application selected is also used to influence the performance of the routing protocol similarly the selected traffic type also influence the performance of routing protocol that may be reactive or proactive that is used throughout the MANET. The issues related to these MANETs are discussed in many existing studies and researches which also includes the comparison of performance of routing protocols in various aspects which are done mostly among the selected routing protocols when compared to the selected kind if traffic. 

                         The comparison of performances between various traffics should be focused more in the research studies when compared to the focus on the routing protocol across the MANETs, and this is very much necessary. Across every Mobile Adhoc Network, the selected level of mobile transmission and selected type of application decides the type of traffic used in the application and the original source of traffic in the application. In the studies of many research scholars the TCP traffic is compared against the other source of traffics is selected as the research area. Generally in MANETs implementation of comparison of traffics sources is not so perfect when compared to the other networks which are sophisticated, for example the networks like WLAN and ATM. It is already discussed the MANETs supports a wide range of applications, and various metrics will decide the performance of the application used. These metrics are also called as parameters which depend on the selected routing protocol and selected number of mobile nodes used across the network.

                       The performance of any application is decided in terms of the metrics like throughput, delay, packet delivery ratio maintained across the networks. Many studies have included the knowledge about the performance of application attained by the routing protocol used in the network but are not cover the generation of traffic using these applications. In this project, the research area considered is the selection of traffic type across the MANET and the impact of this traffic on the performance of the network, particular in this project, Adhoc Distance Vector routing protocol is used to evaluate the performance of TCP and CBR traffic and also for comparing these traffics across the MANETs.

This paper is written and submitted by Sujana Priya V

Abstract:

Presented an optimal routing and data aggregation scheme for maximizing the network Lifetime of sensor networks. By exploiting the special structure of the sensor networks and Proposed some functions to overcome the non differentiability of optimization problem so that The distributed solution is possible. The optimality conditions are derived and the distributed Algorithm is designed accordingly. This scheme significantly reduces the data traffic and      improves the network lifetime. The distributed algorithm can converge to the optimal value efficiently.

Outcomes:

Presented an optimal routing and data aggregation scheme for maximizing the network lifetime of sensor networks.

This paper is written and submitted by sandeep k