TABLE OF CONTENTS
Title page – – – – – – – – – – – – i
Approval page – – – – – – – – — – – ii
Certification- – – – – – – – — – – – iii
Declaration- – – – – – – – – – – – iv
Dedication – – – – – – – – – – – – v
Acknowledgement – – – – – – – – – – – vi
Abstract – – – – – – – – – – – – vii
Glossary- – – – – – – – – – – – viii
List of Figures- – – – – – – – – – – x
List of Tables – – – – – – — – – – – xi
Table of contents- – – – – – – – – – – xii
CHAPTER ONE: INTRODUCTION
1.0 Introduction- – – – – – – – – – – 1
1.1 Description of MANET Transport Protocol- – – – – – – 1
1.1.1 The Transmission Control Protocol (TCP)– – – – – – 1
1.1.2 The User Datagram Protocol – – – – – – – – 2
1.2 Description of MANET Routing Protocols- – – – – – – 3
1.2.1 Proactive (table driven) Routing Protocols- – – – – – 3
1.2.2 Reactive (On Demand) Routing Protocol – – – – – – 4
1.3 Problem Statement- – – – – – – – – – 5
1.4 Objective of the study- – – – – – – – – – 6
1.5 Significance- – – – – – – – – – – 6
1.6 Scope of the Study- – – – – – – – – – 6
1.7 Methodology- – – – – – – – – – – 6
1.8 The Structure of the Work- – – – – – – – 7
viii
CHAPTER TWO: LITERATURE REVIEW
2.0 Introduction- – – – – – – — – – – 8
2.1 Relationship between Protocol Performance and Mobility Model. – – – 8
2.2 Network Dynamics of Manets- – – – – – – – – 10
2.3 Node Density and Mobility Rate- – – – – – – – – 11
2.4 Protocol Performance Analysis in Manhattan Grid Mobility Model- — – – 13
CHAPTER THREE: METHODOLOGY
3.0 Introduction- – – – – – – – – – – 14
3.1 The Matthew Mathis Model- – – – – – – – – 14
3.2 Experimental Validation- – – – – – – – – – 17
3.2.1 Experimental Requirements- – – – – – – – – 17
3.2.2 Network Performance Tools- – – – – – – – – 18
3.3 TCP Connections – – – – – – – – – – 18
3.3.1 Configuring the System Using Iperf- – – – – – – – 19
3.3.2 TCP Test Section A- – – – – – – – – – 20
3.3.3 TCP Test Section B- – – – – – – – – – 20
3.4 Conducting a UDP Speed Test in Iperf – – – – – – – 20
3.4.1UDP Test Section A- – – – – – – – – – 21
3.4.2 UDP Test Section B- – – – – – – – – – 22
CHAPTER FOUR: RESULTS, ANALYSIS AND DISCUSSION
4.0 Introduction- – – – – – – – – – – 23
4.1 Presentation of Results- – – – – – – – – – 23
4.1.1 The Mathis Model- – – – – – – – – – 23
4.2 Result Analysis and validation – – – – – – – – – 24
4.2.1 Simulation of the Modified Mathis Model for Throughput and Delay- – – 30
4.2.2 The Throughput (Mb) as a Function of Speed (m/s) for Client in Motion And
Server Stationary- – – – – – – – – – 30
4.2.3 The Throughput (Mb) as a Function of Speed (m/s) for Client in Motion and
Server Stationary- – – – – – – – – – 30
4.2.4 Delay in UDP Data Transfer as a Function of Speed when the
Clientis inMotion at Varying Speeds and the Server Stationary- – – – 30
ix
4.2.5 Delay in UDP Data Transfer as a Function of Speed when the
Server is in Motion at Varying Speeds and the Client Stationary- – – – – 31
4.3. Discussion- – – – – – – – – – – 31
CHAPTER FIVE: CONCLUSIONS AND RECOMMENDATIONS
5.0. Introduction- – – – – – – – – – – 32
5.1. Conclusions- – – – – – – – – – – 32
5.2. The Significance of the Study- – – – – – – – – 33
5.3. Recommendations- – – – – – – – – – 33
5.4. The Limitations of the Study- – – – – – – – – 33
5.5. Further Work- – – – – – – – – – – 33
References- – – – – – – – – – – – 34
Appendix A– — – – – – – – – – – – 39
Appendix B- – – – – – – – – – – – 42
Appendix C- – – – – – – – – – – – 45
Appendix D- – – – – – – – – – – – 49
CHAPTER ONE
Introduction
1.0 Introduction
Wireless networks are becoming more and more popular among recent network technologies as
compared to the traditional wired network. Wireless networks are connected through the wireless
channel. Generally there are two kinds of wireless networks. The cellular networks and WLANs that
have a wired backbone in which the base stations are the boundary nodes and the extended connections
between mobile users and the base stations are wireless channels. The other is wireless ad hoc network,
which is an infrastructureless self-configuring network [3] with more than one hop wireless channels in
the connection. This kind of topology is not widely implemented yet, but it is useful in emergency
search such as rescue operations, and military applications [2].
1.1Description of MANET Transport Protocol
This section describes the transport protocols used in the evaluation of MANET; it includes the
Transmission Control Protocol and the User Datagram Protocol.
1.1.1 The Transmission Control Protocol (TCP)
The Transmission Control Protocol (TCP) [31] is a connection-oriented and reliable protocol, which
provides reliable host-to-host data transmission in packet-switched computer communication networks.
TCP uses a sliding windowmechanism in combination with timers in order to adapt to network
conditions and retransmit lost packets to provide reliability. In TCP, the window size determines
the number of bytes of data that can be sent before an acknowledgment from the receiver must arrive. T
CP establishes a full-duplex virtual connection between two endpoints which is defined by the IP
address and the port number of each endpoint. The byte stream is transferred in segments. Typically,
TCP is the best transport layer protocol for applications that require guaranteed delivery of data [50].
The TCP algorithm as a whole is quite complex and there are many different versions and extension
proposals available [30].
TCP congestion control: Slow start and congestion avoidance phase.
To avoid network overload and resolve network congestion, TCP provides window-based congestion
control. It assumes network congestion upon the detection of packet loss. The figure below illustrates
the different phases of TCP’s congestion control.
2
Figure 1.0 TCP congestion window illustration of [30].
When a new TCP connection is established between two end points, the Slow-Start (SS) mechanism
takes place in order to probe the capacity of the network: Starting from a congestion window (cwnd)
equal to one, the sender increases the cwnd by the current segment size when it receives a new ACK
(i.e., non-duplicated) acknowledgment (ACK). Keeping to this principle, the window size increases to
an estimated capacity, termed SS-threshold. Figure 2.2 illustrates the slow start algorithm for the first
four packets (i.e., between packet number zero and three). When the SS-threshold is reached, TCP
enters the Congestion Avoidance (CA) phase, as shown between packet numbers four and seven. In
CA, the cwnd increases linearly up to the receiver’s maximum advertised window or until packet loss is
detected. Regular TCP [30] assumes a packet loss when the retransmission timer expires before the
respective segment is acknowledged. In this case, SS is unavoidable. In Figure 1.0, the sender detects a
packet loss upon the transmission of packet number seven. Consequently, it enters SS: When the sender
transmits packet number eight, the cwnd is reset to one and the SS-threshold drops by half of the
current cwnd.
1.1.2 The User Datagram Protocol (UDP)
Unlike TCP, the user datagram protocol (UDP) is a connectionless transport protocol that is usually
employed on top of packet switched IP networks. UDP assumes that Internet Protocol (IP) is used as
3
the underlying protocol. It offers a minimal transport service, which allows applications to directly
access the datagram service of the IP layer. UDP has some limitations which results in its inability to
provide reliability or error recovery, but it provides services like checksum calculation and
multiplexing by port number. The speed of data transmission in UDP is relatively high because it
introduces only minimal overhead but offers no guaranty for data delivery and duplicate protection.
Real-time applications like IP telephony or video conferencing with some specific requirements make
use of UDP. These applications do not require reliability or congestion control, but rather aggressively
use the network according to their bandwidth and delay requirements [30]. For real-time applications,
packet loss up to a certain limit is tolerable because human perception is not sensible to small
interruptions (i.e., depending on the codec, assuming that the codec is able to cope with packet loss). In
real-time communications, the retransmission of a lost packet would just waste bandwidth and increase
cost because the receiver drops the packet when it is too late (i.e after the data stream has been
presented to the receiver). In addition, real-time applications require a constant transmission rate, which
conflicts with transport layer services, such as congestion control. A drop in throughput due to
congestion control could result in a quality decrease or even link failure.
1.2 Description of MANET Routing Protocol
In an ad hoc network, mobile nodes communicate with each other using multi-hop wireless links. There
is no stationary infrastructure such as base stations. Each node in the network also acts as a router,
forwarding data packets to other nodes. A central challenge in the design of ad hoc networks is the
development of dynamic transport and routing protocols that can efficiently find routes between two
communicating nodes [6]. The routing protocol must be able to keep up with the high degree of node
mobility that often changes the network topology drastically and unpredictably. The routing protocols
for mobile adhoc network that are available in the literature are classified into two categories:
Proactive (table driven) and Reactive (on demand) protocols [1].
1.2.1 Proactive (table driven) Routing Protocols
Each node in the network has routing table for the broadcast of the data packets and want to establish
connection to other nodes in the network. These nodes record all the presented destinations, number of
hops required to arrive at each destination in the routing table. The routing entry is given a sequence
number which is created by the destination node. To retain the stability, each station broadcasts and
modifies its routing table from time to time. How many hops are required to arrive that particular node
and which stations are accessible is result of broadcasting of
4
packets between nodes [1]. Each node that broadcasts data will contain its new sequence number and
for each new route, node contains the following information:
i. The number of hops required to arrive at that particular destination node.
ii. A new sequence number marked by the destination node
iii. The destination address
The proactive protocols are appropriate for networks with less number of nodes because its periodic
update over floods the network with much more Routing overhead resulting in more bandwidth
consumption [1]. An example of Proactive Routing Protocol is Destination Sequenced Distance Vector
(DSDV).
Destination-Sequenced Distance-Vector (DSDV)
Destination-Sequenced Distance-Vector Routing protocol is a proactive table driven algorithm based on
classic Bellman-Ford routing. In proactive protocols, all nodes learn the network topology before a
forward request comes in. In DSDV protocol each node maintains routing information for all known
destinations. The routing information is updated periodically. Each node maintains a table, which
contains information for all available destinations, the next node to reach the destination, number of
hops to reach the destination and sequence number. The nodes periodically send this table to all
neighbors to maintain the topology, which adds to the network overhead. Each entry in the routing table
is marked with a sequence number assigned by the destination node. The sequence numbers enable the
mobile nodes to distinguish stale routes from new ones, thereby avoiding the formation of routing loops
[2].
1.2.2 Reactive (On Demand) Routing Protocol
Reactive Protocol has lower overhead since routes are determined on demand. It employs flooding
(global search) concept [1]. Constant updating of route tables with the latest route topology is not
required in on demand concept. Reactive protocol searches for the route in an on-demand manner and
set the link in order to send out the packet from a source node to destination node [6]. Route discovery
process is used in on demand routing by flooding the route request (RREQ) packets throughout the
network. Examples of reactive routing protocols are Dynamic Source Routing (DSR) and Ad hoc Ondemand
Distance Vector routing (AODV).
Dynamic Source Routing (DSR)
Dynamic Source Routing protocol is a reactive protocol i.e. it determines the proper route only when a
packet needs to be forwarded [6]. The node floods the network with a route-request and builds the
required route from the responses it receives. DSR allows the network to be completely self5
configuring without the need for any existing network infrastructure or administration. The DSR
protocol is composed of two main mechanisms that work together to allow the discovery and
maintenance of source routes in the ad hoc network. All aspects of protocol operate entirely on-demand
allowing routing packet overhead of DSR to scale up automatically.
Route Discovery: this is the mechanism used by a node to find out the route leading to the destination
node that will receive the packet it wants to send. Route Discovery is used only when Source node
attempts to send a packet to Destination node and has no information on a route to the Destination
node.
Route Maintenance: When there is a change in the network topology, the existing routes can no longer
be used. In such a scenario, the source node can use an alternative route to the destination node, if it
knows one, or invoke Route Discovery. This is called Route Maintenance.
Adhoc On-demand Distance Vector Routing (AODV)
AODV uses a very special technique to maintain routing information. AODV protocol is both an ondemand
and a table-driven protocol. It adopts flat routing tables, one entry per destination unlike DSR,
which can maintain multiple route cache entries for every one destination. Unlike DSR The packet size
in AODV is uniform [1]. In AODV there is no need for system-wide broadcasts due to local changes,
unlike DSDV. AODV has multicasting and unicasting routing protocol property within a uniform
framework. Source node, destination node and next hops are addressed using IP addressing. AODV
builds routes using a route request /route reply cycle. AODV uses sequence numbers maintained at
each destination to determine freshness of routing information and to prevent routing loops. Sequence
number for both destination and source are used. These sequence numbers are carried by all routing
packets. Maintenance of timer-based states in each node, regarding the use of individual routing table
entries is an important feature of AODV. If routing table entry is not used recently then routing table
entry is expired [6]. When the next-hop link breaks nodes are notified with repeat request (RERR)
packets. Each predecessor node, forwards the RERR to its own set of predecessors, thus effectively
erasing all routes using the broken link. Route error propagation in AODV can be visualized
conceptually as a tree whose root is the node at the point of failure and all sources using the failed link
as the leaves. It is loop free, self-starting, and scales to large numbers of mobile nodes.
1.3 Problem Statement
Among recent research works, wireless mobile adhoc network has received a major attention from
different researchers because of its unique operations and importance. It is self-configuring and uses no
centralized control among member nodes. The mobility of the member nodes is expected to have a
6
significant effect on the performance of the networkresulting in difficulty in achieving high data
throughput and low packet delivery delay as widely suggested from simulation results in the literature.
Hence, this research work is intended to experimentally validate the impact of varying node velocity on
the performance of TCP and UDP of a wireless mobile adhoc data network using data throughput and
packet delivery delay as the performance metric. In general, throughput and packet delivery delay are
the most important performance metric for any wireless network systems [1]. The capacity represents
the throughput (bits per second) of the whole system including all nodes, and the delay represents the
average time duration of a packet transmitting in the network from a source to the destination.
1.4 Objective of the study
Generally, it is difficult to achieve both high throughput and low packet delay because as the node
velocity increases link failures and network overhead increases [2]. Furthermore, our objective is to
evaluate and validate the impact of mobility rates on the performance of MADNET transport protocols
TCP and UDP using Uniform mobility model and make recommendations based on the findings so as
to achieve high throughput and low packet delay.
1.5 Significance
This study will be employed to improve capacity and performance of such computer to computer
networks employed in areas struck by disaster for the rescue of the victims. Also, this kind of network
is useful for sharing data among members in a conference. Furthermore this study will help in decision
making when designing new protocols or when improving an existing one.
1.6 Scope of the Study
There are many network performance metrics which can be employed to evaluate mobile adhoc data
network performance. The evaluation in this study will use network throughput, and packet delays to
evaluate the performance of MADNET using TCP and UDP at different nodal velocities.
1.7 Methodology
In this research we intend to carry out an experiment using two identical laptops, designating one as our
sender and the other as the receiver. We put the sender in motion and the receiver stationary during
which we measured the amount of TCP and UDP data sent and the delay experienced at different
velocities. We also reversed the process by keeping the sender stationary and the receiver in motion at
corresponding nodal speeds. We adopted Iperf as our performance tool which we installed in the
7
laptops and configured one as sender and the other as client using Iperf commands in each section of
the experiment. We plotted the graph of throughput with speed and delay with speed from our records
from where we made useful recommendations based on our observations.
1.8 The Structure of the Work
This study will address the problem in these chapters: chapter one will introduce the study,
chapter 2 will review some literatures available on the topic, Chapter 3 will describe the methodology,
chapter 4 will give out the main results and explain their meanings to the capacity and the packet delay
based on the MADNET
performance. Finally in chapter 4 we conclude and make necessary recommendations.
IF YOU CAN'T FIND YOUR TOPIC, CLICK HERE TO HIRE A WRITER»
