This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
Mobile Ad hoc networking (MANET) is a collection of nodes which forms temporary wireless network it does not require any base stations or routers as networking services or infrastructure. It is easily deployable, therefore Ad hoc network is very useful for relief operation such as search and rescue and military purposes in battle field. In this Ad hoc network, nodes operate as intermediaries such as routers which are used for communicating with other nodes, as this can operate on single hop and multiple bases. Nodes are operated with limited power batteries and constrained to limited bandwidth.
Ad hoc network is used in communication between computers with out router, the actual meaning of Ad hoc is `` this is for special purpose ``. As the Ad hoc network is dynamic there was some need to develop some routing protocols, some of the prominent protocols are Ad hoc On Demand Vector (AOVD), Dynamic Source Routing (DSR), Temporarily Ordered Routing Algorithm(TORA) and Optimized link state Routing (OLSR). In this project we analyze the performance of AOVD.
Ad hoc on demand vector is routing protocol that has been recently introduced in the OPNET simulator that is in 2004 in 10.5th version. In order to understand the behavior of the AODV routing protocol we had done many simulations on the transport layer of the protocol while using both TCP and UDP in AODV networks. Analyzing the results are taken separately for both TCP and UDP. To analyze the UDP communication we have used video streaming which is unidirectional in analyzing the TCP performance we had used file transfers via File transfer protocol (FTP).
A Mobile Ad hoc network (MANET) is a mobile network which is wireless and does not require any infrastructure such as router or base station, this MANET network has nodes which acts as both router and a host. As transmission range is limited between the wireless networks, we need multiple hops where the data needs to be exchanged from one node to another node of the network. This dissertation is about the performance analysis of AODV routing protocol using OPNET. Here we first discuss about the importance of Ad-hoc routing and its main routing protocols AODV, DSR, TORA and OLSR, form all the routing protocols we analyze the performance of AODV.
AODV is a reactive type protocol and is analyzed using UDP and TCP based protocols.
OPNET simulator is most leading simulators used for doing research in networks, development is and also used to study and design communication networks, devices, protocols and used by the applications with are at grater flexible. For physical layer modulator to process the applications, OPNET simulator provides a graphical editor interface for building various models and networks.
The dissertation consists of the following chapters
Chapter 2 gives the basic introduction of OPNET and its uses and how we can develop a network in OPNET
Chapter 3 gives the overview of AOVD protocol, which is dealing with the operation of AODV and wireless properties of IEEE 802.11b wireless technology which is used in this simulation.
Chapter 4 gives the details of generating AODV network and its analysis using TCP and UDP protocols, it also discuss results while using these test based applications
Chapter 5 gives comparison of simulation results with the test bed results that had already been taken
Chapter 6 concludes the results and presents the future work report
OPNET is very large and powerful software which enables to simulate wide variety of networks with different protocols, it was initially developed for military purpose but was reformed as a world leading simulator tool. OPNET developed an enhanced environment for modeling and evaluating communication networks and distributed systems. These tools can be categorized into three phases of modeling and simulation projects they are specification, data collection, analysis and simulation.
All these phases are all sequential they follow a return to specification analysis, before going to these topics we will discuss about the basics of OPNET.
OPNET is tool which is mainly used for simulation and operates at packet level, which is initially created for analyzing the networks, it has a large library of accurate models that are fixed network hardware and protocols which are commercially available in the market. There are many various tools of OPNET they are, source coding editing environment which is very useful, a editor which is used for, a model which is used for editing node, a process where we can edit the model, a pattern of antenna where we can edit, an editor where we can modulate the curve, an editor where we can format the packet, a tool where we can analyze the configuration, simulation tool, an editor where we can edit the information about interface control, a model which is used for linking, and a model which is used to set the path, and finally the animation viewer which is an integral part of OPNET.
Now here is a small example on creating a workspace from start up wizard in OPNET by just clicking the next entry,
If we follow the above table the workspace will created and a separate table will be created for the object palette that we have selected and are shown as below
We can create the network by three steps they are we can directly drag objects from palette to the workspace, or we can use the rapid configuration from topology or we can import the topology.
MANET model architecture in OPNET:
There are many models such as AODV, DSR, TORA and OLSR and all are available in OPNET 11th version. Here we explain about all the things that AODV network uses in the process and the node models of the palette and header
The MANET model can be clearly explained form the above figure; WLAN workstation could be operated in Ad hoc mode. Manet_mgr is the command which maintains the overall routing protocols that are present in OPNET and is created by the main two process function which are ip_dispatch and ip_encap and also created by a very specialized process for ad hoc routing.
MANET models which includes nodes:
There are some nodes that are included in MANET object palette that are shown below
Workstations and servers which are Wireless LAN: there are some of the nodes the nodes under this model that can be used to generate application traffic such as file transfer protocol, email, http, TCP, IP, WLAN.
MANET stations: raw packets are generated by these node models over IP over WLAN, these node models can also be used as routing protocols which are configured to run AODV, which acts as a traffic source or destination.
In certain period of time profiles describes the patterns of the users or users group in term of applications, if we take a LAN or a workstation there are many several different profiles running under them. All these profiles can also represent different user groups such as sales, administration etc.
We can construct a profile using two different application definitions, and we can specify different usage parameters for different applications such as duration of the start time and repeatability, and we can also have same applications but with different usage parameters and specify different names as they perform different applications.
Receiver group configuration:
This groups tells us how many possible receivers are used for communication in that node, this unity node will eliminate the receivers that do not mach and speeds up the simulation process
Task configuration: by using this task configuration node we can customize the applications.
Mobile configuration: mobile configuration is the profile that is used to define individual nodes which refer model mobility and based on the configured parameters, and the configuration node which is controls the node movement in the network.
Model Files in AODV:
AODV network defines some files in MANET they are stated as below .
Model called process:
This model are used for generation and processing of AODV control packets, maintains routing protocols of AODV and updates the IP common routing table. This model is stated in the MANET directory which is named as <opnet_dir>/std/manet directory and it is aodv_rte.
These are the files which shows are defines the data structure and the constants that the route requires for connectivity table, these files can be found in the directory <opnet_dir>/std/manet.
Aodv_ pkt_support.h is a file that the structured packet options and types of packets are defined such as rrep.rrep.rerr and aodv.ptypes.h files are function prototypes functions files.
External source files:
These files are available in the directory <opnet_dir>/std/manet, and are written in C - code
Configuring AODV in OPNET:
In order to configure the AODV parameters we should right click on the nodes in the project editor. There opens a new window for editing the attribute values of different variables
Taking results of AODV:
We can take the results by choosing individual discrete event simulation (DES) statistics we will have many statistics which is going to be stimulated.
Ad hoc On-demand Distance Vector routing (AODV) and it overview
From the last decade there was a drastic change in the way of communication by people; mobility has become the major part of communication as there was huge development in technology, devices like smart phones which are providing wireless abilities with low battery consumption. MANETS are wireless networks that do not require any infrastructure for communication between nodes and these nodes establish a direct communication between them.
There were many algorithms that have been introduced to analyze and solve the need for routing algorithm in Ad hoc networks.
The above figure shows the different routing protocols of MANET and in this dissertation we discuss and analyze the performance of AODV routing protocol.
Ad hoc On-demand Distance Vector (AODV) routing is the dynamic reactive protocol which is designed specially to reduce the overheads produced by proactive routing protocols and keeps the information for active routes only. This is a self starting network which is very flexible by the users in using ad hoc networks. Each mobile node in this network operates as a specialized router and with in the periodic intervals we can obtain the routes on demand.
AODV routing protocol activates the multi hop routing between the nodes, this routing runs on a distance vector algorithm. As the AODV routing is reactive it requests the node only when it is required for communication. When the endpoints have valid routes for communication AODV does not play any role. This protocol also includes link breakages and loop freedom, which used to send immediate notifications to the affected set of nodes. AODV supports the multicast routing but does not support counting to infinity problem.
In order to discover and maintain links this AODV algorithm uses different messages, while finding the routes to establish a connection between two nodes its sends a broadcasting message Route Request (RREQ) to all its neighbors. This message RREQ travels through the network until it finds the destination node, then unicasting message is send back to the source stating that the route is made available.
First we will discuss about the properties of WLAN (IEEE 802.11X) which is used to make connection in Ad hoc networks, then we will discuss about the operation of AODV.
Properties of WLAN (IEEE 802.11b):
The communication system where it does not require any cables for connection between two computers can be considered as a WLAN.
WLAN is a data transmitting system designed specially to help the mobile user to connect the local area network through radio waves. The property of WLAN connection differs from other LAN connections. For instance, WLAN requires a special MAC sub-layer protocols to connect the LAN with computer. The protocol of conventional LAN is not suitable for wireless LAN because this connection gets interference mostly on the receiver side than sender. These conventional LAN protocols in wireless LAN create two types of problem first one is Hidden station problem and another one is exposed station problems.
Hidden station problem:
The hidden station problem can be seen clearly through this figure the A,B,C,D are the four independent station in where the distance between the two stations A and B is very short, so that communication between A and B can be made very effective. In the same way the station C can communicate easily with B and D, but it can't access with A because it far from the radio range.
Consider if A transmits the information or data to station B as per fig (3.2) and if the station C can not receive the message properly even though it finds a medium. Because station A is situated far away from station C and it falsely thinks that it can transmit the message. If the station C starts transmitting the message then Station B will get it clearly than A. The problem behind this station is not being detected by potential competitors because the competitors are placed out of range. This type of problem is called as a problem with hidden station problem.
Problem when station is exposed
Suppose station B is transmits a message to station A according to the fig 3.3. If station C finds the channel, it catches the transmissions and falsely assume that it cannot transmit to D, where such kind of transmission in the area of stations A and B might crate a bad reception, where both the receivers were located. This type of problem is called as a problem when station is exposed.
3.1.3 Basic Access Method: CSMA/CA
This is a basic access method which is used in WLAN, the abbreviation of CSMA is Carrier sense multiple access is a basic access mechanism used in Wireless LAN. This CSMA/CA combines with distributed co-ordination function (DCF). The distributed coordinative system uses CTS (clear to send) and RTS (request to send) signals to minimize the possibility of clash occurred by hidden station problems. In order to reduce the problems when the station is exposed, there is a standard time period called a random back-off time between two successive packets transmission times where a node should wait. The process how it works is explained below.
The station which transmits data through wireless network shows medium of carrier sense. If the medium which transmits a message is idle for a certain period of time, it should wait for a certain time period called as DCF inter frame spacing (DCIF) till it sense the medium again. After DIFS process, if the medium falls again in idle then the message or data that is to be transmitted will be sent immediately. The receiver of the data waits for a time period called SIFS (Short Inter frame spacing), which transmits and replies back the acknowledgement ACK to all sending stations when it has a direct access to the medium. Or else, if the transmitting station recognizes the medium in busy state then it has to wait for DIFS and after the collision that each station will create a random back-off time and wait for the time period. When the medium falls idle, the collision time is calculated in form of slot, as soon as the slots finishes value of back-off is counted down.
Carrier sense which is virtual:
If a station plans to send a packet of data it will first send a short packet which is called as RTS (Request to send) which is called a control packet. This involves the Starting point where the message starts, the destination point and time duration of transactions. The station to which the data is directed and responds only to controlled packet called as CTS (Clear to send). It involves the same duration of information when medium is free. The station which receives information from RTS or CTS will set their Carrier Sense which is virtual indicator named as Network Allocation Vector (NAV) for a particular period of time duration and shows the time period that a medium is reserved.
AODV and its Basic Operations :
Here in this section we explains about the process that is involved in creating the AODV network, deleting the AODV network and maintaining the routes of the network.
Discovering the Path:
Path discovery is a process of communication where nodes try to communicate with each other which do not have any information about routing in its table. Every node carry two sections that carry with them they are broad casting ID and node sequence number. The process of discovering the path can be started by the source of nodes by sending a route request packet to all its neighbours. There are some fields that the RREQ packet contains they are address of the source, sequence number of the source, identification of the packet for broadcasting, address of the destination, sequence number for destination and hop count.
The RREQ is identified uniquely by broadcasting id and source nodes. Whenever the source node issues a new RREQ the broadcasting id is incremented every time. Each neighbour is satisfied either by sending the route reply back to the source or by broadcast again the Route request to its own neighbour after raising the hop counts. A node can receive many copies of same broadcast request packets from many neighbours. If the RREQ is received before with same broadcast ids and source nodes then it will automatically stop sending it again.
Reverse path setup:
There are two sequence numbers which are highly integrated with RREQ. One is sequence number and another one is last destination sequence number. The sequence number maintain the freshness information about the reverse route to the source and destination sequence number identifies how fresh a route to the destination must be before it can be accepted by the source.
Based on the figure when a source node of S decides to route with the destination node D and it won't have any available roots. With this the node S starts broadcasting the RREQ message to its available neighbours in search for the route of destination. The nodes 1 and 4 will receive the RREQ message from node S for being its neighbour. The 1 and 4 nodes create a reverse link from which it received the RREQ message. While the 1 and 4 nodes were not aware of the link to connect D node and it simply re-broadcast the RREQ message to its neighbour nodes like 5 and 2. So this is clear that RREQ travel from source to various destinations whenever it travels it automatically create a reverse link or path from all neighbouring nodes to its source. These reverse path route is very much need for the nodes to receive RREQ back to the source where it was originated. The source nodes will buffers the IP address and RREQ ID. If the nodes receive RREQ back from the neighbours then it will not re-forward or process the packets.
Forward Path setup:
The RREQ will appear on a source node which actually possesses the current route to destination. The receiving source from this route will have a check whether the RREQ message are received from bi-directional links. Checking the current route is process by first determining the route that compares the destination sequence number with the entry route of the sequence number in RREQ. The route which is intermediate does not use the recorded route for responding RREQ if the intermediate node is lower than the sequence number of this RREQ. Moreover instead of rebroadcasting the RREQ, the intermediate node can depend only on the route with a sequential number which are greater or equal to the number contained in RREQ. If the RREQ are not processed previously and if it doesn't have any current route to destination then the nodes will unicast the RREP (Route reply packet) back to the neighbour from where it received the RREQ. The process of RREP includes the following factors, they are address of the source packet, address of the destination packet, sequence number of the destination, counting the no of hops in the routing and the time that took for the packet to travel form source to destination which is called life time of the packet
As soon as the source node receives a broadcast packet it can easily supply a clear route to destination, and also it will assign a reverse path automatically to the source of RREQ,
There fore the RREP goes back to its original address that is the main source of its origin. It also updates the time-out data's from the entry route to the destination and source and from the requested destinations the last destination sequential number is recorded. The Figure 2-2 explains the settings of the forward path setup of RREP which travels through the 1,2,3 nodes from the destination station D to its source node S. we had determined by RREP that Nodes 4 and 5 were not along the path, after time out check it will delete the pointers which are reverse, this process is called active route time out check. If the node received a RREP broadcast then the first RREP message travels for a given source node. If it receives additional RREP's, it automatically updates its routing information and broadcast RREP only if it contains a higher destination sequence number than the older RREP, or with the same destination sequence number with a lesser hop count. As soon as the first RREP is received the source node S can start transmitting the data. Later it can also update its routing information if it learns of a better route.
Route Table Management
The timer which is connected with the reverse part of routing entries is known as route request expiration timer. The main idea of this timer is to clear all reverse path routing entries from the nodes and make it not to lie on the route from source to destination. The size of ad-hoc network determines the expiration time. Another important factor connected with routing entries is the route caching timeout by which the router will be considered invalid after a period of time. Each routing table entry records and maintains the address of active users or neighbours by which the packet data's are received. The user or neighbour is considered to be active for the destination if it originate or relies on at least one packet for that destination with in the most recent active timeout period. These data's are maintained mainly to notify all active source nodes when a link along with a path to destination breaks. The entry route is considered to be active if it is been used by active neighbours. The mobile nodes maintain the table entry for each destination. the entry table contain the following information. They are
Number of hops
Sequence number for the destination
Active neighbours for this route
Expiration time for the route table entry
When data's are transmitted from source to its destination the time frame is reseted to the current time plus active route timeout. If any mobile node is offered then the destination sequence number of new route differs from current route. Usually the route with higher sequential numbers will be selected in case if the sequence numbers are same then the new routes are selected only if it holds low metric to destinations
If any link breakage occurs then the node will invalidate the existing route connections from routing table entry. The nodes will list the affected destinations and also determines which neighbour will be affected by this link breakage. After this the node will send a error message (RERR) to the connected neighbours. The error messages (RERR) are broadcasted to a number of neighbours who are connected with that route. The host can also relatively unicast the message to its neighbours if broadcast are not possible.
3.2.4 Path Maintenance
The nodes movement which are not on the active path will not affect the routing to path the destination. During an active session if any source nodes move it may reinitiate the route discovery procedure to introduce a new route to its destination. RREP will be sent to the affected source nodes if any intermediate or destination nodes moves. Timely messages like HELLO can be used to certify symmetric link and also to detect the failure in link. The failure in link can also be detected through Link-layer acknowledgement (LLACKS). Sometimes the failure in link is also detected when the packet forwarded to next hop fails. If the hops become unavailable or unreachable then the nodes upstream broadcast an unwanted RREP with fresh sequence number and bringing the hop counts to infinity to all neighbours. These nodes still relay the messages to active neighbour and this process continues till all source nodes are notified. The source nodes restart the discovery process in order to receive the notification of broken link and also to route the destinations. In order to know that the route is still required the source nodes will check whether the routes are used frequently and also inspect the upper level protocol control blocks to know whether connection remains open in using specific destinations. If the source nodes try to build the route for destination, it actually transmits a RREQ with higher destination sequence number to ensure that it builds a new route and no nodes will be responded if they think that the before route is valid
3.2.5 Local Connectivity Management
AODV- Ad hoc On-Demand Distance Vectoris is a protocol which sends the HELLO message periodically to all the active neighbours in order to inform them that the link towards the host server is alive. The message HELLO is broadcasted with TTL equal to 1 so the messages are not forwarded again and again. When this HELLO message is received buy host it will automatically update the information about the host in routing table. In case if the host didn't receive any message from neighbours for authorized hallo loss* hello interval of time, then the information's on routing table will be notified as lost. The RRER messages are generates to inform the host neighbours about the link breakage. The local connection administration with HELLO message can be used to make sure whether the nodes with bi-directional connectivity are considered to the neighbours. The HELLO messages are sent with a node list which is designed to be heard by the neighbours. Each of these nodes are checked wisely to assure that it use only routes to neighbours who heard the message HELLO. In order to save local bandwidth, such checking can also be made only if the nodes are configured clearly.
3.2.6 Local repair
If any link breakage occurs due to fewer amounts of hops then the host will try to repair it locally. For repairing the link the host maximises the sequence number of destination and also broadcast the RREQ message to the host. Time to live for internet protocol header should be measured and calculated, so that the local repair settings will not affect the entire network. The host server waits for RREP message to its RREQ message for a particular period of time. In case if the host didn't receive the RREP message properly then the routine table status for entry goes to invalid. On the other hand if the host receives the RREP message then the hop metric counts are compared. If the message has more hop metric than the previous one then the RERR with N field setup is broadcasted. The N field setup indicates that host has repaired the link locally and the entries in the table will not be deleted. The received RREP message is handled as original message.
Modeling of AODV networks and analysis
Chapter 4 clearly analyse the model of Ad hoc on demand vector network with a parameter of RFC3561 by a magnetic OPNET model. The performance and results of User datagram protocol and Transmission control protocol are also in this study. File transfer protocol was used to analyse Transmission control protocol performance and video streaming is used to analyse the performance of User datagram protocol.
4.1 Modeling an Ad hoc on demand distance vector
The networks with 5 wireless LAN nodes were positioned in campus area which constitutes around 3990m x4000m wide and long. The entire node in this network is constructed with AODV method. Except one all four nodes were fixed with AODV node (3 -Node, 4-Node, destination and source) and other node is portable /mobile. During the simulation period, the portable node were started to move after 3 minutes with a specified path by the trajectory
AODV was used as routing protocol for all nodes. Fig 4.1 and 4.2 displays the constructed AODV network in campus area. Fig 4-1, shows the five nodes of Wireless LAN stations though Fig 4.2 use only 4 WLAN stations and a Wireless LAN server. Wireless LAN server was applied mainly to produce the traffic of transmission control protocol by working as a file transfer protocol.
4.2 Creating UDP traffic
This section describes how the uni-directional traffic of user datagram protocol is produced in the model of AODV network.
4.2.1 Configuring the Task Configuration Object
The uni-directional user datagram protocol traffic was not possible to generate in standard applications it can be generated only by configuring certain custom applications. The traffics which has to be configured is called as task configuration objectit can be done by giving a special task name as "udp-task" and configuring manual tariff are extracted mainly by sending more than 5060 requested packet from the source to destination with a period of time constitutes of 1 second. In order to make uni-directional User datagram protocol traffic the responses from destinations were disabled. Each requested packets has more than 1470 bytes, each of it were transmitting only one packet at a time. Figure 4.3 explains how to configure unidirectional traffic using Task Configuration object in OPNET.
4.2.2 Configuring the Application Configuration Object
If main task are explained once, then it should build the custom application in application configuration objects. The data transmitting protocol used for application are explained in the custom table.
4.2.3 Configuring the Profile Configuration Object
If custom applications are described in this process then it will be used later in profile configuration objects. Based on the above Ad hoc on-demand distance vector model, the profile will be named as "udp_iperf_user" will start by nearly 1 min 40 seconds and lasts till the simulation process ends. The application like "udp_application" are positioned in the Application Table of Profile Configuration Table as per in the fig 4-4. This type of application will begin to run without any offset and lasts long till the profile end
The above stated profile will be positioned in creating an application for node, therefore like source of node. The Fig4.6 explains how to add the above stated defined UDP application in the source node.
4.3 Generating TCP traffic
The generating procedures for Transmission control protocol traffic are nearly similar in generating the User datagram protocol traffic. The originating nodes for Transmission control protocol sends only one request for File transfer protocol and destination nodes send continuous information to source. Compared to source the destination transmits additional tariff. Whereas the source file sends only one request packet for downloading file. The generations of tariff in Transmission control protocol are displayed in Figs 4.7 and 4.8.
Transmission control protocol as the transport protocol for the application is defined in the custom table of Application Definition Table in Application Configuration Object as shown.
The configuration object for transmission control protocol tariff generation is explained in User datagram protocol traffic generation. The configurations of destination nodes are very different in transmission control protocol. The supported services of destinated attributes should support the applications (transmission control protocol application) which are defined in Application Configuration Object. Configuring this destination is explained clearly..
4.4 Configuring the mobile node
In order to define the mobility patterns the mobile node is directly attached to the trajectory and it can also be made by choosing a pre-defined trajectory stated as per Fig 4-11.
The mobile node waits only for 160 seconds and during simulation process it starts moving its way as per trajectory during the process of simulation. Fig 4.12 shows how this trajectory attributes are defined
4.5 Analysis of UDP Results
This UDP application model was run for 220 seconds on the AODV five nodes of the network. Then the final result must been shows the performance of the AODV network, all this process must be used when the personalized video streaming based on the indirect method of the UDP traffic network application
4.5.1 AODV Routing Traffic
In the above fig 4-13, in this process that the source node sending two RREQs router requests packets to router detection to find final destination , RREQs first packet have been send at 100 seconds through mobile node and second packet also sent at 165 seconds. These packets are sent from node3 and node4 when mobile nodes are moving close to find the trajectory. The destination node sends corresponding reactions to requesters by its source nodes.
ADOV route application traffics send packets to the every node in the fig4-8. In the destination that the acknowledge was suddenly raised the traffics to 100 and 165 seconds, and it continuously transmits the packets of RREP packets during rest time that traffics are because of hello messages. In mobile node that the traffics signal suddenly rises at 100 and 165 seconds by the messages of RREQ and REPP and reaming is for hello message. In both 3&4 nodes that full traffics is appropriate for hello messages but expect at 165 seconds. In case that the traffics are raised because of the broadcast of the second packet messages of RREQ.
In the fig -15 the Ad hoc on demand distance vector the traffics application is obtained by a node in network model, then all the traffics obtained in all nodes mainly because of the messages of RREQs, hello and RREPs. In the figure 4-16 it shows number of hops per router in ADOV application. The nodes of source and destination nodes have mainly 2hops per route before then node in mobile node start to move at that time the mobile nodes are gone through the nodes3 and 4 then it will call count 3hopes per router. Node3 and node4 all ready have the 1hopes per route of the mobility of the mobile nodes, after this process it counts the 2 hops per route & mainly the counts in mobile like 1 hops per route
4.5.2 IP Traffic
The internet protocol includes both the user datagram protocol traffic and AODV routing traffic. The AODV & Internet protocol traffics transmitted through destination nodes are precisely equal. While the destinations are not transmitting the user datagram protocol traffic. The destinations of internet protocol and user datagram protocol traffic are expected to be same in pocket counts. The traffic of internet protocol is received and sent by source, nodes of mobile and nodes 3 and nodes 4 are very parallel to the tariff of AODV received and sent. However this may vary in number of packet data sent though the internet protocol which involves user datagram protocols. The traffic of internet protocol received and sent as per in figs 4.17 and 4.18
4.5.3 UDP Traffic
According to UDP traffics it is uni-directional from destination and source in the model of network, this User datagram protocol application traffics is transmitted by destination, it's shown nothing in the fig 4-19. The node3 and node4 of mobile are functioning as a router along the traffic is sent from the destination and source, and also from UDP application the traffic can send and receive the information from the nodes it will be shown the zero in the fig4-19 and 4-20. The main concept is source node initiate sending the UDP applications along with traffic of 100 sec and it lasted till the profile end. The tariff received from destination was suddenly reduced from 160 sec with a breakage in link of portable nodes and it also moves continuously with trajectory.
Test bed result VS simulation results
Here we briefly describe the comparison between the set of results that had been taken in real time environment with the result above chapter. In this real time environment AODV network routing also had five nodes and where nodes in this network are notebooks. By using iperf traffic generator we have generated raw UDP and TCP traffic.
From the above table it clearly shows that both the results are same that is test-bed result and simulation results. In order to investigate the performance we have extended the simulation further, while changing the values of HELLO Interval and hello loss.
As per the observation from the TCP model by changing the hello loss from3 to 4, we can compare the two link breakages from the values with the two different scenarios that have been taken; they are hello loss 3 and hello loss 5. From the figure it is clearly seen that Hello loss 5 scenario has less hops than the hops that are in the scenario of hello loss 3
The scenario with hello loss 5 had also less RERRs that had been sent and the packets had been dropped are less than the other scenario in hello loss 3. it is clearly shown in the below figures.
The above figure shows the throughputs variation in the TCP model, if we use the hello loss values a bit higher than there is a gradual decrease in the values of the throughput while in the movement, and if the transmission of normal data is less the throughput falls down to lower level.
Link brakes can be detected based on hello messages by AODV, as it is in OPNET. After receiving the at least one message while the duration of ``hello loss or hello interval`` each node should wait while detecting the link breaks. In the process of updating the life time of the route the AODV should process each node. Hello packets might be lost at a very high rate, if we have more data at link layer.
This study has deeply analyzed the performance of AODV networks by using the OPNET simulator. Recently, the OPNET simulators were started coming with AODV model (in the year 2004 the version 10.5 was released in market). Many models were involved in this study to examine the behaviour of TLP, with User datagram protocol & Transmission control protocol in Ad hoc on demand vector networks. The output of both User datagram protocol & Transmission control protocol were analyzed individually. The video streaming was also used to analyse the communication of User datagram protocol, though the performance of Transmission control protocol performance were examined by downloading files through FTP. Based on the analysis of simulation results the main conclusions were derived.
In common, Transmission control protocol & user datagram protocol behaviour in simulation and real test-bed Ad hoc on demand vector are nearly same.
When hops are more in number then the throughput are degraded because of higher delay in Round trip time.
If the application traffic is more in number then the throughput are disturbed again due to link layer loss. The Link layer loss can be caused because of wireless collision and hidden nodes problem.
Transmission control protocol throughputs are disturbed by changing some parameters of Ad hoc on demand vector. It is identified by shifting the 'Hello-Loss' values from 3 - 5, Transmission control protocol throughput was improved for the Files transfer protocol rates of 500 Kbps, reducing the time duration for detecting the medium and increasing the sequential number of break in routes.
Identifying the breakage in link by using link layer information will avoid some of the route breaks while using more number of loads in AODV networks.
The below stated work will be carried in future by analyzing the OPNET simulators:
The main function and performance of Ad hoc on demand distance vector routing networks will be analysed with various network parameters.
A variety of AODV protocols and its optimum Performance will be analysed in future
The performance of TCP in Ad hoc network will also be analysed for implementing the best available wireless protocols in future