25.01.2014 Views

Performance Evaluation of AODV and ADV Protocols in VANET ...

Performance Evaluation of AODV and ADV Protocols in VANET ...

Performance Evaluation of AODV and ADV Protocols in VANET ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Kusum Dalal et al,Int.J.Comp.Tech.Appl,Vol 3 (1), 50-55<br />

ISSN:2229-6093<br />

available or route expired [10].When RREQ reaches<br />

the particular dest<strong>in</strong>ation <strong>and</strong> if source node receives<br />

RREP then by us<strong>in</strong>g unicast<strong>in</strong>g, <strong>in</strong>formation is<br />

forwarded to the source node <strong>in</strong> order to ensure that<br />

route is available for communication.<br />

DSR protocol [11] uses source rout<strong>in</strong>g, that is, the<br />

source <strong>in</strong>dicates <strong>in</strong> a data packet’s the sequence <strong>of</strong><br />

<strong>in</strong>termediate nodes on the rout<strong>in</strong>g path. In DSR, the<br />

query packet copies <strong>in</strong> its header the IDs <strong>of</strong> the<br />

<strong>in</strong>termediate nodes that it has traversed. The dest<strong>in</strong>ation<br />

then retrieves the entire path from the query, <strong>and</strong> uses it<br />

to respond to the source. As a result, the source can<br />

establish a path to the dest<strong>in</strong>ation.<br />

TORA rout<strong>in</strong>g [12] belongs to a family <strong>of</strong> l<strong>in</strong>k<br />

reversal rout<strong>in</strong>g algorithms where a directed acyclic<br />

graph (DAG) toward the dest<strong>in</strong>ation is built based on<br />

the height <strong>of</strong> the tree rooted at the source. When a node<br />

has a packet to send, it broadcasts the packet. Its<br />

neighbor only broadcasts the packet if it is the send<strong>in</strong>g<br />

node’s downward l<strong>in</strong>k based on the DAG.<br />

Thus, among these three reactive protocol strategies<br />

<strong>AODV</strong> is preferred <strong>in</strong> <strong>VANET</strong>s because <strong>in</strong> <strong>AODV</strong> data<br />

packets carry the dest<strong>in</strong>ation address, whereas <strong>in</strong> DSR,<br />

data packets carry the full rout<strong>in</strong>g <strong>in</strong>formation. This<br />

means that DSR has potentially more rout<strong>in</strong>g overheads<br />

than <strong>AODV</strong>. Furthermore, as the network diameter<br />

<strong>in</strong>creases, the amount <strong>of</strong> overhead <strong>in</strong> the data packet<br />

will cont<strong>in</strong>ue to <strong>in</strong>crease. Also, TORA provides a route<br />

to all the nodes <strong>in</strong> the network, ma<strong>in</strong>tenance <strong>of</strong> these<br />

routes can be overwhelm<strong>in</strong>gly heavy, especially <strong>in</strong><br />

highly dynamic <strong>VANET</strong>s.<br />

The load carry<strong>in</strong>g capacity <strong>of</strong> <strong>AODV</strong> is much better<br />

than proactive rout<strong>in</strong>g protocols like DSDV, OLSR etc.<br />

thus <strong>AODV</strong> is preferred for this study.<br />

3.3. Hybrid Rout<strong>in</strong>g<br />

Hybrid rout<strong>in</strong>g comb<strong>in</strong>es characteristics <strong>of</strong> both<br />

reactive <strong>and</strong> proactive rout<strong>in</strong>g protocols to make<br />

rout<strong>in</strong>g more scalable <strong>and</strong> efficient [9]. Mostly hybrid<br />

rout<strong>in</strong>g protocols are zone based; it means the number<br />

<strong>of</strong> nodes is divided <strong>in</strong>to different zones to make route<br />

discovery <strong>and</strong> ma<strong>in</strong>tenance more reliable for MANETs<br />

or <strong>VANET</strong>s.<br />

The most recently developed <strong>ADV</strong> hybrid rout<strong>in</strong>g<br />

protocol starts with DSDV proactive rout<strong>in</strong>g approach<br />

by attach<strong>in</strong>g sequence numbers to rout<strong>in</strong>g entries <strong>and</strong><br />

then gradually shifts to on-dem<strong>and</strong> approach <strong>in</strong> order to<br />

reduce the overhead related with proactive approach.<br />

This feature is achieved us<strong>in</strong>g the follow<strong>in</strong>g dual<br />

strategy:-<br />

1. Vary<strong>in</strong>g the number <strong>of</strong> active routes ma<strong>in</strong>ta<strong>in</strong>ed:-<br />

This is achieved by advertis<strong>in</strong>g <strong>and</strong> ma<strong>in</strong>ta<strong>in</strong><strong>in</strong>g<br />

routes for active receivers only, which are receivers<br />

<strong>of</strong> any currently active connection.<br />

2. Vary<strong>in</strong>g the frequency <strong>of</strong> rout<strong>in</strong>g updates: -<br />

Accord<strong>in</strong>g to this approach a node should trigger an<br />

update under three conditions only:-<br />

1. if it has some buffered data packets due to lack <strong>of</strong><br />

route.<br />

2. if one or more <strong>of</strong> its neighbors make a request for<br />

fresh routes it is a forward<strong>in</strong>g node that <strong>in</strong>tends<br />

to advertise any fresh valid/<strong>in</strong>valid route to the<br />

dest<strong>in</strong>ation so as to keep the route fresh.<br />

4. Research Methodology Used<br />

To carry out the experiment discussed <strong>in</strong> this paper<br />

NCTUns-6.0 simulation tool is used. The scenarios<br />

used for analysis, simulation setup, performance<br />

metrics used for mak<strong>in</strong>g various comparisons are<br />

discussed <strong>in</strong> this section.<br />

4.1 Simulation Tool Used<br />

In order to carry out a simulation work for vehicular<br />

networks two basic simulator types are required<br />

namely-network simulators <strong>and</strong> traffic simulator. But <strong>in</strong><br />

this study a hybrid simulator is used which provides an<br />

<strong>in</strong>tegration <strong>of</strong> both network <strong>and</strong> traffic simulator.<br />

The hybrid simulator used is NCTUns-6.0(National<br />

Chiao Tung University Network Simulator) which is<br />

the latest version <strong>and</strong> whose core technology is based<br />

on the novel kernel re-enter<strong>in</strong>g methodology <strong>in</strong>vented<br />

by Pr<strong>of</strong>. S.Y. Wang [13]. The various features <strong>of</strong><br />

<strong>VANET</strong> supported by NCTUns-6.0 makes it an<br />

obvious choice for this study.<br />

Figure 3: Strength <strong>of</strong> Traffic, <strong>VANET</strong>, <strong>and</strong> Ns-2 [3]<br />

4.2. <strong>Performance</strong> Metrics<br />

For this study three performance metrics are<br />

selected namely:-<br />

1. Throughput: - Throughput describes as the total<br />

number <strong>of</strong> received packets at the dest<strong>in</strong>ation out <strong>of</strong><br />

total transmitted packets [14].Throughput is calculated<br />

<strong>in</strong> bytes/sec or data packets per second.<br />

Total number <strong>of</strong> received packets at dest<strong>in</strong>ation* packet size<br />

T = ------------------------------------------------------------<br />

Total simulation time<br />

2. Packet Drop:-It shows total number <strong>of</strong> data packets<br />

that could not reach dest<strong>in</strong>ation successfully. The<br />

IJCTA | JAN-FEB 2012<br />

Available onl<strong>in</strong>e@www.ijcta.com<br />

52

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!