• Share
  • Email
  • Embed
  • Like
  • Save
  • Private Content
Jh2516461655
 

Jh2516461655

on

  • 97 views

 

Statistics

Views

Total Views
97
Views on SlideShare
97
Embed Views
0

Actions

Likes
0
Downloads
1
Comments
0

0 Embeds 0

No embeds

Accessibility

Categories

Upload Details

Uploaded via as Adobe PDF

Usage Rights

© All Rights Reserved

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Processing…
Post Comment
Edit your comment

    Jh2516461655 Jh2516461655 Document Transcript

    • Sharnjeet Kaur, Dr. Gurpreet Singh Josan / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 5, September- October 2012, pp.1646-1655Performance Evaluation Of Topology Based Routing Protocols In Vanet Sharnjeet Kaur*, Dr. Gurpreet Singh Josan** *(Department of Computer Science, Punjabi University, Patiala) ** (Assistant Professor, Department of Computer Science, Punjabi University, Patiala)ABSTRACT Vehicular Ad-hoc Networks (VANET) telecommunication and computer networks beganare formed between moving vehicles equipped with a strong emphasis on wires, both for thewith wireless interfaces, which are attracting a communications infrastructure and for the last hopgreat deal of interest. VANET is a new where the actual connection towards the userscommunication paradigm that enables the terminals takes place [1]. In the last decade this trendcommunication between mobile nodes having has shifted towards wireless networks, especially atdynamic topology on roads. There are still the user side.VANET (Vehicular Ad-hoc Network)several areas of VANETS, such as congestion is a new technology that has to be taken enormouscontrol, security, traffic engineering, traffic attention in the recent years. Due to rapid change inmanagement, dissemination of emergency topology and frequent disconnection makes itinformation to avoid hazardous situations and difficult to design an efficient routing protocol forrouting protocols, which lack large amounts of routing data among vehicles, called V2V or vehicleresearch. Here in this research we have to use to vehicle communication" [2]. VANET is aOMNeT++ i.e. freely available simulator is used technology that uses moving cars as nodes in awith traffic simulator (SUMO) that are uses the network to create a mobile network. It turns everyTraCI (Traffic control interface) module to participating car into a wireless router or node,couple the simulators which works in sync. It uses allowing cars approximately 100 to 300 meters ofthe UDP Basic Burst Notification application. In each other to connect and, in turn, create a networkthis paper basically to evaluate the proactive and with a wide range. As cars fall out of the signalreactive routing protocols that are commonly range and drop out of the network, other cars canused in mobile ad-hoc networks, which will apply join in, connecting vehicles to one another so that ato VANETs. Optimized Link State Routing Mobile Internet is created. It is estimated that the(OLSR), Dynamic Source Routing (DSR) and first systems that will integrate this technology areDynamic MANET On-demand (DYMO) are police and fire vehicles to communicate with eachinitially simulated in a city, main road, and other for safety purposes. Other purposes includecountry environment in order to provide an essential alerts and accessing comforts andoverall evaluation. entertainment. VANETs are a kind of MANETs provide vehicle to vehicle (V2V) and vehicle toKeywords: OLSR, DSR, DYMO, VANETs, roadside wireless communications, this means thatOMNeT++, SUMO, IEEE 802.11b. every node can move freely within n/w coverage and stay connected. Vehicles are equipped with wireless1. INTRODUCTION transceivers and computerized control modules are A vehicular network is a kind of wireless used.networks that has emerged thanks to advances inwireless technologies and the automotive industry[1]. Vehicular networks are formed between movingvehicles equipped with wireless interfaces that couldbe of homogeneous or heterogeneous technologies.These networks, also known as VANETs (VehicularAd-hoc Networks) are considered as one of the ad-hoc network real-life applications, enablingcommunications among nearby vehicles as well asbetween vehicles and nearby fixed equipment(roadside equipment). The development of communicationnetworks was a significant step for mankind,undoubtedly facilitating everydays tasks andimproving the quality of life. Both Figure 1: Vehicular Ad hoc Network [3] 1646 | P a g e
    • Sharnjeet Kaur, Dr. Gurpreet Singh Josan / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 5, September- October 2012, pp.1646-1655 receiver is an important factor; the more theVANETs integrates multiple Ad hoc networking distance, the smaller the probability of reception oftechnologies such as Wi-Fi IEEE 802.11 b/g, packets. In infrastructure-based technologies,WiMAX 802.16, Bluetooth, IRA, ZigBee for easy, handoff between base stations is also relevant [5].accurate, effective and simple communication  Permanent access: Permanent access tobetween vehicles on dynamic mobility[3]. the network is one of the main drawbacks of vehicular communications. In VANET designs, a Vehicular networks consist of large no. of physical infrastructure is not necessary, due to thenodes, approximately no. of vehicles exceeding 750 inherent decentralized design.million in the world today, these vehicles will  Location Awareness: Next generationrequire an authority to govern it, and each vehicle vehicles are expected to exchange information notcan communicate with other vehicles using short only beyond their immediate surroundings and line-radio signals of 2.4 GHz at bit rate of 11Mbps. This of-sight with other vehicles, but also with the roadcommunication is an Ad Hoc communication that infrastructure and Internet databases [6].means each connected node can move freely, no  Time Awareness: Vehicular applicationswires required, the routers used are called road side often require a reliable communication channel thatunit (RSU). supports time-critical message transmissions.  Penetration rate dependency: This rate is2. BACKGROUND defined as the percentage of vehicles equipped with The growing mobility of people and goods the necessary on board data unit (OBU) on the road.incurs in high social costs: traffic congestion,fatalities and injuries. Around the globe each year 3. ROUTING PROTOCOLS IN VANETabout 1.2 million people die because of traffic Routing is the act of moving informationaccidents [4]. The traffic accidents place as the across an internetwork from a source to afourth cause of mortality by this statistics in the destination. Along the way, at least one intermediateworld. Also, this high number of fatalities and node typically is encountered. Routing occurs atinjuries high healthcare costs, more than any other Layer 3 (network layer) of the OSI model. Thetype of injury or disease. routing protocols are further divided into number of categories but here focus onto the OLSR, DSR, It is in this context that Vehicular Ad-hoc DYMO protocols mainly, that are evaluated on theNetworks (VANETs) have emerged. A VANET is behalf of throughput and latency.based on smart cars and base-stations, which shareinformation via wireless communications. This The topology based routing protocols areinterchange of data may have a great impact on further divided into two different categories for ad-safety and driving, reducing the number of accidents hoc data networks, according to [7]: Proactive andand helping to optimize transport. While the original reactive.motivation for VANETs was to promote trafficsafety, recently it has also become increasingly The first is a proactive routing protocol,obvious that VANETs open new vistas for Internet which relies on the periodic broadcast of dataaccess, distributed gaming, and the fast-growing network topology. Popular proactive protocol ismobile entertainment industry. The importance and OLSR (Optimized Link State Routing) and FSRpotential impact of VANETs have been confirmed (fisheye state routing). The second category,by the rapid proliferation of consortia involving car reactive routing protocols, can be viewed as amanufacturers. solution to proactive routing protocols because they only search for a route when one is needed. Some2.1 Factors that affect communication in VANET popular reactive protocols are DSR and DYMO. High velocity of the vehicles. Environment factors: obstacles, tunnels, 3.1 Optimized Link State Routing (OLSR)traffic jams, etc. Protocol Determined mobility patterns that depend OLSR is the proactive routing protocol thaton source to destination path and on traffic is evaluated in this paper. OLSR achieved RFCconditions. status in 2003 (T. Clausen (Ed.), and P. Jacquet (Ed.) High congestion channels (e.g. due to high Oct. 2003) [8]. Basically OLSR is an optimization ofdensity of nodes). the classical link state algorithm adapted for the use in wireless ad hoc networks. First, few nodes are 2.2 Network requirements in VANETs selected as Multipoint Relays (MPRs) to broadcastapplications the messages during the flooding process. Mobility: Wireless network technologiesallow devices to move freely. In 802.11 Second level of optimization is achieved bytransmissions the distance between the sender and using only MPRs to generate link state information. 1647 | P a g e
    • Sharnjeet Kaur, Dr. Gurpreet Singh Josan / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 5, September- October 2012, pp.1646-1655This results in minimizing the “number” of control further, To avoid the possibility of infinite routemessages flooded in the network. As a final level of discoveries, node E will piggyback the route replyoptimization, an MPR can chose to report only links on the new route request [9].between itself and those nodes which have selectedit as their MPR. MPRs play a major role in the In the event that a node cannot successfullyfunctionality of the protocol[8]. transmit a packet to the next hop, it needs to perform route maintenance. When forwarding packets along OLSR is designed to support large and a source route, each node on the route is responsibledense wireless networks. It is also suitable for for the successful transmission of the packet to thescenarios, where the communicating pairs change next hop. This reply can either be done by using anover time. Once the communicating pair changes, a existing part of the MAC protocol in use or doneroute to new pair is readily available, and no control passively [10]. In the Figure 2, node B can confirmtraffic or route discovery process is needed as in the the receipt of the packet to node C by listening to seecase of reactive protocols. This can be beneficial for if node C tries to forward the packet again. A routesituations where time critical or safety related data error message is sent out in the event that a nodeneeds to be delivered with minimum possible delay. does not receive a successful transmission.3.2 Dynamic Source Routing 3.3 Dynamic MANET On-Demand Routing Dynamic Source Routing (DSR) is another (DYMO)routing protocol that was specifically designed for The final routing protocol that was used foruse in multi-hop mobile ad-hoc networks [9]. Like evaluation is the Dynamic MANET On-DemandOLSR, DSR is a completely self organizing (DYMO) routing protocol [11]. The DYMO routingprotocol. protocol is another protocol that is designed for useIt has two mechanisms: in mobile wireless ad-hoc networks. Unlike the work in [12], this implementation will be integrated right into the network layer and not as part of the application layer. Just like DSR, DYMO consists of two main operations: Route Discovery and Route Maintenance.Figure 2: Route Discovery for DSR Routing DYMO route discovery is performedAlgorithm similar to the DSR routing algorithms. When a node• Route Discovery: The process of discovering a needs to send out a packet to another node it willroute from a source to a destination. first search its route cache or routing table to see if• Route Maintenance: Allows for the topology of an up to date route exists. If one does, the sourcethe network to change and a nodes routing table to node uses that route to send the packet to itsremain fresh. DSR does not use any type of periodic destination. However, if a route does not exist thepackets or messages at any level. The purely on node must go through a process to find a path to thedemand behaviour of the DSR routing algorithm destination, called route discovery. The source nodeallows it to cut down network overhead that do use it creates a route request (RREQ) message to send out[9]. to all neighbouring nodes. The RREQ contains the following information [13]: Route discovery is the process that the DSR • Destination Addressalgorithm uses to find a route to send a packet from • Sequence Numbersource to destination. When no route is present the • Hope Countsource node transmits a route request (RREQ). Each • Next Hopnode broadcasts the message until it reaches the • Valid Timeoutdestination. Once at the destination node, that node • Delete Timeoutwill send back a route reply (RREP) to the source. The source node will then send out theAs shown in the Figure 2 node A wishes to send a RREQ via broadcast to all of the surrounding nodes.packet to node E. Node A initiates a route discovery The receiving node will look at the packet to makewith an RREQ that contains node A as the initiator, sure that it has not seen it before and if it has thean empty route record list, and a unique request ID. packet will be discarded. If it has not been seenAs each node broadcasts the request to all nodes before, the node will then start to look at thewithin range, the route is built because each node information contained inside of the RREQ. Lastly, ifappends itself to the route record in the RREQ. the sequence number indicates there is new information in the RREQ then the data in the routing Once the route request reaches the table is updated and the RREQ is passed on. Onceintended destination, the destination node will send a the RREQ reaches the destination node, that noderoute reply back to the initiator. Node E request 1648 | P a g e
    • Sharnjeet Kaur, Dr. Gurpreet Singh Josan / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 5, September- October 2012, pp.1646-1655will then form a route reply (RREP) that contains the created in order to fully simulate a VANET. Thenew route and is sent back along the reverse path. wireless network was setup to be based on the IEEE802.11b standard, which is commonly used in4. IMPLIMENTATION FRAMEWORK VANET simulations [19]. The maps for the traffic4.1 OMNeT++ simulation consisted of three different environments OMNeT++ [14] is an open-source (city, main road, country). The purpose of usingsimulation environment. The primary simulation multiple traffic environments was to expose theapplications are Internet simulations, mobility, and routing protocols to a variety of topologies ratherad-hoc simulations. This simulator has a component- than just one. Figure 3 shows an example of how thebased design, meaning that new features and client and server interact.protocols can be Performance evaluation ofvehicular ad-hoc networks over high speedenvironment supported through modules. In order to provide large scale simulationswith reusable models, OMNet++ uses modules todevelop the different components of the simulation.Simple modules are the most basic modules as theyprovide extremely basic functionality. Compoundmodules are created by grouping simple modulestogether to create an object with a complexfunctionality, such as a vehicle equipped with anIEEE802.11b radio. Modules in Omnet++ areconnected to each other’s input and output gates Figure 3: TraCI Connection Examplewith the use of simple ’connection’ modules. Theseconnections are all defined in a file that uses the 4.4 simulationsNED language. NED models are reusable and The simulations were run in different sets.designed to work with other NED files to create For each set of simulation runs there was up to 10much larger models. A nice feature of Omnet++ is TxRx pairs. The number of TxRx pairs was variedthat it uses a two way editor to create and modify the between one and ten sync pairs. Also, the trafficNED files that make up the environment. density change. The only difference between each We use a text editor or the IDE’s graphical set was the number of transmit and receive (TxRx)editor to create the network. Modules in the network pairs that existed in the network. The reason to varycontain a lot of unassigned parameters, which need the number of TxRx pairs is to create more datato be assigned before the simulation can be run. The network traffic that would have in impact on howname of the network to be simulated, parameter quickly a message could get delivered betweenvalues and other configuration option need to be TxRx pairs. The final parameter that changes in thespecified in the omnetpp.ini file. simulations was the density of the traffic, which varied from high density (460 vehicles) to medium4.2 Traffic Simulation density (250 vehicles) and to low density (165 Simulation of Urban Mobility (SUMO) is a vehicles).C++ application developed to simulate the From each simulation, the averagemovement of objects along a road network. It is a throughput and latency were recorded forfree and open sourced simulator. Along with being comparison of the protocols. This will allow for anable to model small areas, SUMO is also capable of in-depth analysis of the strengths and weaknesses ofmodelling traffic in large networks, such as cities or the routing protocols based on scenario and traffichighway networks, without any changes. SUMO density.simulations are considered to be multimodal,meaning that every object in the simulation is 5. RESULTS AND ANALYSISsimulated [16]. The routing protocols used for evaluation are OLSR, DSR, and the DYMO protocols. All these4.3 Simulator Coupling protocols are designed by various technologies; To accurately model a VANET, Omnet++ however, they are all designed for mobile ad-hocand SUMO are connected with a technique to networks that have to play an important role within asynchronize node movement between two VANET. Tables 1, 2, 3 consists the evaluated valuessimulators. The Traffic Control Interface (TraCI) is of throughput using three different environments.used to couple the simulators. VANET simulationapproaches used mobility traces that the network 5.1 Throughputsimulator read in [17, 18]. TraCI works in a client- The throughput is calculated by monitoringserver manner [17]. A wireless network, traffic map, the channel that determines the speed at which theand obstacles in the wireless environment were packets are successfully transmitted by the 1649 | P a g e
    • Sharnjeet Kaur, Dr. Gurpreet Singh Josan / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 5, September- October 2012, pp.1646-1655application. The results for each simulation scenario 1 3 0 50 2 0 98 2 0 731are shown in figures 4(A, B, C), 5(A, B, C) and 6. 3 2.9 1 9.9 0 .99High Traffic Density- Throughput (bits/sec) 3 5 0 5 5 OLSR DSR DYMO 8. 5. 4.Tx Ci C Ma Ci C Ma Ci C Ma 2 2 4 3 5 1Rx ty ou in ty ou in ty ou inpai nt Ro nt Ro nt Ro 2 3 0 50 2 0 49 2 0 652r ry ad ry ad ry ad 0 4.1 7 2.5 4 .981 23 38 58 18 38 67 19 19 34 5 5 1 5 9 7. 5. 1. 39 5. 2.5 51 0. 1.9 23 0. 5.8 9 3 5 .6 25 2 .2 55 5 .1 15 5 5 1 4 5 3 1 3 2 0 73 2 0 52 2 0 9052 19 36 66 22 14 76 20 15 36 5 5.1 3 3.4 1 .15 99 5. 5.5 65 5. 5.0 15 6. 6.5 5 5 2 5 4 .3 15 9 .7 52 2 .5 46 5 6 9 5 0. 5. 5. 4 1 73 19 31 86 18 39 11 13 13 57 2 2 4 35 3. 9.5 15 9. 65. 65 7. 7.3 4 2 0 95 1 0 47 1 0 902 .4 55 3 .3 77 55 .6 28 7 5 1 5 0 9.2 7 5.6 5 .56 8 3 9 5 44 13 90 67 22 18 46 18 38 12 5. 1. 2. 95 .0 8.4 19 9. 5.4 45 5. 35. 6 2 1 .6 3 6 .7 18 5 .9 65 56 5 6 2 1 8 9 5 1 0 66 1 0 53 1 0 8825 16 36 75 17 45 10 13 16 71 7 5.5 6 5.3 5 .11 85 6. 3.8 45 8. 36. 22 5. 2.9 5 5 0 2 5 .4 55 5 .3 95 35 .1 23 8 4. 1. 6. 5 5 1 1 26 17 31 76 19 20 93 12 27 11 2 2 5 35 3. 5.1 01 1. 6.9 53 5. 20. 6 1 0 43 1 0 42 1 0 965 .0 18 8 .4 28 9 .9 12 35 8 8.6 1 3.5 3 .95 5 3 2 3 5 3 4 67 16 17 46 14 41 93 11 29 12 9. 9. 5. 65 0. 5.8 63 4. 5.6 17 4. 88. 1 2 2 .3 52 5 .2 65 5 .5 12 12 5 1 5 5 1 3 7 2 0 54 1 0 56 1 0 1028 14 33 66 18 30 10 12 30 14 0 5.4 2 5.1 1 9.1 86 3. 5.2 35 9. 12. 67 8. 38. 6 6 3 2 6 2 .3 65 3 .2 65 96 .5 35 32 1. 8. 9. 8 5 6 0 1 29 16 23 82 18 38 10 13 33 15 1 2 4 81 2. 3.8 15 9. 55. 05 4. 85. 8 1 0 60 1 0 49 1 0 108 .8 68 4 .9 95 11 .2 02 05 5 2.4 1 8.5 2 5.1 9 8 3 9 5 8 5 8 610 13 19 82 17 32 10 13 34 17 8. 1. 8. 58 5. 2.5 80 0. 98. 85 4. 51. 5 3 6 .4 12 5 .9 31 53 .9 32 95 6 5 5 8 1 9 9 1 0 69 1 0 48 7 0 113Table 1 Throughput results for high traffic density 4 5.3 1 4.1 8 5.4scenario 0 4 7 1 5. 6 5. 4. 1Medium Traffic Density- Throughput (bits/sec) 3 1 6 OLSR DSR DYMO 5 5Tx Ci C Ma Ci C Ma Ci C 10 1 0 73 1 0 51 5 0 118Rx ty ou in ty ou in ty ou Ma 4 5.2 3 2.5 7 5.9pai nt Ro nt Ro nt in 7 4 5 6 5. 6r ry ad ry ad ry Ro 8. 8. 5 ad 5 1 3 5 4 1650 | P a g e
    • Sharnjeet Kaur, Dr. Gurpreet Singh Josan / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 5, September- October 2012, pp.1646-1655Table 2 Throughput results for Medium trafficdensity scenarioLow Traffic Density- Throughput (bits/sec) OLSR DSR DYMOTx Ci C Ma Ci C Ma Ci C MaRx ty ou in ty ou in ty ou inpai nt Ro nt Ro nt Ror ry ad ry ad ry ad1 18 0 23 17 0 44 17 0 12 23 8.6 85 5.1 05 35. (A) .1 5 .1 3 .5 74 2 6 52 15 0 52 15 0 97 16 0 37 28 8.1 28 7.2 62 4.1 .9 3 .9 3 .8 4 9 9 53 11 0 61 24 0 80 11 0 90 85 8.6 35 3.9 55 5.6 .1 5 .1 9 .1 5 5 5 54 16 0 55 13 0 46 24 0 43 03 1.1 45 2.1 56 5.1 (B) .0 5 .2 2 .1 2 2 9 45 11 0 70 15 0 59 23 0 66 12 6.9 39 8.1 12 8.6 .0 9 .1 6 .0 5 5 2 26 16 0 79 15 0 59 24 0 70 99 1.1 23 2.4 65 5.1 .1 5 .1 4 .5 4 1 4 87 14 0 87 14 0 57 18 0 71 85 5.1 55 2.9 63 6.2 (C) .9 1 .1 9 .6 5 9 2 4 Figure 4(A, B, C): Throughput in bits/second for8 11 0 88 17 0 55 16 0 68 city environment 31 8.0 35 5.7 42 5.6 .0 1 .6 2 .9 5 2 1 99 93 0 88 13 0 52 16 0 64 2. 2.1 21 3.9 31 3.5 13 6 .1 6 .1 6 2 410 85 0 86 12 0 51 15 0 70 5. 5.1 53 2.1 99 6.1 16 5 .1 3 .1 8 4 7 (A)Table 3 Throughput results for Low traffic densityscenario (B) 1651 | P a g e
    • Sharnjeet Kaur, Dr. Gurpreet Singh Josan / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 5, September- October 2012, pp.1646-1655 The graphs in figure 5(A, B, C) show the results obtained from the high, medium, and low density main road simulations. The DYMO protocol shows as the number of TxRx pairs increases the throughput is increases. As to increase in number of TxRx pairs that cause to increase the number of nodes communicate with each other, which cause to increase the chances to make the new routes. When the less dense spacing of the vehicles, OLSR actually gains throughput as the number of TxRx pairs increase.(C) 5.2 Latency ResultsFigure 5(A, B, C): Throughput in bits/second for The latency results are important formain road environment applications that are time sensitive, such as collision avoidance or emergency vehicle warning. These latency measurements are calculated by determining the time between the message send by the sender and to receive by the receive node. Tables 4, 5, 6 consist the values of latency by using three different environments. High Traffic Density-Latency (s) OLSR DSR DYMO Tx C C Ma C C Ma C CFigure 6: Throughput in bits/second for Rx it ou in it ou in it ou MaiCountry environment pai y nt Ro y nt Ro y nt n r ry ad ry ad ry Roa The DSR routing protocol is to prove the dhigher throughput transmitting protocol in the city 1 0 1. 0.8 3 10 23. 0 0. 0.09simulations of high traffic density. The high traffic . 53 66 . .0 24 . 00 1density simulations in the country environment results 1 8 9 22 9 0 4in some communication between the nodes. The 7 2 7DYMO protocol remains the most stable throughout 1 4 5the all 10 TxRx pairs, rather the DSR routing protocol 2 0 0. 0.7 1 14 14. 0 0. 0.00has the highest throughput with 5 TxRx pairs. OLSR . 00 83 . .0 85 . 02 8do not remain as stable as DYMO because over the 10 5 5 5 54 6 0 5simulations the throughput oscillations from high to 3 2 0low. The high traffic density city simulations, the all 1 2 5three routing protocols with only one transmit receive 3 0 0. 0.8 1 9. 15. 0 4. 0.04pair started with the throughput of 1850 bits/sec to . 01 78 . 71 02 . 24 42350 bits/sec. The OLSR routing protocol had the 3 5 8 7 5 0 5most drastic change, about 100-150bits/s, as the 0 8 0number of TxRx pairs were increased. The DSR 5 8 3routing protocol remains stable with throughput 4 0 0. 2.0 2 20 11. 0 0. 0.00between 1700 and 2250 bits/s. The DYMO protocol . 06 51 . .1 52 . 00 8also shows the dip in throughput as on the OLSR that 1 6 0 05 8 0 6are of 900bits/sec drop. The similar effects show the 6 2 0medium traffic density environment for each of the 2 5 2routing protocols. In figure 4 look at the second line 5 0 0. 1.0 2 2. 8.9 0 0. 0.01graph, each of the routing protocols with the single . 00 12 . 55 03 . 00 1TxRx pair started at a high throughput between 3 5 5 4 0 42000bits/sec to 3500 bits/sec and finished with 10 0 5 6TxRx pairs at a level that are between 400 and 1600 2 4 5bit/s. DSR showed the lower loss in throughput for 6 0 0. 0.4 2 8. 5.1 0 0. 0.00the 10 simulations. The DYMO protocol has an . 85 81 . 59 42 . 05 9increase in throughput from six to seven TxRx pairs. 3 3 8 5 0 5However, the throughput drops to similar level as the 5 1 2number of TxRx pairs increased. 5 4 5 1652 | P a g e
    • Sharnjeet Kaur, Dr. Gurpreet Singh Josan / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 5, September- October 2012, pp.1646-16557 0 0. 1.0 3 7. 6.4 0 0. 0.14 5 6 1 . 05 28 . 55 65 . 89 5 6 6 7 1 8 1 1 0 1 9 0. 0 0.6 5. 0 4.1 0. 0 0.0 8 3 2 7 16 1 88 0 61 8 8 4 8 6 18 0 0. 0.9 2 6. 7.2 0 0. 0.07 8 4 7 . 65 99 . 80 13 . 15 4 10 0. 0 0.5 3. 0 3.9 0. 0 0.0 6 8 2 9 0 6 6 95 5 15 0 63 5 1 4 5 4 1 5 6 5 4 5 89 0 0. 0.7 2 5. 5.4 0 0. 0.04 Table 5 Latency results for Medium traffic density . 58 35 . 76 85 . 06 5 scenario 6 1 7 1 0 5 6 8 2 Low Traffic Density-Latency (s) 1 5 8 OLSR DSR DYMO10 0 0. 0.6 3 8. 5.9 0 0. 0.05 . 65 95 . 42 95 . 12 4 Tx Ci C Ma Ci C Ma Ci C Ma 7 3 1 4 0 2 Rx ty ou in ty ou in ty ou in 7 3 3 pai nt Ro nt Ro nt Ro 3 6 8 r ry ad ry ad ry adTable 4 Latency results for high traffic density 1 0. 0 0.0 3. 0 16. 0. 0 0.0scenario 06 05 62 78 00 02 25 81 98 00 45 25 2 1 1 2Medium Traffic Density- Latency (s) 2 0. 0 0.4 3. 0 10. 0. 0 0.0 OLSR DSR DYMO 34 88 75 12 00 06Tx C C Ma C C Ma C C Ma 07 91 99 16 15 89Rx it ou in it ou in it ou in 5 9 3 6pai y ntr Ro y nt Ro y nt Ro 3 0. 0 0.2 3. 0 7.9 0. 0 0.0r y ad ry ad ry ad 67 78 45 06 00 041 0. 0 1.1 3. 0 6.4 0. 0 0.0 04 82 60 05 50 28 2 45 6 45 0 93 5 1 1 5 6 0 4 0. 0 0.5 3. 0 8.2 0. 0 0.0 1 5 4 53 24 06 36 00 132 0. 0 0.4 1. 0 14. 0. 0 0.0 53 15 55 01 26 52 0 55 9 877 0 05 5 2 5 6 4 0 5 0. 0 0.8 3. 0 7.1 0. 0 0.0 5 5 2 35 32 21 25 00 043 0. 0 0.5 2. 0 4.5 0. 0 0.0 86 95 89 52 83 28 3 52 9 08 0 12 5 2 4 1 1 3 6 0. 0 0.7 2. 0 6.5 0. 0 0.0 5 2 8 46 57 96 78 00 094 0. 0 1.0 5. 0 8.1 0. 0 0.0 55 61 50 12 65 55 5 65 7 75 0 18 75 2 1 5 3 0 7 0. 0 0.9 2. 0 6.5 0. 0 0.0 1 5 9 28 45 08 67 00 105 0. 0 3.2 4. 0 3.9 0. 0 0.1 70 95 55 14 70 56 2 95 8 95 0 06 2 1 5 9 0 0 8 0. 0 1.0 2. 0 6.1 0. 0 0.0 5 9 6 28 65 65 86 00 116 0. 0 1.0 2. 0 5.0 0. 0 0.0 05 99 39 42 75 57 2 65 6 05 0 53 02 9 8 6 3 1 9 0. 0 1.1 2. 0 5.7 0. 0 0.0 5 1 5 42 91 45 87 00 127 0. 0 0.6 4. 0 6.8 0. 0 0.0 82 00 34 86 81 67 2 12 8 22 0 55 05 2 5 3 0 1 1 10 0. 0 1.3 2. 0 5.4 0. 0 0.0 1 6 6 43 21 32 07 00 138 0. 0 0.5 4. 0 4.4 0. 0 0.0 95 56 29 02 86 75 9 98 1 58 0 58 65 5 1 1653 | P a g e
    • Sharnjeet Kaur, Dr. Gurpreet Singh Josan / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 5, September- October 2012, pp.1646-1655Table 6 Latency results for low traffic densityscenarios (b)(a) (c) Figure 8(a, b, c): Latency in seconds for main road environment(b) Figure 9: Latency in seconds for country environment(c) In figures 7(a, b, c), 8(a, b, c), and 9 shows theFigure 7(a, b, c): Latency in seconds for city results of latency in different traffic environmentsenvironment with multiple TxRx pairs. The most noticeable thing in all plots the DSR routing protocol has the highest latency between 5 to 10 seconds in all simulations, and the DYMO having the lowest latency between 0.005 to 0.1 seconds. The city environment resulted in some of the lowest latencies for DSR that ranged between 2 seconds and 7 seconds. The main road and country scenarios are the environments that resulted in much higher latencies. For each of the main road environments, when the TxRx pair’s remains low the DSR routing protocol has the highest latency. The OLSR is the second lowest latency protocol under 3 to 3.5.(a) 6. CONCLUSION AND FUTURE WORK In this conclusion, the DYMO routing protocol is to be the best choice for a routing protocol because of its very low latencies and 1654 | P a g e
    • Sharnjeet Kaur, Dr. Gurpreet Singh Josan / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 www.ijera.com Vol. 2, Issue 5, September- October 2012, pp.1646-1655throughput comparable to other protocols. The Multi-Hop Wireless Ad Hoc Networks,”second choice is the OLSR routing protocol is also Discovery, vol. 5, pp. 1–25, 2001.good because its average latency values are [10] J. Jubin and J. Tornow, “The DARPAreasonable even though higher than DYMO but are packet radio network protocols,”the accepted range. In the end it is concluded that Proceedings of the IEEE, vol. 75, no. 1, pp.traditional approach of using proactive routing 21–32, 1987.protocols in VANETs is not justifiable as reactive [11] P. C. Chakers I., “Dynamic MANET On-routing protocols have performed better than demand Routing (DYMO),” no. 21, 2010.proactive routing protocols in variety of scenarios. [Online] Available: http://tools.ietf.org/id/draft-ietf-manet- To create a wider variety of traffic dymo-21.txtenvironments for simulations is the other possibility [12] C. Sommer, I. Dietrich, and F. Dressler,for future work. Even though a city, a main road and “Simulation of Ad Hoc Routing Protocola country environment are simulated, not all traffic using OMNeT++,” Mobile Networks andenvironments are the same as defined for these Applications, pp. 786-801, Jun. 2009.simulations. Other future work could expand upon [13] S. Wang, “Implementing and evaluatingthis research by constructing multiple VANET three routing protocols in dual-radio-dualsystems that could be placed into a vehicle and mode IEEE 802.11(b) wireless meshtested on real roads. networks,” Computer Communications, vol. 31, no. 10, pp. 2596–2606, Jun. 2008.REFERENCES [14] Hassan, A., VANET simulation, Masters [1] H.F¨ußler, M. Torrent-Moreno, M. Transier Thesis in Electrical Engineering, and H. Hartenstein. "Thoughts on a Halmstad University, Sweden, 2009. Protocol Architecture for Vehicular Ad- [15] A. Varga, “THE OMNET ++ DISCRETE Hoc Networks", In Proc. of WIT 2005, EVENT SIMULATION SYSTEM,” The pages 41–45, Hamburg, Germany, March European Simulation Conference ESM 39, 2005 pp. 319–324, 2010. [2] Vehicular ad-hoc network, [16] D. Krajzewicz, et. al. “The ”Simulation of [Online]Available: Urban MObility” package : An open source http://en.wikipedia.org/wiki/Intelligent_veh traffic simulation,”European Simulation icular_ad-hoc_network and Modelling Conference, 2003. [3] Vehicular network and its applications, [Online].Available:http://elib- [Online]Available: .dlr.de/21385/1/dkrajzew ESMc2003.pdf http://ers.hclblogs.com/2011/03/vehicular- [17] A. K. Saha and D. B. Johnson, “Modelling networking- and-its-applications/ mobility for vehicular ad-hoc networks,” [4] Real Automóvil Club de España, Proceedings of the first ACM workshop on http://www.race.es/seguridad_vial/estudios Vehicular ad hoc networks - VANET ’04, p. _informes/datos_siniestralidad/ 91, 2004. [5] Kevin C. Lee, Uichin Lee, Mario Gerla, [18] D. R. Choffnes and F. E. Bustamante, “An “Survey of Routing Protocols in Vehicular integrated mobility and traffic model for Ad Hoc Networks”, Advances in Vehicular vehicular wireless networks,” Proceedings Ad-Hoc Networks: Developments and of the 2nd ACM international workshop on Challenges, IGI Global, Oct, 2009. Vehicular ad hoc networks - VANET ’05, p. [6] Suresh Kumar, K. D. Narayan, Jogendra 69, 2005. Kumar,“Qualitative Based Comparison of [19] A. L. Santos and F. Milagro, “Experimental Routing Protocols for VANET” Journal of demonstration of the viability of IEEE,” Information Engineering and Applications, Proceedings of the 4th International ISSN 2224-5758, Vol 1, No.4, 2011 Conference on Test beds and research [7] C. Sommer and I. Dietrich, “Simulation of infrastructures for the development of Ad Hoc Routing Protocols using networks communities, 2008. OMNeT++ A Case Study for the DYMO Protocol,” Springer Science Business, 2009. [8] T. Clausen (Ed.), and P. Jacquet (Ed.), “Optimized Link State Routing Protocol (OLSR)” (RFC 3626), Oct. 2003. [Online] Available: http://www.ietf.org/rfc/rfc3626.txt [9] D. B. Johnson and D. A. Maltz, “DSR : The Dynamic Source Routing Protocol for 1655 | P a g e