Research on performance of routing protocols in manet

Australian Journal of Asian Country Studies
SCIE Journals
Australian Society for Commerce Industry & Engineering
www.scie.org.au
72
Research on Performance of Routing Protocols in MANET
Chia-Sheng Tsai 1* Chih-Yang Lu 1
Department of Computer Science and Engineering, Tatung University, Taipei, Taiwan
* E-mail of the corresponding author: icstsai@gmail.com
Abstract
MANET is a kind of network that can transmit data without fixed infrastructures. Nodes can come in or out freely. If nodes which are in the network want to send messages to destination nodes, they can use nodes which are in the network as intermediate nodes. In a network, a routing protocol plays an important role for the performance. Not only routing protocols, we also added another factors：Speed. In this paper, we will introduce MANET first, and then we will explain how AODV, DYMO, OLSR, ZRP, DSR and LAR1 work. We will describe our simulation parameters in section 3, such as routing protocols, node speed and map size, etc. We will put our simulation results in section 4. Our mainly compare targets are Total Number of Packet Received and Average End-to-End Delay. We hope users can refer our simulation results to choose an appropriate routing protocol in the future.
Keywords: MANET, Routing Protocols
1. Introduction
Wireless network can be divided into two types：infrastructure mobile networks and non-infrastructure mobile networks. MANET (mobile ad hoc network) belongs to non-infrastructure mobile networks. There is no specific router and server in MANET, and the network constructed by a group of mobile nodes. Every node could be a router. Data are sent by using multi-hop. The mobile node can transmit data with the help of other in the absence of the base station. It can not only maintain the reliability of communications between mobile nodes but also retain the characteristic of random moving to increase the application of mobile nodes.
In this paper, comparing MANET‘s transmission performance in several kinds of situation is the main purpose. We considered many situations. Such as speed of nodes, number of nodes and routing protocols. Using these factors we mentioned above to simulate. Then we compared MANET‘s performance in several kinds of situation with results of simulations.
2. Related Work
2.1 MANET
MANET has following features：
 Autonomous and infrastructureless
 Multi-hop routing
 Dynamic network topology
 Device heterogeneity
 Limited physical security
 Network scalability
 Self-creation, self-organization and self-administration
Table 1. MANET applications
Tactical networks
 Military communication and operation
 Automated battlefield
Emergency service
 Search and rescue operations
 Disaster recovery
 Replacement of fixed infrastructure in case of environmental disaster
 Policing and fire fighting
 Supporting doctors and nurses

SCIE Journals
www.scie.org.au
73
in hospitals
Commercial and civilian environments
 E-commerce：electronic payments anytime and anywhere
 Business：dynamic database access, mobile offices
 Vehicular services：road or accident guidance, transmission of road and weather conditions, taxi cab network, inter-vehicle networks
 Sports stadium, trade fairs, shopping malls
 Networks of visitors at airports
Sensor networks
 Home application：smart sensors and actuators embedded in consumer electronics
 Body area networks (BAN)
 Data tracking of environmental conditions, animal movements, chemical/biological detection
Coverage extension
 Extending cellular network access
 Linking up with the internet, intranets, etc.
2.2 Routing Protocols
2.2.1 AODV
The full name is Ad-hoc On-demand Vector Routing (Bhatt et al. 2014). It is reactive routing protocol. When source node wants to send messages to destination node source node will broadcast Rout REQest packet (RREQ) to other nodes which are nearby. These nodes will record the route that back to the source node then continue broadcasting RREQ. When the destination node is found destination node will send back Route REPly (RREP) packet. When source node receives RREP packet a route between the source node and the destination node will come out.
2.2.2 OLSR
The full name is Optimizes Link State Routing (Jain & Shiwani 2014). It is the one of proactive routing protocol. The basic principle is similar to broadcast link-state message to every nodes in the network. That makes it possible to construct the entire network. OLSR gets the goal of optimize message flooding by choosing Multipoint Relaying (MPR) and broadcasting route maintaining messages to reduce unnecessary repeat transmission.
2.2.3 DYMO
The full name is Dynamic MANET On-demand Routing (Gupta et al. 2013). It is evolved from AODV, so it is also called AODVv2. DYMO also has RREQ and RREP, but DYMO has an additional message packet：Route Error (RRER). RRER is mainly used to mark invalid paths. Operations of DYMO can be divided into Route Discovery and Route Maintenance.
Route Discovery：
DYMO is almost the same with AODV basically. But, DYMO‘s RREQ adds sequence numbers of nodes and then compares merits of sequence numbers. DYMO has a special feature. When a node has low power, the node will not forward RREQ to make sure this node will not become the one of path‘s nodes.
Route Maintenance：
If a link of the path is disconnected, the node would broadcast RRER and delete this link. If the source

SCIE Journals
www.scie.org.au
74
node wants to send messages when the link is disconnected, it must do Route Discovery again. When a link between node 2 and node6 is disconnected, it will look for a new path. After comparing merits of sequence numbers, the new path will transmit from node 2 to node 6 through node 5.
2.2.4 ZRP
ZRP (Zone Routing Protocol) is the one of hybrid protocols (Kachal & Suri 2014). ZRP is constructed by three sub-protocols：IARP (Intrazone Routing Protocol), IERP (Interzone Routing Protocol) and BRP (Bodercast Resolution Protocol).
IARP：
IARP is primarily responsible for transmission within the zone. Every node will update its routing table by periodic broadcast. When a packet will be send, it just check routing table. It doesn‘t need to find a new path again. Moreover, IARP will periodically update Link-State Table to ensure the path which is recorded in the table is new and valid. But there will be a broadcast storm, if nodes are too close. It will waste network resources.
IERP：
IERP is mainly used to transmit between zones. When the source node cannot find the destination node in the zone, ZRP will use IERP to transmit. The source node will send IERP RouteRequest, and the source node will use the table of IARP and BRP to send the packet to the peripheral node of the zone. When IERP finds the destination node, destination node will send IERP Route Reply back to the source node. Then the path finding will finish and start transmitting data.
BRP：
To get more efficiency to broadcast, ZRP uses BRP to help transmitting IERP RouteRequest when using IERP to transmit messages between zones.
BRP will build a Bordercast tree. BRP can use Bordercast tree to transmit messages to all peripheral nodes by multicasting.
2.2.5 DSR
The full name is Dynamic Source Routing (Shankar et al. 2014)(Sharma & Rani 2014). DSR is similar to DYMO, but it has Route Cache. Route Cache is proposed for optimizing DSR. For example, A->X- >B, when A wants sending data to B. The path to B is already stored in the cache of X, and then X will reply the path to A. The advantage is that can reduce transmitting number of RREQ to speed up the searching time of path. But on the other hand, the size of RREQ may be too big, because it contains sequence number. And the path which is in the cache may not be correct.
2.2.6 LAR1
The full name is Location Aided Routing Scheme 1 (Kumar & Dr. Kumar 2013). It is a kind of routing protocol which uses location information of nodes to limit flooding area. LAR1 gets the location information by GPS. As shown in, LAR1 uses two geographic areas as basis for control packets sending： Expected Zone and Request Zone.
Expected Zone：
Expected Zone is the area where the destination node is predicted by the source node. Assuming the source node S gets the location of the destination node D in time t0. Now, the time is t1, and the average speed of D is v. Then the radius of the Expected Zone is v(t1-t0). But, if the source node S doesn‘t know the location of the destination node D, the source node S will put the entire network considered the Expected Zone.
Request Zone：
Request Zone is the smallest rectangle area that contains the source node S and the Expected Zone. Its sides parallel to X axis, Y axis respectively. Nodes which are only in this area are allowed to transfer RREQ, otherwise it will ignore these packets.
3. Simulation

SCIE Journals
www.scie.org.au
75
Table 2. Simulation Parameters Parameters Values
Routing Protocol
AODV, OLSR, DYMO, ZRP, DSR, LAR1
Number of Nodes
10, 20, 50, 100
Simulation Time(s)
3600
Area Size(m*m)
3000*3000
Movement Model
Random Waypoint Model
Speed(m/s)
Min=0, Max=2.78, 13.89 & 27.78
Traffic Type
CBR
4. Results
4.1 Total Number of Packets Received
We discuss total number of packets received at destination. We sent packets from second 1 and every packet is 512 bytes. So in the ideal situation, the total number of packets must be 1842688 bytes.
Figure 1. Total number of packets in 2.78(m/s)

SCIE Journals
www.scie.org.au
76

SCIE Journals
www.scie.org.au
77
4.2 End-to-End Delay
End-to-End Delay refers to time taken for a packet to be transmitted from source to destination. Packets sometimes will take a few extra seconds. Because the transmission path might be reselect. The lower value means better performance.
Figure 4. End-to-End Delay in 2.78(m/s)

SCIE Journals
www.scie.org.au
78
5. Conclusion
We divided into two area to discuss：Total packet received and average End-to-End delay.
There is no difference between total packets received when the number of nodes is not much and speed is not so fast. As the number of nodes and speed are increased, reactive routings have significant different from proactive routings in these three area. But there have two exceptions：DSR and LAR1. The performance of DSR is similar to AODV and DYMO in low speed, but the performance of DSR is significantly reduced when the the number of nodes is up to 100. The performance of LAR1 is low in reactive routings.
In average End-to-End delay, we do not discuss DSR and LAR1 because the value of average End-to- End delay of DSR and LAR1 almost are greater than one second. Compare to other routings, these values are too big. Therefore, we didn‘t consider them. We can see performance of proactive routing： OLSRv2 and hybrid routing：ZRP are better than reactive routing：AODV and DYMO.
We defined three scenarios： 2.78(m/s), 13.89(m/s) and 27.78(m/s). Our paper‘s contribution is comparison of appropriate routing protocols to meet each situation requirement. Also, it could be a useful referred material for future other dedicated studies.
Acknowledgement
Part of this work was supported by Tatung University, Taipei, Taiwan, under grant B103-I01-031.
References
U. R. Bhatt, A. Dangarh, A. Kashyap and A. Vyas, ―Performance Analysis of AODV & DSR Routing Protocols for MANET,‖ 2014 Fourth International Conference on Communication Sstems and Network Technologies, pp.254-258, April 2014
T. Jain and S. Shiwani, ―Analysis of OLSR, DSR, DYMO Routing Protocols in Mobile Ad-Hoc Networks using Omnet++ Simulation,‖ Global Journal of Computer Science and Technology, Vol. 14, No. 1-E, 2014
A. K. Gupta, H. Sadawarti and A. K. Verma ―Implementation of DYMO Routing Protocol,‖ Modeling and Computing(IJITMC) International Journal of Information Technology, Vol1.1, No.2, May, 2013
R. Kachal and S. Suri, ―Comparative Study and Analysis of DSR, DSDV and ZRP in Mobile Ad-Hoc Networks,‖ International Journal of Computer Siences and Engineering, Vol. 2, pp. 148-152, May 2014
S. Shankar, G. Varaprasad and H. N. Suresh, ―Importance of On-demand Modified Power Aware Dynamic Source Routing Protocol in Mobile Ad-hoc Networks,‖ Antennas & Propagation IET Microwaves, Vol. 8, Issue 7, pp.459-464, May 2014

SCIE Journals
www.scie.org.au
79
G. Sharma and M. Rani, ―Advancement in Dynamic Source Routing Protocol for MANETs,‖ International Journal of Computer Science and Mobile Computing, Vol. 3, Issue 5, pp. 428-433, May 2014
P. Kumar and Dr. A. Kumar, ―Simulation Based Analysis of DSR, LAR and DREAM Routing Protocol for Mobile Ad Hoc Networks,‖ International Journal of Computer Science & Information Technology, Vol. 3, No. 2, pp. 58-62, August 2013

Research on performance of routing protocols in manet

More Related Content

What's hot (19)

Similar to Research on performance of routing protocols in manet (20)

More from Australian Society for Commerce Industry Engineering (16)

Recently uploaded (20)

Research on performance of routing protocols in manet