Provisioning of Quality of Service in MANETs Performance Analysis & Comparison (AODV and DSR )

Nidhi Sharma1 (Student , M.tech)

Sanjeev Rana2 (Associate Professor) 1,2 Deptt. of Computer Engg. 1,2 MMEC, M.M.U, Mullana, Ambala ,Haryana,India sharma.sunidhi2000@gmail.com sanjeevrana@rediffmail.com
AbstractToday wireless networks are becoming popular because of their 3 Anys - Any person, Any where, Any time. Mobile ad-hoc networks (MANETs) is an emerging research area with practical application. Dynamic and reliable protocols are required in MANETs, as they have no infrastructure (base stations) and their Network topology changes frequently. There are different protocols for handling the routing problem in MANETs. In this paper we focused on the two popular algorithms Ad-hoc on Demand Distance Vector (AODV) and Dynamic Source Routing (DSR), both being reactive routing protocols. We analyze and compare their performance through simulation using NS2 simulator. We province Quality of Service (Qos) in aspects of packet delivery rate , average time delay and routing load overhead by varying network size and transmission range of the respective nodes .The general observation from the simulation is that for application oriented metrics such as average delay and packet delivery rate, DSR outperforms AODV in less dense situations. AODV, however, outperforms DSR in more dense situations. However, DSR consistently generates less routing load than AODV. Keywords Routing, MANET, AODV, DSR, Comparison, Simulation, Analysis I. INTRODUCTION

R.M. Sharma3
Associate Professor Deptt. of computer Engg. National Institute of Technology, Kurukshetra, Haryana,India. rmsharma123@rediffmail.com
Mobile ad hoc network is a collection of independent mobile nodes that can communicate to each other via radio waves. The mobile nodes can directly communicate to those nodes that are in radio range of each other, whereas others nodes need the help of intermediate nodes to route their packets. These networks are fully distributed, and can work at any place without the aid of any infrastructure. This property makes these networks highly robust. In Figure 1.1 nodes A and C must discover the route through B in order to communicate II. BACKGROUND

MANET is a kind of wireless ad-hoc network and it is a selfconfiguring network of mobile routers (and associated hosts) connected by wireless links the union of which forms an arbitrary topology. The routers, the participating nodes act as router, are free to move randomly and manage themselves arbitrarily; thus, the network's wireless topology may change rapidly and unpredictably. Such a network may operate in a standalone fashion, or may be connected to the larger Internet [1].

Routing is the act of moving information from a source to a destination in an network. At least one intermediate node within the network is encountered during the transfer of information. Basically two activities are involved in this concept: determining optimal routing paths and transferring the packets through an internetwork. The transferring of packets through an internetwork is called as packet switching which is straight forward, and the path determination could be very complex. Routing protocols use several metrics as a standard measurement to calculate the best path for routing the packets to its destination that could be number of hops, which are used by the routing algorithm to determine the optimal path for the packet to its destination. The process of path determination is that, routing algorithms find out and maintain routing tables, which contain the total route information for the packet. The information of route varies from one routing algorithm to another. The routing tables are filled with entries in the routing table are ip-address prefix and the next hop. Destination/next hop associations of routing table tell the router that a particular destination can be reached optimally by sending the packet to a router representing thenext hop on its way to the final destination and ip-address prefix specifies a set of destinations for which the routing entry is valid. Routing is mainly classified into static routing and dynamic routing. Static routing refers to the routing strategy being stated manually or statically, in the router. Static routing maintains a routing table usually written by a networks administrator. The routing table doesnt depend on the state of the network status, i.e., whether the destination is active or not [2]. Dynamic routing refers to the routing strategy that is being learnt by an interior or exterior routing protocol. This routing primarily depends on the state of the network i.e., the routing table is affected by the activeness of the destination.

Fig 1.1 Example of a simple ad-hoc network with three participating nodes

978-1-4244-6349-7/10/$26.00 c 2010 IEEE

V7-243

III.

PROBLEMs IN ROUTING WITH MANETS [2]

using the broken link. In contrast to DSR, RERR packets in AODV are intended to inform all sources using a link when a failure occurs. Route error propagation in AODV can be visualized conceptually as a tree whose root is the node at the point of failure and all sources using the failed link as the leaves. B. DSR: For DSR [8,15], this protocol has two mechanisms: Route Discovery and Route Maintenance. The source route is needed when some node originates a new packet destined for some node by searching its route cache or initiating route discovery using ROUTE

The major problems for routing in mobile ad-hoc networks are as follows: i). Asymmetric links: Most of the wired networks rely on the symmetric links which are always fixed. But this is not a case with ad-hoc networks as the nodes are mobile and constantly changing their position within network. ii). Routing Overhead: In wireless ad hoc networks, nodes often change their location within network. So, some stale routes are generated in the routing table which leads to unnecessary routing overhead. iii). Interference: This is the major problem with mobile ad-hoc networks as links come and go depending on the transmission characteristics, one transmission might interfere with another one and node might overhear transmissions of other nodes and can corrupt the total transmission. iv). Dynamic Topology: Since the topology is not constant; so the mobile node might move or medium characteristics might change. In ad-hoc networks, routing tables must somehow reflect these changes in topology and routing algorithms have to be adapted. For example in a fixed network routing table updating takes place for every 30sec. This updating frequency might be very low for ad-hoc networks. The purpose of this paper is to study, understand, analyse and compare, two mobile ad-hoc routing protocols DSR and AODV. Both are reactive protocols, they find a route to a destination on demand [4], whenever communication is needed. The routing mechanism in DSR uses source routing [13], while AODV uses a table driven routing framework (hop by hop) or destination routing and destination sequence numbers [6].The effectiveness of our work is illustrated by means of extensive simulations using ns-2 simulator. . IV. LEADING MANETS CONTENDERS DSR and AODV are two most widely used protocols for routing in mobile ad-hoc networks. These protocols are described in the following sections: A. AODV: For AODV [5,6,7,14], it performs Route Discovery using control messages route request (RREQ) and route reply (RREP) whenever node wishes to send packet to destination. To control network wide broadcasts of RREQs, the source node use an expanding ring search technique. The forward path sets up in intermediate nodes in its route table with a lifetime association using RREP. When either destination or intermediate node moves, a route error (RERR) is sent to the affected source nodes. When source node receives the (RERR), it can reinitiate route discovery if the route is still needed. Neighbourhood information is obtained from broadcast Hello packet. An important feature of AODV is the maintenance of timer-based states in each node, regarding utilization of individual routing table entries. A routing table entry is expired if not used recently. A set of predecessor nodes is maintained for each routing table entry, indicating the set of neighbouring nodes that use that entry to route data packets. These nodes are notified with RERR packets when the next hop link breaks. Each predecessor node, in turn, forwards the RERR to its own set of predecessors, thus effectively erasing all routes

REQUEST and ROUTE REPLY messages. On detecting link break, DSR sends ROUTE ERROR message to source for new route. DSR is a simple and efficient routing protocol designed for use in multi hop wireless ad hoc networks of mobile nodes. Network is self-organizing and self-configuring when using DSR. When the nodes in an ad hoc network move and join the network while forwarding packets, all routing is automatically determined and maintained by DSR. DSR allows nodes to dynamically discover a source route across multiple network hops to any destination. Each data packet sent then carries in its header the complete, ordered list of nodes through which the packet must pass, allowing packet routing to be trivially loop free and avoiding the need for up-to-date routing information in the intermediate nodes. So for DSR, the overhead incurred by performing a new Route Discovery can be avoided when the caching of multiple routes to a destination occurs. C. Comparison on AODV and DSR : The DSR and the AODV protocol are two dynamic routing protocols that initiate routing activities for ad hoc networks on an on demand basis [12-13]. These protocols were designed for reducing the routing RREP. When either destination or intermediate node loading in networks. The routing mechanism in DSR uses source routing, while AODV uses a table driven routing framework and destination sequence numbers. AODV relies on certain timer-based activities while DSR does not rely on such options. In DSR the sender knows the hop-by-hop route to the destination because of the use of source routing in the protocol. In DSR, the routes are stored in route cache. The packet header As routing table entry is expired if not used recently. A contains the source route. Route Discovery is used to dynamically determine a route when the route is not known. Route Request packets are send to flood the network and Route Error packets are send when any notified with RERR packets when the next hop link in the source route is broken. DSR makes use of source routing and route caching. DSR uses Route Maintenance mechanism to repair routes that get broken when sending packets from the sender to destination. Packet Salvaging, Automatic Route Shortening, Increased Spreading of Route Error Messages are some of the mechanisms, which are used under Route Maintenance. AODV discovers routes on an on-demand basis using a similar route discovery process as in DSR. AODV uses traditional routing tables, one entry per destination for maintaining routing information. DSR on the other hand maintains multiple route cache entries for each destination. To propagate route reply back to the source and to route data packets to the destination, AODV relies on routing table entries. Sequence numbers at each destination determines freshness of routing information and prevents routing loops. Routing packets carry the sequence

V7-244

2010 2nd International Conference on Computer Engineering and Technology

[Volume 7]

numbers. The maintenance of timer-based states in each node for utilization of individual route entries is an important feature of AODV protocol. Sets of predecessor nodes are maintained for each routing table entry, which indicates the set of neighbouring nodes. Route Error packets are used to notify these nodes when the next-hop link breaks. All the routes using the broken link are erased when the route error packets are sending to its own set of predecessors. Route Error packets in AODV are intended to inform all sources using a link when a failure occurs. In AODV, Route Error propagation is visualized as a tree structure where the root is the node at the point of failure and all sources using the failed link as leaves. An optimizing technique in AODV to control Route Request flood in the route discovery process is to use an expanding ring search to discover routes to unknown destination. DSR with the influence of source routing and promiscuous listening of data packet transmissions has access to a significantly greater. In all, DSR allow cache more paths from a source to a destination, while AODV just use amount of routing information than AODV. AODV can gather only limited amount of routing information. DSR uses route caching aggressively by replying to all requests reaching a destination from a single request cycle. In AODV, the destination replies only once to the request arrive first. The rest of the requests are ignored. Since DSR does not have any mechanism for the expiration of stale routes stored in the cache, some of these stale routes may start polluting other caches. AODV when faced with the choice of stale routes would choose the fresher one. The entry in the routing table if not used recently gets expired [16-17]. the path first discovered. Thus, DSR have significant greater amount of routing information than AODV. Meanwhile, DSR has access to many alternate routes which saves route discovery floods, the performance then will be better if they are actually in use. DSR does not contain any explicit mechanism to expire stale routes in the cache. Choosing stale routes under stressful situation is normal thing in DSR. So, AODV is more conservative while DSR is more aggressive in using the past information. Which is better is hard to say [3,16,17]. Accord to the discussion on AODV and DSR above, our performance evaluation mode should give out the expected result on the following matrix: For routing load, DSR should be lower than AODV (in terms of number of packets) due to its aggressive caching (though more replies and errors) and another reason is that AODV is dominated by RREQ packets. For packets delivery rate and delay, DSR should be better if less stressful traffic load. Since caching only provide significant benefits to a certain extent. Stale caches are chosen in high loads which will cause unnecessary bandwidth consumption and pollution of caches in other nodes under high load with DSR. For average delays, which may determine by network congestion, bandwidth cost by routing overhead, and routing error, so it is a little hard to say but DSR may be better in most cases since the routing overhead is extremely great with AODV. V. EXPERIMENTAL SETUP

and the packet-sending rate in each pair is varied to change the offered load in the network. The mobility model uses the random waypoint model in a rectangular field. The field configurations used is: 800 m x 800 m field with total 50 nodes. Here, each packet starts its journey from a random location to a random destination with a randomly chosen speed (uniformly distributed between 1 20 m/s). Once the destination is reached, another random destination is targeted after a pause. The pause time, which affects the relative speeds of the mobiles, is varied. Simulations are run for 100, 200, 300, 400, 500, 600, 700, 800 simulated seconds. Identical mobility and traffic scenarios are used across protocols to gather fair results. B. Performance Metrics

Some important performance metrics can be evaluated:Number of nodes: Performance testing usually needs to be scalable in the number of nodes and network transmitting packets, Transmission Range: The effect of varying the transmission range of mobile nodes (Total packets received / total packets lost), Packet delivery rate: This equals to Total packets successfully received/ Total packets send, Average delay: This means Sum (for each i equal to packet number, (packet i received time- packet i send time)/ Total packets transmitted), Average routing overhead: This means Total routing control packets/Simulation time C. Essential parameters to be varied include For effective tests on the above metrics, the following parameters are varied. Network size--measured in the number of nodes, Network density--average degree of a node, i.e. the average number of neighbours of a node, Mobility--the most appropriate model for simulating node mobility in a MANETS, Transmission Range: circular extent of the networking capacity of individual node. D. Scenario for implementation of AODV & DSR -----------------------------------------------------------------------Parameter Value Number of nodes 10 and 45 Simulation Time 1000sec Pause Time 100s Environment Size 800x800 Transmission Range 250m and 850 m Traffic Size CBR (Constant Bit Rate) Packet Size 512 bytes Packet Rate 5 packets/s Minimum Speed 1 m/s Maximum Speed 20 m/s Queue Length 50 Mobility Model Random Waypoint
Table I: Scenario for Implementation of AODV & DSR

A. Traffic & Mobility Model VI. Constant bit rate (CBR) [10] traffic sources are used. The sourcedestination pairs are spread randomly over the network. Only 512byte data packets are used. The number of source-destination pairs A. PROVISIONING QoS with SIMULATION

Calculated Metric in tabular format

[Volume 7]

2010 2nd International Conference on Computer Engineering and Technology

V7-245

Protoco Number Data l of Packet Node receive DSR 45 170000 AODV 45 480000 DSR 10 478200 AODV 10 170000 B.

Total Packets lost 80000 560000 560000 80000

Total Control packets 246500 214000 386000 995000

Total Packets Transmitted 2715000 22440000 1424200 1245000

Table II: Comparison of Routing Protocols ( AODV & DSR)

Packet Delivery Rate Comparisons

Following Figures shows the experiment results of packets delivery rate. We can see, under low traffic load as 10 sources, DSR outperform the AODV, but when the sources become more, DSR with small pause time which means high mobility began to perform worse than AODV. This result is what we want to see, since AODV have more routing control packets but may always choose the fresh route, while DSR with smaller number of routing packets under stressfully situation with the network topology continue to change from time to time, will be inclined to choose wrong routes, thus lower the packets delivery rate

Figure 6.2:Average Delay comparisons with TR=250m

D.

Average routing overhead comparisons

Since AODV always has more routing control packets than DSR, the routing overhead of AODV than will always be higher even in stressful environment. Under heavy load, through DSR may incline to choose wrong route, however, under such situation, AODV will also generate far more control packets than DSR.

Figure 6.1: Packet delivery rate comparisons with TR=250m

C.

Average Delay Comparisons

Following figures shows about the average delay of AODV and DSR. Since AODV has much more routing packets than DSR, and those routing packets will consume more bandwidth, AODV then will have more delay than DSR. In this test, under normal load which is 10 sources, AODV cost become higher, this is because routing packets are more with AODV which may cause some latency of sending data packets. When load become heavy which is 45 sources, DSR with small pause time have more delay time since stale routes often be choose, which lost many delivery time. AODV also generate many more RREQ packets, so the RREQ packets in AODV is indeed a great bottle neck

Figure 6.3: Average Routing Load comparisons with TR=250m

E.

Transmission Range Comparison with Sources=10

In the X graphs of AODV, we see that as the simulation start the packet received and packet loss is initially zero, because initially

V7-246

2010 2nd International Conference on Computer Engineering and Technology

[Volume 7]

there is no CBR connection and nodes taking their right place. As the CBR connections establish between the nodes the number of packet received increases but no packet loss is there, it means all generated packets are being received by the nodes. But the packet loss increases substantially on the transmission range increases. Finally the packet received is more than the packet loss. In the X graphs of DSR, we see that as the simulation start the packet received and packet loss is initially zero, because initially there is no CBR connection and nodes taking their right place. As the CBR connections establish the number of packet lost increases very much as compare to packet received. It shows that mostly generated packets are being dropped by the nodes. But the packet loss decreases substantially as the transmission range increases, and number of packet received increases substantially as the transmission range increases

Figure 6.7: X Graph of DSR with TR=850m

The above figures shows the same behaviour of AODV & DSR in fact of packet receiving and packet loss, initially in AODV no packet loss and in DSR very high packet loss. But as the transmission range increases the packet loss goes down and packet receiving increases. VII. CONCLUSION & FUTURE SCOPE We have compared two On-demand routing protocols, namely, Ad hoc On-Demand Distance Vector Routing (AODV) and Dynamic Source Routing (DSR). It is observed that the packet loss is very less in case of AODV, initially but it increases substantially on the simulation time increases. In case of DSR simulation the packet loss is very high initially but it decreases substantially on the simulation time increases. So, we can conclude that if the MANET has to be setup for a small amount of time then AODV should be prefer due to low initial packet loss and DSR should not be prefer to setup a MANET for a small amount of time because initially there is packet loss is very high. If we have to use the MANET for a longer duration then both the protocols can be used, because after some times both the protocols have same ratio of packet delivering. But AODV have very good packet receiving ratio in comparison to DSR In particular, DSR uses source routing and route caches and does not depend on any periodic or timer-based activities. DSR exploits caching aggressively and maintains multiple routes per destination. AODV, on the other hand, uses routing table, one route per destination, and destination sequence numbers, a mechanism to prevent loops and determine freshness of routes. The general observation from the simulation is that for application oriented metrics such as average delay and packet delivery rate, DSR outperforms AODV in less stressful situations. AODV, however, outperforms DSR in more stressful situations. However, DSR consistently generates less routing load than AODV. The poor delay and throughput performances of DSR are mainly attributed to aggressive use of caching, and lack of any mechanism to expire stale route or to determine the freshness of routes when multiple choices are available. Aggressive caching, however seems to help DSR at low load and also keeps its routing load down. We believe that mechanisms to expire routes and /or determine freshness of routes in the route cache will benefit DSR performance significantly. In addition, one interesting observation is that the delays for both protocols increase with 45 sources with very low mobility. This is due to a high level of network congestion and multiple access interferences at certain regions of the ad hoc network. Neither protocol has any mechanism for load balancing, i.e, for choosing routes in such a way that the traffic can be more evenly distributed in the network. This phenomenon is less visible with higher mobility where traffic automatically gets more evenly distributed due to source movements. In future, it would be interesting to observe the behaviour of these two protocols by varying other network parameters .Since we now only explore some important fields of the trace file, in the future, we still need to provide the measurement with other fields of trace file and analysis more details on the things what we can get in the trace file. ACKNOWLEDGEMENT

Figure 6.4: X Graph of AODV with TR=250m

Figure 6.5 X Graph of DSR with TR=250m

Figure 6.6 X Graph of AODV with TR=850m

[Volume 7]

2010 2nd International Conference on Computer Engineering and Technology

V7-247

R.B.G. thanks to: Rajneesh Gujral, Associate Professor, Deptt. of Computer Engg., M.M.Engg.College, M.M.University, Mullana, Ambala, Haryana, India. REFERENCES MANET,.http://www.ietf.org/html.charters/manetcharter Krishna Gorantala, Routing in Mobile Ad-hoc Networks Umea University, Sweden, June-2006. [3] Geetha Jayakumar and Gopinat Ganapathy,Performance Comparison of Mobile Adhoc Network Routing Protocol, International Journal of Computer Science and Network Security (IJCSNS), VOL.7 No.11, pp. 77-84 November 2007. [4] Laura Marie Feeney, A taxonomy for routing protocols in mobile ad hoc networks, Technical report, Swedish Institute of Computer Science, Sweden, 1999. [5] C.E.Perkins and E.M. Royer,Ad-hoc On-Demand Distance Vector Routing, Proceedings of the 2nd IEEE Workshop on Mobile Computing Systems and Applications, New Orleans, LA, pp. 90-100, February 1999. [6] Charles E. Perkins, Ad Hoc On Demand Distance Vector (AODV) Routing. Internet draft, draft-ietfmanetaodv-02.txt, November 1998. [7] C.E Perkins, E.M. Royer, and S. Das, Ad hoc Ondemand Distance Vector (AODV),RFC 3561, July 2005. [8] D. B. Johnson and D. A. Maltz, Dynamic Source Routing in Ad Hoc Wireless Networks, Mobile Computing, Chapter 5, pp. 153-181, Kluwer Academic Publishers, 1996, [9] The VINT Project, The network simulator - ns-2. http://www.isi.edu/nsnam/ns/ [10] The CMU Monarch Project, The CMU Monarch Projects Wireless and Mobility Extensions to ns.http:// www .monarch. cs.cmu.edu/(1998-11-29).Work in progress. [11] V. Davies,Evaluating mobility models within an ad hoc network , Masters thesis, Colorado School of Mines, 2000. [12] M. Zonoozi and P. Dassanayake, User mobility modeling and characterization of mobility patterns. IEEE Journal on Selected Areas in Communications, 15(7):1239-1252, September 1997. [13] Azzedine Boukerche , Khalil El-Khatib , Li Xu , Larry Korba, Performance evaluation of an anonymity providing protocol for wireless ad hoc networks, Performance Evaluation, v.63 n.11, p.1094-1109, November 2006 [14] C. Perkins, E. Belding- Royer , Adhoc on- demand distance vector (AODV) routing,RFC 3561, Experimental, July 2003. [15] David B. Johnson , David A. Maltz, and Yih -Chun Hu, The Dynamic source routing protocol for mobile ad hoc networks (DSR), nternet-Draft, draft-ietf-manets dsr08.txt,February24 2003. [16] Bai, R., Singhal, M.: DOA: DSR over AODV Routing for Mobile Ad Hoc Networks. IEEE Transactions on Mobile [1] [2]

[17]

Computing 5, 14031416 (2006) Tafazolli, H.: A Survey of QoS Routing Solutions for Mobile Ad Hoc Networks. IEEE Communications Surveys & Tutorials 9, 5070 (2007)

V7-248

2010 2nd International Conference on Computer Engineering and Technology

[Volume 7]