JOURNAL OF TELECOMMUNICATIONS, VOLUME 5, ISSUE 2, NOVEMBER 2010 18
A Comparison of Proactive and Reactive Multi-hop Routing Protocols for Mobile Ad Hoc Networks Foez Ahmed, Md. Shariful Islam, Nayeema Islam, M. Rahmathullah and Muntasir Ahmed Abstract—Wireless communication technology is increasing daily; with such growth sooner or later it would not be practical or simply physically possible to have a fixed architecture for this kind of network. Ad hoc wireless network must be capable to selforganize and self-configure due to the fact that the mobile structure is changing all the time. Mobile hosts have a limited range and sending the message to another host, which is not in the sender’s host transmission range, must be forwarded through the network using other hosts which will be operated as routers for delivering the message throughout the network. To accomplish this communication, a routing protocol is used to discover routes between these hosts. This paper presents a variety of comprehensive simulation results for two popular protocols – Dynamic MANET On-demand (DYMO) and Optimized Link State Routing (OLSR) using NS2. Both two protocols are provided with identical traffic load and mobility patterns. We have considered TCP as transport protocol and FTP as traffic generator. The analysis is significant because we considered near about all the metrics as suggested by RFC 2501 and till to-date there are a few comparisons based on TCP.
Index Terms—MANET Routing Protocols, OLSR, DYMO, TCP, FTP, Throughput, Routing Overhead, NS2.
—————————— ——————————
1 INTRODUCTION
I
N a mobile ad hoc network (MANET), mobile nodes communicate using wireless links without a fixed infrastructure such as base stations (access points) or centralized control. A typical mobile ad hoc network is a group of hosts or nodes operating in wireless ad hoc mode. Each mobile node acts as a router to enable multihop communication. A node is free to move around randomly and as a result, the topology formed by the nodes is highly dynamic and unpredictable. A MANET can operate in a stand-alone fashion, or can be connected to a fixed internetwork, for example, the Internet.
node A to be able to communicate with node C and D that both are out of the transmission range of A, the intermediate node B must forward packets to the other nodes.
The need for multi-hop routing arises when some nodes are out of transmission range of others. A node only handles traffic within a local cloud of wireless devices. As an example consider figure 1.1, in which the circle around the nodes indicates the transmission range. For ————————————————
• Foez Ahmed is with the Dept. of Computer Networks and Communication Engineering, King Khalid University, Kingdom of Saudi Arabia. • Md. Shariful Islam is with the Department of Computer Science and Engineering, Rajshahi University,Bangladesh. • Nayeema Islam is with the Department of Information and Communication Engineering,Rajshahi University,Bangladesh. • M. Rahmatullah is with the Computer Engineering Department, King Khalid University, Kingdom of Saudi Arabia. • Muntasir Ahmed is with the Dept. of Information & Communication Eng., Pabna University of Science and Technology, Bangladesh.
Fig. 1.1: Illustration of multi-hop routing. Every node must act as a router. The circle around each node shows the transmission range. If A needs to reach C and D, node B must forward packets from A.
Routing protocols are divided into two categories based on how and when routes are discovered, but both find the shortest path to the destination. Proactive routing protocols are table-driven protocols; they always maintain current up-to-date routing information by sending control messages periodically between the hosts which
19
update their routing tables. When there are changes in the structure then the updates are propagated throughout the network. The proactive routing protocols use link-state routing algorithms which frequently flood the link information about its neighbors. Other routing protocols are on-demand routing protocols, in other words reactive, ones which create routes when they are needed by the source host and these routes are maintained while they are needed. Such protocols use distance-vector routing algorithms; they have vectors containing information about the cost and the path to the destination. When nodes exchange vectors of information, each host modify own routing information when needed. The ad hoc routing protocols are usually classified as a pure proactive or a pure reactive protocol, but there are also hybrid protocols [7].
If the originator entry in the RREQ is found to be stale or disregarded, the RREQ is dropped. For other nodes, the information is removed from the RREQ. If an RREQ is not dropped, each node processing the RREQ can create reverse routes to all the nodes for which addresses are accumulated in the RREQ. When the RREQ reaches the destination node, it processes the packet similar to nodes that have forwarded the packet, and uses the information accumulated in the RREQ to add route table entries. An RREP message is then created as a response to the RREQ, containing information about destination node, i.e., address, sequence number, prefix, and gateway information, and the RREP message is sent back along the reverse path using unicast. 2.2 Route Maintenance
This paper will compare the two most prominent ad hoc routing protocols: reactive Dynamic MANET Ondemand (DYMO) and proactive Optimized Link State Routing (OLSR) protocols. The reminding part of this paper is organized as follows. Sec. 2 will give information about DYMO protocol and section 3 will discuss about OLSR protocol. In section 4 the possible information about performance metrics and simulation environment will be given. Actual comparison will be done in Sec. 5. Finally, Sec. 6 will conclude this paper.
2 DYNAMIC MANET ON-DEMAND ROUTING PROTOCOL (DYMO) The Dynamic MANET On-demand DYMO routing protocol is a newly proposed protocol currently defined in an IETF Internet-Draft [1] in its sixth revision and is in progress. As is the case with all reactive ad hoc routing protocols, DYMO consists of two protocol operations: route discovery and route maintenance. These operations are discussed briefly in below.
2.1 Route Discovery Route discovery is the process of creating a route to a destination when a node needs a route to it [1]. When a source node wishes to communicate with a destination node, it initiates a Route Request (RREQ) message. When a node receives an RREQ, it processes the addresses and associated information found in the message. •
•
•
If the routing table does not contain an entry for the originator, one is created. The next hop entry is the address of the node from which the RREQ was received. Likewise, the next hop interface is the interface on which the RREQ was received. If an entry exists, the sequence number and hop count found in the RREQ is compared to the sequence number and hop count in the table entry to check if the information in the RREQ is stale or should be disregarded. If an entry exists and is not stale or disregarded, the entry is updated with the information found in the RREQ.
Route maintenance is the process of responding to changes in topology that happens after a route has initially been created [1]. To maintain paths, nodes continuously monitor the active links and update the Valid Timeout field of entries in its routing table when receiving and sending data packets. If a node receives a data packet for a destination it does not have a valid route for, it must respond with a Route Error (RERR) message. When creating the RERR message, the node makes a list containing the address and sequence number of the unreachable node. In addition, the node adds all entries in the routing table that is dependent on the unreachable destination as next hop entry. The purpose is to notify about additional routes that are no longer available. The node sends the list in the RERR packet. The RERR message is broadcasted. When a node receives an RERR, it compares the list of nodes contained in the RERR to the corresponding entries in its routing table. If a route table entry for a node from the RERR exists, it is invalidated if the next hop node is the same as the node the RERR was received from and the sequence number of the entry is greater than or equal to the sequence number found in the RERR. If a route table entry is not invalidated, the corresponding entry in the list of unreachable nodes from the RERR must be removed. If no entries remain, the node does not propagate this RERR further. Otherwise, the RERR is broadcasted further. The sequence number check mentioned, is performed to only invalidate fresh routes and to prevent propagating old information. The intention of the RERR distribution is to inform all nodes that may be using a link, when a failure occurs. RERR propagation is guaranteed to terminate as a node only forwards an RERR message once.
20
3 OPTIMIZED LINK STATE ROUTING PROTOCOL (OLSR) Optimized Link State Protocol (OLSR) [5] is a proactive routing protocol, so the routes are always immediately available when needed. OLSR is an optimization version of a pure link state protocol. So the topological changes cause the flooding of the topological information to all available hosts in the network. To reduce the possible overhead in the network protocol uses Multipoint Relays (MPR).
tric neighbors from which it is possible to reach all the symmetrical strict two hop neighbors. The host must have the information about one and two hop symmetric neighbors in order to start the needed calculation for the MPR set. All the exchange of information are broadcasted using Hello messages. The neighbor must be symmetric in order to become an MPR.
3.4 Topology Information
OLSR uses two kinds of the control messages: Hello and Topology Control (TC). Hello messages are used for finding the information about the link status and the host’s neighbors. TC messages are used for broadcasting information about own advertised neighbors which includes at least the MPR Selector list. The TC messages are broadcasted periodically and only the MPR hosts can forward the TC messages. There is also Multiple Interface Declaration (MID) messages which are used for informing other host that the announcing host can have multiple OLSR interface addresses [3][5][6][8].
In order to exchange the topological information and build the topology information base the host that was selected as MPR need to send the topology control (TC) message. The TC messages are broadcasted throughout the network and only MPR are allowed to forward TC messages. The TC messages are generated and broadcasted periodically in the network and the size of the TC message can be quite big [5]. The TC message includes the own set of advertised links and the sequence number of each message. The sequence number is used to avoid loops of the messages and for indicating the freshness of the message.
3.1 Neighbor Sensing
3.5 Routing Table Calculations
The link in the ad hoc network can be either unidirectional or bidirectional so the host must know this information about the neighbors. The Hello messages are broadcasted periodically for the link sensing, neighbor’s detection and MPR selection process. Hello message contains: information how often the host sends Hello messages, willingness of host to act as a Multipoint Relay, and information about its neighbor. Information about the neighbors contains: interface address, link type and neighbor type. The link type indicates that the link is symmetric, asymmetric or simply lost. The neighbor type is just symmetric, MPR or not a neighbor. The MPR type indicates that the link to the neighbor is symmetric and that this host has chosen it as Multipoint Relay. The Hello messages are only broadcasted one hop away [5].
The host maintains the routing table. The routing table is changed if the changes occur in the following cases: neighbor link appear or disappear, two hops neighbor is created or removed, topological link is appeared or lost or when the multiple interface association information changes. But the update of this information does not lead to the sending of the messages into the network. For finding the routes for the routing table entry the shortest path algorithm is used [4][5][8].
3.2 Multipoint Relays (MPR) The Multipoint Relays (MPR) is the key idea behind the OLSR protocol to reduce the information exchange overhead. Instead of pure flooding the OLSR uses MPR to reduce the number of the host, which broadcasts the information throughout the network. The MPR is a host’s one hop neighbor which may forward its messages. Each host must have the information about the symmetric one hop and two hop neighbors in order to calculate the optimal MPR set. The MPR set of host is kept small in order for the protocol to be efficient. In OLSR only the MPRs can forward the data throughout the network [5].
3.3 Multipoint Relays Selection The algorithm proposed by [5] constructs the MPR set, which includes minimum number of the one hop symme-
4 SIMULATION Our protocol evaluations are based on the simulation using ns2 [2]. NS2 is a discrete event simulator developed by the University of California at Berkeley and the VINT project. Simulation environment consists of 20 wireless nodes (here also used different no. of nodes), forming an ad hoc network, moving about over a 600 X 600 flat space for 200 seconds of simulated time. The physical radio characteristics of each mobile node’s network interface, such as the antenna gain, transmit power, and receiver sensitivity, were chosen to approximate the Lucent WaveLAN direct sequence spread spectrum radio. In order to enable direct, fair comparisons between the protocols, it was critical to challenge the protocols with identical loads and environmental conditions. Each run of the simulator accepts as input a scenario file that describes the exact motion of each node and the exact sequence of packets originated by each node, together with the exact time at which each change in motion or packet origination is to occur. We pre-generated some different scenario files with varying movement patterns and traffic loads (FTP), and then simulate both two routing protocols against each of these scenario files.
21
4.2 Metrics In comparing the protocols, we chose to evaluate them according to the following metrics: • Throughput: It is defined as total number of packets received by the destination. It is a measure of effectiveness of a routing protocol. • Routing overhead: The total number of routing packets transmitted during the simulation. To achieve a given level of data routing performance, two different protocols can use differing amounts of overhead, depending on their internal efficiency. • Packets lost/dropped: it is a measure of the number of packets dropped by the routers due to various reasons. The reasons we have considered for evaluation are Collisions, time outs, looping, and errors.
Throughput (Mbps)
0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0
DYMO OLSR
2
6
10
14
20
50
Num ber of Nodes
Fig. 5.2: Variation of throughputs due to different number of mobile nodes.
Total Received Packets
4.1 Movement Model Nodes in this simulation move according to a model that we call the “random waypoint” model. The movement scenario files we used for each simulation are characterized by a pause time. Each node begins to move after remaining stationary for a specific time which is defined by pause time (seconds). It then selects a random destination in the 600 X 600m flat rectangular space and moves to that destination at a speed distributed uniformly between 0 and some maximum speed. Upon reaching the destination, the node pauses again for the defined pause time, selects another destination, and proceeds there as previously described, repeating this behavior for the duration of the simulation. Each simulation ran for 200 seconds of simulated time. We simulate our simulations with movement patterns generated for 8 different pause times: 0, 10, 20, 30, 50, 100, 150, 200 seconds and 5 different speeds: 2, 5, 10, 15, and 20. A pause time of 0 seconds corresponds to continuous motion, and a pause time of 200 (the length of the simulation) corresponds to no motion.
25000
DYMO
20000
OLSR
15000 10000 5000 0 0
10 20 30 50 100 150 200 Pause Time (s)
Fig. 5.3: Total no. of packets received at various levels of mobility.
45000 40000 35000 30000 25000 20000 15000 10000 5000 0
DYMO OLSR
2
6
10
14
20
50
Num ber of Nodes
Fig. 5.1: Variation of routing overhead with different number of mobile nodes.
Total Dropped Packets (%)
Routing Overhead (pkt)
5 ANALYSIS OF SIMULATION RESULTS 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0
DYMO OLSR
0
10
20
30
50 100 150 200
Pause Time (s)
Fig. 5.4: Total number of packets dropped with variation in mobility.
0.5 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0
DYMO OLSR
0
10
20
30
50 100 150 200
Total Received Packets
25000
DYMO OLSR
20000
10000 5000 0 10
15
20
Speed (m/s)
Fig. 5.6: Total number of packets received at various levels of speed.
Total Dropped Packets (%)
4
DYMO
3.5
OLSR
3 2.5 2 1.5 1 0.5 0 5
10
5
10
15
20
15
Fig. 5.8: Variation of throughputs with different speed.
The simulation results bring out some important characteristic differences between proactive (OLSR) and reactive (DYMO) routing protocols. Since OLSR is the proactive routing protocol, it updates its routing table periodically throughout the entire network and therefore more routing overheads are need over DYMO. Form Fig. 5.1 it can be observed that the routing overhead of OLSR is enough large than DYMO in case of different dynamic dense network. Consequently in case of dense dynamic network topology, DYMO also offers high throughput over OLSR which is illustrated in Fig. 5.8.
15000
2
OLSR
Speed (m/s)
Fig. 5.5: Variation of throughputs at different levels of mobility.
5
DYMO
0.5 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 2
Pause Time (s)
2
Throughput (Mbps)
Throughput (Mbps)
22
20
Speed (m/s) Fig. 5.7: Total number of packets dropped with variation in speed.
In wireless mobile ad hoc network speed and mobility of nodes is so much effective. At different level of mobility the total numbers of received application layer data packets are better while using DYMO as a routing protocol. On the other hand for OLSR the numbers of dropped packets are more with different level of mobility. Fig.5.2 and Fig. 5.3 illustrate those scenarios respectively. From Fig. 5.4, it can be inferred that DYMO throughput is better than OLSR in case of node mobility and dynamic network topology. Though OLSR is a proactive routing protocol in most cases it uses already established route and tries to get rid of the packets immediately but in this scenario it performs poor over DYMO. The reason is that due to the dynamic network topology OLSR needs more topology control message to update routing tables throughout the entire network. Hence, DYMO performs better than OLSR. The variation of performance of the two protocols: DYMO and OLSR with speed is similar to the variation in the performance with mobility as shown in Fig. 5.5 to Fig. 5.7. DYMO performs better than OLSR in all cases.
23
6 CONCLUSION This paper compares the two prominent wireless routing protocols for MANET namely DYMO and OLSR. Proactive routing protocols are intrinsically not suitable for any mobile network because they involve periodic exchange of information for updating the routing tables resulting in consumption of energy of battery operated nodes. The significant observation is that simulation results agree with expected results based on theoretical analysis. As expected, reactive routing protocol DYMO performs better than OLSR in case of node mobility and dynamic network topology.
ter Wireless Sensor Networks and Sparse Signal Processing in Wireless Communication.
Md. Shariful Islam obtained his M.Sc degree in Computer Science and Engineering from Islamic University, Bangladesh during the year 2006. He has joined at Dutch-Bangla Bank Limited, Bangladesh and working as an senior officer in the department of Netwroks and System. Currently he is doing his M.Phil program in the department of Computer Science and Engineering, University of Rajshahi. His research interests include Mobile Ad-hoc network, Sensor Network, MIMO, OFDM, and MCCDMA.
REFERENCES [1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
Ian D. Chakeres and Charles E. Perkins. “Dynamic MANET ondemand (DYMO) routing protocol,” Internet-Draft Version 4, IETF, March 2006. draft-ietf-manet-dymo-04.txt, (Work in Progress). Kevin Fall and Kannan Varadhan, editors, “NSDocumentation,” http://www.isi.edu/nsnam/ns/nsdocumentation.html, December 2006. P.Jacquet, A. Laouiti, P. Minet and L. Viennot “Performance Analysis of OLSR Multipoint Relay Flooding in Two Ad Hoc Wireless Network Models,” Research Report-4260. INRIA, September 2001. RSRCP journal special issue on Mobility and Internet. P.Jacquet, P. Mühlethaler, T Clausen, A. Laouiti, A. Qayyum and L. Viennot “Optimized Link State Protocol for Ad Hoc Networks,” IEEE INMIC Pakistan 2001. T. Clausen and P. Jacquet “Optimized Link State Routing Protocol (OLSR),” RFC 3626, IETF Network Working Group, October 2003. T.H. Clausen, G. Hansen, L. Christensen and G. Behrmann “The Optimized Link State Routing Protocol, Evaluation through Experiments and Simulation,” IEEE Symposium on "Wireless Personal Mobile Communications". September 2001. Xiaoyan Hong, Kaixin Xu and Mario Gerla “Scalable Routing Protocols for Mobile Ad Hoc Networks,” Computer Science Department, University of California, Los Angeles, August 2002. Ying Ge, Thomas Kunz and Louise Lamont “Quality of Service Routing in Ad-Hoc Networks Using OLSR,” Proceeding of the 36th Hawaii International Conference on System Science (HICSS’03).
Foez Ahmed joined as a Lecturer in the Dept. of Electronics and Communication Engineering, Northern University Bangladesh (NUB), Dhaka, Bangladesh in the 2008. At present he is on leave from Northern University and working as a Lecturer with the Dept of Networks and Communication Engineering, College of Computer Science, King Khalid University, Kingdom of Saudi Arabia. He did his under graduation and post graduation in Information and Communication Engineering in 2007 and 2009 respectively from Rajshahi University, Bangladesh. He has received various Awards and Scholarships for the under graduation and post graduation results. His research interests include Mobile Ad-hoc Networks and Routings, Cognitive Radio Networks, Cooperative Communications, Underwa-
Nayeema Islam was born 1976 in Bangladesh. She received her M.S. in Telecommunication Engineering from Asian Institute of Technology, Thailand in 2004. Also she received her M.Sc. and B.Sc. in Computer Science and Technology from Rajshahi University, Bangladesh in 1998 and 1997 respectively. She is presently working as Assistant Professor in the department of Information and Communication Engineering, University of Rajshahi since 2004. Her fields of interest include Telecommunications, computer networking, mobile ad-hoc networks and QoS routing.
Mohammed Rahmathullah completed Bachelors of Engineering in Electronics and Communications Engineering in the year 2002 and completed Master of Engineering in 2008 with Systems and Signal Processing as his major. He receieved many medals and certificates from various organizations for his strong credentials in academics. From February 2003 - May 2009 he served as Assistant Professor at Muffakham Jah College of Engineering and Technology, Hyderabad, India and since May 2009 he is working as lecturer in Computer Engineering Department of King Khalid University, Abha, Kingdom of Saudi Arabia. Since many years he is teaching Data Communications and Computer Networks to undergraduate engineering students. His major research interests are in Wireless Sensor Based Networs, Adhoc Networks, speech signal processing. He is the life member of Indian Society of Engineers.
.
Muntasir Ahmed joined as a Lecturer in the Dept. of Electronics and Communication Engineering, University of Information Technology & Science, Rajshahi, Bangladesh in the 2010. At present, he is working as a Lecturer with the Dept of Information and Communication Engineering, Pabna University of Scinece and Technology, Bangladesh. He did his under graduation and post graduation in Information and Communication Engineering from Rajshahi University, Bangladesh. His research interests include Mobile Ad-hoc Networks and Routings, Cognitive Radio Networks, Cooperative Communications, and Sensor Networks.
