Elsevier

Computer Communications

Volume 30, Issue 9, 30 June 2007, Pages 1931-1941
Computer Communications

Performance evaluation of multicast ad hoc on-demand distance vector protocol

https://doi.org/10.1016/j.comcom.2007.02.003Get rights and content

Abstract

Multicast Ad hoc On-demand Distance Vector routing protocol (MAODV) is an on-demand multicast routing protocol that constructs a shared delivery tree to support multiple senders and receivers in a multicast session. Existing research does not address the performance evaluation of MAODV for short-lived connections. Existing research have studied the protocol assuming only long-lived connection. MAODV performance may be affected when most of the connections are short-lived ones. We show that the efficiency of delivering data packets for short-lived connection is greater than that for long-lived connections. Different performance parameters have been considered such as mobility speed, number of senders, and multicast group size. The average Latency for long-lived connection is greater than that for short-lived connections. The channel access efficiency of MAODV over long-lived connections is better than that for short-lived connections. The scalability of MAODV in short-lived situation is better than that in long-lived connection with respect to the number of senders.

Introduction

Routing is the process of finding a path between two or more nodes that want to interact with each other. Many types of transmission in computer networks exist such as unicast, multicast and broadcast. Some applications, such as distributed database systems and audio/video teleconferencing, need multicast communication. Multicasting is the transmission of data packets from one sender to multiple receivers [27]. Multicasting can be implemented in several ways: one-to-all unicast, explicit multicast, or application-level multicast [14]. Explicit multicast uses multicast routing algorithms. It is better than one-to-all unicast according to the simulation results in [5].

A Mobile Ad hoc Network (MANET) is a wireless network composed of mobile nodes that has a dynamic nature. There are no routers, servers, cables, or access point, i.e. no infrastructure. The mobile nodes of MANET should function as a router to forward packets over MANET. The ad hoc topology may change with time according to the mobile nodes movement. Some applications of MANET relate to temporary networks in airports, military battlefield, and search and rescue operations [31].

The multicast routing protocols proposed for MANET are classified into four categories according to the route construction or route topology [6], [32]: tree-based approaches, mesh-based approaches, stateless multicast, and hybrid approaches. In the tree-based approach, which is used in MAODV protocol, multicast routing uses a source-based or shared tree among sources and receivers. Only one path exists between any pair of nodes. The amount of bandwidth and resources required for initializing a tree is usually less than that required for other structures [31].

Multicast Ad hoc On-demand Distance Vector routing protocol (MAODV) follows directly from Ad hoc On-demand Distance Vector (AODV) routing protocol. It discovers multicast routes on demand using a broadcast route discovery mechanism employing the same control messages that exist in the unicast AODV protocol. It also employs other messages to serve the multicast operation of the protocol. The control messages; Route Requests (RREQ), Route Replies (RREP), Multicast Activations (MACT) and Group Hellos (GRPH) are employed by MAODV protocol. These messages are used to establish the multicast delivery tree and maintain it. The RREQ is broadcasted to establish the reverse path from source to destination, and the RREP is unicasted form destination back to source in order to establish the forward path. A MACT message is sent from source to destination to activate the path [6], [25].

The first member of the multicast group becomes the leader for that group, which also becomes responsible for maintaining the multicast group sequence number and broadcasting this number to the multicast group. This update is done through a GRPH message. The GRPH contains extensions that indicate the IP addresses and sequence numbers (incremented every Group Hello) of all multicast groups for which the node is group leader.

The major role of MAODV protocol after building the multicast delivery tree is the maintenance of that tree. MAODV tries to reconnect the multicast tree if it was partitioned using the control messages mentioned earlier. If reconnection fails, a new multicast tree construction phase begins from scratch.

The major goal of the proposed multicast routing protocols for MANETs is to allow group communication among network nodes in the ad hoc environment. These protocols need to be accurate and reliable [37]. Researchers make every effort to test these protocols’ performance [3], [13], [15], [16], [20] in order to make sure that they will perform well. They tried to prove the reliability [17], [19], [4], [22], [29] of these protocols in addition to other aspects such as scalability [9], [10], [11], security [2], [7], [38], power control [12], [21], [36], [35], address configuration and MANET applications [18].

Many papers about the performance evaluation of MAODV have been published [4], [5], [24], [30], [8]. These papers have addressed reliability, scalability, and other criteria regarding long-lived connections. Long-lived connections mainly evaluate the route maintenance mechanism because the construction of the multicast delivery tree is done at the beginning of the simulation, and the rest of the operation of the protocol depends on detecting link breakages and maintaining them [24].

As mentioned earlier, long-lived connections evaluate mainly route maintenance. MAODV performance may be affected with short-lived connections, and the weight (i.e. most of the protocol operation) could be on route discovery process. The main goal of this paper is to study MAODV routing protocol when short-lived connections are used. In other words, the performance of MAODV protocol regarding various connection models will be investigated and evaluated.

Section snippets

Connection models

Two types of connection models are presented here, long-lived and short-lived connection models. In short-lived, the session lifetime of multicast group transmissions is considered short. Such applications that need this type of connections are small queries, short transactions, message exchange, tactical information exchange in military/emergency environments, and resource discovery [1], [33]. Short-lived connections also mean that small amount of data are transferred within small amount of

MAODV specification

The Network Simulator NS-2 includes the standard implementation of AODV protocol. MAODV protocol is implemented according to AODV implementation in [5]. The initial MAODV implementation has two limitations. Firstly, only group members can send multicast traffic to multicast group. Secondly, the multicast data packets are unicast, which waste bandwidth [40], [41].

The version used in this paper, which is implemented by Yufang Zhu and Thomas Kunz overcomes the previous limitations. In this

Performance evaluation metrics

The following are some of the performance metrics that IETF recommends to use to judge the merit of any routing protocol [44]. Firstly, The packet delivery ratio (PDR) is the total number of data packets received divided by the total number of data packets supposed to be delivered. This number represents the effectiveness of a protocol in delivering data to the intended receivers within the network. Secondly, The Control Overhead (CO) is the total number of control packets transmitted divided

Simulation environment and network parameters

The packet size that has been used in the following experiments is 256 bytes. The transmission range was 250 m. The Radio propagation model was the Two-ray Ground Reflection model [26], [34].

The Random Way-point mobility model was considered in our experiments [43]. Each mobile node moves randomly from a random location to a random destination with a randomly chosen speed (uniformly distributed between 0 and some maximum speed). Once the destination is reached, another random destination is

Determining the traffic rate for connection models

The description of the connection, whether it is long-lived or short-lived, depends on the amount of sent data packets per second. In order to determine this value, the first group of experiments was carried out. The aim of these experiments was to determine the amount of data packets, which should be sent every second, so the connection could be described as long-lived or short-lived.

Three experiments were done in order to select the traffic rate for both connection models. The number of

Conclusions

We have evaluated MAODV protocol for both connection models (long-lived and short-lived). While varying mobility speed, number of senders, and group size. Four metrics have been used to evaluate the effects of changing the previous parameters on MAODV protocol in both connection models.

The effectiveness of MAODV protocol in delivering data packets to their destinations in long-lived situation is decreased when the number of senders increased from 1 to 10 senders. The efficiency of delivering

Future work

We recommend to investigate further the relationship between the number of senders, receivers, and network nodes. To what extend the number of senders can be increased in a network consisting of (N) nodes and (M) receivers (where M  N). What would be the recommended number of senders for that network? What would be the maximum number of receivers that can be served by MAODV protocol in that network without exceeding a threshold (determined by MAODV users)? What would be the largest number of

Wesam A. AlMobaideen ([email protected]): Born in 1975 in Amman – Jordan. Ph.D. Degree in Computer Networks at the University of Bologna – Bologna – Italy – 2003. Master degree in computer science at the University of Jordan – Amman – Jordan – 1999. B.S. degree in computer science at Mu’ta University – Karak – Jordan – 1997. Current Position: Assistant.

References (44)

  • E.K. Al-Tarawneh, Mobile Ad hoc Network’s Route Maintenance Analysis in Reactive Routing Protocols. Master thesis, the...
  • M. Asfand-e-Yar et al.

    Secure route path formation in ad hoc on-demand distance vector routing protocol

    Information Technology Journal

    (2004)
  • J. Broch, D.A. Maltz, D.B. Johnson, Y. Hu, J. Jetcheva, A performance comparison of multi-hopwireless ad hoc network...
  • R. Chandra, V. Ramasubramanian, K.P. Birman, Anonymous gossip: improving multicast reliability in mobile ad-hoc...
  • E. Cheng, On-demand multicast routing in mobile ad hoc networks. Master thesis. Carleton University, Ontario, Canada,...
  • C. Cordeiro et al.

    Multicast over wireless mobile ad hoc networks: present and future directions

    IEEE Network

    (2003)
  • X. Du, Y. Wang, J. Ge, Y. Wang, XiDian. A method for security enhancements in AODV protocol, in: Proceedings of the...
  • T. Gopalsamy, M. Singhal, D. Panda, P. Sadayappan, A reliable multicast algorithm for mobile ad hoc networks, in:...
  • C. Gui, P. Mohapatra, Scalable multicasting in mobile ad hoc networks, in: Proceedings of the IEEE INFOCOM’2004, Hong...
  • X. Hong et al.

    Scalable routing protocols for mobile ad hoc networks

    IEEE Network Magazine

    (2002)
  • A. Iwata et al.

    Scalable routing strategies for ad-hoc wireless network

    IEEE Journal of Selected Areas on Communications

    (1999)
  • C.E. Jones et al.

    A survey of energy efficient network protocols for wireless networks

    Wireless Networks

    (2001)
  • T. Kunz, E. Cheng, On-demand multicasting in ad-hoc networks: comparing AODV and ODMRP, in: Proceedings of IEEE...
  • J.F. Kurose et al.

    Computer Networking, A Top-Down Approach Featuring the Internet

    (2002)
  • S. Lee, M. Gerla, C. Chiang, On-demand multicast routing protocol, in: Proceedings of IEEE Wireless Communications and...
  • S. Lee, W. Su, J. Hsu, M. Gerla, R. Bagrodia, A performance comparison study of ad hoc wireless multicast protocols,...
  • W. Liao et al.

    Family ACK tree (FAT): supporting reliable multicast in mobile ad hoc networks

    IEEE Transactions on Vehicular Technology

    (2003)
  • T. Lin

    Mobile Ad-hoc Network Routing Protocols: Methodologies and Applications. Doctoral dissertation in Computer Engineering

    (2004)
  • J. Luo, P.T. Eugster, J. Hubaux, Route driven gossip:probabilistic reliable multicast in ad hoc networks, in:...
  • M.K. Marina, S.R. Das, Routing performance in the presence of unidirectional links in multihop wireless networks, in:...
  • T. Ozaki, J.B. Kim, T. Suda, Bandwidth-efficient multicast routing for multihop ad-hoc wireless networks, in:...
  • G. Pei

    Scalable Routing Strategies for Large Ad Hoc Wireless Networks. Doctoral Dissertation

    (2000)
  • Cited by (6)

    Wesam A. AlMobaideen ([email protected]): Born in 1975 in Amman – Jordan. Ph.D. Degree in Computer Networks at the University of Bologna – Bologna – Italy – 2003. Master degree in computer science at the University of Jordan – Amman – Jordan – 1999. B.S. degree in computer science at Mu’ta University – Karak – Jordan – 1997. Current Position: Assistant.

    Hani Mahmoud Abd Al-Raheem Al-Mimi (E-mail: [email protected]) MS Computer Science from, University of Jordan, B.Sc. Computer Science from Al-Zaytoonah University of Jordan, currently is an instructor at Al-Zaytoonah University of Jordan.

    Fawaz A.M. Masoud (E-mail: [email protected]) is an Associate Professor at the University of Jordan in the Department of the Computer Information Systems. He has earned his Ph.D. from the University of Liverpool, UK on 1987 in Quality Metrics in Software Engineering. He has published over 40 research papers in an international refereed journals and conferences. His research interests are in: Software Systems Development Methodologies, Software Engineering, Algorithms Reengineering, Computer Networks, Software Systems Retrieval, Computer Arabization, Quality Metrics and Quality Control, Databases, Applications of Artificial Intelligence, and Computer Education and Distance Learning.

    Dr. Emad A. Qaddoura (E-mail: [email protected]) Emad Qaddoura is currently the dean of Faculty of Computer Science and Information Technology at Applied Science University (ASU) in Jordan. Emad has been in the academia for about 5.5 years. Prior to joining the Academic field, Emad has worked with Nortel Networks in Texas for about 13.5 years in a research and development environment in areas related to Internet evolution such as Internet Mobility, wireless voice and data, and data communication technologies. Emad has held several positions at Nortel Networks such as member of scientific staff and management positions starting from technical manager, senior manager and director and chief architect of research groups. Additionally, Emad has worked with Fujitsu in Texas for about 2 years as a firmware engineer for a fiber optics multiplexer.

    View full text