Anda di halaman 1dari 4

Journal of Computer Applications (JCA) ISSN: 0974-1925, Volume V, Issue 3, 2012

Optimized Multicasting System to Evade Packet Loss in Mobile Adhoc Networks

M.Murali a,*, S.Vijayalakshmi b, 1 Abstract - Group communication takes up an imperative role in MANETs. Multicasting provides this technique in small scale networks Group Membership management is a tough task in dynamic topology. To conquer the problem we suggest a protocol Efficient Geographic Multicasting Protocol (EGMP). EGMP uses virtual two-tier zone based structure for making the membership management and packet delivery ratio an efficient one. It uses bi-directional tree structure for multicast packet delivery. The position information is used to guide the zone structure edifice, multicast tree construction, and multicast packet forwarding, which resourcefully reduces the overhead for route searching and tree structure maintenance. Finally, we design a scheme to handle empty zone problem faced by most routing protocols using a zone structure. The projected architecture reduces the packet loss and gives the higher delivery ratio. EGMP has widely lower header size and it increases the data transmission rate over a large scale network.
Index Terms MANETs, EGMP.

Even though some GCA implementations on the legacy Internet use client/server transactions (e.g., Instant Messaging typically uses central servers to detect user presence and availability), the eventual communication pattern among GCA users is more similar to P2P than to client/server systems. MANETs can be established without any pre-existing infrastructure. This makes plausible to run GCA completely on-demand, as soon as a group of users decide that they want to communicate. Based on these remarks, we believe that Group Communication Applications is an exciting field to be explored to identify valuable applications for MANET users. II. EGMP OVERVIEW A. Related Work Conventional topology-based multicast protocol comprises tree-based protocols (Eg.., [1]-[3]) and mesh-based protocols (Eg.., [4], [6]).In tree-based protocol a tree is constructed for the entire group to forward the packets. Mesh- based protocols in turn enlarges the multicast tree with additional paths to make it more efficient. Even supposing we built multicast tree it is not competent for large scale of network. EGMP instead uses location aware approach for efficient packet forwarding and scalable for both large size and small size network. In SPBM [5], the network terrain is divided into quad tree with L levels. The top level is the whole network and the bottom level is constructed by basic squares. In [7], we proposed efficient and robust geographic multicast protocol for MANET. We introduce zone-supported geographic forwarding to reduce a routing failure and provide mechanism to handle zone partitioning. We further introduce a path optimization process to handle multiple paths packet delivery. B. Proposed System We propose an efficient geographic multicast protocol (EGMP).It uses a hierarchical structure to implement scalable and efficient group membership management. We design a scheme to build and maintain the intrazone and interzone topology for supporting scalable and efficient multicast forwarding. We make use of the position information to implement hierarchical group membership management, and combine location service with the hierarchical membership management to avoid network-range location searches for the group members, which is scalable and efficiently. With nodes self-organizing into zones, a zone-based bi-directional tree is built in MANET environment. We introduce an important concept zone depth, which reflects the relationship between a member zone and the zone where the root of the tree exists. We also design a scheme to handle the empty zone problem, a challenging problem in designing a zone-based protocol. 78

I. INTRODUCTION Efficient support of group communication is critical for most ad hoc network applications. However, MANET group communications issues differ from those in wired environments for the following reasons: The wireless communication medium has variable and unpredictable characteristics and the signal strength and propagation fluctuate with respect to time and environment. Further, node mobility creates a continuously changing communication topology in which routing paths break and new ones form dynamically. A high share of applications running on the current Internet is based on the client-server paradigm, the World Wide Web being the most popular example. Group-Communication Applications (GCA) such as Instant Messaging and Distributed Games are increasingly diffused, and take a completely different communication pattern. In GCA, a group of users want to communicate with each other, instead of issuing requests to a central server.
Manuscript received 20/July/2012. Manuscript selected 14/Aug/2012. M.Murali, Assistant Professor, Department of Information Technology, Sona College of Technology, Salem, India, E-mail: S.Vijayalakshmi, Department of Electronics and Communication Engineering, Sona College of Technology, Salem, India, E-mail:

Optimized Multicasting System to Evade Packet Loss in Mobile Adhoc Networks

Our analysis results indicate that the cost of the protocol defined as the per-node control overhead remains constant regardless of the network size and the group size. Our simulation studies confirm the scalability and efficiency of the proposed protocol. C. EGMP Overview EGMP Overview EGMP provides consistent and scalable membership management and packet forwarding through a virtual two-tier zone based architecture. At the lower level with reference to a virtual origin nodes are self-organized themselves into a set of zones and a leader is nominated to manage a local group membership. At the upper level a leader is responsible for a node to join or leave a multicast group. A local information will be integrated in a design to built a efficient zone based multicast tree. EGMP supports bi-directional multicast packet forwarding along the tree structure. At the upper layer, the multicast packets will forward both upstream to the root zone and downstream to the leaf zones of the tree. At the lower layer, when a zone leader receives the packets, it will send them to the group members in its local zone. There are many issues that need to be concentrate on EGMP to make it more efficient and scalable. The issues related to zone management are zone construction, maintenance, zone leader selection with minimum overhead and potential packet loss when a group members move across a zones. The issues related to packet forwarding are efficient multicast path selection, handling of empty zone problem and tree maintenance during mobility [8]. D. Terms in EGMP Zone: The network terrain is divided into square zones. r: Zone size, the length of a side of the zone square. The zone size is set to r. rt=p2, where rt is the transmission range of the mobile nodes. To reduce intra-zone management overhead, the intra-zone nodes can communicate directly with each other without the need of any intermediate relays. Zone ID: The identification of a zone. A node can calculate its zone ID (a, b) from its position coordinates (x, y) as: a =[ xx0r ], b = [ yy0r ], where (x0; y0) is the position of the virtual origin, which can be a known reference location or determined at network setup time. A zone IDs are always positive. Zone center: For a zone with ID (a,b), the position of its center (xc; yc) can be calculated as: xc = x0 + (a+ 0:5) r,yc = y0 + (b + 0:5) r. A packet destined to a zone will be forwarded towards the center of the zone. zLdr: Zone leader. A zLdr is elected in each zone for managing the local zone group membership and taking part in the upper tier multicast routing. Tree zone: The zones on the multicast tree. The tree zones are responsible for the multicast packet forwarding. A tree zone may have group members or just help forward the multicast packets for zones with members. Root zone: The zone where the root of the multicast tree is located. Zone depth: The depth of a zone is used to reflect its distance to the root zone. For a zone with ID (a; b), its depth is: depth = max(ja0 aj; jb0 bj);

where (a0; b0) is the root-zone ID. the root zone has depth zero, the eight zones immediately surrounding the root zone have depth one, and the outer seven zones have depth two.

Figure 1. Zone structure and multicast session example

With the introduction of virtual zone, EGMP does not need to track individual node movement but only needs to track the membership change of zones, which significantly reduces the management overhead and increases the robustness of the proposed multicast protocol. E. Neighbor Table Generation and Zone Leader Election Each node periodically sends a BEACON message to distribute its position to aid leader election and reduce overhead. EGMP simply insert a flag in a BEACON message to identify a zone leader. To reduce BEACON overhead, instead of broadcasting message in a fixed interval, EGMP follows adaptive time interval. A non-leader node will send a BEACON message for every period of intvalmax and when a node moves to other zone. A leader node has to send a message for every period of intvalmin to broadcast its leadership role. When a node receives a BEACON message, it records a nodeID, position and flag to its neighbor table. When a node enters a zone it broadcast a BEACON message to announce its existence. It wait for Intvalmax to receive a BEACON message from other nodes. For every period of Intvalmin it checks its neighbor table and determine its leader in four different cases:1) The neighbor table contains no other nodes in the same zone, it will announce itself as the leader.2) If the node is closer to the zone center than other nodes, it will announce its leadership role through a beacon message with the leader flag set.3) More than one node in the same zone have their leader flags set, the one with the highest node ID is elected.4) Only one of the nodes in the zone has its flag set, then the node with the flag set is the leader. F. Multicast Tree Construction EGMP, instead of connecting each group member directly to the tree, the tree is formed in the granularity of zone with the guidance of location information, which significantly reduces the tree management overhead. With a destination location, a control message can be transmitted immediately without incurring a high overhead and delay to find the path 79

Journal of Computer Applications (JCA) ISSN: 0974-1925, Volume V, Issue 3, 2012 first, which enables quick group joining and leaving. Using Zone depth concept the multicast tree has been built efficiently. The optimized path is selected to deliver a packet efficiently. When a node M wants to join the multicast group G, if it is not a leader node, it sends a JOIN REQ (M; PosM; G; fMoldg) message to its zLdr, carrying its address, position, and group to join. The address of the old group leader Mold is an option used when there is a leader handoff and a new leader sends an updated JOIN REQ message to its upstream zone. If M did not receive the NEW SESSION message or it just joined the network, it can search for the available groups by queryin its neighbors. If a zLdr receives a JOIN REQ message or wants to join G itself, it begins the leader joining procedure. If the JOIN REQ message is received from a member M of the same zone, the zLdr adds M to the downstream node list of its multicast table.
nodeID 1 10 3 13 position (x1,y1) (x10,y10) (x3,y3) (x13,y13) flag 1 0 1 1 zoneID (1,2) (1,1) (2,1) (2,2)

multicast tree connected. When a leader is moving away from a non-root tree-zone and the zone is becoming empty, it will send its multicast table to its upstream zone. The upstream zone leaders will then take over all its downstream zones, and delete this requesting zone from its downstream zonelist. The new upstream zone needs to send JOIN REPLY messages to all the newly added downstream zones to notify them the change. When receiving the JOIN REPLY messages, these downstream zones will change their upstream zone ID accordingly III. PERFORMANCE EVALUATION The simulations were run with 400 nodes randomly distributed in an area of 2400m 2400m. The nodes moved following the modified random waypoint mobility model. The moving speed of nodes are uniformly set between the minimum and maximum speed values which are set as as 1 m/s (with pause time as 100 seconds) and 20 m/s respectively except when studying the effect of mobility. IEEE 802.11b was used as the MAC layer protocol. Each simulation lasted 500 simulation seconds. Each source sends CBR data packets at 8 Kbps with packet length 512 bytes. The CBR flows start at around 30 second so that the group membership management has time to initialize and stop at 480 second. By default, there is one source, and one multicast group with 100 members. A simulation result was gained by averaging over six runs with different seeds.

Figure 2. The Neighbour table of Node 16 in Fig.1

When a member M wants to leave G, it sends a LEAVE (M; G) message to its zone leader. On receiving a LEAVE message, the leader removes the source of the LEAVE message from its downstream node list or zone list depending on whether the message is sent from an intra-zone node or a downstream zone.

Figure 4. Time Delay Vs Packet Size

Figure 3. Node Leader Sending BEACON Message

G. Empty Zone Problem A zone may become empty when all the nodes move away. The probability that a zone is empty is approximately P =er2 when the node density is and the zone size is r. Assume r = 150m, which is the zone size that allows all the nodes in the same zone to be within the transmission range, the probability of the zone being empty is: P = 0:207 if d = 70nodes=km2, and P = 0:509 if d = 30nodes=km2. We can see that the probability of a zone becoming empty is not negligible and it is critical to address the empty zone problem. In EGMP, if a tree zone becomes empty, the multicast tree will be adjusted correspondingly to keep the 80

Figure 5. Packet Delivery Ratio Vs Node Density

Optimized Multicasting System to Evade Packet Loss in Mobile Adhoc Networks

IV. CONCLUSION There is an increasing demand and a big challenge to design more scalable and reliable multicast protocol over a dynamic Ad hoc network (MANET). In this paper, we propose an efficient and scalable geographic multicast protocol, EGMP, for MANET. The scalability of EGMP is achieved through a two-tier virtual-zone-based structure, which takes advantage of the geometric information to greatly simplify the zone management and packet forwarding. Compared to conventional topology based multicast protocols; the use of location information in EGMP significantly reduces the tree construction and maintenance overhead, and enables quicker tree structure adaptation to the network topology change. We also develop a scheme to handle the empty zone problem, which is challenging for the zone-based protocols. Our results indicate that geometric information can be used to more efficiently construct and maintain multicast structure, and to achieve more scalable and reliable multicast transmissions in the presence of constant topology change of MANET. Our simulation results demonstrate that EGMP has high packet delivery ratio, and low control overhead and multicast group joining delay under all cases studied, and is scalable to both the group size and the network. REFERENCES
[1] E. M. Royer and C. E. Perkins. Multicast operation of the ad hoc on-demand distance vector routing protocol August 1999, pp. 207218. C. Wu, Y. Tay, and C.-K. Toh. Ad hoc multicast routing protocol utilizing increasing id-numbers (AMRIS) functional specification. November 1998. X. Zhang and L. Jacob. Multicast zone routing protocol in mobile ad hoc wireless networks.October 2003. C.-C. Chiang, M. Gerla, and L. Zhang. Forwarding group multicast protocol (FGMP) for multihop mobile wireless M. Transier, H. Fubler, J. Widmer, M. Mauve, and W. Effelsberg. A Hierarchical Approach to Position-Based Multicast for Mobile Ad-hoc Networks M. Gerla, S. J. Lee, and W. Su. On-demand multicast routing protocol (ODMRP) for ad hoc networks. X. Xiang and X. Wang. An Efficient Geographic Multicast Protocol for Mobile Ad Hoc Networks, Buffalo,Newyork,2006 M.Saravanakarthikeyan, M.Murali and Dr.S.sujatha Identifying Performance Metrics To Maximize Manets Throughput, ACE, 2010, IEEE, pp. 357-359


[3] [4] [5]

[6] [7] [8]

M.Murali is working as Assistant Professor in Department of Information Technology at Sona College of Technology, Salem, Tamilnadu, India for the past two years. He received M.E degree in Pervasive Computing Technologies from Anna University, Trichy, Tamilnadu, India and B.E degree in Computer science and Engineering from Kongu Engg College, Erode affiliated to Anna University, Chennai, Tamilnadu, India. He is a member of IAENG, IACSIT and IAOE. S.Vijayalakshmi Research Scholar, Anna University, Coimbatore, Tamilnadu, India working as Assistant Professor in Department of Electronics and Communication Engineering at Sona College of Technology, Salem, Tamilnadu, India for the past five years. She received M.E degree in Communication Systems from Anna University, Chennai, Tamilnadu, India and B.E degree in Electronics and Communication Engineering from Sona College of Technology affiliated to Anna University, Chennai, Tamilnadu, India.