187

Virtual Subnet Internetworking with IPv6 Multicast Membership for Mobile Ad Hoc Networks

Tzu-Chiang Chiang*, Yueh-Min Huang* and Fenglien Lee***

*Department of Engineering Science, National Cheng-Kung University, Taiwan, ROC **Department of Information Management, Hisng-Kuo University of Management

***National Center for High-Performance Computing Taiwan n9892105@ccmail.ncku.edu.tw

ABSTRACT Mobile ad hoc networks need the flexibility to collect more than two devices equipped with wireless communication and networking capability. In recent years, the wireless network has been attracting a lot of attention, and due to this wireless devices have enjoyed a tremendous rise in popularity. However, the broadcast storm becomes a very serious problem for ad hoc networks to migrate into the third generation (3G) telecommunication for the application of group conferences. The main concept of virtual local area network (VLAN) technology is the capability to group users into broadcast domains, which divides a LAN into logic, instead of physical, segments and reduces the traffic overhead. With this characteristic, we propose a novel interoperability network model integrating a self-organizing ad hoc network and Internet/a conventional network with the same virtual local area network. Moreover, we describe a protocol to establish the VLAN broadcast domains by using the IPv6 multicast-membership in ad hoc networks and perform IP-based network communications in a multi-switch backbone. Since the VLAN technology functions by logically segmenting the network into different broadcast domains, packets can only be delivered between fixed/mobile nodes with the same VLAN identity (group member). Therefore we can prevent the broadcast storm problem in MANET. The emulations show the proposed protocol in ad hoc network environment from 20 to 80 nodes with different size of VLANs groups. The result shows that the throughput of our protocol delivering packets to all nodes in the same VLAN group is more than 84% of all the cases. Furthermore, compared with the flooding, the efficiency of communication is improved up to 51%.

Key Words: Virtual Local Area Network, VLAN, IPv6 multicast, ad hoc networks.

1. INTRODUCTION

Ad hoc networks can deploy rapidly and freely comparing with the traditional wireless networking systems, it does not need the pre-existing network infrastructures. In recent years, the wireless network has been attracting a lot of attention, and due to this wireless devices have enjoyed a tremendous rise in popularity. A few of the applications are absorbing greater amounts of bandwidth. Therefore today’s corporations routinely buy the fastest computers and fastest network equipment on the market and want these machines to run on the fastest safest network possible Author and year[1][2][3]. An ad hoc network is a collection of mobile nodes

188

that want to communicate to each other, but has no fixed links like wireless infrastructure network. Each node acts as a router and is responsible for dynamically discovering other nodes it can directly communicate with. The network topology of an ad hoc network changes frequently and unpredictable, so routing and multicast become extremely challenging [4][5].

The goal for developing multicast is that there are applications that need to send a packet from a multicast source to a group of nodes (multicast receivers). There are two types of configurations [6] for an ad hoc wireless multicast protocol. One of the configuration type is a tree-based protocol (e.g., adhoc Multicast Routing (AMRoute) [7], and Ad hoc Multicast Routing protocol utilizing Increasing id-numberS (AMRIS) [8]), and the other is a mesh-based protocol (e.g., On-Demand Multicast Routing Protocol (ODMRP) [9], and Core-Assisted Mesh Protocol (CAMP) [10]). However, the broadcast storm becomes a very serious problem for ad hoc networks to migrate into the third generation (3G) telecommunication for the application of group conferences. The main concept of virtual local area network (VLAN) technology is the capability to group users into broadcast domains, which divides a LAN into logic, instead of physical, segments and reduces the traffic overhead. The network switch was invented to assist networks to improve this situation. The LAN switch network permits users to be combined into as “Virtual LANs.” A VLAN is a logical, rather than a physical connectivity, collection of network devices. In the router-based network devices are recognized by their physical location in the network. The network-layer address is used to inform the router physical segment where must send data to. A VLAN behaves like an ordinary LAN, but connected devices don't have to be physically connected to the same segment. The VLAN allows a flexible mechanism; simply grouping physical ports together, or can combine existing hubs, routers and backbone with dedicated switched ports, wide area networks, and more. While clients and servers may be distributed anywhere on a network, they are grouped together by VLAN technology, and broadcasts are sent to devices within the same VLAN.

With this characteristic, we propose a novel interoperability network model integrating a self-organizing ad hoc network and Internet/a conventional network with the same virtual local area network. Moreover, we describe a protocol to establish the VLAN broadcast domains by using the IPv6 multicast-membership in ad hoc networks and perform IP-based network communications in a multi-switch backbone. Since the VLAN technology functions by logically segmenting the network into different broadcast domains, packets can only be delivered between fixed/mobile nodes with the same VLAN identity (group member). Therefore we can prevent the broadcast storm problem in MANET.

1.1 Research contributions We consider a core network of 802.1Q-liked switches interconnected by trunk lines. This

network can span one or more small towns in sparsely populated areas, interconnecting communities as well as company LANs. The intent is to provide both telephony and data

189

services over the same technology (Ethernet), stack (IP), with seamless integration with the Internet. Our work chooses to distribute the bandwidth according to a hierarchical link sharing model.

The rest of this paper is organized as follows. We describe the Virtual Local Area Network in Section 2. Section 3 in detail for the infrastructure of virtual LAN internetworking in ad hoc networks, Section 4 simulated the constructing VLAN successful rate with NS2 simulation. Finally, section 5 provides our concluding remarks and futures works.

2. PROBLEM DESCRIPTION

The motivation behind our approach is that network partitioning can improve critical functions such media access, routing, mobility management and virtual circuit set-up, while reducing signaling/ control overhead. It can be observed in this type of network that portioning may result also in lower congestion compared to one large network. This paper discuss an architecture based on a specific logical topology superimposed over a physical topology (determined by transmission coverage of network nodes); the architecture selects links to be activated (logical links) out of a pool of physical links. Our main concern is finding an efficient logical topology and a suitable routing procedure which result in high performance and reliability. In this architecture, network nodes are grouped into two types of clusters (subnets): physical and virtual, and may dynamically change their affiliation with these subnets due to their mobility. Each node is allocated an address based on its current subnet affiliation. We consider networks that have several tens to several hundreds mobile nodes. It is assumed that there exists a channel access protocol which resolves contentions and/or interference in the network. [A mobile radio network architecture with dynamically changing topology using virtual subnet] Broadcasting in a shared wireless medium may result in multiple nodes receiving a transmission destined for a single node, and ultimately, in multiple transmission mutually interfering at a single node. Nodes can reduce the chances of interference by separating transmissions in time, space, frequency, or spreading code. By coordinating this separation instead of acting independently, nodes can further reduce the chances of interference and hence increase network throughput. The cluster-based control structure provides a natural organization of network nodes that simplifies coordination of transmission among neighboring nodes. [Ad Hoc Networking book]

With the fast development of wireless technology, ad hoc network is walking out from research papers and becoming more and more closer to the common consumers. For example, in order to save battery power and reuse spectrum, the forth generation (4G) wireless communication is considering using low-powered ad hoc mesh networks instead of traditional star networks. In this case, end-user wireless handsets form a peer-to-peer network in which

190

they act as both end terminals and wireless routers that are part of the overall network infrastructure. Broadcasting is one of the basic communication methods for ad hoc networks and has many applications. Unlike wired networks, ad hoc networks are usually formed and maintained in a dynamic manner. This means a node will have no knowledge about other nodes or services in an ad hoc network when it joins the network. To discover other devices and services, nodes usually have to broadcast query/request packets to find out what is out there in the network. Besides service and device discovery, data distribution across all network nodes can take advantage of broadcasting for efficiency. In network layer, broadcasting is a fundamental component for many routing protocols. [A Border-aware Broadcast Scheme for Wireless Ad Hoc Network] However, the simple broadcasting without a rebroadcasting bounding mechanism at each node may result in an excess of redundancy, channel contention, and collisions. This phenomenon is called the Broadcast Storm Problem [The Broadcast Storm Problem in a Mobile Ad Hoc Network]. Redundancy indicates a situation where a node hears the same messages from more tha n one neighbors. Channel contention is due to the different nodes which are simultaneously trying to rebroadcast the received messages thus contending for the shared media, increasing the probability of collisions. To address redundancy, the decision whether or not rebroadcast must be controlled at each node receiving the message. For contention and collision, all nodes trying to rebroadcast rely on backoff mechanism with randomly selected slots. Y.-C. Tseng et al. [6] suggested several schemes to alleviate the broadcast storm problem, namely: the counter-based scheme, the distance-based scheme, and the location-based scheme. Although the authors indicates the location-based scheme as the best alternative, it requires all nodes to be equipped with a global positioning system (GPS) device to provide the appropriate accuracy of longitude and latitude. The distance-based scheme provides an higher level of reachability with respect to the counter-based scheme but it does not offer the same reduction of rebroadcast as its counterpart. [A Bounding Algorithm for the Broadcast Storm Problem in Mobile Ad Hoc Networks]

2. VIRTUAL LOCAL AREA NETWORK

In a local area network, data link-layer broadcast and multicast traffic is delivered to all devices, but this traffic cannot go beyond the LAN boundary. In the past, shared cabling or hubs were the boundaries for LANs. A VLAN is an administratively configured LAN or broadcast domain. VLANs facilitate easy administration of logical groups of stations that can communicate as if they were on the same LAN. They also facilitate easier administration of moves, adds, and changes in members of these groups [14]. VLANs configured by using Media Access Control addresses can recognize when a device has been moved to another port

191

on a switch. VLAN management software can then automatically reconfigure that device into its appropriate VLAN without the need to change the device's MAC or IP address. Traffic between VLANs is restricted. Bridges forward unicast, multicast and broadcast traffic only on LAN segments that serve the VLAN to which the traffic belongs. So the packets are only switched between ports that are designated for the same VLAN. By confining packet broadcast on a particular LAN only to the LANs within the same VLAN, switched virtual network avoid wasting bandwidth, a fault in traditional switched network where packets are often forwarded to LANs that even do not require them. Hence offers benefits in term efficiency use of bandwidth, flexibility, performance and security [16][17][18][19].

The IEEE 802.1Q standard defines the operation of VLAN Bridges that permit the definition, operation and administration of Virtual LAN topologies within a Bridged LAN infrastructure. The IEEE's 802.1Q standard was developed to address the problem of how to break large networks into smaller parts so broadcast and multicast traffic would not grab more bandwidth than necessary. The mechanisms support VLAN in Bridged LAN environment including the frame format and how the frames are relayed to destinations. Any frame has a VLAN-tagged to associate with the incoming port’s VID (VLAN ID). The Filtering Database (FDB) stored addressing information and frame-filtering information in the form of Mac addresses and VLAN entries.

Switches are one of the core components of VLAN communication. Switches provide the intelligence to make filtering and forwarding decisions by packet and to communicate to other switches and routers within the network. 802.1Q-compliant switch ports can be configured to transmit tagged or untagged frames. The switch implementing the standard is as following typical operations:

First, when a frame enters the switch it is checked for errors. An error-free frame will be associated with a VID as described above so that frame’s learning and forwarding is relative to the VLAN it belongs to. If the Ingress Filtering set to enable and the incoming port is not a member of this VALN, the frame will be rejected.

Second, an accepted frame will be submitted to the forwarding process to be relayed to other ports, and at the same time the switch observes its source MAC address, source port, VID associated and other necessary addressing information and automatic using them to update the FDB. The process of forwarding uses the MAC address and VID of the frame to index into the FDB to find out where the frame shall be relayed. And eventually the frame is sent through those corresponding outbound ports if not filtered out by Egress Filter, which also decide whether the outgoing frame should carry a VID. The logical interrelationships are illustrated in Figure 1.

192

Figure 1. The VALN logical interrelationships It is obvious that LANs switches afford essential improvement in performance and

dedicated bandwidth across the networks, with the intelligence necessary for VLAN segmentation.

3. INFRASTRUCTURE OF VIRTUAL SUBNET INTERNETWORKING

As shown in Figure 2, three groups of mobile nodes form an internetworking wireless ad hoc network with IP-based Internet. The communication between each node in this network infrastructure is established by using VLAN-enabled multi-route equipments and wireless multi-hop paths. Some MNs in this ad hoc network want to access the Internet. 加入一些為何使用 VLAN 的關念與想法 First, in order to access the IP-based Internet and form the virtual network, a set of VLAN-enabled Multi-route Routers (VMR) is established in advance. The VMR are connected to the Internet and communicate with the mobile nodes in wireless ad hoc networks via wireless transceivers. The VMR consists of three components:

1.) IP-based component with traditional IP-based protocol suite installed is designed to connect with IP-based Internet (either wired or wireless);

2.) Ad hoc component with ad hoc related protocols installed is connected with ad hoc networks through a wireless interface;

3.) VLAN component with virtual LAN related core component of VLAN communication providing the intelligence to make filtering and forwarding decisions.

193

Figure 2. An internetworking wireless ad hoc network with IP-based Internet The protocol architecture stacks for mobile nodes and VMRs are represented in Fig. X. In

the physical layer, a mobile node uses by all the 802.11 variants, including infrared, FHSS, DSSS, OFDM and HR-DSSS, provides one of the five permitted transmission techniques to send a MAC frame among those mobile nodes. The standard techniques detail operational specifications for wireless connectivity for infrastructure and ad hoc networks within a local geographical area. The physical layer corresponds to the OSI physical layer, but the data link layer is split into two sublayers. Medium Access Control (MAC) sublayer controls the channel allocation details two operational modes: the Distributed Coordination Function (DCF) and the Point Coordination Function (PCF). The Logical Link Control (LLC) sublayer corresponds to hide the differences between the different 802 variants and makes the network layer transparent. Besides the wireless LAN standards IEEE02.11, some extended versions, HiperLan or Bluetooth could also serve this communication. The network layer is concerned with getting packets from the source to the destination. Besides an IP-based ad hoc routing protocol (e.g., proactive protocols and reactive protocols) is used, we also retool the protocol to enlarged IP address fields from 32 bit to 48 bit or 128 bit for virtual subnet using at layer 2, or handling ad

194

hoc networks of IPv6 addressable computers. In higher layers, conventional IP-based protocols are applied for wireless communications.

The VMR contains protocols and acts as a default gateway to communicate between the

conventional Internet and the wireless ad hoc networks. In response to the purpose for virtual subnet, VMRs are based on specially-designed Virtual-Subnet aware switches by identifying the virtual-subnet field in the frame header. Now let us take a look at the frame format with virtual-subnet tag, which is shown in Fig. #. The virtual protocol ID is adapted from 802.1Q and has the value 0x8100. The Tag field contains three subfields. The virtual-subnet identifier is the low-order 12 bits, representing which virtual-subnet does the frame belong to? On the conventional Internet side, it runs the usual Internet protocol and our adapted virtual-subnet protocol. On the ad hoc network side, it sends and receives packets using an ad hoc routing algorithm (e.g., AODV[ ], DSDV [] , DSR [] and TORA []). To make the virtual-subnet function correctly, two different routing tables are used including configuration tables for mapping virtual-subnet identifier and access links. The VMR may also contain protocols in higher layers, in case there is a need for translation in these layers (e.g., conversion of usual TCP to TCP for wireless channels).

The architecture partitions a mobile ad hoc network into logically independent subnetworks. Network nodes are members of physical and virtual subnets and may change their affiliation with these subnets due to their mobility.

Initially, the network is partitioned into a set of disjoint clusters, which are termed physical subnet and are based on node locality. Members of different physical subnets are clustered together to form virtual subnets, each of which ideally spans all physical subnets and is used to provide communication paths among distant nodes.

Figure #. The Virtual-Subnet frame format To connect the IP-based Internet communication, the mobile nodes need to discover the

existence of the VMRs, which it belongs to, and join one of the VMRs first. This could be done either by listening to a VLAN-enabled Multi-route Router Advertisement (sending by VLAN-enabled Multi-route Router) or sending a VLAN-enabled Multi-route Router Request message (sending by mobile node) at the initial state as shown at Figure 3. After successful

195

VLAN-enabled Multi-route Router discovery, the mobile node registers with one of the discovered VLAN-enabled Multi-route Routers. We use Multicast Listener Discovery (MLD) with virtual LAN tag for IPv6 and Mobile IPv6 [20] to manage the multicast membership and macro-mobility in this internetworking system. Thus, when registers with a certain VMR, mobile node generates an IP care-of-address (CoA) with the IP prefix of the selected VMR. If an mobile node is approaching a new VMR which can serve it better than the currently registered VMR, a handover procedure is performed between old and new VMRs. After the switch to the new VMR is completed, the mobile node generates a new IPv6 CoA with the new network prefix. Thus, the CoA of a certain mobile node always contains the information of its current location.

196

Figure 3. A VLAN-enabled Multi-route Router Advertisement Due to the same IP prefix of CoAs, all of the nodes attaching to the VMR can be regarded

as members of one IP subnet, and the attached VMR can be regarded as the default gateway for virtual subnet . In other words, the entire ad hoc network is logically separated into several virtual subnets (as identified in Figure 4 with different colours). These virtual subnets are connected to the core IP network through the gateway routers.

Unicast routing in the ad hoc side is operated in a hierarchical way based on the subnet partitioning. I.e. packets addressed to a destination in the same subnet can be directly forwarded, while packets address to a destination in a different subnet must be routed through the VMR s (even if there is a direct wireless multi-hop link between the correspondent nodes).

IP multicast routing is carried out among the multicast routers only. Hosts are transparent to the routing protocol. Each multicast routers should obtain the up-to-date information about the existence of any group member on its local link. The multicast tree is then built up to deliver the multicast datagrams among all routers connected to local group members. Hosts may join or leave a multicast group dynamically. To ensure that multicast packets are delivered to all links with active receivers, a protocol called Multicast Listener Discovery (MLD) was developed for membership management between routers and hosts in IPv6. Routers use MLD to learn which multicast address has listeners on each of their attached links. On each link, at least one of the routers keeps a list of multicast addresses having listener on this link. This list

197

is referred as membership list in the rest of this paper. As an asymmetric protocol, different behaviours for multicast listeners and routers are specified in MLD. Three types of MLD messages are defined: Query, Report and Done.

Figure 4: The entire ad hoc network is logically separated into several subnets

198

199

4. SIMULATION ANALYSIS

Numerous simulations of our proposed protocol have been performed using the NS2 simulation package. The mobility model used in each of the simulations is in a random direction. In each simulation, nodes are initially placed randomly within a predefined 600m x 600m grid area. Each node then chooses a random direction between 0 and 360 degrees and a speed from 0 to 20 meter/second. Once the node reaches the boundary of the area, it chooses a period of time to remain stationary. After the end of this pause time, the node chooses a new direction, this time between 0 to 180 degrees, adjusted relative to the wall of the area on which the node is located. This process repeats throughout the simulation, causing continuous changes in the topology of the underlying network.

First, we compare the efficiency of the proposed method to Flooding using computer simulation. Networks are randomly generated within 600m x 600m grid. There are n nodes in the grid with k VLANs and nodes can communicate each other within transmission range. Each VLAN is formed with randomly members. The re-routing adapts the AODV algorithm.

200

In the simulation study, we vary n and k to compare how the network sizes and VLAN sizes affect the efficiency.

Figure 5 shows the result for network sizes from 50 nodes to 80 nodes with 5 to 8 VLANs. As a consequence of the greater node density, our proposed method is sufficient to deliver packets to all nodes in a VLAN group. There are some nodes moving out of transmission range, which cannot re-route successfully. It shows that the percentage up to 84% in the 80 nodes existing 8 VLANs by our proposed method. In the same condition, original flooding achieves only 80%. The main impact of the improvement is that our proposed method re-routes successfully before nodes moving out of transmission range, original flooding has too many redundant transmissions (control packet flood) to reach destination in time.

Similar results for the percentage of the overhead of our proposed method and original flooding are shown in Fig. 6. We assume the flooding overhead to be 100% for comparing basis. It shows that the overhead of our proposed method in 5 VLANs is from 50% to 39% when nodes number from 50 nodes to 80 nodes. And the overhead of our proposed method in 8 VLANs is from 39% to 32% when nodes number from 50 nodes to 80 nodes. It gives a clear result that the overhead improves up to 68% in the 80 nodes existing 8 VLANs. Furthermore, compared to different sizes of VLAN, our proposed method was still efficient when the sizes of VLAN changing.

Figure 7 refers to a 60-node ad hoc network with 5 VLANs. It shows the percentage of successful multicast for nodes whose velocity varies from 0m/s to 20m/s (around 70km/h) using three different routing protocols, and all the nodes of the addressed multicast group received the packet more than 90% when the velocity is less than 10m/s (around 35km/h) for our proposed method. Furthermore, when the velocity of nodes is more than 4m/s, the packet delivery ratio of flooding will drop immensely as the velocity of nodes speeds up.

Delay improvement ratio with respect to flooding is shown in Figure 8. As expected, simulations show that our proposed method improves on the average delay of successful packet completion up to 50%.

201

Figure 5. the percentage of successful deliver packets

0

10

20

30

40

50

60

70

80

90

100

50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80

Nodes size

perc

enta

ge o

f ov

erhe

ad 5 VLAN FCVP6 VLAN FCVP

7 VLAN FCVP

8 VLAN FCVP

Flooding

Figure 6. The percentage of the overhead of our proposed method and original flooding

202

Packet Delivery Ratio

40

50

60

70

80

90

100

0 2 4 6 8 10 12 14 16 18 20

Speed (meters/sec)

Pack

et D

eliv

ery

Rat

io (%

)

FCVPAODVFlooding

Fig. 7: Packet delivery ratio for three different routing protocols

Delay improvement with respect to flodding Ratio

10

20

30

40

50

0 2 4 6 8 10 12 14 16 18 20Speed (meters/sec)

Del

ay im

prov

emen

t Rat

io (%

) .

FCVPAODV

Fig. 8: Delay improvement ratio with respect to flooding

203

Average Route Acquisition Latency

0

10

20

30

40

50

60

70

20 30 40 50 60 70 80 90 100 110 120

Network sizes

Del

ay (m

sec)

.

5 VLANs 10 VLANs15 VLANs 20 VLANs

Fig. 9: Average route acquisition latency

Control Overhead

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

0 2 4 6 8 10 12 14 16 18 20

Speed (meters/sec)

Pack

ets

50 Nodes100 Nodes150 Nodes200 Nodes

Fig. 10: Control overhead

Figure 9 represents the average route acquisition latency for the simulations. It indicates that

there is a significant different between the different network sizes and the numbers of VLAN. The delay is more a factor of the network sizes, topologies and the queuing delays experienced

204

at the individual nodes. Since using the forwarding cache routes can be discovered with minimal delay.

Figure 10 shows that the amount of overhead in terms of CREQ, CREP and ACK message sent. The simulations presented utilize the forwarding cache algorithm with four phases. Each CREQ is broadcast across the entire VLAN network, so there are a large number of RREPs generated, specifically in the larger-sized network.

The results present in this session quantify the improvement in our proposed method that results from the use forwarding cache table to reduce redundant transmissions. However, if the network increases greater nodes density, the improvement increases more obviously.

5. CONCLUSIONS AND FUTURE RESEARCH In this paper, we have described the adapted VLAN protocol for ad hoc networks. The main

objective of our protocol is efficient to progress the behaviours of VLAN in ad hoc networks. We propose a novel interoperability network model integrating a self-organizing ad hoc network and Internet/a conventional network with the same virtual local area network. Moreover, we describe a protocol to establish the VLAN broadcast domains by using the IPv6 multicast-membership in ad hoc networks and perform IP-based network communications in a multi-switch backbone. Since the VLAN technology functions by logically segmenting the network into different broadcast domains, packets can only be delivered between fixed/mobile nodes with the same VLAN identity (group member). Therefore we can prevent the broadcast storm problem in MANET.

We plan to identify the suitable cache table refreshing mechanism on the proposed method in the future works. We will also generalize the clustering method to progress the behaviours of VLAN so that they can be applied in ad hoc wireless networks.

ACKNOWLEDGEMENT This research is supported by NICI IPv6 Infrastructure Development Division, Taiwan, Grant #I-0400.

REFERENCES REFORMAT USING THE TEMPLATE

A. Gupta and D. Ferrari, (1995), Resource Partitioning for Real-time Communication, IEEE/ACM Trans. on Networking, 3.5, 501–518.

[2] N. Kavak, (1995), Data Communications in ATM Networks, IEEE Network, vol.9, no.3,May/June 1995.

205

[3] Rajaravivarma, V.; (1997), Virtual local area network technology and applications, Proceeding of the Twenty-Ninth Southeastern Symposium on , 9-11 March 1997 Pages:49 - 52

[4] Internet Engineering Task Force (IETF) Mobile Ad Hoc Networks (MANET) Working Group Charter. http://www.ietf.org/html.charters/manet-charter.html.

[5] J. Jubin and J.D. Tornow, “The DARPA Packet Radio Network Protocols,” Proceeding of the IEEE, Vol.75, no. 1, Jan. a987, pp. 21-32.

[6] Sung-Ju Lee; Su, W.; Hsu, J.; Gerla, M. and Bagrodia, R.; “A Performance Comparison Study of Ad Hoc Wireless Multicast Protocols.” INFOCOM 2000, Nineteenth Annual Joint Conference of IEEE Computer and communications Societies. Proceedings, IEEE vol:2 :2000, pages 565-574.

[7] E. Bommaiah, M. Liu, A. McAuley, and R. Talpade, “AMRoute: Adhoc Multicast Routing Protocol,” Internet-Draft, draft-talpade-manet-amroute-00.txt, Aug. 1998, Work in progress.

[8] C. W. Wu, Y.C. Tay, and C.-K. Toh, “Ad hoc Multicast Routing protocol utilizing Increasing id-numberS (AMRIS) Functional Specification,” Internet-Draft, draft-ieft-manet-amris-spec-00.txt, Nov. 1998, Work in progress.

[9] S.J. Lee, M. Gerla, and C.-C Chiang, “On-Demand Multicast Routing Protocol,” In Proceedings of IEEE WCNC’99, New Orleans, LA, Sep. 1999, pp. 1298-1304.

[10] J.J. Garcia-Luna-Aceves and E.L. Madruga, “Core-Assisted Mesh Protocol,” IEEE Journal On Selected Areas in Communications, Vol. 17, no. 8, Aug. 1999, pp. 784-792.

[11] D. J. Baker and A. Ephremides, (1981), The architectural organization of a mobile radio network via a distributed algorithm, IEEE Trans.Commun., pp. 1694–1701, Nov. 1981

[12] D. J. Baker, J. Wieselthier, and A. Ephremides, (1982), A distributed algorithm for scheduling the activation of links in a self-organizing, mobile, radio network, in Proc. IEEE ICC’82, pp. 2F.6.1–2F.6.5.

[13] I. Chlamtac and S. S. Pinter, (1987), Distributed nodes organization algorithm for channel access in a multihop dynamic radio network, IEEE Trans.Comput., pp. 728–737, June 1987.

[14] IEEE Std 802.1Q, 2003 Edition. (2003), IEEE standards for local and metropolitan area networks, Virtual bridged local area networks.

[15] J. Jubin and J.D. Tornow, (1987), The DARPA Packet Radio Network Protocols, Proceeding of the IEEE, Vol.75, no. 1, pp. 21-32, Jan. a987.

[16] M. Gerla and J. T.-C. Tsai, (1995), Multicluster, mobile, multimedia radio network, ACM-Baltzer J. Wireless Networks, vol. 1, no. 3, pp. 255–265, 1995.

[17] Royer, E. and Perkins, C., (1999) Multicast Operation of the Ad-hoc On-Demand Distance Vector Routing Protocol, MobiCom ’99, Seattle, WA, Aug, pp. 207-218.

[18] Tzu-Chiang Chiang, Ching-Hung Yeh and Yueh-Min Huang,(2004) A Forwarding Cache VLAN Protocol (FCVP) in wireless Networks, 4th IASTED International Multi-Conference on Wireless and Optical Communications - WOC 2004, July 8-10, 2004,Banff,Canada

206

[19] Wu, C., Tay, Y., and Toh, C.,(1998), Ad Hoc Multicast Routing Protocol Utilizing Increasing id-numbers (AMRIS) Functional Specification, Internet-Draft, Nov. 1998.

[20]D.B. Johnson, C. E. Perkins, and J. Arkko, “Mobility support in IPv6.” IETF internet draft, June 2002.