Abstract:
Network Mobilty (NEMO) means moving the mobility functional from mobile nodes to a mobile network's routers, it is able to change its attachment point to the Internet in a manner that is transparent to attached nodes. With the development of various wireless and mobile technologies, mobile multicast becomes more important because it can provide many applications such as video conference, on-line Interactive game, video/audio broadcast and real time news/applications distribution etc….
A Network Mobility (NEMO) system is built up by at least one Mobile Router (MR) and a set of Mobile Network Nodes (MNNs). The MR handles all routing related tasks to provide connectivity between the MNNs and an access network including mobility management. Correspondingly the MR also subscribes to multicast groups and forwards emerging multicast traffic on behalf of a MNN. For optimize routing of multicast data a hierarchical multicast agent is introduced as an entity providing an anchor to the multicast tree.
The agent based (AB) procedure combines the strengths of Bi-directional Tunneling (BT) procedure in terms of low multicast tree reconstruction and Remote Subscription (RS) procedure in terms of minimum routing path distance, while counteracting their disadvantages.
The multicast anchor agent (MAA) is a multicast router enabled to receive and forward multicast traffic it should be located in between the AR and the Home agent (HA) of the NEMO’s MR. All multicast data aimed at nodes outside the NEMO are tunneled to the MAA to separate mobility issues from the multicast tree. Because the multicast traffic is characterized by session duration, direction to or from the NEMO and service parameters as Quality of service (QoS) requirements, the suitable mechanism to support multicast services in NEMO environments is agent based approach. It could be solve most NEMO multicast problems, When a route optimization schemes is applied to avoid pinball routing problems.
The research discusses the cost functions of the three multicast schemes (BT, RS and AB) that developed from a probability model of engagement between any MNNs in multicast group to a HA. So the performance parameters required taken under continuous session duration constraints, taken in account the effects of nesting topology of network mobility system and movement of it NEMO cluster. The cost functions of providing multicast among BT, RS and AB are compared.
Then we used the probabilities of multicast anchor agent forwarding data to their destinations, to formulate the multicast path function as a 0-1 knapsack problem_ to show the selected items (multicast paths) by MAA in combinatorial optimization selection manner. It was shown that the performance of AB multicast schemes is better than BT and RS, the probability of engagement is increasing function when the number of HAs is fixed and the numbers of nodes joining to multicast tree are increasing, and