Research On P2P Data Exchanging Model Oriented To Visual Steering

With the development of network, it becomes possbile to realize Visual Steering Systems in large grid environment. These distributed systems can be used in many fields, such as Military Simulation, Collaborative Design, E-Business, Telemedicine, MMORPG and Virtual Cruise.Massive data exchanging and processing among nodes is the bottleneck of deploying such distributed systems. With the development of Peer-to-Peer technology, these data streaming can be transmitted by P2P systems on overlay network. Scalability of P2P network can resolve the problems like lack of bandwidth in traditional C-S models.Current P2P Streaming technology is mainly oriented to Media Multicast, to balance the load of networks, whose topology of management is centralized and not very scalable. Considering the requirement of interoperability and real-time interactivity, this paper will analyze typical P2P streaming models and design improved P2P streaming model for visual steering.The main research and innovative work is as follows: First, to distribute the management function to every nodes in P2P system, to make it becomes Decentralized Structured Topology and ensure it can adapt the join, leave and fail of nodes rapidly; Second, to apply concept of QoS, QoS Function and heuristic algorithm for p2p streaming and allow nodes to customize the portion and QoS of request data, in order to optimize the total quality of data transmission according requirments; Third, to build a prototype system upon virtual network environment of BRITE project, to realize new model and its algorithms upon the system.It is indicated by the result of emulational experiments that new model is adapt to visual steering much better than traditional ones.