| Co-Investigator(Kenkyū-buntansha) |
MATSUDA Takahiro Graduate School of Engineering, Osaka University Assistant Professor, 大学院・工学研究科, 助手 (50314381)
TODE Hideki Graduate School of Engineering, Osaka University Associate Professor, 大学院・工学研究科, 助教授 (20243181)
YAMAMOTO Miki Graduate School of Engineering, Osaka University Associate Professor, 大学院・工学研究科, 助教授 (30210561)
|
|---|
| Research Abstract |
For high-reliable multicast communications, high-reliability support technology in IP-level and ATM-level and also integration technology of both levels are necessary. In this research, for ATM-level, self-healing, traffic control and routing problems have been treated. For IP-level, retransmission control, flow control and congestion control issues have been treated. For integration technology of these two levels, server configuration issue and address resolution issue for IP Multicast over ATM have been studied. As research results in ATM-level, we developed effective multicast-switch distribution algorithm. Self-healing technique taking account of communication quality was also developed. For reliable multicast protocol in IP-level, we showed that NAK-based retransmission control is suitable from the viewpoint of scalability. Flow control method suitable for NAK-based retransmission control has been also proposed. And we improved this proposed flow control so that it can adjust transmission rate of the sender to adequate rate when congestion occurs in the network. Our main purpose of this congestion control method is fair-share of network resource with TCP which is widely used in the Internet. We show that our proposed method share bandwidth of bottleneck link fairly with TCP traffic. For integration technique of IP and ATM level, we developed effective VC setup method for multicast server. In this method, multicast server uses multiple point-to-multipoint VC's when it disseminates data to multicast group members. Multiple VC can avoid HOL-like blocking among cells fragmented from different IP datagram.
|
