1-2hit |
Yasushi ICHIKAWA Takashi TOMIMOTO Toshihiko SHIMOKAWA Yuko MURAYAMA
A peer-to-peer (P2P) Contents Delivery Network (CDN) is a system in which the users get together to forward contents so that the load at a server is reduced. Lately, we have high-speed services for an access to the Internet such as the Asymmetric Digital Subscriber Line (ADSL). Some broadcasters may not have such services because they have only dial-up services and wireless services as PHS and a mobile phone to broadcast live. A problem with P2P CDN is its overhead to construct a distribution tree. It becomes a crucial problem when a broadcaster has only a low-speed access to the Internet, and we propose a P2P CDN system which reduces such an overhead. A server peer is the root peer of a distribution tree and provides users with contents. With the existing algorithms, new peers measure a Round Trip Time (RTT) and a throughput from a broadcaster site when they join the distribution tree. With our algorithm, a new peer sends the server peer a Search Request message which is forwarded throughout the distribution tree until a suitable peer which has enough bandwidth to accomodate is found finally so that the new peer will measure a throughput to that peer. The problem with our algorithm is that as the number of users in the tree increases, the new peer will be preoccupied with measurement, because it may find many suitable peers as its parent candidates. To solve this problem, we introduce a Time To Stop Broadcast (TTSB) on the Search Request message in order to reduce the number of measurement. We have compared the traditional algorithm with ours by simulation. From the simulation results, we have found that our method is effective when a server peer has a low-speed access to the Internet, while the users have a high-speed access.
Chenyu PAN Merdan ATAJANOV Mohammad BELAYET HOSSAIN Toshihiko SHIMOKAWA Norihiko YOSHIDA
With the rapid spread of information and ubiquitous access of browsers, flash crowds, a sudden, unanticipated surge in the volume of request rates, have become the bane of many Internet websites. This paper models and presents FCAN, an adaptive network that dynamically optimizes the system structure between peer-to-peer (P2P) and client-server (C/S) configurations to alleviate flash crowds effect. FCAN constructs P2P overlay on cache proxy server layer to distribute the flash traffic from origin server. It uses policy-configured DNS redirection to route the client requests in balance, and adopts strategy load detection to monitor and react the load changes. Our preliminary simulation results showed that the system is overall well behaved, which validates the correctness of our design.