An Effective Replication Technique Using Rateless Codes for Unstructured P2P Networks
Keiichi Endo, Ryosuke Hamabe, Dai Okano, Kaname Amano · Journal of Communications · 2012
It is possible to recover the original data from a certain amount of encoded data by using rateless codes, which are utilized for multicast streaming. In this paper, we propose an effective replication technique using rateless codes for unstructured P2P networks. More specifically, we propose a method to spread and find chunks generated from files efficiently. Through simulation experiments, we show that the proposed technique achieves high search success ratio and shortens the time required to obtain a file. Various Internet services are available today owing to the rapid growth of the Internet. Many of these services use the client/server (C/S) model in which a computer receiving the service (a client) is connected to a dedicated computer providing the service (the server). This model is based on the relationship between a master and a servant. The C/S model allows client data to be centrally managed on the server and thus has the advantage of being easy to design as a business model. In recent years, the rapid migration of everyday household connections to broadband has made it possible for such Internet services to transfer large volumes of content, and high- quality sophisticated services are expected in connection with these large-volume communications in the corporate world. It is extremely costly for the C/S model to provide such advanced services; the costs arise from the need to provide backup servers and to invest in equipment that can keep pace with the newest advances in processing capacity and can support an increasing number of clients. Quite simply, it is becoming more and more difficult to provide stable services using the C/S model. A new network model that is currently attracting much attention is the peer-to-peer (P2P) model. In this network model, computers (peers) on the network communicate and share files on an equal footing with each peer serving as both a server and a client. As a result, traffic is not concentrated only at the server unlike the case in the C/S model, thereby reducing the cost required for investing in new equipment.