Effective content drifting method on content-based network using game theory

Content distribution networks (CDNs) are well known to be able to distribute many kinds of content on the Internet, where caches and mirror servers are used to store and distribute content. In this paper, we propose novel network architecture to distribute content, called a content-based network (CBN). Each unit of content in the CBN has meta-data including keys to retrieve content If meta-data has a parameter that indicates how often the content has recently been retrieved, we can select adequate operation ("store content", "delete content", "move content", "divide content", and "merge content") considering network traffic, capacity in each node, and also the meta-data parameters. To achieve the above, we used the game theory. In this paper, two game players and gain are defined; "content group" player and "neighboring nodes group" player. Computer simulation revealed that the proposed algorithm has the capability of minimizing the number of hops to retrieve target content.