Replication of Personal Contents Following a User with Personalized Access History and Schedule

The advent of ubiquitous computing environments enables users to access their digital contents when they want at any time once their devices are connected to the Internet. For efficient management of remote contents, we need a replication technique among servers. However, th existing approaches focus on shared contents using the access heuristics of groups of users. In this paper, we propose a personalized replication scheme especially for contents owned by one user. The proposed scheme selects target replication servers according to both the user’s past behavior and future movement plan. We periodically analyze each user’s access location in order to figure out where the user usually accesses contents. A user also can register his/her schedule information on personal or business trips to our system, which then expects that the user will be likely to access contents at a different location during the specified dates. With this simple reasoning, our replication scheme additionally replicates contents to another replication server near to the user’s destination. Simulations show that the proposed scheme reduces the expense of storage usage because we replicate contents only to selective servers, and it keeps content access time low because contents are accessed from nearby replication servers with high probability based on the user’s past and future location information. The proposed approach can be applied to distributed storage and personal content synchronization applications. Especially, it is appropriate to storage service for a photographer, who travels all over the world and deals with large amount of photograph data. As personal content management has been one of promising services in cloud computing systems, we expect that the proposed scheme also can be adopted in cloud computing nodes in order to efficiently manage contents as well.