Search

Learning patterns of human behavior from sensor data is extremely important for high-level activity inference. We show how to extract a person's activities and significant places from traces of GPS data. Our system uses hierarchically structured conditional random fields to generate a consistent model of a person's activities and places. In contrast to existing techniques, our approach takes high-level context into account in order to detect the significant locations of a person. Our experiments show significant improvements over existing techniques. Furthermore, they indicate that our system is able to robustly estimate a person's activities using a model that is trained from data collected by other persons. [ABSTRACT FROM AUTHOR]

This paper describes a secure multicast infrastructure for large-scale group communications in Mobile Internet and proposes a key management protocol based on the infrastructure. The multicast communication domain is logically divided into several administrative areas with a key server associated with each area. All the key servers participate in a Public Key Infrastructure (PKI) as trusted entities known by the subgroup members. Therefore, it's efficient to minimize the re-key overhead implemented in the mobile host tier. The simulation results show that the proposed protocol has better performance compared to the centralized protocols without PKI support. The numbers of the real re-key messages and the re-key events are reduced to approximately 30% and 65%, respectively. [ABSTRACT FROM AUTHOR]