A time based markov model for automatic position-dependent services in smart home

A smart home is likely in the near future. An important ingredient in an intelligent environment such as a home is automatic services, which means home system itself could know or predict what the inhabitant want to do, and so provide inhabitant the services automatically. Many researches reveal that most of the services in smart home are location-dependent so the automatic services must be built on the location awareness. In this paper we model inhabitant location pattern as a time based Markov model (TMM). The simulation result shows that compared to the other models, the TMM has a set of benefits such as less time complexity, higher prediction accuracy and faster convergences rate. These benefits make TMM meets the requirements of automatic services in smart home.