Fast On-Line Summarization of RFID Probabilistic Data Streams

RFID applications usually rely on RFID deployments to manage high-level events. A fundamental relation for these purposes is the location of people and objects over time. However, the nature of RFID data streams is noisy, redundant and unreliable and thus streams of low-level tag-reads can be transformed into probabilistic data streams that can reach in practical cases the size of gigabytes in a day. In this paper, we propose a simple on-line summarization mechanism, which is able to provide small space representation for massive RFID probabilistic data streams while preserving the meaningful information. The main idea behind the proposed approach is to keep on aggregating tuples in an incremental way until a state transition is detected. Probabilistic tuples are processed as they arrive, hence avoiding the use of expensive offline disk based operations, and the output is stored in a probabilistic database in such a way that, as we also experimentally prove, a wide range of probabilistic queries can be applicable and answered effectively.