High-performance top-k complex event detection from RFID event stream

RFID technology holds the promise of real-time identifying, locating and monitoring physical objects. Complex Event detection is the key part of the RFID middleware. Most of the current methods of complex event detection can´t support large-scale RFID applications since they have not considered much about large sliding window. On the other hand, sometimes it is not possible to get all complex events from fast RFID stream and it is useful to get k most important complex events. In this paper, a high performance top-k Complex Event detection method is proposed. Based on Nondeterministic Finite Automaton and Active Instance Stacks, this method processes RFID stream parallel to get better performance. Heuristic methods are also used to process top-k complex events with high performance. The experimental study shows that this method is efficient to detect top-k complex events from fast RFID stream.