Operator-Scheduling Using Dynamic Chain for Continuous-Query Processing

Sensor networks are a sort of wireless networks; monitoring and collecting data about the natural phenomena. These sensor data behave very differently from traditional database sources: they are continuous arrival in multiple, rapid, time varying, possibly unpredictable, unbounded streams, and keeping no record of historical information. Continuous-query processing for these data streams must be run using an efficient scheduler; the development of the original chain algorithm for scheduler has focused only on minimizing the maximum run-time memory usage, ignoring the important aspect of output latency. During bursts in input streams, chain suffers from tuple starvation, thereby incurring a high latency for these tuples. In this paper the proposed dynamic chain algorithm which is used in the scheduler is very powerful in reducing memory requirements for the system and has a very good performance in the latency issue, which is the drawback of the other scheduling algorithms.