Sub-queries are also continuous queries. They scan, filter
(select) and integrate the corresponding infinite real-time data streams and
produce continuous partial results to return to the global query processor.
The global query processor processes the results further and returns the fi-
nal result to users. Local query processing is the key of continuous query.
The same as global continuous query processing, the execution of continuous
sub-queries will take a long time. During this period, the sensor nodes, the
characteristics of the data they produce, and the workload of the sensor nodes
are all changed. So it is very important to make the local processor adapt to
the change of the environment.
A Continuously Adaptive Continuous Queries over Streams (CACQ) can
be used in a local query processor in sensor networks. It will be introduced in
two aspects: single continuous query and multiple continuous query. Suppose
316
Chapter 12 Data Management in Sensor Networks
no join operator exists in a query.
Single Query
The query is decomposed into a set of operators that constitute the processing
that must be applied to every tuple flowing into the system. Since we
are not considering join at the moment, the only operators that can exist are
scan operators, which fetch tuples, and select operators, which filter those
tuples based on a user-specified Boolean predicate. For now, we assume that
the queries contain only conjunctions (ANDs) of predicates.
Pages:
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509