Database Reference
In-Depth Information
collects measurements from the environment
receives data from other sensors in its network neighborhood
possibly performs some computations on the received and collected
data
broadcasts data to the WSN
enters the sleep mode until it wakes up for the next cycle
Two major challenges in WSNs refer to the effectiveness of communi-
cation between sensor nodes and the e
ciency of data processing. The
first challenge spawns the design space of network protocols. Various
protocols ensure the smooth and automatic communication between sen-
sor nodes [4]. A network protocol specifies when a sensor node samples
the environment, to whom it reports to and from whom it receives data.
There are three categories of protocols, according to how they organize
the network: i) tree-based topologies, ii) multi-path-based topologies
and iii) the hybrid ones that combine the first two approaches. The sec-
ond challenge, i.e., the eciency of data processing, is mostly the focus
of the Database community. Several observations realize this task from
different aspects. Firstly, in most applications, only a small part of the
data is interested by the user. Using descriptive queries like
Q
1
helps
drop uninteresting data as soon as possible. Secondly, sensor data often
exhibit strong spatial-temporal correlations [40]. It is reasonable to pre-
dict the sensor readings with certain confidence, based on historical and
easily observable sensor readings. Finally, when imprecise answers are
allowed, returning approximated ones (with error bounds) often lead
to large energy savings. This survey focuses on query processing for
WSNs, mostly from the database point of view. The remainder of this
chapter is organized as follows. Section 2 introduces the characteristics
and limitations of sensor nodes, which motivate the challenges involved
in designing WSN systems. Section 3 surveys common topologies of
WSNs; different algorithms optimize certain topologies. Section 4 intro-
duces the data storage techniques in WSNs; some applications require
the data generated by sensor nodes to be stored inside the network for
a period of time. Section 5 discusses data acquisition and aggregation
techniques in WSNs; a topic widely investigated in the Database com-
munity. Section 6 studies the model-based data acquisition, probabilistic
queries and event detection in WSNs. Finally, Section 7 concludes this
chapter.
