Database Reference
In-Depth Information
are surveyed. At the end, Section 6 contains a summary of the chapter
along with conclusions.
2. Model-Based Sensor Data Acquisition
In this section, we discuss various techniques for model-based
1
sensor
data acquisition. Particularly, we discuss pull- and push-based sensor
data acquisition methods. In general, model-based sensor data acquisi-
tion techniques are designed for tackling the following challenges:
Energy Consumption:
Obtaining values from a sensor requires high
amount of energy. In contrast, since most sensors are battery-powered,
they have limited energy resources. Thus, a challenging task is to mini-
mize the number of samples obtained from the sensors. Here, models are
used for selecting sensors, such that user queries can be answered with
reasonable accuracy using the data acquired from the selected sensors
[2, 17, 16, 27, 28].
Communication Cost:
Another energy-intensive task is to communi-
cate the sensed values to the base station. There are, therefore, several
model-based techniques proposed in the literature for reducing the com-
munication cost, and maintaining the accuracy of the sensed values [41,
18, 66, 12].
Table 2.1.
Summary of notations.
Symbol Description
S
Sensor network consisting of sensors
s
j
,where
j
=(1
,...,m
).
s
j
Sensor identifier for a sensor in
S
.
v
ij
Sensor value observed by the sensor
s
j
at time
t
i
, such that
v
ij
∈
R
.
m
.
v
i
Row vector of all sensor values observed at time
t
i
, such that
v
i
∈
R
V
ij
Random variable associated with the sensor value
v
ij
.
2.1 Preliminaries
We start by describing our model of a sensor network and establish-
ing the notation that is utilized in the rest of the chapter. The sensor
network considered in this chapter consists of a set of stationary sensors
S
=
. The value sensed by a sensor
s
j
at time
t
i
is
denoted as
v
ij
, which is a real number. In addition, note that we use
s
j
,
where
j
=(1
,...,m
), as sensor identifiers. In certain cases the sampling
interval could be uniform, that is,
t
i
+1
−
{
s
j
|
1
≤
j
≤
m
}
t
i
is same for all the values of
1
We use
model-based
and
model-driven
interchangeably.
