Database Reference
In-Depth Information
1. Introduction
With the fast development of sensing and wireless communication
technology, wireless sensor networks (WSNs) become popular tools to
capture the physical worlds. A sensor is a device with one or several sens-
ing devices, a radio component, and limited computational resources.
It takes physical measurements of the environment, e.g., temperature,
light, sound, and humidity. A wireless sensor network (WSN) consists
of a base station and a set of sensor nodes. Each node is able to di-
rectly communicate with others within its radio coverage. The base
station, also called the data sink, is equipped with a radio component so
that it is able to communicate with the nearby sensor nodes and collect
their captured data. Sensors far away from the sink will transmit the
data to the sensors near the sink first, and then the data are relayed
to the sink. Depending on the size of a monitoring area, data captured
by the sensors located on the boundary of the monitor area may need
to relay multiple hops (sensors) before they reach the sink. To query
the sensing data, users submit queries at the base station, which then
reports the query results. WSNs were first applied in military and sci-
entific projects. Applications of WSNs flourish as the cost of sensors
drop, while the capabilities increase. In the past few years, WSNs have
attracted considerable interest from both the Network and Database
communities. Assume, for instance, a WSN that monitors the physical
status of a forest. An environmentalist is interested in the temperature
readings, while a biologist is interested in the level of soil moisture. In
order to capture different requirements, they submit queries at the base
station. In many applications, the query language is declarative and sim-
ilar to SQL.
Q
1
shows a typical of an extreme value monitoring query
[4 5 6], which monitors the maximum temperature readings of the forest.
Q1:
SELECT
MAX TEMPARTURE
FROM
SENSORS
WHERE
SAMPLE INTERVAL=5 mins
Q
1
treats the sensor data as a relational table and the temperature
reading as an attribute of this table. The SAMPLE INTERVAL clause
specifies the cycle length of this network, i.e., the interval between two
data collecting activities. Since sensors are battery-powered, it is cru-
cial to minimize their energy consumption in order to prolong the life
time of the network, especially when the sensors are deployed in harsh
or dicult-to-access environments, e.g., wildlife tracking [2] and habit
monitoring [3]. Making sensor nodes working in cycles is a standard way
forenergysavinginWSNs.Withinacycle,asensor
