SEARCH
0-9 A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
Prev | Current Page 455 | Next

Yingshu Li, My T. Thai, and Weili Wu

"Wireless Sensor Networks and Applications"

For example, N sensors are deployed in an area to
measure temperature. To compute the average temperature accurately, the
data from all the N sensors should be obtained and aggregated, which will
cost a lot of power. Random sampling techniques can be adopted in this case.
By satisfying user-requested-accuracy, only the data fromM sensors (M < N)
should be used. Thus, the cost is reduced dramatically. A traditional database
does not support such a technique.
The sixth di?®erence is that, query processing techniques in a traditional
database system are not suitable for sensor networks. The reasons are as
follows:
??? The query optimization technique in a traditional data-base system is
based on a fixed cost model and statistical information. There is no reliable
statistical information about data in sensor networks, and it is very di?±cult
to forecast the action of data streams from sensors. The query plan in
sensor networks must minimize the power consumption and be suitable
for the actions of data streams.
??? A traditional database system locks the data and unlocks them when errors
occur. This is unsuitable for sensor networks since the actions of data
streams are uncertain and the query may last for a long time.
??? A traditional database system operates on an existing and fixed database.
Answers for queries are fixed. In sensor networks, the objects operated on
are infinite and uncertain data streams.


Pages:
443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467