Databases Reference
In-Depth Information
FIGURE 9.2: DHGN distributed event detection framework. (Muhamad
Amin, A.H., Khan, A.I., and Raja Mahmood, R.A. “A distributed event
detection scheme for wireless sensor networks,” In Proceedings of the 7th
International Conference on Advances in Mobile Computing and Multime-
dia (MoMM '09), pp. 295-299. ©2009 ACM, Inc. Reprinted by permission.
http://doi.acm.org/10.1145/1821748.1821804)
Note that the SI module is intended to be deployed in a controlling node,
such as the base station, or a super-node. The DHGN subnet module is lo-
cated in each WSN subnet that is located within a specific sensory region.
The event classification process (evaluation of event/non-event signals) in the
DHGN event detection scheme is a dual-layer process. The first layer focuses
on the subpattern recognition in the DHGN subnet, whereas the second layer
involves pattern classification using a voting scheme that is conducted by
the SI module. Subpattern recognition is the process of determining the re-
call/store status of an input subpattern. This process is conducted in DHGN
subnets. The output of this process is either a recalled index of the stored
subpattern or a new index for the input subpattern. This index is sent to the
SI module for pattern classification. Note that the DHGN considers an event
as a pattern that represents a state of normality or abnormality for the entire
sensory network.
