For example, Li,
et. al., [28, 37] argue that in order to facilitate the distributed detection of a
target, the routing of the data should be based on the geographical location
of the sensor nodes, rather than their hardware identities. The detection of
the target could be done either based on some a priori knowledge about the
sensor signals that the target produces, or it could be done simply based on
thresholding the energy of the sensor signals. The former approach is more
useful in detecting multiple targets and in classifying and maintaining target
identities. Some of these approaches are suggested and studied in [9, 31, 32].
Chen, et. al., [9] provide an overview of the issues involved in using beamforming
techniques to detect a target. They also make a case for using collaborative
signal processing among sensors, in order to accurately localize the
target. In [31, 32], the problem of target detection and classification is considered
at two levels. Firstly, each sensor can use signal processing techniques
based on the signals obtained from all of its physical sensing elements, and
secondly, sensors can also collaborate with neighboring sensors. While target
detection could be performed using a single sensing element, for target
classification, collaboration both within and amongst sensors is very useful.
Once multiple targets are detected and classified, each target identity needs
to be tracked.
Pages:
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69