Towards this end, we motivate
and provide a taxonomy of WSN applications, based on their objectives, data
delivery requirements and tra?±c characteristics. Such a classification enables
us to partition and organize the design space, and to achieve the right mix of
generality and specificity in our designs. However, there are some significant
problems which surface only during experimental studies. Therefore it is necessary
to use the taxonomy-based approach, in conjunction with the lessons
learnt from sensor network experiments. We have provided a brief survey of
28
some of the significant problems in WSNs, in both the above contexts. Such
an approach also enables us to identify future directions for research, some of
which we list below.
1. Event Detection and Reporting: As mentioned in Subsection 5.1.1,
addressing is a major cause of overheads in an event-detecting sensor network.
So it is important to devise smart and scalable addressing schemes,
for the MAC and routing protocols in a WSN. Now, for localizing an event
or a target, the authors in [28, 37] argue that nodes need to exchange information
about their geographical location with their neighbors. Devising
and using addressing schemes based on geographical location, at an appropriate
level of granularity, so that they can be used for MAC, routing
and localization of the event, is an important problem.
2. Data Gathering and Periodic Reporting: Although the problem of
energy-e?±cient routing has been studied extensively, there are other important
issues such as overhearing at MAC layer which results in significant
energy expenditure, and the lack of end-to-end reliability which
results in high packet loss.
Pages:
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85