On the other hand, reactive routing schemes obtain routing information
on demand. In other words, there is no periodic control information that is
exchanged between the nodes. Reactive routing is more suitable for event
detection, since the data transfer from the sensors to the sink takes place
relatively infrequently. Also, due to the many-to-one nature of the data flow, it
is possible to use smart mechanisms to route the data to the sink. For example,
[26] uses an address-light, integrated MAC and routing protocol (AIMRP), in
which the nodes do not maintain any dynamic routing information. Each node
simply keeps track of its ???tier-id??? (a measure of its distance from the sink in
terms of number of hops). Whenever a node detects an event, its report is
forwarded towards the sink by a node with a lower tier-id. Thus, it is possible
to exploit the many-to-one nature of the data flow to reduce the addressing
and routing overheads.
Idle-listening and Power-saving Algorithms
As noted in Subsection 4.3, a considerable amount of energy is consumed by
the radio in idle-listening. This can have a significant impact on the network
lifetime. Due to the fact that the sensor events are rare, the sensor nodes
spend very little time in actually communicating information. It would be
reasonable to turn the radio module of these nodes o?®, and only keep the
sensing module on. However, most sensor networks require the use of multihop
communication, since the communication range of an individual sensor
could be much smaller than the size of the region.
Pages:
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59