Other approaches are clustering
or constructing aggregation trees. Nodes in the network form multiple clusters,
and nodes send their sensing data to their clusterheads or tree root. Packets are aggregated
at these clusterheads or roots and then sent to the sink. In this chapter, we
introduce some aggregation approaches in this current research field.
2 Directed Diffusion
Directed Diffusion [1] is a data centric routing and aggregating protocol for robust
and energy efficient communication in sensor networks. In Directed Diffusion, data
is a named attribute-value pair. The sink injects an interest which describes its data
for a particular event to the sensors in the network. The dissemination of the interest
sets up gradients, which form multiple paths from sensor nodes to the sink. When
sensors sense the events that match the interest, they send the sensing data along
multiple paths to the sink. These data are aggregated opportunistically at the intermediate
nodes. The network reinforces one or more paths on which higher data rate,
or higher quality of data, can be propagated to the sink.
The processes of directed diffusion are illustrated in Figure 2. The sink first
broadcasts the interest describing its interest for data to the network. The interest
is a set of name-value pairs. For example, if the sink is interested in receiving the
data of events for vehicle movement, the interest may look like:
type: vehicle-movement
interval: 1s
duration: 10s
rect: [100, 100, 200, 400]
The interval specifies the data report rate when a sensor senses the event, and the
duration specifies the expiration time for this interest.
Pages:
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532