In directed diffusion [1], the sink selects preferred paths by reinforcing the paths
with lower delay or better packet delivery ration. Packets are aggregated opportunistically
along the paths while being forwarded to the sink. Using this approach, packets
from two sources may not share the same path even if they are close. This reduces
the possibility of aggregation because packets are forwarded on different routes.
Even if sources share some paths, there may be only few instances of aggregation because
aggregation happens opportunistically. This approach favors paths with lower
delay.
Chapter 13 Data Aggregation in Wireless Sensor Networks 339
Greedy Aggregation, on the other hand, favors minimization of energy consumption
over reducing latency by increasing path sharing. As illustrated in Figure 4, the
path sharing number in (b) is higher than (a). This has the benefit of reducing overall
transmitting power consumption if packets can be aggregated when the routes merge
into one path.
(a) (b)
Fig. 4. Increasing path sharing reduces the number of transmissions. Assume packets are always
aggregated at the merging node. (a) requires eight transmissions while (b) requires only
six transmissions.
Like directed diffusion, the sink floods its interests to all nodes in the beginning,
and nodes construct the gradients when the interests are propagated across the network.
When sources detect an event os interest, they send the data back to the sink
along the gradient paths.
Pages:
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542