But no detailed performance comparison
between these schemes is available.
In this chapter, we propose Forest, a new clustering scheme, and compare
its performance with two other well-known clustering schemes, LEACH and
Max-Min D-Cluster. Each of these three algorithms represents a category
of clustering schemes, i.e., energy-based clustering, identifier-based clustering
and topology-based clustering. Each of the three schemes performs well in
some aspects and has certain drawbacks in others. LEACH performs well
with small energy dissipation and appropriate cluster sizes. Yet, LEACH??™s
clusters are not distributed as evenly as Forests does and it doesn??™t perform
well in presence of node mobility. Max-Min D-Cluster performs well only
when the node density of the network is small. Forest performs pretty well
with the least intra cluster energy dissipation in data transmission, the most
uniform cluster distribution and high stability in mobile environment. The
only significant drawback is the high message overhead. However, the cluster
formation consumes much less energy than the regular data transmission, and
it is done less frequently. The benefit of using Forest can be observed over a
long term.
When the network is static, the structure of Max-Min clusters and Forest
clusters will not change. In many scenarios, sensor nodes will be static, so
clusterheads will continue to be clusterheads until they are depleted of energy.
Pages:
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577