Appropriate clustering can reduce the need for global coordination
and restrict most of the sensing, data processing and communication activities
within clusters, thus can improve resource e?±ciency and prolong network
lifetime ([1], [2], [3], [4] and [5]). Clustering can also provide load balancing if
appropriately configured.
The salient advantage of using clusters in a sensor network comes from
in-network data aggregation. Data aggregation has emerged as a basic tenet
in sensor networks ([6], [7] and [8]). The key idea is to combine data from different
sensors to eliminate redundant transmissions. With clustering, sensor
nodes transmit their local information to their clusterheads, and clusterheads
aggregate the received information and forward it to the base station. Periodic
re-clustering can select nodes with higher residual energy to act as clusterheads.
Network lifetime is prolonged by (1) reducing the number of nodes
contending for channel access (2) summarizing network state information and
updates at clusterheads through intra-cluster coordination, and (3) routing
349
Yadi Ma and Maggie Cheng
through an overlay among clusterheads with relatively small network diameters
[9].
In this article, we propose a new clustering scheme, Forest, and compare it
with two other clustering schemes, LEACH (Low-Energy Adaptive Clustering
Hierarchy [10]) and Max-Min D-cluster ([2]). The three schemes each represent
a di?®erent category of clustering method.
Pages:
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557