| Previous | [ 1] | [ 2] | [ 3] | [ 4] | [ 5] | [ 6] | [ 7] | [ 8] | [ 9] | [ 10] | [ 11] | [ 12] | [ 13] | [ 14] | [ 15] | [ 16] | [ 17] | [ 18] | [ 19] | [ 20] | [ 21] |
¡@
John Sum, Jie Wu and Kevin Ho
1Department of Information Management
Chung-Shan Medical University
Taichung, 402 Taiwan
E-mail: pfsum@csmu.edu.tw
2Department of Computer Science and Engineering
Florida Atlantic University
Boca Raton, Florida 33431, U.S.A.
E-mail: jie@cse.fau.edu
3Department of Computer Science and Communication Engineering
Providence University
Taichung, 433 Taiwan
E-mail: ho@pu.edu.tw
While restricted rule-k has been succeeded in generating a connected dominating
set (CDS) of small size, not much theoretical analysis on the size has been done. In this
paper, an analysis on the expected size of a CDS generated by such algorithm and its relation
to different node density is presented. Assume N nodes are deployed uniformly
and randomly in a square of size LN ¡Ñ LN (where N and LN ¡÷ ¡Û); three results are obtained.
(1) It is proved that the node degree distribution of such a network follows a
Poisson distribution. (2) The expected size of a CDS that is derived by the restricted
pruning rule-k is a decreasing function with respect to the node density n?. For n? ? 30.
it is found that the expected size is close to N/n?. (3) It is proved that the lower bound
on the expected size of a CDS for a Poissonian network of node density ? n is given by
1 ?
? 1 ? 1 { n exp( ( ? 1))} .n n n N ? ? ? ? ? The second result is of paramount importance for practitioners.
It provides the information about the expected size of a CDS when the node density
? n is between 6 and 30. The data (expected CDS size) for this range can hardly be
provided by simulations.
Received September 15, 2006; accepted February 6, 2007.
Communicated by Ten H. Lai, Chung-Ta King and Jehn-Ruey Jiang.