Title of article :
Design and analysis of asymptotically optimal randomized tree embedding algorithms in static networks
Author/Authors :
Li، نويسنده , , Keqin، نويسنده ,
Issue Information :
روزنامه با شماره پیاپی سال 2005
Pages :
23
From page :
141
To page :
163
Abstract :
The problem of dynamic tree embedding in static networks is studied in this paper. We provide a unified framework for studying the performance of randomized tree embedding algorithms which allow a newly created tree node to take a random walk of short distance to reach a processor nearby. In particular, we propose simple randomized algorithms on several most common and important static networks, including d-dimensional meshes, d-dimensional tori, and hypercubes. It is shown that these algorithms, which have small constant dilation, are asymptotically optimal for embedding healthy trees. Our analysis technique is based on random walks on static networks. Hence, analytical expressions for expected load on all the processors are available.
Keywords :
Hypercube , Static network , Randomized tree embedding , torus , Asymptotic performance , Dynamic load distribution , MeSH
Journal title :
Performance Evaluation
Serial Year :
2005
Journal title :
Performance Evaluation
Record number :
1569837
Link To Document :
بازگشت