Title :
Fault-Tolerance Schemes for Hierarchical Mesh Networks
Author :
Zurawski, Jason ; Wang, Dajin
Author_Institution :
Montclair State University, NJ
Abstract :
In [3], a hierarchical configuration for mesh was developed. The proposed scheme divides a mesh network into uniform smaller clusters. Each of these clusters contains a leader (or monitor) to communicate with the other members of the group. Leaders are then required to communicate with other leaders to form groups at higher levels. The hierarchical approach has been shown to reduce the communication cost by reducing the overall distance traveled by messages in the network [1, 2, 3]. Experiments were conducted on the hierarchical configuration to simulate different activities that it may be used for. Simulations were designed to test various sizes of the underlying mesh, as well as potential cluster sizes that may be utilized. In efforts to see if additional improvements could be made, a variety of throughputs of data were tested for the system.
Keywords :
Clustering algorithms; Computer science; Computerized monitoring; Condition monitoring; Costs; Fault tolerance; Mesh networks; Potential well; Testing; Throughput;
Conference_Titel :
Parallel and Distributed Computing, Applications and Technologies, 2005. PDCAT 2005. Sixth International Conference on
Print_ISBN :
0-7695-2405-2
DOI :
10.1109/PDCAT.2005.140