In this Thesis, we aim to provide the network designer with a simple view of the network robustness to traffic matrix changes. It is accepted wisdom that changes in the traffic matrix entail capacity over-provisioning, but there is no simple measure of just how much over-provisioning can buy. This value characterizes the loss of the network computing capabilities resulting from the increase of the degree of network degradation. The average number of working processors in the degraded 4-dimensional network with a given order (degree of degradation) for the (1,1│G) resource placement is determined. Two algorithms for determining (m,d│G) – perfect resources placement and the (k│G) – resources perfect placement in a structure G of the hypercube network was given. In the work the notion of (m,d│G) – perfect resources placement in a structure G of the hypercube type is extended to (k│G) – perfect placement concept, that is a such allocation of k resources which minimizes the average distances between the working processors and resource processors in the structure G. The two types of network processors (resource processors - I/O ports and working processors) are taken into consideration. In the paper some properties of the perfect resources placement in the 4 dimensional hypercube processors network with soft degradation are investigated.
0 kommentar(er)
