PROJECT TITLE :

Two-Aggregator Topology Optimization using Multiple Paths in Data Center Networks - 2017

ABSTRACT:

In this paper we have a tendency to specialize in the matter of knowledge aggregation using two aggregators in an exceedingly information center network, where the supply racks are allowed to separate their knowledge and send to the aggregators using multiple paths. We tend to show that the matter of finding a topology that minimizes aggregation time is NP-arduous for k = two, three, 4, where k is the utmost degree of every ToR switch (number of uplinks in a high-of-rack switch) in the info center. We tend to conjointly show that the matter becomes solvable in polynomial time for k = 5 and vi and conjecture the identical for k > 6. Experimental results show that, for k = 6, our topology optimization algorithm reduces the aggregation time by as a lot of as 83.thirty twop.c and reduces total network traffic by as abundant as 99.fivepercent relative to the torus heuristic, proposed in [1], which readily proves the significant improvement in performance achieved by the proposed algorithm.