Enhancing Adaptive Path Isolation Technique for Heterogeneous Data Fow in Data Centers

Abstract

As the resource competition and contentions in data center networks (DCNs) are frequent and to large degrees, this becomes the reason of the evolution of our concept. Hedera, JellyFish, UltraLow, HULL, DCTCP are some of the tools and algorithms which are available for the implementation of this concept. Many of the existed methodologies transact not up to the level, due to high time and space constraints, so this leaves a space for some contribution. Maintenance of the priority queue to manage the traffic and the congestion in network for data of all sizes is the basic concept to be implemented here. Priority queue contains each element having some priority confederated with it. We enforce priority queue to prioritize huge extent of data accessible in data centers. Thus, to enhance the data flow at the high congestion network of data centers proposed methodology put forwards an idea of easing the flow by maintaining the priority queue which is powered with fuzzy classification and decision tree models. The transmission delay at the real time are analyzed and the paths are assigned accordingly with the evaluation of the request matrix. Finally scheduling through recursive genetic algorithm is done and efficiently data is transmitted from the source to the destination.