A New Approach for Maximizing Performance of Network with Heterogeneous Flows

Abstract

Communication networks supporting heterogeneous traffic flows should have performance optimization mechanisms. This is essential as there are flows such delay sensitive and delay insensitive flows. Treating both the flows in the same fashion is not a suitable solution. Both the flows are to be treated separately in order to optimize performance in terms of maximizing throughput and reducing congestion.In order to handle the complex flows over communication networks, the existing design models are inadequate. Recently Li et al. proposed a mathematical framework that could handle heterogeneous flows for optimizing network performance. In this paper we implement that framework through simulations. On top of it we also implement a distributed congestion control algorithm for improving throughput and reducing congestion. The empirical results revealed that the framework is effective in improving performance of communication networks.

Index: Terms–Load balancing, heterogeneous flows, routing, and utility maximization