Abstract
In this paper, we present a novel max-flow min-cut based algorithm to solve the flow routing problem in the Software Defined Network Controller. Routing using traditional shortest path first algorithms often results in bottlenecks that cause performance degradation including higher energy use, reduced throughput, and increased slowdown. Our algorithm uses the max-flow min-cut algorithm to identify potential bottlenecks in order to avoid them in the flow routing decisions. Our simulations show that our max-flow min-cut based algorithm improves the network performance by minimizing the mean wait time by 15.1%, minimizing the mean slowdown by 6.1%, minimizing the maximum completion time by 9.6%, and maximize the mean throughput by 18.3% compared to the Shortest Path algorithm. Explicitly considering congestion in determining routes, such as with our Max-Flow Min-Cut algorithm, is necessary to maximize performance.