Topology Dependent Ant Colony-Based Routing Scheme for Software-Defined Networking in Cloud

TLDR: In this paper , the authors proposed the ant colony-based shortest path routing algorithm (AC*) to illustrate typical topologies, which decouples the control plane from the data forwarding plane, and topology description language (TPDL) files are utilized as prerequisite information to establish initial topology in SDN controllers.

Abstract: The exponential growth of cloud computing has resulted in the rapid creation of data center networks (DCNs). As data center networks have grown in popularity, efficient routing has become a critical problem for maximizing network performance, scalability, and reliability. Traditional link-state algorithms are widely used in data center networks, but they take a longer convergence time. In DCN, topology-aware routing algorithms were recently discovered to be efficient. This paper proposes the ant colony-based shortest path routing algorithm (AC*) to illustrate typical topologies. To guarantee unified control of the whole network, this approach decouples the control plane from the data forwarding plane, and topology description language (TPDL) files are utilized as prerequisite information to establish initial topology in the software-defined network (SDN) controllers. Unlike other topology-aware routing algorithms, the ant colony-based shortest path routing algorithm (AC*) is designed to work on a variety of standard topologies. The AC* algorithm outperforms traditional link-state routing methods according to the results of the experiments.