ReTra: Reinforcement based Traffic Load Balancer in Fog based Network
V. Divya, R. Leena Sri · 2019
As the trend moves along the use of IoT devices and greater dependence on the internet which creates a huge amount of data, the traditional architecture consisting of cloud and the networking devices fail to provide real-time decisions. Thus to satisfy the need of the real-world applications, a paradigm of computing and an intelligent routing environment was introduced. The era of fog computing and Software-defined networking(SDN) increased the performance of the traditional cloud-based architecture. With the increasing data bombardment from various sources, without proper load balancing, the refined architecture also cannot render maximum performance to the application. In view of this our paper deals with the real-time fog computing environment created with the support of SDN. On top of this platform, a novel approach for load balancing based on reinforcement learning has been proposed. The algorithm understands the behavior of the network and balances the load to provide the maximum possible availability of the resources. The distributed nature of this architecture also makes the network resilient. This paper gives an overview of the built architecture and the proposed novel architecture.