A Novel Dynamic Software-Defined Networking Approach to Neutralize Traffic Burst

Abstract

Software-defined networks (SDN) has a holistic view of the network. It is highly suitable for handling dynamic loads in the traditional network with a minimal update in the network infrastructure. However, the standard SDN architecture control plane has been designed for single or multiple distributed SDN controllers facing severe bottleneck issues. Our initial research created a reference model for the traditional network, using the standard SDN (referred to as SDN hereafter) in a network simulator called NetSim. Based on the network traffic, the reference models consisted of light, modest and heavy networks depending on the number of connected IoT devices. Furthermore, a priority scheduling and congestion control algorithm is proposed in the standard SDN, named extended SDN (eSDN), which minimises congestion and performs better than the standard SDN. However, the enhancement was suitable only for the small-scale network because, in a large-scale network, the eSDN does not support dynamic SDN controller mapping. Often, the same SDN controller gets overloaded, leading to a single point of failure. Our literature review shows that most proposed solutions are based on static SDN controller deployment without considering flow fluctuations and traffic bursts that lead to a lack of load balancing among the SDN controllers in real-time, eventually increasing the network latency. Therefore, to maintain the Quality of Service (QoS) in the network, it becomes imperative for the static SDN controller to neutralise the on-the-fly traffic burst. Thus, our novel dynamic controller mapping algorithm with multiple-controller placement in the SDN is critical to solving the identified issues. In dSDN, the SDN controllers are mapped dynamically with the load fluctuation. If any SDN controller reaches its maximum threshold, the rest of the traffic will be diverted to another controller, significantly reducing delay and enhancing the overall performance. Our technique considers the latency and load fluctuation in the network and manages the situations where static mapping is ineffective in dealing with the dynamic flow variation.

A Temporal Deep Q Learning for Optimal Load Balancing in Software-Defined Networks

“…Several comparisons were conducted for three network scenarios: mild, moderate, and heavy. Initially, the results of the tDQN were compared with the previously proposed dSDN [ 19 ] to analyze the network performance. Subsequently, a comparison was made between the tDQN, traditional network, and eSDN.…”

“…The NetSim simulation framework is configured to assess the efficiency of the tDQN model when compared to traditional networks, eSDNs, and dSDNs. Our earlier research [ 19 ] presented the simulation setup, and the tDQN is integrated into the dSDN controller within the same configuration. The setup involves creating network scenarios with varying traffic loads, including mild, moderate, and heavy network traffic conditions.…”

“…In our prior research work [ 19 ], we initially developed reference models for mild, moderate, and heavy network traffic using a standard SDN in traditional networks, followed by proposing and implementing an eSDN [ 19 ], which has better QoS than standard SDNs due to the priority scheduling and congestion control algorithms. However, the deployment of controllers in an eSDN is static, which fails to accommodate the growing devices by adjusting their loads in real time.…”

“…Consequently, it leads to a lack of load balancing among the controllers. To address this limitation, we proposed and implemented a dSDN [ 19 ] aiming to sustain the network load with the dynamic flow fluctuation in heterogeneous IoT devices. The dynamic controller mapping in a dSDN enhances the overall network QoS, particularly in a heavy traffic network that emulates the futuristic network.…”