A Survey of Software-Defined Networks Based on Advance Machine Learning Based Techniques

Abstract

Presently, networks have opportunities to build systems with improved and intelligent solutions which can be easily tailored for different users. In conventional networking, the software defined networking (SDN) work separately for its control plane and data plane, which makes it better in managing, more secure and affordable. Since ML is the most fundamental branch of AI when amalgamated with SDN it will provide an efficiency and effectiveness in the management of resources such as bandwidth mapping, flow control, error control and security on the network. In this paper you will find the use of network alongside ML implemented on SDN concepts in two ways. First it describes how appropriate ML algorithms are integrated with SDN based networks following assessment. On the other hand, it provides reasonable recommendations for various network applications based on SDN. And toward the end, it discusses the extra development needed for Machine Learning (ML) algorithms and SDN concepts. The common point of AI, Big data, computer networking and similar fields is discussed in this paper. Researchers from different professions and ages have their findings with regards to AI for various uses since it is a young and intricate area. Thus, this paper will assist these researchers to point out these main faults more accurately.

Keywords:  Artificial intelligence, machine learning, network management, software-defined networking.