Data analytics in SDN and NFV: Techniques and Challenges

Abstract

Software defined networking and network function virtualization are drawing huge attention from researchers both in industry and academia. NFV reduces the capital and opera- tional expenditure of the organization by decoupling the network functions from physical hardware on which they run, which poses new challenges in the perspective of network management such as data management, resource management and performance analysis. Consequentially, novel techniques and strategies are required to address these challenges in efficient way. This paper discusses the most widely used data analytics techniques like machine learning and time series data analysis. Further it describes the review of data mining tools and frameworks. Machine learning helps to overcome the challenges of network management by providing intelligence in network. Hence, in this paper we describe an overview of high level architecture of machine learning analysis framework, the challenges of applying machine learning algorithms in virtual environment and also some of the interesting problems of network management which can be solved by using machine learning.