Integrating Context Awareness and SDN for a Lightweight Approach to Adaptive Networking

Abstract

With computer networks everywhere, managing them is becoming increasingly complex, leading to higher costs, longer response times, and more human errors in decision making. Given that, the search for a Smart Network Management, with less human intervention and more automation, becomes indispensable. Several technologies have been proposed to enable “Smart Management,” including machine learning, notably for more complex cases. The Software-Defined Networking (SDN) paradigm is one of the most promising. Literature shows, however, that there are still several research challenges and opportunities for the automation of many network management tasks. This paper proposes to combine context awareness with SDN in a lightweight approach to Smart Management, or Adaptive Networking, to handle in a simplified way everyday events such as link-down and traffic congestion. A comparative analysis shows that the management approach combining context awareness with SDN is between 40% and 73% faster than the support of SDN to a human-driven management.