Akademska digitalna zbirka SLovenije - logo
E-viri
Celotno besedilo
Recenzirano
  • Intent-based networking wit...
    Rafiq, Adeel; Afaq, Muhammad; Song, Wang-Cheol

    Journal of ambient intelligence and humanized computing, 11/2020, Letnik: 11, Številka: 11
    Journal Article

    Network establishment and management in the data center need a considerable investment of time and money. To increase the cost productivity of networks, maximizing the utilization of network resources is mandatory. Many network inconsistencies may occur due to the utilization of maximum link capacity. Nowadays, Fat-Tree topology is a popular network architecture in data center networks therefore many researchers are designing routing algorithms for fat-tree topology by adopting load balancing methods. The inefficient way of obtaining network traffic statistics from network devices is a major problem to achieve load balancing traditionally. Software-Defined Network (SDN) platform provides researchers, an environment to develop a user-defined routing algorithm that can be flexible as well as cost-efficient. In this paper, an SDN based system has been proposed for Fat-Tree topology that provides a solution that enables networks to make effective resource utilization and minimize the maximum link capacity utilization of network using intent-based networking (IBN). The proposed system provides a permanent unique path from source to destination that allows traffic to flow uniformly and enables all the links to be offloaded by providing them with equal traffic share proactively. By taking a proactive approach which is inherence of an IBN, the proposed system provides the secure channel to provision the host in network with load balancing capability in the running network by restricting illegal in-premise access to the network.