DLVisor: Dynamic Learning Hypervisor for Software Defined Network

Software Defined Network (SDN) is one of the modern networking technologies that provide network flexibility and simplifies management. Virtual SDN (vSDN) enhances sharing physical resources by multiple slices representing tenants or services where each tenant has control over their applications (VN). virtualization gives service providers more to offer new innovative with extra efficiency reliability. Running virtual networks a given infrastructure creates challenges for efficient resource allocation mechanisms avoid congestion starvation maintain Service Level Agreement (SLA), management in vSDN carried out hypervisors. Few studies have addressed dynamic domain. Therefore, efficiently utilize virtualized infrastructure, hypervisors must be proactive self-reconfiguration capabilities assign highly adaptable react changing future demands. Thus, learning-based are improve hypervisor operations. Based on that, this study aims enhance technology an enhanced slice mechanism, traffic delivery utilization. This can fulfilled proposing intelligent forecasting model utilization based improved statistical Machine Learning (ML) techniques. The proposed will dynamically concept drifts then utilized develop mechanism allocation. verified through available real traces datasets from various sources. DLVisor its Dynamic Framework (DLF) reduce overutilization and, consequently, 100% compared related benchmark.