Cognitive network optimization via network virtualization

Cognitive networks are designed to sense, monitor and extract valuable data from their physical environment, and adapt quickly to support complex network applications in order to satisfy fast changing service demands. Virtualization technologies such as Software-Defined Network (SDN) and Network Function Virtualization (NFV) can be combined to create new frameworks offering the advantages of both SDN and NFV. These include dynamic resource reservation and flexible virtual network creation via NFV, and programmability of these resources and easy network management via SDN. These new combined frameworks could be leveraged to optimize and manage cognitive network architectures. However, optimizing cognitive networks using combined SDN/NFV frameworks requires new network management techniques and fast virtual network provisioning algorithms to replace the legacy manual algorithms. In this study, a new system called AUTOVNET introducing automation in the management and provisioning of virtualized software-defined networks is designed and implemented. AUTOVNET simplifies the manual configurations from the network administrator and increases the flexibility and adaptability of virtual networks. In addition, AUTOVNET performs pre-virtualization fault detection and deep packet analysis to determine the best healthy routing option between the source and destination hosts in the network. This approach allows the network administrator to easily and rapidly create, configure and manage virtual networks in larger complex network topologies.