A Novel Technique of Dynamic Resource Allocation in Software Defined Network

Nowadays, Software Defined Networking (SDN) is playing a vital role in advancement of classical network infrastructure. The focus is to make computer networks centralized and programmable by separating the data plane and control plane. The data plane forwards network traffic whereas the control plane deals with the computational aspects of the network such as determining traffic flow rules. The SDN provides more flexibility, control and better troubleshooting capabilities than traditional networks. In this paper, we aim to provide a heuristic approach to dynamically allocate the resources in the SDN. We propose a technique for the dynamic allocation of the resources which utilizes Rendezvous Hash and Bellman-Ford Algorithm to determine optimal resources for Virtual Network (VN) data flow requests and assigns it to Open Virtual Switches (OVSs). This technique consists of two successive phases. First one calculates shortest paths of VN request and weights of switch ports according to their traffic load. The second phase determines optimal resource path of VN requests thorough the Rendezvous Hash algorithm and updates the flow table of the relative substrate switches. Path with the lowest random weight (LRW) value is used for allocating resources. Results and comparison with other resource allocation methods proves that the proposed technique utilizes resources more efficiently.