Efficient Bandwidth Management Using Geo Nearest Datacenter for Inter DCN and Optimal Path for Intra DCN to Optimize Bandwidth Cost

Cloud services are getting much attraction these days among different types of stakeholders. To satisfy their demand, cloud service providers often establish geographically distributed DCs or make contract with third party DC providers. In this work, we consider the problem of resource allocation for enterprise customers considering both inter and intra data center network (DCN) jointly. Therefore, cloud service providers need to find an optimal solution to establish a connection from Point of Presences (PoPs) to the DCs (inter DCN) and from the entry points to the servers within each DC (intra DCN) to minimize the bandwidth cost. Here, a novel mixed-integer linear programming (MILP) formulation is presented to minimize the inter DC bandwidth cost by selecting a DC which is geo-nearest from the demand generating PoP and the intra DC bandwidth cost by using the optimal routes from the entry points to the servers. Efficiency of our approach is analyzed through inter and intra DC bandwidth cost and number of links used.