Hamiltonian Graph Analysis – Mixed Integer Linear Programming (HGA-MILP) Based Link Failure Detection System in Optical Data Center Networks

Abstract

Abstract The internet services have become more popular rapidly over the decades. This is because of the inspiration created by new content-hungry applications on end-user devices like smart phones. Wavelength division multiplexed mesh networks are the optical backbone, working as the aggregating point of huge volume traffic, interconnecting several access networks and end users. The need of virtual data center is to retain services in case of failures such as optical link failures and resource or power outage in data centers that occurs in the cloud infrastructure and improve the survivability. The physical infrastructure may disconnect the virtual machines and physical data centers. This leads to the reduction in computational as well as communicational capabilities of the cloud. To overcome these issues, an integer linear programming model called hamiltonian graph analysis – mixed integer linear programming (HGA-MILP) is proposed in this work by utilizing hamiltonian graph model. The main motive of this research is to design a survivable data centre network mapping problem against physical link failures. Use of hamiltonian path methodology is helpful to organize data and improve network performance.