Bandwidth Squeezed Restoration using p-Cycle together with Minimum Power Consumption in Elastic Optical Networks

Abstract

Due to flexible bandwidth structure, affected services can be partially restored in elastic optical networks (EONs). Bandwidth Squeezed restoration (BSR) approach maximizes restoration level with limited frequency slots (FSs) in each span. However, the power consumption is one of the great challenges in EONs. EONs with flexible bandwidth and spectrally efficient, are also major contributors in the power consumption. There is extremely large amount of data transmission in the network since network failure causes huge data loss. In order to achieve this high reliability requirement, network architectures are laid down with a large number of redundant resources taking into account the resource requirement for backup protection. These redundant resources are also major contributors to the power consumption. Shortest path routing (SPR) is generally considered in power saving strategies. This paper focuses on BSR using preconfigured cycles (p-cycles) in EONs with power conscious routing (PCR) against single link failure. An integer linear programming (ILP) model has been formulated to minimize power consumption. Minimization of power consumption is based on PCR by maximizing number of idle network resources into sleep mode. Numerical results indicate that the PCR is more power efficient than SPR during low traffic but at the cost of additional frequency slots (FSs). BSR provide the maximum restoration level with available FSs in each span after the occurrence of failure. In addition, it is seen that as the traffic increases, the power consumption saturates.