Joint port-cost and power-consumption savings in hybrid hierarchical optical networks.

Abstract: In optical networks, with the number of wavelengths in fibers increasing, the size and cost of Optical Cross-Connects (OXC) have been increased. Waveband merging technology, which groups several wavelength routes into a single waveband tunnel with at least two hops, could perform port-cost savings by reducing the size and cost of OXC since switching one waveband in the All Optical (OOO) domain only consumes two ports at each bypass node. On the other hand, the absence of wavelength conversion restricts the capability of OOO–OXC. Therefore, designing a hybrid hierarchical node to transmit wavebands in an OOO switch and add/drop local requests or perform wavelength conversion in an Optical-Electrical-Optical (OEO) switch has been one of the promising solutions for next-generation optical networks. However, due to the increasing effects of greenhouse gases and the drastic consumption of power, port-cost and power-consumption savings are becoming the new and twin objectives to further develop hybrid hierarchical optical networks. In this paper, we jointly consider port-cost and power-consumption in a hybrid hierarchical optical network and propose efficient approaches including Integer Linear Programming (ILP) models and heuristics. Simulation results of two heuristic approaches, a Cost and Power savings algorithm based on an End-to-End waveband Merging strategy (CPEEM) as well as a Cost and Power savings algorithm based on a Sub-Path waveband Merging strategy (CPSPM), show that the cost of Transmitting Ports (TPs) used in an OOO switch is in contrast to the use of the power consumed by the OEO part of a hybrid hierarchical node. Moreover, comparing with CPEEM, CPSPM has a better performance for port-cost savings but the performance improvement is at the cost of higher power consumption in an OEO switch. On the contrary, CPEEM has a better performance for power-consumption savings but the performance improvement is achieved at the expense of a higher cost of transmitting ports comparing with CPSPM. Based on the parameter, Boundary of Consumed Power (BCP), a New Integrated Algorithm (NIA) that could effectively choose between CPEEM or CPSPM according to the current network state is proposed to obtain the trade-off performances. [Copyright &y& Elsevier]