Energy-Efficient Resilience in Translucent Optical Networks With Mixed Regenerator Placement

Abstract

In this paper, we investigate energy-efficient resilience designs for the translucent optical networks using mixed regenerator placement (MRP). We consider both static and dynamic traffic scenarios and aim to provide 100% restoration against single-link failures while minimizing the total energy-cost on regenerators. For static traffic scenarios, we formulate an integer linear programming (ILP) model to solve the energy-efficient p-cycle design for translucent optical networks with MRP. The ILP model optimizes the allocation of working and protection resources jointly under the quality of transmission constraint, with the objective to minimize the total energy cost. A heuristic that can sequentially optimize the p-cycle designs for connections is proposed afterward. We use simulations to evaluate the performance of the algorithms with both uniform and nonuniform traffic models. For dynamic traffic scenarios, we design an algorithm that can handle the setting-up and tearing-down of p-cycles dynamically, according to the time-variant connections. We enhance this algorithm by introducing a reoptimization procedure, which can reassemble existing p-cycles for higher energy efficiency. An ILP model is formulated and solved for the p-cycle reoptimization. We then consider two reoptimization scenarios: (1) on demand (p-cycle-oDRO) and (2) on schedule (p-cycle-oSRO). In addition to the link-based p-cycle, we also study the path-based shared backup path protection scheme.