Using Genetic Algorithm to Optimize Mixed Placement of 1R/2R/3R Regenerators in Translucent Lightpaths for Energy-Efficient Design

We propose an evolutionary approach to efficiently optimize the mixed placements of 1R/2R/3R regenerators in translucent lightpaths. The genetic algorithm encodes the placements of mixed types of regenerators as genes, and incorporates adaptive genetic operators for effective search strategy. We consider both end-to-end signal transmission quality and regeneration energy cost, and design the fitness function for multi-objective optimization. The simulation results show that the proposed approach converges within 50 generations even when the lightpath under investigation is as long as 31 hops. To improve the flexibility of network designs, multiple energy-efficient mixed regenerator placement (MRP) results can be obtained simultaneously.