Design of Distributed Optical-Fiber Raman Amplifiers using Multi-objective Particle Swarm Optimization

Journal of Microwaves, Optoelectronics and Electromagnetic Applications 12/2011; 10:323-336. DOI: 10.1590/S2179-10742011000200003

ABSTRACT A novel method is presented to design the configuration of pumping lasers of Raman amplifiers using a multi-objective particle swarm optimizer. The goal is to obtain the pump laser wavelengths and powers that maximize the amplifier on-off gain, while maintaining the flatness of the gain over the used bandwidth. We used an algorithm called Multiple Objective Particle Swarm Optimization with Crowding Distance and Roulette Wheel to generate the non-dominated solutions, considering the average on-off gain and the ripple of the amplifier over the transmission bandwidth as the objectives in the optimization process. We designed amplifiers using three, four and five pump lasers. The experimental results showed that our proposal was able to design Raman amplifiers with a gain ripple lower than 0.2 dB and with an average on-off gain around 16.7 dB, when 20 signal channels and a total pump power of 1 W were considered. Moreover, we demonstrated that it is possible to allow the decision maker to choose among many possible non-dominated solutions depending on the application requirements.

  • Source
    Transactions on Combinatorics. 08/2013; 2(1):89-101.
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Support Vector Machine (SVM) is a supervised technique, which achieves good performance on different learning problems. However, adjustments on its model are essentials to the SVM work well. Optimization techniques have been used to automatize this process finding suitable configurations of parameters which attends some learning problems. This work utilizes Particle Swarm Optimization (PSO) applied to the SVM parameter selection problem. As the learning systems are essen-tially a multi-objective problem, a multi-objective PSO (MOPSO) was used to maximize the success rate and minimize the number of support vectors of the model. Nevertheless, we propose the combination of Meta-Learning (ML) with a modified MOPSO which uses the crowding distance mechanism (MOPSO-CDR). In this combination, solutions provided by ML are possibly located in good regions in the search space. Hence, using a reduced number of successful candidates, the search process would converge faster and be less expensive. In our work, we implemented a prototype in which MOPSO-CDR was used to select the values of two SVM parameters for classification problems. In the performed experiments, the proposed solution (MOPSO-CDR using ML) was compared to the MOPSO-CDR with random initialization, obtaining pareto fronts with higher quality on a set of 40 classification problems.
    SMC2012; 01/2012


Available from
Jun 3, 2014