Multiobjective optimization in integrated photonics design

Abstract

We propose the use of the parallel tabu search algorithm (PTS) to solve combinatorial inverse design problems in integrated photonics. To assess the potential of this algorithm, we consider the problem of beam shaping using a two-dimensional arrangement of dielectric scatterers. The performance of PTS is compared to one of the most widely used optimization algorithms in photonics design, the genetic algorithm (GA). We find that PTS can produce comparable or better solutions than the GA, while requiring less computation time and fewer adjustable parameters. For the coherent beam shaping problem as a case study, we demonstrate how PTS can tackle multiobjective optimization problems and represent a robust and efficient alternative to GA.

4 pages, accepted for publication in Optics Letters

References in corpus (2)

• Beam shaping using genetically optimized two-dimensional photonic crystals
• Coherent beam shaping using two-dimensional photonic crystals

Cited by in corpus (5)

• Design of Compact and Efficient Silicon Photonic Micro Antennas with Perfectly Vertical Emission
• Ab initio investigation of lasing thresholds in photonic molecules
• Illustrated tutorial on global optimization in nanophotonics
• Lorenz-Mie theory for 2D scattering and resonance calculations
• Optimization of integrated polarization filters