Designing High-Performance Multimode Fibers Using Refractive Index Optimization

The rich design landscape of optical fibers offers many opportunities for refractive index optimization. In particular, the refractive index profiles of multimode fibers (MMFs) and multicore fibers (MCFs) govern the behavior of spatial and polarization modes, including their bandwidth, mode count, mode coupling, modal dispersion, chromatic dispersion, and mode-dependent loss. In this article, we obtain update equations to optimize the shapes of fiber refractive index profiles for various applications using gradient descent. Starting with an initial fiber designed according to standard best practices, our methods iteratively modify the refractive index profile to improve upon the initial design. Interestingly, we see that optimization can sometimes yield large improvements in key fiber properties, such as a 20-fold reduction in the root-mean-square modal group delay spread, even though the index profile is changed very little from the initial design. We show that our methods can be used to tailor the mode coupling, modal dispersion and chromatic dispersion properties to desired values, and that optimization can be successfully applied over a large bandwidth. We provide illustrative design examples, including an optimization of a graded-index MMF with low group delay spread for long-haul mode-division-multiplexed transmission. Our algorithms can be used to design novel fibers or optimize existing fibers for next-generation transmission systems.