Signal processing and incoherent-MIMO for multimode optical fibers
Abstract
Multimode fibers (MMF) are generally used in short and medium haul optical networks owing to the availability of low cost devices and inexpensive packaging solutions. However, the performance of conventional multimode fibers is limited primarily by the presence of high modal dispersion owing to large core diameters. While electronic dispersion compensation methods improve the bandwidth-distance product of MMFs, they do not utilize the fundamental diversity present in the different modes of the multimode fiber. This thesis draws from developments in wireless communication theory and signal processing to motivate the use of multiple-input multiple-output (MIMO) and signal processing techniques in MMF links. MIMO techniques that utilize the diversity of modes present in the fiber increase data rates and link reliability. Theoretical models for propagation effects in MMF systems are used to analyze and design the geometry of laser and detector arrays for MIMO-MMF links, and study how the design of these arrays impacts link performance. These models are also used to develop and evaluate low-complexity algorithms that efficiently utilize dense detector arrays, with "greedy subset selection" based on submodular optimization. Experimental evaluation of 1 × 1, 2 × 2, 3 × 3 and 4 × 4 MIMO systems have been conducted over various MMF media, including 100 m - 3 km silica MMF with externally modulated distributed feedback lasers and directly modulated vertical cavity surface emitting lasers (VCSELs), as well as with Fabry Perot lasers over 10 m - 100 m plastic MMF. The use of off-the-shelf components as well as the role of axial offset coupling in enhancing modal diversity has been experimentally quantified. The experimental techniques discussed in this thesis have enabled an increase of over 25× in the bandwidth-distance product of the MMF link, when compared to currently deployed MMF systems, such as 10GBASE-SR.