Three-dimensional self-organized microoptoelectronic systems for board-level reconfigurable optical interconnects-performance modeling and simulation

Self-organized microoptoelectronic system (SELMOS) built from three concepts - scalable film optical link multichip-module (S-FOLM), three-dimensional (3-D) microoptical switching system (3D-MOSS), and self-organized lightwave network (SOLNET)-is proposed. The feasibility of SELMOS for board-level reconfigurable optical interconnects is studied by the beam propagation method/finite difference time domain simulation focusing on three key issues; reducing size/cost of electrical to optical (E-O) and optical to electrical (O-E) signal conversion devices, tolerating alignment accuracy for optical coupling, and miniaturizing high-speed massive optical switching. S-FOLM, which consists of film-waveguide-based 3-D structures with embedded optoelectronic active elements and optical Z-connections for interplane links, enables drastic size/cost reduction of E-O and O-E conversion devices. 3D-MOSS, which is an S-FOLM with embedded microoptical switches, has a potentiality of 1024 × 1024 switching with a system size of ∼1.4 × 0.6 cm² and an insertion loss of 29 dB. The switching rate of the 3D-MOSS is determined by the heat releasing speed to be ∼2 × 10⁵ 1/s when PLZT waveguide-prism-deflector microoptical switches are used. By using advanced electrooptic materials, rates higher than 10⁸ 1/s are expected. Twenty-five percent misalignment in waveguide assembly raises the insertion loss of the 3D-MOSS to 73 dB. The loss is reduced to 32 dB in SELMOS-based 3D-MOSS, where a self-organized 3-D microoptical network is implemented using SOLNET. Further loss reduction is expected by structural optimization of loss-inducing parts. Thus, SELMOS is found to be a solution of the three key issues for board-level reconfigurable optical interconnects. In addition, photolithographic packaging with selectively occupied repeated transfer (PL-Pack with SORT), which integrates different types of active elements into one substrate in desired arrangements using an all-photolithographic process, can contribute to cost and the coefficient of thermal expansion-mismatching reduction.