Multi-Objective Laser Rate Equation Based Parameter Extraction Using VCSEL Small Signal Response and RIN Spectra

Rate equation parameters used to simulate semiconductor lasers must be carefully chosen in order to accurately model high-speed laser characteristics in optical link simulations. Previous parameter extraction methods have largely characterized these lasers through their small signal frequency response. We show that capturing both the small signal frequency response and relative intensity noise (RIN) spectrum of the laser is crucial for accurate parameter extraction. We present a multi-objective parameter extraction method alongside a stochastic rate equation model to account for a vertical cavity surface emitting laser's (VCSEL's) bandwidth and RIN characteristics. A genetic algorithm extracts VCSEL parameters from a set of experimentally measured small signal responses and RIN spectra at multiple drive currents. Through Pareto sorting we can extract a single set of parameters which accurately captures the laser's BW and RIN characteristics.