Direct measurements of time- and size-resolved particulate matter (PM) concentrations are of major importance in air quality studies and pollution monitoring. Low-cost, compact optical particle counters (OPCs), which provide accurate PM measurements independent of the particle complex index of refraction (CRI), can be useful in personal exposure monitoring and distributed sensor network studies applications. A methodology is presented for the optimization of the sensor design and operation parameter space aimed at reducing the effect of the CRI on particle sizing errors. The Monte Carlo numerical simulation, which utilizes Mie scattering calculations, is used to determine the optimal detector angle for the specific set of constraints described by the weighting coefficients. The optimized detector position (theta= 48 degrees) has the lowest dependency on CRI over the entire particle size range of 0.5-10 mu m. The near-forward, optimized, and perpendicular detector angles are compared experimentally using monodisperse 2 mu m and 4 mu m particles of silica, PSL, and alumina; the light collection cone angle is set at alpha = 20 degrees in all experiments. The data agree well with the numerical results for all tested scenarios. Overall, the perpendicular detector location has the best precision and worst accuracy related to the CRI variations. The optimized detector position has the best accuracy for both silica and alumina particles. The use of low-cost components, such as laser diodes, photodiodes, miniaturized integrated electronics, and simple component layouts allows for the development of compact OPCs capable of accurately sizing PM. The number of sizing bins, sizing accuracy and precision, and other parameters of interest can be used as an input to an optimization algorithm.