Abstract
Waveguide electrooptic optical modulators and switches, such as those based on LiNbO3, have many potential applications. As the desired modulation and switching frequencies increase, limitations in available drive power and power dissipation demand that these devices operate at the highest possible electrical, as well as optical, efficiency. The tradeoffs faced when designing the waveguides, electrodes, and buffer layer of devices to attain low optical insertion loss and low drive voltage for dc operation are well known [1-3]. For high-speed operation, the traveling-wave mode of operation is preferred because it permits a lower drive voltage for a given bandwidth in comparison to lumped operation. Use of traveling-wave electrodes itself, however, introduces additional tradeoff relationships. These have also been examined quantitatively [3-7], but to date the designer has generally weighed these tradeoffs in a qualitative fashion because the effects of all relevant variables have not been considered within a single computational framework.

This publication has 3 references indexed in Scilit: