Optical modulation is one of the key determinants to the operating speed of a network. In this work, we report an accurate methodology to study high-speed eye diagram from electrical and optical simulation data of individual modulators. The methodology constitutes electrical parameters such as capacitance, conductance and transitioning times to model time response and effective complex refractive index from optical simulations of phase shifter arms and in turn model the phase change and resultant loss induced by each arm. This methodology is suitable for interferometer-based optical devices and has been applied to silicon-based depletion mode modulators at 10-, 40-Gbps.

In recent years, high speed silicon optical modulators have garnered significant research interest, along with rapid performance improvements [1]. A large number of such reported devices operate with multiple Gbps regime, typically in excess of 10Gbps. These devices rely largely on the carrier depletion effect commonly embedded into, but not entirely limited to, Mach-Zehnder Interferometer (MZI). A key factor to determine the performance of such devices, especially in GO/NO GO test is the high-speed eye diagram which reveals key information such as rise/fall times, extinction ratios, and jitter characteristics. However, in order to achieve realistic eye diagrams, actual fabricated device details such as topology and carrier induced losses must be encapsulated in the simulations.

Testing of these devices in telecommunication framework requires high-end equipments such as pseudo-random bit sequence (PRBS) generator, high speed oscilloscope etc. The cost of these testing equipments is about $300K for 10 Gbps system, which increases exponentially with high data rates, e.g. over $3 million for 40 Gbps system. Consequently, there is an entry barrier into device testing in high speed telecommunication industry. However, if optical device performance testing is aided by computational tools it will lead to huge cost savings and reduction of entry barrier into high speed telecommunication domain. Spurred on by the possibility of huge cost savings to research capital investment expenditure, we have developed a method to accurately analyze the optical signal output from Mach-Zehnder-Interferometer (MZI) configuration of modulators, allowing an unprecedented ability to predict electrical eye diagrams from electrical and optical simulation characteristics of individual modulators.

Here, we evaluate the performance of a silicon-based MZI device by simulating the eye diagram based on its inherent electrical and subsequent optical modeling of individual silicon depletion modulator. This methodology directly takes into account the characteristics of a modulated optical beam, constituting electrical parameters such as capacitance, conductance, and transitioning times to model time response and to obtain effective complex refractive index from optical simulations of the phase shifter arms of the MZI. In turn this simulates the phase change and resultant loss induced by each arm. This methodology is suitable for interferometer-based optical devices and has been applied to silicon-based depletion modulators at 10-, 40-Gbps and demonstrated good agreement with experimental data. This development enables rapid design iterations with full accuracy and can be extended to other optical devices such as detectors and ring resonators.

The proposed method is applied to the depletion type of SOI modulators operating at 10 Gbps; the results of which show an excellent comparison with the measured eye diagrams. Thus, this development can help designers and researchers to perform design iterations with a high level of accuracy. The present developed code is tailored for generating eye diagrams in a MZI comprising two phase-shifter arms using carrier refraction mechanism. With further enhancement, this method can be extended to include other types of high speed optical devices such as ring resonators, and detectors with different length and driving bias conditions.

Tagged in : AccuSpeed, AccuMap, silicon modulators

Jason received his doctorate (PhD) in silicon photonics under Intel Photonics consultant, Professor GT Reed and has more than 13 years experience in the field of photonics. His research so far has resulted in more than 50 patents filed/pending, a silicon photonics book, journals and conference papers. He won a string of awards, including the coveted Royal Academy of Engineering Prize and IET Innovation Prize in Software Design (Highly Commended).