Optimization of InP APDs for High-Speed Lightwave Systems

Abstract

Abstract-Calculations based on a rigorous analytical model are carried out to optimize the width of the indium phosphide avalanche region in high-speed direct-detection avalanche photodiode-based optical receivers. The model includes the effects of intersymbol interference (ISI), tunneling current, avalanche noise, and its correlation with the stochastic avalanche duration, as well as dead space. A minimum receiver sensitivity of 28 dBm is predicted at an optimal width of 0.18 m and an optimal gain of approximately 13, for a 10 Gb/s communication system, assuming a Johnson noise level of 629 noise electrons per bit. The interplay among the factors controlling the optimum sensitivity is confirmed. Results show that for a given transmission speed, as the device width decreases below an optimum value, increased tunneling current outweighs avalanche noise reduction due to dead space, resulting in an increase in receiver sensitivity. As the device width increases above its optimum value, the receiver sensitivity increases as device bandwidth decreases, causing ISI to dominate avalanche noise and tunneling current shot noise.

Temperature dependence of avalanche multiplication and breakdown voltage in Al0.52In0.48P

“…Since band to band tunneling is inversely proportional to the exponential of the band gap, using a wider-gap semiconductor as the multiplication region in an APD should lead to a lower tunneling current. 4 The temperature dependence of avalanche multiplication characteristics and breakdown voltage in several widegap semiconductors have been previously reported. Ma et al 5 reported that Al x Ga 1-x As with an Al composition of 60% has the smallest C bd in the Al x Ga 1-x As material system for a given avalanche region thickness, w. Groves et al have reported a C bd of 3.18 mV/K for a w ¼ 0.09 lm Al 0.6 Ga 0.4 As p þ -i-n þ diode.…”

Sensitivity Calculations of High-Speed Optical Receivers Based on Electron-APDs

“…The model in this paper is based on the bit-error-rate (BER) model developed in [1] and [2], although there are some significant differences and modifications which need to be discussed. The unique impulse response of an e-APD means that many of the equations previously developed cannot be used, and a revision of the model must be carried out.…”

“…We emphasize, however, that the limitation of the model in [2] arises only when an analytical formula for the receiver sensitivity is sought; the model in [2] remains valid for the numerical calculation of the sensitivity for electron APDs, though this would require significantly more computing resources to match the accuracy provided by a more analytical model. Although the analytical expressions for the mean and variance of the APDs stochastic impulse response function from [2] are no longer valid, the defining integrals for the mean and the variance are still valid and can still be used.…”

“…The formulae for the mean and variance of the ISI contributions (μ ISI,n and σ 2 ISI,n ) to receiver output from the nth past bit with duration T are given by equations (A3) and (A7) from [2]:…”

“…which are (9) and (11) in [2]. Here a n (I j ) = 1 when the nth bit in the pattern I j is a "1," and 0 otherwise.…”