Comments, with reply, on 'comparison of multiquantum well, graded barrier, and doped quantum well GaInAs/AlInAs avalanche photodiodes: a theoretical approach by K. Brennan

It is pointed out that the lucky drift model complements the Monte Carlo simulation used in a previous paper (ibid., vol.QE-23, no.8, p.1273-82, 1987) to predict the performance of avalanche photodiodes containing multilayer structures. In his reply, the author agrees with the basic promise raised by the commenter, that models such as luck drift serve to complement detailed, first principles models like the Monte Carlo approach, aiding in their interpretation. >