I. Introduction
At present, the Fourier transform (FT) method has been successfully applied to extract net modal gain from the amplified spontaneous emission (ASE) spectrum of Fabry–Pérot (FP) semiconductor lasers [1]. Further work using this method to measure the linewidth enhancement factor was also carried out [2]. Compared with the well-known Hakki–Paoli method [3], the resolution bandwidth influence on the FT method can be alleviated by a deconvolution process [4]. In this letter, we extend the functionality of the FT method to extract both quasi-Fermi level separation and internal loss from the ASE spectrum. The internal loss and quasi-Fermi level separation can be derived from the relationship between the spontaneous emission and gain [5], [6]. This requires the unamplified spontaneous emission to be measured, for example, from the side of the FP laser [5], which is quite difficult in practice. The emission from the end facet of the FP laser is easy to measure but is intrinsically modulated by the FP cavity. This modulation can be easily decomposed by the FT method, which transforms the quasi-periodic spectrum into a series of separated peaks corresponding to harmonics of the optical round-trip of the cavity [1], [4], [7]. The round-trip gain can be derived from the ratio of the inverse Fourier transformed zeroth and first peaks [1], [4]. As described below, the zeroth harmonic also contains the spontaneous emission information [6] which, however, is omitted in the standard FT method. In this letter, we use a minimum search process to extract the internal loss and quasi-Fermi level separation from the spontaneous emission and the net modal gain spectrum, both of which are obtained from the FT method. The gain extraction always needs a high resolution bandwidth, otherwise the underestimation of the gain will have a negative influence on the accuracy of the extracted internal loss and quasi-Fermi level separation. We also show that by using a deconvolution process, the gain underestimation can be reduced [4] so that the requirement for high resolution bandwidth can be eased.