Trade‐off between single and double pass amplification schemes of 1480‐nm pumped EDFA

Abstract

INTRODUCTIONThe development of erbium-doped fiber amplifiers (EDFAs) has had a profound impact on the evaluation of optical-communication systems. This device has become an important component for long-haul communications for single-mode fiber in the 1550-nm wavelength region [1]. Its inclusion in optical-fiber transmission networks has already resulted in a capacity increase of more than two orders of magnitude, and compensation for signal attenuation resulting from distribution or component-insertion losses. The erbium-doped fiber (EDF) length, pump wavelength, EDF characteristics, signal power, and pump power are important factors for optimizing the performance of EDFAs [2]. Traditionally, a singlepass amplification utilized in EDFAs (SP-EDFA) was intensively investigated. A single pass with a midway isolator was proposed to filter out the backward-amplified spontaneous emission (ASE) that degrades the signal amplification at the input end of EDF, hence reducing the effect of ASE self-saturation in the amplifier system [3]. In addition to improving noise figures, high gain can also be achieved using single-pass amplification [4]. Filters can also be used together with the midway isolator to further suppress the forward ASE and improve the signal gain [5, 6]. Another approach to obtain high gain involves incorporating double-pass amplification (DP-EDFA) in the EDF [7, 8]. This approach improves the signal gain due to having twice the number of signal amplifications while propagating in the EDF. However, the double-pass amplifier produces higher noise figure values as well. Therefore, a trade-off is needed between these two amplifier configurations in terms of gain and noise figure.This paper describes the trade-off performance between singlepass and double-pass amplifiers which affects the signal-power range. An optimization of these EDFA configurations using practical comparative analysis is proposed in this work. The critical input power needs to be determined in order to capitalize upon the performance of the amplifier structures based on the intensity of the input signal. EXPERIMENTAL SET-UPThe conventional SP-EDFA is shown in Figure 1(a). The two typical optical circulators, used as isolators before and after the active material (4.5-m EDF), have an isolation of over 50 dB. The DP-EDFA configuration, shown in Figure 1(b), consists of the same components of SP-EDFA, with the only difference being the usage of the second circulator to recycle back the forward C-band ASE and signal into the active material. The name DP-EDFA is given with respect to the input-signal light amplified in both forward and backward through the Erbium-doped fiber.The active material of EDF has absorption of 12 dB/m at 1480 nm, numerical aperture of 0.21, and cutoff wavelength of 1400 nm. A 1480-nm pump laser is used to provide optical energy in a forward pumping direction. A 1480/1550-nm wavelength division multiplexer (WDM) is used to multiplex and demultiplex the signal and pump lights. For the DP-EDFA, the output from port 3 of c...