Non-invasive characterization of cascaded SOAs on InP-based photonic integrated circuits

Semiconductor optical amplifiers (SOAs) are widely used as active elements in optical switching networks for their features of loss compensation and fast reconfiguration time. To optimize the SOA chain operation, it is essential to know the gain response of each SOA. However, the performance of the individual SOA in an on-chip network is challenging to characterize without additional splitters, which increases path insertion losses and, therefore, power consumption. In this work, to address these issues through a feasible and non-disruptive solution, a non-invasive pre-monitoring for N-cascaded SOAs on InP chips is proposed to assess the SOA gain and path losses between cascaded SOAs. This solution utilizes each SOA as both an optical amplifier (OA) and photodetector (PD), characterized by both light propagation directions, forward and backward. We experimentally demonstrate the non-invasive characterization method on a photonic integrated chip (PIC) containing 3-cascaded SOAs, which are part of a photonic integrated SOA-based optical neural network chip. Experimental results show that both losses between SOA and noise can be retrieved and used to obtain the SOA gain curves. The obtained SOA curves are in good agreement with the reference SOA measurements with errors <1.0 dB in gain and <1.5 dB in noise response. These results pave the way to automated non-invasive screening of N-cascaded SOAs in large-scale and packaged PICs.