Blind Identification of the Shaping Rate for Probabilistic Shaping QAM Signal

Abstract

With the emergence of probabilistic shaping (PS) technique, both approaching the Shannon capacity and achieving the rate-adaptive transmission become possible for the fiber optical communication at a fixed forward error correction (FEC) code rate. Therefore, elastic optical network can utilize not only various modulation formats but also variable shaping factors of PS quadrature amplitude modulation (PS-QAM). Consequently, blind identification of the shaping rate for the PS-QAM signals is in an urgent demand for the flexible transceiver. Here, we propose a scheme to identify shaping rates of the received PS-QAM signals, by digitally loading a frequency offset (FO) to the data block at the transmitter-side (Tx), according to a pre-defined codebook. Then, the FO estimation (FOE) is implemented for each block at the receiver-side (Rx) to obtain the information of shaping rates. The feasibility of our proposed scheme is experimentally verified by 32G-baud PS-16/64QAM signals under the condition of a rate-adaptive back-to-back (B2B) transmission. Meanwhile, almost the same transmission results for 2G-baud PS-16/64/256QAM signals are presented to verify its correct function with a narrow FOE range induced by the low baud-rate.