Modulation Format Independent Joint Polarization and Phase Tracking for Coherent Receivers

TL;DR

The paper tackles the challenge of tracking random polarization drift and carrier-phase noise in coherent optical receivers by proposing a jointly optimized, model-based tracking algorithm that operates in a data-driven, decision-directed fashion for arbitrary modulation formats. It unifies Jones, Stokes, and 4D real formalisms to track phase and SOP through gradient-based updates on a dynamic channel model, achieving fast convergence with reduced complexity compared to state-of-the-art methods. Numerical results across PM-QAM formats at 28 Gbaud show superior tolerance to phase noise and polarization drift, with minimal degradation from additive noise and rapid convergence (≈2.5k symbols). The approach is highlighted as suitable for elastic optical systems, though practical deployment would require integration with channel equalization and consideration of non-idealities like dispersion and PMD.

Abstract

The state of polarization and the carrier phase drift dynamically during transmission in a random fashion in coherent optical fiber communications. The typical digital signal processing solution to mitigate these impairments consists of two separate blocks that track each phenomenon independently. Such algorithms have been developed without taking into account mathematical models describing the impairments. We study a blind, model-based tracking algorithm to compensate for these impairments. The algorithm dynamically recovers the carrier phase and state of polarization jointly for an arbitrary modulation format. Simulation results show the effectiveness of the proposed algorithm, having a fast convergence rate and an excellent tolerance to phase noise and dynamic drift of the polarization. The computational complexity of the algorithm is lower compared to state-of-the-art algorithms at similar or better performance, which makes it a strong candidate for future optical systems.

Figures (4)

• Figure 1: Receiver block diagram with elementary DSP modules.
• Figure 2: Tracked channel parameters using the proposed algorithm with $\Delta \nu=1$ MHz and $\Delta p=1$ kHz at 28 Gbaud PM-16-QAM are shown. As can be seen, the algorithm has excellent tracking capabilities without exhibiting cycle slips.
• Figure 3: The achievable performance of the three tracking schemes for PM-16-QAM, PM-64-QAM, and PM-256-QAM is shown. Each column corresponds to a modulation format, whereas the rows present different performance metrics. The polarization-noise tolerance is shown in the first row by plotting SER versus $\Delta p \cdot T$. The penalty compared to the AWGN curve at low $\Delta p \cdot T$ is due to the applied phase noise. The tolerance to phase noise is plotted on the second row, where $\Delta \nu \cdot T$ is varied, whereas the noise sensitivity is shown in the third row by varying the SNR. The round markers shown in the first three rows correspond to the same channel conditions, i.e., the same $\Delta p \cdot T$, $\Delta \nu \cdot T$, and SNR. The convergence rate is compared on the fourth row, where the SER is plotted versus the symbol index $k$.
• Figure 4: Proposed algorithm