Scalable Wavelength Arbitration for Microring-based DWDM Transceivers

TL;DR

The paper tackles scalable initialization of microring-based DWDM transceivers by separating arbitration policy from algorithm through an ideal wavelength-aware model and a practical wavelength-oblivious algorithm. It introduces three policy options (LtD, LtC, LtA) and develops a two-phase LtC algorithm (Record and Matching) with a variation-tolerant enhancement, showing near-perfect alignment to the ideal model under realistic device variabilities. Through AFP and CAFP metrics, the work demonstrates robustness advantages of the proposed schemes over sequential tuning, and provides FSR design guidelines to minimize arbitration failures. The findings offer a holistic framework for scalable, autonomous wavelength arbitration in silicon photonics, with practical implications for large-scale deployment and system-level optimization.

Abstract

This paper introduces the concept of autonomous microring arbitration, or wavelength arbitration, to address the challenge of multi-microring initialization in microring-based Dense-Wavelength-Division-Multiplexed (DWDM) transceivers. This arbitration is inherently policy-driven, defining critical system characteristics such as the spectral ordering of microrings. Furthermore, to facilitate large-scale deployment, the arbitration algorithms must operate independently of specific wavelength information and be resilient to system variability. Addressing these complexities requires a holistic approach that encompasses the entire system, from device-level variabilities to the transceiver electrical-to-optical interface - this system-wide perspective is the focus of this paper. To support efficient analysis, we develop a hierarchical framework incorporating an ideal, wavelength-aware arbitration model to examine arbitration failures at both the policy and algorithmic levels. The effectiveness of this approach is demonstrated in two ways: by analyzing the robustness of each policy in relation to device variabilities, and by developing an algorithm that achieves near-perfect alignment with the ideal model, offering superior robustness compared to the traditional sequential tuning method. The simulator code used in this paper is available at https://github.com/wdmsim/wdm-simulator.

Figures (16)

• Figure 1: Overview of the system model. (a) System block diagram. (b) Spectral ordering enforcement level, an arbitration policy used to classify the arbiter type. Lock-to-Deterministic (LtD) allows only a single spectral ordering. Lock-to-Cyclic (LtC) allows any cyclic equivalent of the specified ordering. Lock-to-Any (LtA) does not impose any restrictions on the final spectral ordering. These enforcement levels have an inclusive relationship, with LtA being the most permissive and LtD being the most restrictive. MWL stands for Multi-Wavelength Laser, and MRR stands for Microring Resonator.
• Figure 2: Multi-Wavelength Laser (MWL) and Microring Resonator (MRR) row models in the wavelength domain. L.V. denotes local variation, G.V. denotes global variation, and sampled variation parameters are prefixed by $\Delta$. The tuning range (TR) and free spectral range (FSR) of the MRR are indicated in the figure. Nomenclatures are explained in Table \ref{['tab:model_params']} and further detailed in the text.
• Figure 3: Simulation setup for measuring the robustness of wavelength arbitration through two metrics: Arbitration Failure Probability (AFP) and Conditional Arbitration Failure Probability (CAFP). The experiments sample multi-wavelength lasers (MWL) and microring resonator (MRR) rows and subject them to arbitration tests under different policies. Policy-level evaluation uses the ideal arbitration model, which assumes wavelength-awareness and calculates AFP based on failure statistics. Algorithm-level evaluation, on the other hand, employs a wavelength-oblivious arbitration model to reflect operational constraints, computing CAFP to assess how closely the algorithm approximates ideal arbitration success.
• Figure 4: General shmoo trend of Arbitration Failure Probability for different arbitration policies. Model parameters are shown in Table \ref{['tab:model_params']}. The choice of sweep ranges for local resonance variation ($\sigma_{lLV}$) and tuning range ($\bar{\lambda}_{TR}$) are explained in Section \ref{['sec:model-laserring']}.
• Figure 5: Comparison of minimum tuning range for different DWDM and arbitration parameters. Different colors represent specific sets of DWDM parameters: channel count ($N_{ch}$) of 8 or 16 (wdm8/16) and channel spacing ($\lambda_{gS}$) of 200 or 400 (g200/400). The arbiter parameters are detailed in Table \ref{['tab:arb_cfg']}. The minimum tuning range is defined as the smallest microring tuning range required for complete arbitration success. (a-d) shows the minimum tuning range trend under different arbitration policies and (e-h) plots the trends of (a-d) normalized by channel spacing.