Parameter Estimation for Jump-Diffusion Stochastic Master Equations

This paper investigates parameter estimation for open quantum systems under continuous observation, whose conditional dynamics are governed by jump-diffusion stochastic master equations (SMEs) associated with quantum nondemolition (QND) measurements. Estimation of model parameters such as coupling strengths or measurement efficiencies is essential, yet in practice these parameters are often uncertain. We first establish the existence and well-posedness of a reduced quantum filter: for an N-level system, the conditional evolution can be represented in an O(N)-dimensional real state space rather than the full O(N^2) density-matrix state space. Building on this, we extend the stability theory of quantum filters, showing that exponential convergence persists not only under mismatched initial states but also in the presence of parameter mismatch. Finally, we develop an estimation framework for continuous parameter domains and prove almost sure consistency of the estimator in the long-time limit. These results yield a rigorous treatment of parameter estimation for jump-diffusion SMEs, combining structural reduction with stability and identifiability analysis, and thereby extend the mathematical theory of parameter estimation for open quantum systems.