High-fidelity acoustic rendering in modern full-flight simulators requires real-time, state-consistent sound generation. Conventional sample-based systems often suffer from perceptual discontinuities and limited extrapolation. This paper proposes a practice-oriented grey-box framework for real-time aircraft cabin noise synthesis, designed for Level-D simulators. By retaining classical aeroacoustic scaling laws as functional basis functions and identifying coefficients from cockpit recordings, the model implicitly captures complex transmission effects without expensive structural-acoustic modeling. The framework integrates engine and airframe sources with a novel energy-dissipation proxy for transient dynamics. Validated against Boeing 737-800 cockpit data, the model achieves spectral errors below 2.8 dB and a processing latency under 6 ms. This methodology provides a scalable, physically consistent solution for immersive flight simulation.
Yang et al. (Sun,) studied this question.