This study investigates a nonlinear time-domain filtering approach for inertial sensor drift, utilizing a heuristic bound derived from the Unruh-Bogoliubov thermal spectrum formulation in quantum field theory. The method maps phase increments to a local frequency proxy to evaluate signal-noise discrimination. Preliminary evaluation on synthetic non-Markovian noise demonstrates O(N) computational complexity and performance comparable to standard Extended Kalman Filter (EKF) baselines. Empirical results show comparable estimation error and bias, with reduced computational latency due to a single-pass structure. No claims are made.
Priyal Bhagwanani (Thu,) studied this question.