Top-Down Derivation of Performance Requirements for Inertial Navigation Systems

This work describes a mathematically sound approach to deriving requirements on component level from top-level navigation performance requirements. Considerations focus on inertial navigation for aerial vehicles as a prominent application. Firstly, a probabilistic analysis establishes the general mathematical relations between error distributions in initialization, inertial measurements and navigation solution in a bottom-up manner. Secondly, top-level requirements on the navigation errors are formulated as inequality constraints for the relevant marginal probabilities. The accordingly reduced design space serves as the basis for a top-down requirement decomposition method. In a numerical example, the proposed method is illustrated for a glide bomb with tactical grade inertial sensors.