Architectures, principles, and strategies of rotational modulation inertial navigation systems: a review

Abstract

Rotational modulation inertial navigation systems (RMINSs) provide an effective system-level approach for improving inertial navigation accuracy without external aiding. By mounting the inertial measurement unit on one or multiple rotating axes, rotational modulation transforms selected constant or slowly varying sensor errors into periodic components, thereby reducing their long-term accumulation in navigation integration. Although extensive studies have investigated specific rotation schemes, calibration methods, and application-oriented implementations, the intrinsic relationship among rotational architecture, error modulation principle, and strategy design has not been sufficiently clarified from a unified system perspective. In fact, the achievable performance of an RMINS is jointly determined by these tightly coupled aspects: architecture defines the available modulation geometry and excitation capability, principle explains how inertial and structural errors evolve under rotation, and strategy determines how effectively the theoretical modulation potential is translated into practical navigation performance. Motivated by this gap, this paper presents a systematic review of RMINSs from the perspectives of architectures, principles, and strategies. Representative single-, dual-, and tri-axis RMINSs are first compared based on structural characteristics, modulation capability, engineering complexity, and typical applicability. The fundamental principles of rotational modulation are then summarized, with emphasis on the propagation and suppression of constant bias, scale factor error, misalignment error, nonorthogonality, and Earth-rate-coupled error terms. Typical rotation strategies are further critically reviewed with respect to modulation effectiveness, engineering burden, and suitability under different sensor grades and dominant error conditions. A representative-study comparison matrix and design-oriented conclusions are further provided to clarify when single-, dual-, and tri-axis architectures should be selected, and which error sources can be mainly handled by rotational modulation or must rely on calibration and compensation. Finally, current challenges and future research directions are discussed. This review aims to provide a structured reference for the analysis, design, evaluation, and practical development of high-performance RMINSs.