Deeply Integrated GNSS-INS with CRPA to Constrain Attitude Biases in Robust Navigators

This paper analyzes the impact that a sub-wavelenth antenna array has on the performance of a deeply coupled navigator. The use of sub-wavelength antenna arrays have received a large amount of focus in terms of interference mitigation, but has overshadowed the potential for attitude integration. The effects of constraining the attitude with a filter update and beamsteering from the antenna array with two, three, and four elements are investigated. Sub-wavelength antenna arrays are more robust in design due to their effectiveness in harsher signal to interference ratio environments additionally allowing for attitude constraints at lower carrier-to-noise density ratios (C/N0). Performance enhancements to the inertial sensor biases are also investigated. In these scenarios, IMUs would continue to receive both position and orientation constraints from the deeply integrated receiver and promote better behavior, especially for low quality IMUs. This additionally improves performance over the conventional deep integration since orientation misalignment is accounted for. The study reveals that the inclusion of a two, three, or four element antenna arrays provides massive improvements over the traditional deeply integrated architecture. For the lower (automotive) grade sensor, multiple individual states, including the attitude and IMU biases, receive more than 40 dB of relative improvement. The results quantify how architectural differences affect the deeply integrated performance.