The AUTONAV feature of the Global Positioning System (GPS) Block IIR satellites (SVs) provides both improved accuracy and survivability of the navigation function for 180 days without ground contact. These improvements are accomplished by means of a UHF crosslink measurement and data communication system and an onboard computational capability. The Operational Control Segment (OCS) uploads a long term prediction of its ephemeris to each SV. Each satellite inputs SV-to-SV pseudorange measurements and communicated data to its onboard estimator to improve the OCS ephemeris prediction and to estimate its clock. The improved ephemeris and clock estimates are then downlinked to users.Unfortunately there is insufficient information content in the crosslink measurements to fully estimate the ephemeris states. This initially cast doubt on the feasibility of the AUTONAV concept. Suppose an estimator has produced a set of satellite positions for the constellation, at a given time, that is a rotation of the set of true satellite positions at that time. The intersatellite ranges and, hence, pseudoranges that would be obtained from rotated truth are the same as those obtained from truth. There is no information content in the measurements available at an instant to distinguish a rotation of truth from truth. Under Keplerian orbital dynamics, given a constellation, A, there is a three-parameter family of constellations, B, with the property that, at any given time, the positions of B are a rotation of those of A. The AUTONAV measurements cannot distinguish any such B from A. This paper formally develops the concept of unobservable rotation error. It quantifies the system error due to rotation error in the OCS upload that cannot be corrected by the onboard estimators. It also presents an algorithm that prevents the imperfectly modeled onboard estimators from accumulating additional spurious rotational errors in their estimates. Perhaps most importantly, it provides a basis for an intuitive understanding of why AUTONAV can work.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.