Search

Inertial measurement units (IMUs) are key components of various applications including navigation, robotics, aerospace, and automotive systems. IMU sensor characteristics have a significant impact on the accuracy and reliability of these applications. In particular, noise characteristics and bias stability are critical for proper filter settings to perform a combined GNSS/IMU solution. This paper presents an analysis based on the Allan deviation of different IMU sensors that correspond to different grades of micro-electromechanical systems (MEMS)-type IMUs in order to evaluate their accuracy and stability over time. The study covers three IMU sensors of different grades (ascending order): Rokubun Argonaut navigator sensor (InvenSense TDK MPU9250), Samsung Galaxy Note10 phone sensor (STMicroelectronics LSM6DSR), and NovAtel PwrPak7 sensor (Epson EG320N). The noise components of the sensors are computed using overlapped Allan deviation analysis on data collected over the course of a week in a static position. The focus of the analysis is to characterize the random walk noise and bias stability, which are the most critical for combined GNSS/IMU navigation and may differ or may not be listed in manufacturers’ specifications. Noise characteristics are calculated for the studied sensors and examples of their use in loosely coupled GNSS/IMU processing are assessed. This work proposes a structured and reproducible approach for working with sensors for their use in navigation tasks in combination with GNSS, and can be used for sensors of different levels to supplement missing or incorrect sensor manufacturers’ data. [ABSTRACT FROM AUTHOR]

The present contribution evaluates how the European Geostationary Navigation Overlay System (EGNOS) meets the International Maritime Organization (IMO) requirements established in its Resolution A.1046 for navigation in harbor entrances, harbor approaches, and coastal waters: 99.8% of signal availability, 99.8% of service availability, 99.97% of service continuity and 10 m of horizontal accuracy. The data campaign comprises two years of data, from 1 May 2016 to 30 April 2018 (i.e., 730 days), involving 108 permanent stations located within 20 km of the coast or in islands across the EGNOS coverage area, EGNOS corrections, and cleansed GPS broadcast navigation data files. We used the GNSS Laboratory Tool Suite (gLAB) to compute the reference coordinates of the stations, the EGNOS solution, as well as the EGNOS service maps. Our results show a signal availability of 99.999%, a horizontal accuracy of 0.91 m at the 95th percentile, and the regions where the IMO requirements on service availability and service continuity are met. In light of the results presented in the paper, the authors suggest the revision of the assumptions made in the EGNOS Maritime Service against those made in EGNOS for civil aviation; in particular, the use of the EGNOS Message Type 10. [ABSTRACT FROM AUTHOR]