Comparison of Tightly Coupled and Deeply Coupled GPS/INS Integration for Automotive Applications Using a Software Defined GNSS Receiver Framework

M. Langer, G.F. Trommer

Abstract:	In this paper a tightly coupled GPS/INS system is compared to a non-coherent deeply coupled GPS/INS system using a software defined GNSS receiver framework. Both navigation systems are used to evaluate the performance in an urban environment in a moving automotive platform during real world test drives in an urban environment. The navigation solutions are compared to a RTK reference solution. In a tightly coupled GPS/INS System the GNSS receiver can be seen as a black box and only the receiver outputs are processed. In this paper a 17 state error state space kalman filter is used to process range and delta range measurements. The inertial part of the system consists of an automotive grade MEMS IMU. In an urban environment it can be challenging to receive and process GPS signals. Due to multipath and shadowing the visible GNSS constellation and quality of the GNSS signals can change rapidly. Even complete outages of GNSS signals are possible, for example in deep urban canyons or tunnels. In contrast to this a deeply coupled GPS/INS integration has several advantages. A deeply coupled GPS/INS system can be described as a coupling of signal tracking of single GPS receiver channels through a position, velocity and timing solution with simultaneous support of the navigation solution with dynamic information of an inertial navigation system. The signal tracking of each satellite benefits of the other and is improved by the complementary characteristics of inertial sensors. Therefore a satellite which is blocked by an obstacle does not need to be reacquired because the satellite signal can be tracked by the inertial navigation system and the signal can be used to update the navigation solution instantly after it reappears. Even short complete GPS outages can be bridged depending on the quality of the inertial components. In a standard GNSS receiver no access to the signal processing is available. To realize a deeply coupled system a software defined GNSS receiver framework has been developed. The framework is built up as a modular C++ implementation so that different filter and sensor configuration can be used. Therefore a dead reckoning sensor could easily be added or as previously shown [1], step length updates from a pedestrian navigation system. The receiver framework is able to track GPS and GALILEO signals in real-time. For evaluation the raw GNSS and sensor data are stored for post processing. Several test drives were recorded, including urban and suburban environments, tunnels with complete GPS outages and urban canyons with severe multipath influence and shadowing. In this environment the two types of GPS/INS integration are evaluated and compared. The deep integration approach is evaluated with respect to the influence of C/N0 and multipath on the feedback of the signal tracking to the GNSS receiver and its advantages and disadvantages is compared to a tightly coupled approach in urban canyons and tunnels. It is shown that the deeply coupled GPS/INS integration approach has significant advantages in weak signal conditions. The results show that due to the feedback of the navigation solution to the signal tracking of the GNSS the availability and continuity of GPS measurements is improved. [1] Deeply Coupled GPS/INS Integration in Pedestrian Navigation Systems in Weak Signal Conditions, Markus Langer, Stefan Kiesel, Christian Ascher, Gert F. Trommer, 2012 International Conference on Indoor Positioning and Indoor Navigation, 13-15th November 2012
Published in:	Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013) September 16 - 20, 2013 Nashville Convention Center, Nashville, Tennessee Nashville, TN
Pages:	1308 - 1316
Cite this article:	Langer, M., Trommer, G.F., "Comparison of Tightly Coupled and Deeply Coupled GPS/INS Integration for Automotive Applications Using a Software Defined GNSS Receiver Framework," Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013), Nashville, TN, September 2013, pp. 1308-1316.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In