Real-Time Navigation System for Ultra-Tight Integration of GNSS and Multi-Sensors

T. Li, J. Georgy, Z. Syed, C. Goodall

Abstract:	High Sensitivity Global Navigation Satellite Systems (HSGNSS) receivers with improved signal acquisition and tracking capabilities, and better navigation performance are usually preferred to satisfy the increasing demands of today’s vehicle and personnel navigation. However, in urban canyon or indoor areas where the signals are severely attenuated, jammed, reflected or completely blocked, even current HSGNSS receivers would suffer degraded performance or may completely fail to operate. In integrated navigation, sensors including inertial sensors, barometers, and magnetometers can be integrated with GNSS to provide a continuous, always-available and reliable navigation solution even when GNSS signals is blocked or degraded. The commercially available GNSS and multi-sensors integrated navigation systems are mainly focused on loose and tight integration strategies. Although the loosely and tightly coupled integration can improve navigation accuracy and bridge short GNSS gaps to some extent, the GNSS receiver cannot benefit from the aiding sensor information. In ultra-tight integration strategy, the feedback from the navigation sensors is used to aid the receiver’s tracking loops. Previous research has demonstrated improved tracking sensitivity, increased multipath mitigation capability and better navigation performance with ultra-tight integration. However, most of the previous work applied post-processing strategy for the performance analysis. The limited real-time results presented by some researcher were PC-based. To date, the real-time performance of an ultra-tight GNSS/multi-sensor system on an embedded hardware platform has not been fully investigated. Trusted Positioning presents a robust real-time ultra-tight integrated HSGNSS/multi-sensors navigation system. First, the paper presents the general architecture of the ultra-tight HSGNSS/multi-sensor system used in this work. The HSGNSS applies long coherent integration time to improve the tracking sensitivity. However, the maximum coherent integration time for a conventional Phase Locked Loop (PLL) is limited by the product of the tracking loop bandwidth and the integration time (typically <0.4 for a rate only feedback PLL), otherwise the loop will be unstable. Although, Digital PLL or Kalman filter (KF)-based tracking loops are designed in the digital domain to solve this loop stability issue, a longer integration time still does not guarantee improved tracking performance. The reasons are: (i) the underlying assumption of Gaussian distribution for both processing strategies is not always valid because of the signal fading and the multipath; (ii) although both strategies can be designed in the digital domain to solve the stability issues associated with longer integration time, improved tracking performance is not guaranteed because of the corresponding lower channel update rate; and (iii) the linear regions of the discriminator or the equivalent linear regions of the two tracking strategies are too narrow due to the limited number of correlators used. Thus the high sensitivity block processing strategies for the carrier frequency and code phase estimation is used in this work. Unlike conventional closed-loop tracking which utilizes only three correlators: prompt, early and late, the block processing strategy uses a block of correlators to estimate frequency and code phase. It allows for long coherent integration time without the stability issue inherited from the conventional closed loops. The ultra-tight integration used in this work is based on vector tracking loop. Compared with standard tracking loops, the ultra-tightly coupled receiver replaces the standard discriminator and loop filter with a block estimator-based “channel estimator” that estimates the tracking error. The navigation solution is used to directly control local signal generators with the code phase and carrier Doppler being respectively adjusted using the calculated position and velocity solutions. The navigation solution is generated using both GNSS and multi-sensor measurements. The Trusted Positioning Inc.’s hardware platform used in this work utilizes an ARM-based processor and an FPGA platform. The ARM processor is used for the integrated navigation processing and the FPGA is used for the intense signal processing for the HSGNSS receiver. The platform provides interface for inertial sensors, barometers, magnetometers, optionally odometer?and WiFi positioning. Due to the modular design concept, all the sensor measurements can be ultra-tightly integrated with the ultra-tight HSGNSS receiver using a generalized interface. The system can be used either as: (i) tethered navigator for machine or vehicles; or (ii) portable navigator whether strapped or in free motion configurations in vehicle or on-foot with automatic mode switch. So the presented ultra-tight capabilities are added to Trusted Positioning’s navigator series: (i) the Trusted Machine Navigator (T-MN); (ii) Trusted Vehicle Navigator (T-VN); (iii) Trusted Portable Navigator (T-PN) on embedded systems. In terms of the real-time processing, an advanced technique is used to synchronize the tasks of the front-end data acquisition module, the multi-sensor data acquisition module, the signal processing module, and sensor integration module. The proposed system has an alignment phase and it can be initialized with or without HSGNSS. If absolute navigation information is not available from HSGNSS or WiFi positioning, the system can start from the last saved position or from a user given position. The system can be initialized static or in-motion. The alignment routine also calculates the initial attitude without any a-priori information. Finally, the system transfers to the ultra-tight navigation mode when the alignment phase is finished. Field tests were conducted under different operational environments to evaluate both the real-time ability and the performance of the proposed system. The performance is analyzed in terms of the reacquisition speed, tracking sensitivity, positioning accuracy in both urban canyon and indoor environments for the real time embedded system.
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:	2269 - 2275
Cite this article:	Li, T., Georgy, J., Syed, Z., Goodall, C., "Real-Time Navigation System for Ultra-Tight Integration of GNSS and Multi-Sensors," Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013), Nashville, TN, September 2013, pp. 2269-2275.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In