ION GNSS 2012
Session B1: GNSS Simulation, Testing & Timing Applications 1

Title: Multi-Level GNSS Signal Simulator for Multi-System GNSS Receiver Developing
Author(s): X. Zhang, W. Mou, G. Sun, G. Ou, National University of Defense Technology, China
Date/Time: Wednesday, September 19, 2012, 9:20 a.m.
Room: 103/104 (NCC)

Since Global Positioning System (GPS) keeps running for almost 20 years, modernized Global Navigation Satellite Systems (GNSS) has become one of the most important demands for navigation and location services. New GNSS systems like Galileo and COMPASS are making their efforts to feed the demand of providing higher measurement accuracy, better tracking robustness and more tracking sensitivity. These improvements are achieved through new signal structure, more frequency bands, system interoperation and so on. Thereby, testing multi-system GNSS receivers in the developing phase becomes a complicated work, including new architectures designed for new system navigation signal acquisition, tracking and data demodulation processes, especially for COMPASS and Galileo navigation system which is under construction. GNSS signal simulator provides a comprehensive simulation platform for verifying and assessing the multi-system GNSS receiver before the corresponding new satellites are deployed. The traditional GNSS simulator can really help in analyzing and developing new receiver algorithms, but when it comes to find the detail of the errors or comparing the performance of the receiver in customized signal situation, we can get little help from it. Thereby, a GNSS simulator providing multi-level testing method can speed up the self-developed multi-system GNSS receiver development greatly. This paper presents a Software Defined Radio (SDR) architecture multi-level testing GNSS signal simulator based on PXI bus, which can generate GPS, Galileo and GLONASS navigation signal, even the candidate signal for COMPASS, on the common hardware platform. First of all, the system implementation, frequency plan and PXI bus structure are addressed in this paper detail. Before introducing the division of the device function modules and hardware design, we analyze the signal structure of all the four navigation system, although the COMPASS just gives the candidate. Then we chose DSP and high-capacity high-performance FPGA structure as the basic signal generating hardware module, which can easily adapt to the signal structure changing. Besides the basic signal generating module, we need time-frequency module, bus control module and the RF module to build the whole simulator. All these modules are connected with PXI bus, which is one of the most popular bus specifications for instrument developing. When it comes to frequency plan, we don´t consider the plan for every system independently; we take the spectrum block as the plan standard instead. After brief introduction of the implementation and design information, we have our attention focus on explaining how to divide testing level, and what we want to do at different testing levels. Generally, the multi-level testing GNSS signal simulator can provide three testing levels, signal testing level, information testing level and system testing level. These three levels give effective testing method for the developer of the multi-system GNSS in different developing phase, and make it easier to distinguish the causes of errors resulting from different level. Accordingly, we define three working modes for the multi-level testing GNSS signal simulator: signal testing mode, information testing mode and system testing mode. When working at signal testing mode, multi-level testing GNSS signal simulator outputs the signal with only the same modulation method as GNSS system signal, but not the same navigation message. Take Galileo E1 signal for example, the simulator outputs the RF signal, which is also CBOC(6,1,1/11) with the same pseudo random code and the same message frame, but the data in the message frame can be customized. The self-developed receiver can receive the "pure signal", and verify the signal acquisition and tracking. Also we can make a little modification on the signal, such as getting rid of the weak pilot signal, and compare the receiver performance using the signal with pilot signal and without pilot signal. Certainly, it also need the receiver developer to make some modifications for the receiver´s working mode, but it must be worthy. Information testing mode usually starts after the signal testing has been finished successfully. This working mode helps the developers to validate their information processing algorithms, such as Multi-frequency ionosphere correction algorithm, multi-side integrity monitoring method which combined system integrity information and (Receiver Autonomous Integrity Monitoring) RAIM, the differences of GDOP and WGDOP in the application of mixed constellation, and so on. System testing mode is the top level testing mode, and it is a special mode for multi-system GNSS receiver. When we develop the multi-system GNSS receiver, we may be interest in some question:(1).does our receiver always have better performance when using multi-system signal?(2)if better performance can be achieved when the receiver is taking multi-system signal other than single system, then how many systems it need to use?(3)if it is no need for our receiver to use all the viable and future viable system, can we pick 2 or 3 system randomly and get the according performance? Or can we find a best combination of the different systems at different location and different time? The system testing mode is designed for testing the solution for all the question above. At last, the testing results of the self-developed multi-system GNSS receiver at different testing level are presented, and it looks like that the multi-level GNSS signal simulator really gives a lot of help to the developer of the receiver.

Previous Abstract Return to Session B1 Next Abstract