STARx -- A GPU Based Multi-System Full-Band Real-Time GNSS Software Receiver

B. Huang, Z. Yao, F. Guo, S. Deng, X. Cui, M. Lu

Abstract:	As the construction of global navigation satellite systems (GNSS) and the development of the signals modernization process, in the near future, the four main systems (GPS, BDS, GLONASS, GALILEO) as well as some regional systems such as QZSS and IRNSS will broadcast many signals in multi-frequency. In this process of development, for the systems construction and the ground operation and maintenance, overall monitoring and positioning test of these navigation signals especially of the modernized signals is needed. At the same time, because of the modernization of GNSS signals and the change of the GNSS application area, there are many challenges in GNSS receiver designs. The traditional real-time GNSS receivers have high integration density and low flexibility which are not suited for signals monitoring especially for some new signals. Meanwhile, they cannot handle so many navigation signals in real-time at the same time. Under this application background, the GNSS software receivers which based on software radio get more and more concerned during the last years. On the one hand, the software receivers offer a great level of flexibility in data processing and allow a low-cost implementation with few additional components which easy to change and upgrade for user's needs. On the other hand, the receivers almost based on general high-level programming language (such as C/C++ and Matlab) which makes the receivers can provide PVT results and easy to provide many intermediate data (such as pseudo range, correlation values) which convenient for all aspects of the signals monitoring. However, compare with hardware receivers (such as based on FPGA, ASIC), the software receivers also have some shortcoming. The throughput capacity of the software receivers is not as powerful as hardware receivers. With the rapid development of the computer in recent years, there have been some dual frequency real-time software receiver which using a relatively high configuration computer, but for now, the real time processing of multi-system and multi-frequency is still a challenge for software receivers. Although some techniques which can be used to reduce the computational complexity appear, such as SIMD, bit-wise operation, pre-stored local code and carrier signals, they inevitably have a tradeoff in terms of speed and the quality of the code and phase measurements which cannot match the quality as traditional hardware GNSS receivers and the demand for high precision of the signals quality monitoring. In recent years, with graphics processing unit (GPU) rapid development, there have been some software receivers based on GPU architecture. But to the authors' knowledge, the performance of the GPU has not been fully tapped in existing results. In this paper, in consideration of the demand to process multi-system, multi-frequency, multi-signal, as well as some possible changes in the future, a flexible receiver architecture is shown firstly. In addition to a wideband antenna, a RF front-end and a data acquisition card to fulfill down-conversion and AD conversion, the rest of receiver are implemented in an ordinary PC with a GPU. Under this architecture any GNSS signal can be received and processed easily. After that, the efficient GPU-based channel processing method is presented. The running efficiency of the proposed method independent of hardware configuration are evaluated. With this novel efficient method and the flexible architecture, a high-flexibility multi-system full-band real-time GNSS software receiver is developed. The proposal receiver can process all civil signals from four main systems and some of the regional systems in real-time, including GPS L1C/A, L2C, L5, L1C, BDS B1, B2, GLONASS G1, G2, GALILEO E1, E5a, E5b, and QZSS L1C/A, L2C, L5, and L1C signals. Meanwhile, the GLONASS modernized signals as well as BDS Phase III signals can be quickly added to current receiver after the release of their interface control document (ICD). By using the proposed software receiver, the real-time tracking states and the measured positioning results of each navigation signal (including GPS L1C/A, L2C, L5, BDS B1, B2, Galileo E1, E5a, E5b and GLONASS G1, G2), and the monitoring results of viable GNSS satellites of all civil signals are shown. Finally, since the receiver can handle all the civil GNSS signals in real-time and has a flexible structure, it can be used as a dedicated development platform for various types of navigation research.For example, the software-based architecture makes it can be used as a platform to facilitate developers to test a variety of acquisition, tracking and positioning algorithms, and for the receiver can process across multi-system, multi-frequency and multi-signal, so some testing and validation algorithms of the multi-mode multi-band and the interoperability and compatibility research of multi-system can be implemented on it.
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:	1549 - 1559
Cite this article:	Huang, B., Yao, Z., Guo, F., Deng, S., Cui, X., Lu, M., "STARx -- A GPU Based Multi-System Full-Band Real-Time GNSS Software Receiver," Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013), Nashville, TN, September 2013, pp. 1549-1559.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In