This paper presents a novel technique to improve data demodulation performance of global navigation satellite system (GNSS) signals with data and pilot signal components (e.g. suitable for Galileo E1B/C open service (OS) signals [1]). This technique is called Pilot-Aided data demodulation because it uses the coherence between residual phase errors on the data and the pilot components of the GNSS signal to perform channel estimation. Differently from standard techniques [2], where the carrier phase of the received signal is tracked by means of a phase locked loop (PLL) on the pilot component and such estimation is used to control the data correlators, the pilot aided technique does not necessarily requires a PLL, but it can work as well with a simple and more robust frequency locked loop (FLL) or just with open loop frequency estimation. This is particularly valuable in challenging environments, such as urban scenarios, where the phase lock is hard to be maintained for long periods. The main advantage of this technique is clearly visible in terms of page error rate under fades and multipath. In open sky scenarios and additive white Gaussian noise (AWGN) assumption, it exploits the data plus pilot combination, doubling de-facto the power of the received signal with respect to the single data-less component, but it also introduces additional noise (similar to squaring losses) due to the non-linear operations between data and pilot prompt correlators. The output of such technique can then be used to feed the proper decoder according to the coding scheme adopted in the GNSS navigation message (e.g. Viterbi decoder for E1B). The results presented in this paper focus on the Galileo E1 OS signal and the data demodulation performance achievable in realistic and representative channel models and in on-field tests. The Galileo radio frequency signals, indeed, provided as first a set of signals on several frequencies including Data and Pilot components. Anyway, looking at the first attempts to introduce Galileo into commercial low-cost receivers, it was tried to adapt the Galileo signal structure to the standard and mature processing used for the GPS L1 coarse/acquisition (C/A), e.g. by considering the Data-only component. This clearly cannot exploit the advantage of the Galileo signal structure and it brings to performance degradation not only in terms of tracking but also of data demodulation. In order to verify the advantage brought by this novel technique, it was firstly implemented in a navigation software receiver and it was tested in several environments. Then, the same scenarios were used as well to test an off-The-shelf Galileo ready mass-market receiver and finally the results of the commercial chip and the SW receiver implementing Pilot-Aided data demodulation were compared. The improvement shown is always medium or very high, but it strongly depends on the testing environment and on the propagation channel. In order to better understand such improvement, several live scenarios were considered, by replaying real radio-frequency signals. These signals were acquired by a van equipped with different antennas, GNSS receivers and bit-grabber devices. The live scenarios include a database of 90 hours collected in urban, sub-urban and motorway environments which were replayed in the navigation laboratory at ESTEC.
Pilot-Aided GNSS data demodulation performance in realistic channels and urban live tests
DE OLIVEIRA SALGUEIRO, FILIPE;Luise, Marco;
2016-01-01
Abstract
This paper presents a novel technique to improve data demodulation performance of global navigation satellite system (GNSS) signals with data and pilot signal components (e.g. suitable for Galileo E1B/C open service (OS) signals [1]). This technique is called Pilot-Aided data demodulation because it uses the coherence between residual phase errors on the data and the pilot components of the GNSS signal to perform channel estimation. Differently from standard techniques [2], where the carrier phase of the received signal is tracked by means of a phase locked loop (PLL) on the pilot component and such estimation is used to control the data correlators, the pilot aided technique does not necessarily requires a PLL, but it can work as well with a simple and more robust frequency locked loop (FLL) or just with open loop frequency estimation. This is particularly valuable in challenging environments, such as urban scenarios, where the phase lock is hard to be maintained for long periods. The main advantage of this technique is clearly visible in terms of page error rate under fades and multipath. In open sky scenarios and additive white Gaussian noise (AWGN) assumption, it exploits the data plus pilot combination, doubling de-facto the power of the received signal with respect to the single data-less component, but it also introduces additional noise (similar to squaring losses) due to the non-linear operations between data and pilot prompt correlators. The output of such technique can then be used to feed the proper decoder according to the coding scheme adopted in the GNSS navigation message (e.g. Viterbi decoder for E1B). The results presented in this paper focus on the Galileo E1 OS signal and the data demodulation performance achievable in realistic and representative channel models and in on-field tests. The Galileo radio frequency signals, indeed, provided as first a set of signals on several frequencies including Data and Pilot components. Anyway, looking at the first attempts to introduce Galileo into commercial low-cost receivers, it was tried to adapt the Galileo signal structure to the standard and mature processing used for the GPS L1 coarse/acquisition (C/A), e.g. by considering the Data-only component. This clearly cannot exploit the advantage of the Galileo signal structure and it brings to performance degradation not only in terms of tracking but also of data demodulation. In order to verify the advantage brought by this novel technique, it was firstly implemented in a navigation software receiver and it was tested in several environments. Then, the same scenarios were used as well to test an off-The-shelf Galileo ready mass-market receiver and finally the results of the commercial chip and the SW receiver implementing Pilot-Aided data demodulation were compared. The improvement shown is always medium or very high, but it strongly depends on the testing environment and on the propagation channel. In order to better understand such improvement, several live scenarios were considered, by replaying real radio-frequency signals. These signals were acquired by a van equipped with different antennas, GNSS receivers and bit-grabber devices. The live scenarios include a database of 90 hours collected in urban, sub-urban and motorway environments which were replayed in the navigation laboratory at ESTEC.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.