GNSS Signal Processing and Spatial Diversity Exploitation

Global Navigation Satellite Systems (GNSS) signals are broadly used for positioning, navigation and timing (PNT) in many different applications and use cases. Although different PNT technologies are available, GNSS is expected to be a key player in the derivation of positioning and timing for many future applications, including those in the context of the Internet of Things (IoT) or autonomous vehicles, since it has the important advantage of being open access and worldwide available. Indeed, GNSS is performing very well in mild propagation conditions, achieving position and time synchronization accuracies down to the cm and ns levels, respectively. Nevertheless, the exploitation of GNSS in harsh propagation conditions typical of urban and indoor scenarios is very challenging, resulting in position errors of up to tens or even hundreds of meters, and timing accuracies of hundreds of ns. This thesis deals with the processing of GNSS signals for positioning and timing in harsh propagation conditions. In particular, the focus is on signal processing techniques exploiting the spatial diversities present both at transmission and reception levels when multiple GNSS satellites are in view by multiple receiver antennas, which form a multiple-input multiple-output (MIMO) system. In this context, three problems or research areas open in the GNSS literature are targeted. The first research area is the unambiguous estimation of and positioning with high-order binary offset carrier (BOC) signals. The second research area is the time synchronization in indoor conditions. And the third research area is the positioning with co-located and distributed receiver antennas. In the first research area, this thesis shows that the robust unambiguous positioning with high-order BOC signals in harsh propagation conditions is possible when jointly exploiting these signals in the position domain and taking advantage of the spatial diversity introduced by arrays of antennas. The proposed estimators introduce an important benefit with respect to single-satellite-based unambiguous techniques (operating at pseudorange level) thanks to the processing gain introduced by the MIMO-GNSS system formed. Indeed, when multiple antennas are featured by the receiver, the proposed approach allows the exploitation of high-order BOC signals even in indoor conditions, achieving positioning accuracies of few meters in propagation conditions for which BPSK(1) signals can only achieve accuracies of tens of meters. Moreover, the proposed solutions are implementable based on state-of-the-art multi-correlator receiver architectures, and allow a drastic reduction of the computational burden with respect to the typical implementation of the so-called direct positioning estimation (DPE) techniques. In the second research area, this thesis proposes a joint time and channel estimation approach for static indoor GNSS receivers featuring an array of antennas in order to improve the timing accuracy in indoor propagation conditions. This approach exploits both the structure of the diffuse multipath components of the indoor channel and the MIMO system formed by all the GNSS signals received via an array of antennas. Simulation results with a wideband satellite-to-indoor channel model show that the proposed timing estimators allow an important mitigation of the dominant indoor multipath conditions. Therefore, the joint time and channel estimation approach proposed is considered an appealing solution for indoor applications with tight synchronization requirements, as can be the case in indoor small cells for 5G. Finally, in the third research area, this thesis proposes the exploitation of co-located and distributed receiver antennas for positioning in harsh propagation conditions. In order to improve the performance achieved with co-located antennas, a distributed array processing approach for collaborative GNSS-based snapshot positioning is proposed in the MIMO-GNSS framework. In this solution, one of the receivers is used as anchor and a distributed array is formed, allowing to transform the positioning problem into an angle estimation problem in order to reduce the computational burden. The exploitation of the spatial diversity introduced by multiple receivers enables the derivation of an improved position solution for receivers in degraded propagation conditions, taking in particular advantage of those in better propagation conditions.