Cognitive Opportunistic Navigation in Private Networks With 5G Signals and Beyond

A receiver architecture is proposed to cognitively extract navigation observables from fifth generation (5G) new radio (NR) signals of opportunity. Unlike conventional opportunistic receivers which require knowledge of the signal structure, particularly the reference signals (RSs), the proposed cognitive opportunistic navigation (CON) receiver requires knowledge of only the frame duration and carrier frequency of the signal. In 5G NR, some of these RSs are only transmitted on demand, which limits the existing opportunistic navigation frameworks to signals which are on
always-on
; hence, limiting the exploitable RS bandwidth. To exploit the full available bandwidth and improve ranging accuracy, the proposed CON receiver is designed to estimate all the RSs contained in the transmitted signals corresponding to multiple 5G base stations, (i.e., gNBs). Navigation observables (pseudorange and carrier phase) are subsequently derived from the estimated RSs. The proposed receiver operates in two stages: (i) acquisition and (ii) tracking. The acquisition stage of the CON receiver is modeled as a sequential detection problem where the number of gNBs and their corresponding RSs and Doppler frequencies are unknown. The generalized likelihood ratio (GLR) test for sequentially detecting active gNBs is derived and used to estimate the number of gNBs and their RSs. In order for the receiver to refine and maintain the Doppler and RS estimates provided by the acquisition stage, tracking loops are designed. A sufficient condition on the Doppler estimation error to ensure that the proposed GLR asymptotically achieves a constant false alarm rate (CFAR) is derived. The output of the tracking loops, namely carrier phase and code phase, are then used to estimate the receiver’s position. Extensive experimental results are presented demonstrating the capabilities of the proposed CON receiver with real 5G signals on ground and aerial platforms, with an experiment showing the first navigation results with real 5G signals on an unmanned aerial vehicle (UAV) navigating using the CON receiver over a 416 m trajectory with a position root mean-squared error (RMSE) of 4.35 m.