A Semi-Cognitive Localization Approach with Always-On and On-Demand 5G Downlink Signals

A semi-cognitive receiver for opportunistic localization using 5G downlink signals is presented. The receiver starts with “always-on” reference signals (RSs) and cognitively acquires all available “on-demand” RSs in the 5G orthogonal frequency division multiplexing (OFDM) frame. Navigation observables are subsequently derived from the estimated RSs in the receiver. This receiver operates in two main stages. The first is the acquisition stage to detect the 5G base stations (known as gNBs) and to estimate the initial delay, Doppler, and the receiver’s local RS replica. The second is a Kalman filter (KF)-based tracking stage, where the entire set of RSs are used as the local replica. Experimental results demonstrate that the cognitively estimated replica, comprising “always-on” and “on-demand” RSs: (i) exploits nearly the entire bandwidth, leading to a narrower auto-correlation function; (ii) spans all OFDM symbols, allowing for a longer integration time within a frame and higher frequency-domain resolution; and (iii) achieves higher processing gain, which increases the carrier-to-noise density ratio by more than 15 dB, enabling reliable acquisition and tracking under low signal-to-noise ratio (SNR) conditions.