Optimizing waveforms for positioning in 5G

Today's mobile radio systems deploy reference signals which can be used or which are even dedicated for signal propagation delay-based mobile terminal positioning. Usually, the signal power of such reference signals is uniformly distributed over the available spectrum. It is known from estimation theory that such a uniform power distribution of reference signals is not optimal for signal propagation delay estimation. In this paper we consider mobile terminal positioning based on signal propagation delay estimation in the uplink case. For positioning, we introduce a parametric waveform. This waveform provides a scalar parameter for controlling the distribution of the available signal power over the spectrum. Using this waveform parameter we aim to minimize the positioning error. For optimization, we require a functional dependency between the waveform parameter and the positioning error we can expect. For the derivation of this function we combine the approaches of the Cramér-Rao and Ziv-Zakai bounds for position and propagation delay estimation. As an exemplary environment we consider a mobile terminal located in an area surrounded by 3 base stations. For this environment we show that the optimized waveform spends a significant part of the available power at the spectrum edges, leading to a performance gain of 37.3% at the center of the area between the base stations.