RIS-Aided Joint Initial Access and Localization at Millimeter Wave, BWAC Core Project

More Details

MIMO communication at mmWave and sub-THz provides the angle and delay resolvability that enables accurate channel estimates and localization of users and objects in the environment. In this context, we consider the problem of joint localization and communication in cellular networks, where the received signal at the device or BS can be processed to provide localization information as a by-product of communication.
In this project, we aim to obtain position information of the device during initial access. The problem of position tracking under mobility once the link is established is different from single shot initial localization, and it is left to future work. From a signal processing perspective, several algorithms for joint communication and localization have been proposed for mmWave bands in the last few years. One class of algorithms exploits an ultradense deployment to perform beam-based measurements of different parameters – direction of departure (DoD), direction of arrival (DoA), received signal strength (RSS), or time of arrival (ToA) – with reference signals transmitted from multiple BSs [1], [2], [3]–[5]. While those that exploit DoA and ToA from multiple BSs provide a good accuracy [6], they require very precise network synchronization and do not perform well in current 3GPP networks [7]. Another class of algorithms are based on a two-stage procedure [8-11], which includes a compressive channel estimation phase and a second stage that maps the estimated channel parameters (angles and delays) to the position and orientation of the user exploiting their geometric relationship. That work assumes it is possible to estimate the absolute value for the delay (ToA) of the line-of-sight (LoS) component, obtained by capturing the received signal when the transmitter starts its transmission, as if triggered by an oracle device. Even when making unrealistic assumptions, unfortunately, current approaches do not provide the required localization accuracy for some usecases envisioned in future wireless networks (1 cm for indoorand 10 cm for outdoor) [12].
We propose leveraging mmWave communication aided by reconfigurable intelligent surfaces (RIS) to address the challenges of joint localization and communication for use cases that require very high accuracy. We will derive a localization algorithms aided by a single RIS for single user localization assuming realistic MIMO communication architectures for mmWave and sub-THz bands, i.e. a hybrid beamforming MIMO architecture or a low resolution digital architecture [13]. If time permits, we will also investigate solutions for the multiuser setting, considering multiple RIS either by spatially multiplexing the signal towards multiple RIS at the same time or by making it bounce on multiple RIS to increase the communication and localization range. The developed algorithms will be evaluated using ray tracing simulations.