Privacy-Aware Localization Enhancement Based on Intelligent Reconfigurable Surfaces

Abstract

5G networks enhance the localization performance for both legitimate, yet also for the misuse by adversarial devices. To distinguish between legitimate and adversarial devices, an intelligent reconfigurable surface (IRS) can be configured to only optimize the channel for localization when a user equipment (UE) is localizing itself. The channel comprises a direct part between the new-generation node base (gNB) and the UE, and an indirect part over the IRS. We aim at maximizing the mutual information between the channel and the received signal by adjusting the transmit signal of the gNB and the IRS configuration alternately. In order to allow the UEs to localize themselves on behalf of its own received signal, a publicly known data-set of fingerprints is transmitted to the UEs. For the actual localization, the optimized IRS configuration is enabled for a short time instant and the optimized transmit signal broadcasted. Both parameters are optimized for the channel statistics of a subset of the cell area, in which the UE is situated, rather than an actual channel. Hence, adversaries can only obtain a coarse localization based on the channel statistics. The proposed technique enables an enhanced localization performance compared with the IRS-free case.