Localization has recently become a key functionality in cellular networks to support position-based services. Current techniques
used in 4G provide an estimate of the location using a two-step process. The signal time of arrival (ToA) from multiple base
stations is first estimated using a dedicated positioning reference signal (PRS). The position is then determined based on the observed
time difference of arrival (OTDoA) using a multilateration technique. However, rich multipath channels typical in urban or indoor
environments highly degrade the ToA estimation resulting in significant location errors (20m to 50m). This problem is only
partially addressed in the literature.

This project aims at developing a localization system for 5G networks delivering a high
position accuracy in urban and indoor environments (error target of the order of the meter). The first phase of the project targets the
development of ToA estimation schemes adapted to the new modulation formats foreseen for 5G. New reference signals will be designed
to optimize the ToA estimation. The project will secondly investigate how the TDoA positioning multilateration algorithms can be
improved for urban/indoor environments by exploiting the knowledge of the channel response acquired based on the reference
signals. The last phase of the project investigates the possibility to iterate between the two successive steps generally applied when
localizing a mobile terminal (ToA estimation/multilateration) to better take benefit from all the information present in the
observed signals.