Theoretical limits of drone positioning in cellular networks : Estimation limits of time-of-arrival, angle-of-arrival, and Doppler measurements.

Sammanfattning

Using 5G and future 6G mobile networks for detection and positioning of Unmanned Aerial Vehicles (UAVs) is an ongoing research field termed Integrated Sensing and Communication (ISAC). Before such systems can be deployed fully, the fundamental performance limits of UAV positioning from cellular signals must be understood. This thesis derives the Cramér-Rao Lower Bound (CRLB) of channel parameters as well as analytical Position Error Bound (PEB) and Velocity Error Bound (VEB) on UAV target estimation under a line-of-sight propagation assumption using the standardised 5G numerology. The CRLB is derived in four scenarios of increasing complexity (monostatic Time of Arrival (ToA); bistatic ToA; bistatic ToA+Angle of Arrival (AoA); bistatic ToA+AoA+Doppler) using a two-stage Slepian-Bangs and geometric-Jacobian methodology. We find that using standard 5G FR1 numerology, sub-metre positioning is possible using only monostatic ToA at typical ranges of 300–500 m and moderate signal-to-noise ratio (SNR). The PEB is found to scale as 1/𝑊 in the bandwidth 𝑊 when using ToA, while AoA measurements tighten the PEB further through the receiving rectangular antenna array of size 𝑀3 × 𝑀4 (where 𝑀1 and 𝑀2 are reserved for transmitting antennas). The final scenario finds that the VEB scales as 1/𝐾3/2 in the number of Orthogonal Frequency-Division Multiplexing (OFDM) symbols 𝐾 using Doppler measurements. These bounds are a necessary first step toward assessing whether existing cellular infrastructure can plausibly support UAV surveillance; deployment-level questions such as coverage density, multipath, clutter rejection, and non-line-of-sight (NLoS) regions are outside the scope of this thesis.