Direct Wideband Coherent Localization by Distributed Antenna Arrays

Area of Science:

• Signal Processing
• Wireless Communications
• Localization Techniques

Background:

• Accurate radio transmitter localization is crucial for modern wireless systems, especially in dense urban environments and indoor scenarios.
• Multipath propagation and line-of-sight (LoS) signal components present significant challenges for traditional localization methods.
• Millimeter-wave (mmWave) frequencies, common in small cells, exhibit unique propagation characteristics.

Purpose of the Study:

• To develop and evaluate wideband direct coherent localization algorithms for distributed antenna arrays.
• To assess performance in realistic multipath scenarios with spatially-coherent LoS components.
• To compare proposed algorithms against established methods like SCM-MUSIC.

Main Methods:

• Development of two machine learning (ML)-type algorithms for transmitter waveform localization (known and unknown waveforms).
• Implementation of a subspace-based SCM-MUSIC algorithm for position estimation.
• System modeling with collocated, time/phase-synchronized front-ends and 3D distributed antennas.
• Signal modeling considering spherical wavefronts.

Main Results:

• Monte Carlo simulations demonstrate high localization accuracy, orders of magnitude better than the carrier wavelength.
• Proposed ML-type algorithms and SCM-MUSIC show robust performance in multipath environments.
• Methods not requiring waveform knowledge achieve mean-squared errors close to the Cramér-Rao bound.

Conclusions:

• Wideband direct coherent localization is feasible with distributed antenna arrays, even in complex multipath conditions.
• The proposed algorithms offer significant improvements in localization accuracy for mmWave systems.
• Future research can focus on further optimizing algorithms for real-world deployment and varying channel conditions.