Fast and Accurate Estimation of Angle-of-arrival in Millimetre-wave Large-scale Hybrid Arrays

Publication Type:
Thesis
Issue Date:
2019
Full metadata record
Hybrid array is able to leverage array gains, transceiver sizes and costs for massive multiple-input-multiple-output (MIMO) systems in millimetre wave frequencies. Challenges arise from estimation of angle-of-arrival (AoA) in hybrid arrays, due to the array structure and the resultant estimation ambiguities and susceptibility to noises. In this thesis, we study the unambiguous and non-iterative AoA estimation in two types of hybrid arrays | Localized Hybrid Array of Phased SubArrays (LHA-PSAs) and Lens Antenna Arrays (LAAs). For each type, two AoA estimation approaches are proposed for narrowband and wideband, respectively. The main innovation of the approaches include: • The deterministic sign rules and patterns in LHA-PSAs are unprecedentedly discovered, and exploited to eliminate the estimation ambiguities; • The optimal trade-off between different error sources is achieved, minimising the wideband AoA estimation error in LHA-PSAs; • A new wide beam synthesis method is developed for LAAs, which substantially improves the AoA estimation efficiency in LAAs; • New spatial-frequency patterns are unveiled exploiting the spatial-wideband effect, leading to fast and accurate wideband AoA estimation in LAAs. Performance analysis is provided for all the approaches with closed-form estimation (lower) bounds derived. Corroborated by simulations, our approaches are able to dramatically improve AoA estimation accuracy while reducing complexity and the number of training symbols, as compared to the state of the art. The estimation errors of our methods asymptotically approach the (lower) bounds.
Please use this identifier to cite or link to this item: