Landmobile communication-channel modelling & microphone-array source localization

Abstract

The dissertation has three distinct but loosely complimentary components, which are briefly concluded below. (1) Geometric modeling of landmobile radiowave propagation channels. “Geometric modeling idealizes the spatial geometric relationships among the transmitter, the scatterers, and the receiver in a wireless propagation channel -to produce closed-form formulas of various channel-fading metrics, (such as the distribution of the direction of arrival (DOA) and distribution of the time of arrival (TOA)) using only a very few degrees-of-freedom. In Chapter 2, we thoroughly assessed geometric models in terms of their DOA-distributions against all empirical data available from the open literature. In Chapter 3 a new model for the uplink/downlink multipahs’ TOA-distribution is proposed, the proposed TOA-distribution is compared against some certain empirical data and can better fit them than the customary geometric models can. (2) Near-field measurement model of a microphone-array called acoustic vector-sensors The acoustic vector-sensor is a practical and versatile sound-measurement system in-room, open-air, or underwater. It consists of three identical but orthogonally oriented velocity-sensors plus a pressure-sensor, all spatially collocated. Though its far-field measurement-model has been known for over a decade, we, in chapter 4, pioneer its near-field measurement-model, based on rigorous acoustic physics. Section 4.1 to 4.3 derived the near-field model without any boundary near the acoustic vector-sensor, the closed-form CRB is derived and analyzed. Section 4.4 extends the measurement model from being without boundary to being with a boundary case. (3) Microphone array source localization algorithms In chapter 5, we propose a new algorithm to geolocate a source in 3D near-field space using only one spatially spread acoustic vector-sensor. This algorithm requires no prior knowledge of the temporal structure of the impinging signal, nor any iterative solution. However, this method can allow only one incident source with constant emitting power - a limitation common to basically all received signal Strength Indication (RSSI) methods of geolocation. A new adaptive beamforming signal-processing algorithm is developed in chapter 6 to locate noise-sources aboard a rail-car that passes by a track-side immobile microphone-array. This proposed microphone-array beamformer tracks the rail-car’s spatial movement, with the aid of two inaudible acoustic beacons placed abroad the rail-car. The proposed scheme then localizes the noise-sources with reference to the rail-car’s coordinates. No auxiliary infrastructure (e.g., no radar nor video-camera) is needed besides the onboard beacons. Monte Carlo simulations and anechoic chamber experiments verify the proposed scheme’s efficacy.