Array signal processing using spatial domain infinite impulse response structure and volterra model

• Main Author: Wen, Fuxi
• Other Authors: Ng Boon Poh
• Format: Theses and Dissertations
• Language: English
• Published: 2013
• Subjects: DRNTU::Engineering::Electrical and electronic engineering::Antennas, wave guides, microwaves, radar, radio
• Online Access: https://hdl.handle.net/10356/51128
• Institution: Nanyang Technological University

In this thesis, we discuss array processing using infinite impulse response (IIR) and Volterra signal model in spatial domain. Array beamforming can be interpreted as a spatial filter which allows signals from specific directions to pass through it while suppressing the interference signals. This is analogous to a time-domain digital filter which allows signals to pass through within its passband. The equivalence arises because a digital filter samples the signal in time domain whereas an array samples the signal in spatial domain. With such close relation between array snapshots and discrete-time signals, it is interesting to explore the application of concepts and algorithms developed in different streams of research for array processing. It is well acclaimed that digital filters with desired frequency response can be implemented with either finite impulse response (FIR) structure or IIR structure. In array processing, a one-to-one correspondence of these filter structures with array structures has not been exhaustively studied. Following the analogy between FIR filter structure and a uniform linear array (ULA), the question in hand is: “what is the array structure in spatial domain that is analogous to IIR filter structure?”. We first present the possibility of using an approximate IIR structure for DOA estimation and beamforming applications. We propose a new DOA estimation technique which incorporates virtual feedback resulting in an IIR structure. While the time delay between the taps of the discrete-time IIR spatial structure in the feedback and feedforward networks is a constant proportional to the sampling interval, the analogous delay in spatial domain corresponds to the relative propagation delay across the sensor array which is proportional to the source direction. In the IIR array structure, array snapshots can be directly used for the moving average part of the filter. In order to develop the feedback section of the IIR structure, we propose a virtual structure with an approximate propagation delay. The performance of the proposed technique is observed to have improved performance over the Capon method. Furthermore, we propose a primitive beamformer with an IIR array implementation. Although the configuration presented in this work is not optimal, the scope of this contribution is to prove the concept of IIR structured beamformer. In order to implement a genuine IIR array, it is necessary to have spatial feedback which can be realized using the concept of subarrays. The output of previous subarrays can be used for spatial feedback.