Summary

IEICE Information and Communication Technology Forum

2014

Session Number:COMM3

Session:

Number:COMM3-2

Acoustic Beamforming in Complex Dynamic Systems

Bamidele Adebisi,  

pp.-

Publication Date:2014-08-10

Online ISSN:2188-5079

DOI:10.34385/proc.19.COMM3-2

PDF download (500.7KB)

Summary:
Dynamic systems such as a shopping mall, conference auditorium, underwater operations, airport, seaport, warehouses, production yards, factories and process plants present different noise fields for sound detection and/or perception. Consequently, beamforming operations that perform signal processing deterministically have been investigated for sensor arrays including microphone arrays and sonar arrays. Microphone arrays have found tremendous applications in improving the quality of received audio signals. Uniform linear arrays of microphones have been deployed for spatial filtering targeting interference and/or noise rejection in order to enhance the quality and gain of the desired signal. Acoustic beamforming in complex dynamic systems is a non-trivial task given the time-variant nature of system variables. For a time-invariant system, the dynamics of the system can be modelled and utilised to estimate the system behaviour in real-time. In this presentation a new beamforming algorithm that improves the signal gain for acoustic characterisation of dynamic systems in real-time will be discussed. This beamformer utilizes an adaptive algorithm that selectively enhances the desired signal based on the noise conditions of the individual sensors thereby avoiding noisy signal leaks through the FIR filters of each sensor. The proposed technique, which can be implemented in both endfire and broadside array configurations, will be shown to require less sensors and filter taps for FPGA-based embedded real-time implementations. This holds a great promise for designing, developing and deploying a reliable, sustainable, cost-effective, compact and high fidelity acoustic sensor system for mobile engineering systems and applications.