PhD Proposals

Supervisors: P. Atkins, T. Collins and M. Miller (Medical School)

Introduction

Consultants and GPs require a rapid method of detecting the existence of a constriction within the throat. Current techniques require a general anaesthetic and that the head is forced back into an unnatural and painful position. The aim of this programme is to non-invasively determine the diameter of the throat from the returns from an air acoustic sonar system.
Some work has been published in the literature on the deconvolution of incident and reflected acoustic waves entering the mouth from a loudspeaker. Such deconvolution techniques tend to be very sensitive to noise contamination and small measurement or prediction errors. It would thus appear sensible to use conventional multi-sensor microphone arrays to discriminate between the incident and reflected propagation directions. Such techniques also have the potential to improve the signal-to-noise ratio of the measurements thus aiding the deconvolution process.
The majority of the existing literature assumes that the patient has stopped breathing during the measurement process. The patient would obviously prefer to breath normally and the measurements obtained would therefore be more representative of the state of the throat. However, the air speed in the throat will increase in the vicinity of a constriction with a corresponding change in the velocity of propagation of sound. The deconvolution process should obviously include these effects, the magnitude of which may be extracted from an on-line measurement of sound speed.
It is expected that some experimental equipment that will be constructed based on a PC and National Instruments data acquisition card. The experimental test rig will include a broad-band sound source and an array of microphones. Models of a throat will be used in the first instance, as rigorous external assessment procedures are required before experiments can be carried out on live patients.

Proposed Work Plan

Year 1

• General reading and understanding of the respiratory tract in humans.
• Literature review of associated medical papers, Ware-Aki algorithm and the general inverse problem.
• Formulation of the problem in terms measurement requirements and the necessary signal-to-noise ratio for adequate performance.
• Determination of the requirement for the observation time required to obtain suitable measurements.
• Determination of the number of transmitters and receivers required to uniquely determine the diameter of the throat.

Year 2

• Development of the measurement and deconvolution algorithms.
• Development of a laboratory demonstrator.
• Transmission pulse design requirements, receiver signal processing implications and instrument size limitations.
• The applicability of phase conjugate transmission signals.

Year 3

• Reformulation of the problem in terms of theoretical performance limits (Cramer Rao lower bounds, etc)
• Adaptation of algorithms and laboratory tests - write thesis.

Back to Opportunities

Supervisors: P. Atkins and T. Collins

Introduction

The vast majority of large commercial and military vessels constructed within Europe are tested on a noise range prior to customer acceptance. Typical reasons for this might include passenger comfort on fast ferries and the reduction of radiated noise from fishing vessels that may scare away the fish and reduces the catch. Similarly, road vehicles and trains are required to meet a variety of international standards for radiated noise.
Modern ocean vessels are now becoming so quiet that it is becoming increasingly difficult to accurately measure the radiated noise levels using a single hydrophone in a open-water noise range. The obvious answer to this problem is to use multiple hydrophones accurately located with respect to a sea bed reference. However, the cabling costs could run in to millions of pounds and the ocean currents acting on the tethered hydrophones could move their position by many metres.
An alternative solution would be to use a synthetic aperture in order to simulate the angular resolution of a large array of hydrophones. Such a system would require an active beacon to be located on the vessel under test in order to measure the acoustic path-length.
The same system could be used to locate ‘hot-spots’ of radiated noise on a train, car or lorry. In this case the beacon might be something as simple as multi-frequency whistle mounted on the platform for the duration of the test. Such a system could be of significant interest to vehicle manufacturers and validation centres.
The aim of this project is to perform a performance evaluation of such a system and to demonstrate the concept either using facilities such as the anechoic chamber, the main sonar tank facility operated by the School or an existing coastal noise range.

Proposed Work Plan

Year 1

• General reading and understanding of the inverse synthetic aperture concept.
• Literature review of associated noise ranging and measurement papers.
• Formulation of the problem in terms measurement requirements and the necessary signal-to-noise ratio for adequate performance.
• Determination of the requirement for the observation time required to obtain suitable measurements.
• Determination of the number of receivers required to uniquely determine the radiated noise field.

Year 2

• Development of the measurement and inverse synthetic aperture algorithms.
• Development of a laboratory demonstrator.
• Receiver signal processing implications.
• The applicability of measuring general propagation conditions and transducer positions from the data.

Year 3

• Reformulation of the problem in terms of theoretical performance limits (Cramer Rao lower bounds, etc)
• Adaptation of algorithms and laboratory tests - write thesis.

Back to Opportunities

Supervisors: P. Atkins and T. Collins

Introduction

The Acoustics and Sonar section have recently developed an ultra-wideband sonar system for the classification of fish and zooplankton. One of the problems with this system is that it uses a single transducer for transmit and receive phases and so if the fish or zooplankton is not located directly in the boresight of the transducer, the echo strength will vary as a function of frequency and angle of incidence.
An alternative solution might be to use a separate transmit transducer and a multi-element PDVF or 1:3 composite receive array. In this way a split-beam sonar system could be implemented in order to measure the angle of incidence of a single target within any given range cell. There still remains the problem of calibrating the system which now can now be represented by a multi-variable transfer function.
The aim of this project is to assess the performance of a split-beam sonar based on a ceramic composite receiver array and covering the frequency range 25 kHz to 110 kHz, or a PVDF based device operating in the 1 – 4 MHz band. Standard calibration spheres may be used to assess the performance of the system and techniques should be developed for calibrating the system in a typical ocean environment. The sonar system may be implemented using the latest generation of digital signal processors or by suitable interface cards located in a PC. However, the 1 – 4 MHz system may require the use of large FPGA devices.

Back to Opportunities