Stephen (Steve) McLaughlin - Personal Home Page

Personal Home Page of Steve McLaughlin

I would be glad to hear of any ideas, suggestions or criticisms you have. steve.mclaughlin@ed.ac.uk

Research

My generic research interest is the study of signals and their interaction with systems. These signals are in general electrical signals, as in communication systems, or digitally sampled naturally occurring signals, for example in speech and geophysics. In particular, my focus has been signal processing. Signal Processing can be defined as transformations applied to measurements to maximise their usefulness by an observer or a computer. Typically this is for the purposes of: signal-to-noise enhancement, i.e. optimal extraction of a signal from background noise; detection, i.e. determining if a particular signal is present; parameter estimation, i.e. estimation of the characteristics of a signal or underlying system; classification, i.e. of a signals source or of its nature.

Signal processing, as the above definition implies, is the enabling technology for a wide variety of applications in the modern world, ranging from third generation mobile communication systems to digital multi-media systems, speech synthesis and recognition systems, radar systems, geophysical prospecting systems and biotechnology. The devices that permit the implementation of this technology first became available in the early 1980's.

Consequently the main thrust of my research has been to develop, understand and investigate techniques for a range of practical signal processing problems in real world applications. The continuing increases in computational power which can be expected in the future will enable the development of highly sophisticated new methods for solving problems in signal processing that could not previously have been seriously contemplated. It will then be feasible to build accurate models for the data which are motivated by physical reasoning rather than analytic convenience. In this way, all of our prior information about a problem can be accurately incorporated, thus extracting maximum information about the data.

In particular, my work has concentrated upon developing techniques for; the prediction, regeneration and analysis of naturally occurring signals (e.g. speech,ECG) generated by inherently nonlinear or nonstationary mechanisms. Some examples of these applications are communication systems analysis, techniques for seismic deconvolution and sonar detection as well as image processing problems in biomedical applications such as ultrasound imaging.

My entry in the school publications database provides access to more detailed descriptions of these topics.