History
The Signal Processing and Communications Research Group at Cambridge was founded in 1969 by Dr (now Prof) Peter Rayner to address the rather new area of digital signal processing (DSP), which had in the mid 1960s received a dramatic boost from the invention of the Fast Fourier Transform algorithm by Cooley and Tukey. Also at that time digital computers were rapidly making the transition from occupying whole rooms to being able to fit in a single rack of equipment, a few feet high. The group started with just two PhD students under Peter Rayner in 1969, and they were joined by a further three students in 1970, including Nick Kingsbury. By 1980, the Group had expanded to three members of staff (faculty) and their students, and by 1990 it had reached around 5 staff and 20 students. In the early days there was emphasis on topics related to improving the efficiency of both algorithms and DSP hardware, which evolved during the 80s into signal-adaptive processing for both audio and video signals. During the 1990s many of the projects developed a strong Bayesian flavour, in addition to an increasing trend towards multi-dimensional signals and their transforms. There is a strong emphasis on Bayesian computational methodology, for example sequential Monte Carlo (particle filters) and Markov chain Monte Carlo (MCMC), as well as geometric algebra and multi-resolution (wavelet) methods, for signal analysis and reconstruction. The Group is particularly proud of its strong links to Cambridge-area hi-tech firms, including Cambridge Consultants, Cedar Audio, Autonomy, BluGnome, Alphamosaic, Geomerics, FeatureSpace, Pneumacare and a number of others. Since the 1980s, Signal Processing and Communications have been increasingly taught in years 2 to 4 of the undergraduate Master of Engineering degree course at Cambridge. Current post-graduate research areas are listed under the Research pages of this site.