Maharatna, Koushik, Ahmadi, Arash and Mangieri, Eduardo
Biologically inspired analogue signal processing: some results towards developing next generation signal analysers.
In, International Symposium on Integrated Circuits (ISIC2009),
Full text not available from this repository.
With the more demand of intensive signal processing for providing better quality of life burden on the traditional signal processing architecture is increasing in terms of power and silicon area. To accommodate advanced signal processing features the main reliance is still on device scaling rather then invention of alternative signal processing approach. Given the fact that the device scaling introduces significant variability at the circuit level the performance of the traditionally built signal processing architectures are running into danger of becoming unreliable in terms of accuracy and functionality. This paper discusses the possibility of doing signal processing taking inspiration from biology which may provide a route to the next generation signal analysis system development. Biological organisms excel in information processing by employing coupled non-linear oscillatory phenomena. In this paper we look at the circuit level development of such non-linear oscillators using analogue design approach. The results show the possibility of reduction in silicon area as well as power and at the same time increasing efficiency of a signal analysis system which could be brought to the main-stream signal processing approach in hardware. The open issues to make the approach commercially viable are also discussed in this paper.
Actions (login required)