Neural Signal Processing With Programmable Cortical Arrays

Arfan Ghani

Research output: Book/ReportBook

Abstract

Recent years have witnessed some neuro-inspired engineering developments, to name a few (Neuromorphic chip design, adaptable and autonomous sensor networks, neuroprosthesis and neuro-electronics based on nanotechnology) but the huge challenge remains where rapid developments involving the interplay of biology, computer sciences, mathematics and engineering have yet to be seen. The book first explores different biologically realistic mathematical models of cortical spiking neurons and time-efficient ways for their mapping on reconfigurable hardware. This book then turns to the complexity of neural representation with a thorough literature review and state-of-the-art in neural hardware/software co-design. Next, it examines one of the few existing theoretical frameworks for modelling cortical circuits/columns and proposes solutions for their implementation on reconfigurable hardware, termed as Field Programmable Cortical Arrays. Finally, the book looks at design partitioning on hardware/software platforms, including the use of cortical columns in the context of speech recognition application. This book is written for researchers and professionals both in academia and industry.
Original languageEnglish
PublisherVDM Verlag
Number of pages212
ISBN (Print)978-3-639-21630-1
Publication statusPublished - 2009

    Fingerprint

Keywords

  • neurocomputing
  • reconfigurable hardware
  • cortical computations
  • spiking neural networks
  • area-efficient architectures
  • FPGAs
  • Hardware/Software Co-design

Cite this