Fast, cognitive, and compact integrated systems are needed for signal processing and perception applications. Bio-inspired systems utilizing embedded neural networks (NNs) and genetic algorithms (GAs) shall be presented. EMBEDDED NEURAL NETWORKS with their stable solutions, and learning abilities provide for discovery, reinforcement, and classification of signal information. EMBEDDED GENETIC ALGORITHMS provide wide range solutions for optimization and characterization problems. Hardware artificial NN algorithms are divided into the following categories: Unsupervised, Supervised, Reinforcement, and those utilizing NN Plasticity. Sample applications with integrated circuits including: Bio-Chemical detection (E-Nose), and multi-strategy adaptive communication networks are emphasized, including continuous recurrent dynamic neural networks (RDNN) for noisy signals, GAs for signal characterization and priority optimization, self-organizing feature maps (SOFM) to discover ordered relationships among communication signals, and reinforcement neural networks (RNNs) for bio-chemical detection. Due to the enormous synapse routing requirements and chip area limitations a novel NN-plasticity-based approach for a pseudo bloodhound nose has been developed and will be discussed.

Biography

Hoda Abdel-Aty-Zohdy received the B.A.Sc. degree (with First Class Honors) from Cairo University, the M.A.Sc. and the Ph.D. degrees from the University of Waterloo, all in Electrical Engineering. She is the Founder and Director of the Microelectronics System Design Laboratory, Coordinator of the Engineering Physics Program, and a Full Professor in the Department of Electrical and Systems Engineering at Oakland University. She has been selected as an IEEE/Circuits and Systems Society (CASS) 2004-2005 DISTINGUISHED LECTURER. She has over 135 refereed publications, and more than 100 technical presentations.

Dr. Abdel-Aty-Zohdy has been awarded an AFOSR/IFTA Faculty Research Fellowship in 2003 and 2002; a National Academy of Science/National Research Council Fellowship at the WPAFB, in 2001 and 2000; a Faculty Internship at the Chrysler Technology Center, Advanced Manufacturing Engineering, 1998 and 1997; Consultant to Fanuc-Berkeley MEMS Lab, 1996; DARPA supported Visiting Associate Professor at The University of Michigan, Ann Arbor/Center for Integrated Sensors and Circuits, 1995; Consultant to General Motors Research Labs, ITT; and a summer visiting professor at the Institute for Computer Research, at the University of Waterloo.

Dr. Abdel-Aty-Zohdy presented keynote and plenary lectures at the IEEE/IEE International Conference on Telecommunications ICT05 (South Africa), IEEE NEWCAS 2004 (Canada), IEEE ECCTD 2003 (Poland), and IEEE ICCC 2001 (Tunisia). She has been the elected Chair for the IEEE/South East Michigan Section-Chapter-I on Circuits and Systems (CAS), Signal Processing (SP), and Information Theory (IT) since 2000. She served on numerous IEEE committees including: the Chair of the 2005-2006 IEEE G. R. Kirchhoff award Committee; Chair of future IEEE/MWSCAS symposia committee, general co-chair, track chair, session organizer, and speaker for the IEEE Midwest Symposium on Circuits and Systems (1991-present) ; and Organizer and General Chair of the Educational program for the IEEE Custom Integrated Circuits Conference (CICC88). She served as the General Chair for the first Collaborative Technologies Workshop in 1999, at Oakland University. She has been selected as The Speaker for the President's Colloquium Series, at OU, for 2004-05.

She is a Member of the IEEE, Eta Kappa Nu, Sigma Xi, ACM SIGDA, AWIS Detroit Area chapter, IASTED, CSEE, EIC, and a Senior Member/Faculty Advisor (1982-98) for the Society of Women Engineers, and Faculty Adviser to Eta Kappa Nu (HKN) the EE/CE Honor Society at OU (2005-present).