Skip to main content
Artificial Neural Networks - Memristor Circuit Modeling
Undergraduate Research Conference, Boise State University (2013)
  • Mikaela Cekalski, Boise State University
  • Dustin Koser, Boise State University
Artificial Neural Networks (ANNs) are a biologically-inspired tool for pattern recognition and learning systems. Software implementations of ANNs have been used with measurable success in applications ranging from robotics to medical fields. A prospect of creating hardware implementations has emerged with the realization of the memristor by Boise State University's Neuromorphic Group. Ground work has been successfully done with system-level simulation for basic logic operations utilizing an ideal memristor model. The research being done explores new ways of adapting current setups to the fabricated memristor's physical device characteristics. The system-level simulation setup and associated algorithms are currently being adjusted to better suit these characteristics. Several circuit alterations are being simulated to assess prospective layouts. These changes will be thoroughly tested with varying parameters to provide thorough data for statistical analysis. These setups and resultant data will prove useful for the Neuromorphic group as their research continues towards implementing ANNs on-chip.
Publication Date
April 15, 2013
Boise, ID
Faculty Sponsor: Elisa Barney Smith
Citation Information
Mikaela Cekalski and Dustin Koser. "Artificial Neural Networks - Memristor Circuit Modeling" Undergraduate Research Conference, Boise State University (2013)
Available at: