Title :
Parametric stability in a silicon model of the leech swim oscillator
Author_Institution :
Sch. of Sci., Eng., & Technol., Pennsylvania State Univ., Harrisburg, PA, USA
Abstract :
Parametric sensitivities of the leech swim network are characterized in a cell-by-cell model implemented in analog VLSI hardware. Each hemiganglion of the swim network contains 11 tonically active cells arranged in 34 distinct oscillator subnetworks. Initial circuit tests showed reciprocally inhibitory networks to be sensitive to synaptic fatigue and refractory period, yet immune to cell threshold, while cyclically inhibitory networks were found to be sensitive to cell threshold, yet immune to refractory period. In subsequent tests, a model of the entire hemiganglion was found to to be significantly less sensitive to all parameters than either suboscillator type alone
Keywords :
biomechanics; cellular biophysics; neural chips; neurophysiology; parametric oscillators; physiological models; very high speed integrated circuits; analog VLSI hardware; cell threshold; cell-by-cell model; circuit tests; cyclically inhibitory networks; distinct oscillator subnetworks; hemiganglion; leech swim oscillator; parametric stability; reciprocally inhibitory networks; refractory period; silicon model; synaptic fatigue; tonically active cells; Biomembranes; Cells (biology); Circuit testing; Hardware; Immune system; Intelligent networks; Oscillators; Silicon; Stability; Very large scale integration;
Conference_Titel :
Bioengineering Conference, 1996., Proceedings of the 1996 IEEE Twenty-Second Annual Northeast
Conference_Location :
New Brunswick, NJ
Print_ISBN :
0-7803-3204-0
DOI :
10.1109/NEBC.1996.503244
