Journal of Mathematical Biology

, Volume 29, Issue 3, pp 195–217

Multiple pulse interactions and averaging in systems of coupled neural oscillators


  • G. B. Ermentrout
    • Department of MathematicsUniversity of Pittsburgh
  • N. Kopell
    • Department of MathematicsBoston University

DOI: 10.1007/BF00160535

Cite this article as:
Ermentrout, G.B. & Kopell, N. J. Math. Biol. (1991) 29: 195. doi:10.1007/BF00160535


Oscillators coupled strongly are capable of complicated behavior which may be pathological for biological control systems. Nevertheless, strong coupling may be needed to prevent asynchrony. We discuss how some neural networks may be designed to achieve only simple locking behavior when the coupling is strong. The design is based on the fact that the method of averaging produces equations that are capable only of locking or drift, not pathological complexity. Furthermore, it is shown that oscillators that interact by means of multiple pulses per cycle, dispersed around the cycle, behave like averaged equations, even if the number of pulses is small. We discuss the biological intuition behind this scheme, and show numerically that it works when the oscillators are taken to be composites, each unit of which is governed by a well-known model of a neural oscillator. Finally, we describe numerical methods for computing from equations for coupled limit cycle oscillators the averaged coupling functions of our theory.

Key words

OscillationsNeuronsAveragingNeural circuits
Download to read the full article text

Copyright information

© Springer-Verlag 1991