Controlling neuronal spikes.

We propose two control strategies for achieving desired firing patterns in a physiologically realistic model neuron. The techniques are powerful, efficient, and robust, and we have applied them successfully to obtain a range of targeted spiking behaviors. The methods complement each other: one involves the manipulation of only a parameter, the applied soma current, and the other involves the manipulation of only a state variable, the membrane potential. Both techniques have the advantage that they are not measurement-intensive nor do they involve much run-time computation, as knowledge of only the interspike interval is necessary to implement control.