Synchronization of Hindmarsh Rose Neurons

Area of Science:

• Computational neuroscience
• Hardware implementation of neural models
• Biologically inspired computing

Background:

• Understanding brain function requires accurate modeling of biological neurons and their interactions.
• Neuronal synchronization is crucial for neural signal processing but challenging to study in vivo.
• High biological accuracy in neuron models increases computational demands and hardware resource requirements.

Purpose of the Study:

• To present a computationally efficient, two-coupled hardware implementation of the Hindmarsh-Rose neuron model.
• To synchronize these neuron models using an exponential coupling function and its approximation.
• To evaluate the hardware cost and performance of the approximated coupling function for neural network applications.

Main Methods:

• Mathematical modeling and simulation of the Hindmarsh-Rose neuron model.
• Hardware implementation using Field Programmable Gate Arrays (FPGAs).
• Synchronization using an exponential coupling function and a cost-approximated version.

Main Results:

• The coupled Hindmarsh-Rose neuron system exhibits diverse dynamical behaviors based on model and coupling parameters.
• Hardware implementation of the approximated exponential coupling function on FPGA demonstrates synchronization with acceptable error.
• The approximated coupling function significantly reduces hardware implementation costs compared to the full function.

Conclusions:

• The approximated exponential coupling function offers a viable and cost-effective method for hardware implementation of synchronized neuronal networks.
• FPGA implementation of the coupled Hindmarsh-Rose neuron model with approximated coupling is suitable for creating functional spiking neural networks.
• This approach facilitates the study of neural dynamics and the development of practical applications like signal encoding and decoding.