FPGA implementation of motifs-based neuronal network and synchronization analysis

• 13 kinds of motifs are implemented in FPGA and be clarified into different categories with synchronization analysis.
• The pipeline structure is adopted to implement motif-based small-world networks of different scales and categories.
• The synchronization properties of motif-based small-world networks are analyzed from three aspects: network size, rewiring probability and coupling strength.

Motifs in complex networks play a crucial role in determining the brain functions. In this paper, 13 kinds of motifs are implemented with Field Programmable Gate Array (FPGA) to investigate the relationships between the networks properties and motifs properties. We use discretization method and pipelined architecture to construct various motifs with Hindmarsh–Rose (HR) neuron as the node model. We also build a small-world network based on these motifs and conduct the synchronization analysis of motifs as well as the constructed network. We find that the synchronization properties of motif determine that of motif-based small-world network, which demonstrates effectiveness of our proposed hardware simulation platform. By imitation of some vital nuclei in the brain to generate normal discharges, our proposed FPGA-based artificial neuronal networks have the potential to replace the injured nuclei to complete the brain function in the treatment of Parkinson’s disease and epilepsy.