Synchronization phenomena in dual-transistor spiking oscillators realized experimentally towards physical reservoirs

• Published in: Chaos, solitons and fractals solitons and fractals, 2022-09, Vol.162, p.112415, Article 112415
• Main Authors: Minati, Ludovico, Bartels, Jim, Li, Chao, Frasca, Mattia, Ito, Hiroyuki
• Format: Article
• Language: English
• Subjects: Chaos generation; Pattern formation; Reservoir computing; Spiking dynamics; Synchronization; Transistor
• ISSN: 0960-0779; 1873-2887
• DOI: 10.1016/j.chaos.2022.112415

Transistor-based chaotic oscillators are known to realize highly diverse dynamics despite having elementary circuit topologies. This work investigates, numerically and experimentally using a ring network, a recently-introduced dual-transistor circuit that generates neural-like spike trains. A multitude of non-trivial effects are observed as a function of the supply voltage and coupling strength, including pattern formation under incomplete synchronization and sensitivity to additional long-distance links. Globally-applied noise exerts a synchronizing effect that interacts with the other control parameters. When the network is partitioned in halves at different levels of granularity, their interplay gives rise to adversarial route-to-synchronization phenomena. These results highlight the generative ability of this circuit and motivate its consideration towards the future realization of physical reservoirs. •The spiking dynamics of a ring of transistor-based oscillators are considered.•Experiments show multiple pattern-formation phenomena via synchronization.•Noise injection, additional links and node patterning confer additional complexity.•The system has several features desirable for use in physical reservoir computing.