Synchronization in memristively pulse-coupled oscillator networks – experiments

忆阻脉冲耦合振荡器网络中的同步 â 实验

• 批准号: 392855750
• 负责人: Professor Dr. Hermann Kohlstedt
• 金额: --
• 依托单位: Arbeitsgruppe Nanoelektronik
• 依托单位国家: 德国
• 项目类别: Research Units
• 财政年份: 2017
• 资助国家: 德国
• 起止时间: 2016-12-31 至 2020-12-31
• 项目状态: 已结题
• 来源: https://gepris.dfg.de/gepris/projekt/392855750?language=en
• 关键词: Synchronization; memristively; pulse; coupled; oscillator

Temporal synchrony of firing neuronal ensembles is considered as a key ingredient to explain higher brain functions, such as perception or consciousness. The underlying basic feature is known as the binding problem in neuroscience. The spatial and temporal mechanism for the "binding" of different features of an object, such as color and shape, leads to an entire object representation in the brain, but is not fully understood and explored. Spiking neurons can be electronically represented by relaxation type oscillators: based-on programmable unijunction transistors (PUTs) and operational amplifiers the non-linear dynamics, i.e. phase portraits, bifurcation, phase response curves of two and ensembles of relaxation type oscillators are investigated. Neuronal justified mechanisms, like Poisson distributed firing rates of individual neurons, as well as signal retardation and variable coupling strength will be implemented, respectively, by delay lines and memristive devices. While discrete oscillator networks will be investigated to understand the fundamental dynamical aspects of signal delays and variable coupling strengths, higher integrated networks will be built on a 250 nm CMOS technology Finally, likewise topological aspects, as known from the Hippocampus, as well as synchrony mechanisms will be studied in the framework of time-varying networks. For this purpose, a network emulator is also available, which allows to examine individual memristive devices within different circuit topologies and network induced dynamics.

放电神经元集合的时间同步性被认为是解释高级大脑功能的关键因素，如知觉或意识。潜在的基本特征在神经科学中被称为绑定问题。物体不同特征(如颜色和形状)之间的空间和时间绑定机制导致了整个物体在大脑中的表征，但尚未得到充分的理解和探索。脉冲神经元可以用弛豫型振荡器来表示：基于可编程单结晶体管(PUT)和运算放大器，研究了非线性动力学，即两个神经元的相图、分叉、相响应曲线和弛豫型振荡器的系综。神经元的合理机制，如单个神经元的泊松分布放电率，以及信号延迟和可变耦合强度，将分别通过延迟线和记忆器件来实现。虽然离散振荡器网络将被研究以了解信号延迟和可变耦合强度的基本动力学方面，但更高的集成度网络将最终建立在250 nm CMOS技术之上，同样，从海马体中已知的拓扑方面以及同步机制将在时变网络的框架中被研究。为此，还提供了网络仿真器，它允许检查不同电路拓扑和网络感应动力学中的单个记忆器件。