Temporal Information Processing With an Integrated Laser Neuron
Temporal Information Processing With an Integrated Laser Neuron
复制标题
DOI：
10.1109/jstqe.2019.2927582
复制
发表时间：
2020
影响因子：
4.9
通讯作者：
Hsuan-Tung Peng;Gerasimos Angelatos;T. F. de Lima;M. Nahmias;A. Tait;S. Abbaslou;B. Shastri;P. Prucnal
中科院分区：
文献类型：
--
作者：
Hsuan-Tung Peng;Gerasimos Angelatos;T. F. de Lima;M. Nahmias;A. Tait;S. Abbaslou;B. Shastri;P. Prucnal
脉冲神经网络能够实时高效地处理信息。可激发激光器能够呈现超快的脉冲动力学特性，并且当在光/电/光链路中位于光电探测器之前时，它能够处理不同波长的光脉冲，从而在大型神经网络中相互连接。在此，我们通过实验演示并数值模拟了在光子集成电路中制造的激光神经元的脉冲动力学。我们的脉冲激光神经元被证明能够以纳秒级的时间分辨率进行符合检测，并且我们观察到约0.1纳秒的不应期。我们提出了一种使用我们的激光神经元实现异或分类的方法，并且对由此产生的动力学的模拟表明对定时抖动具有很强的耐受性。
Spiking neural networks enable efficient information processing in real-time. Excitable lasers can exhibit ultrafast spiking dynamics, and when preceded by a photodetector in an O/E/O link, can process optical spikes at different wavelengths and thus be interconnected in large neural networks. Here, we experimentally demonstrate and numerically simulate the spiking dynamics of a laser neuron fabricated in a photonic integrated circuit. Our spiking laser neuron is shown to perform coincidence detection with nanosecond time resolution, and we observe refractory periods in the order of 0.1 ns. We propose a method to implement XOR classification using our laser neurons, and simulations of the resultant dynamics indicate robust tolerance to timing jitter.