Analog-to-spike encoding and time-efficient RF signal processing with photonic neurons

High-speed parallel processing with photonic feedforward reservoir computing

High-speed photonic reservoir computing (RC) has garnered significant interest in neuromorphic computing. However, existing reservoir layer (RL) architectures mostly rely on time-delayed feedback loops and use analog-to-digital converters for offline digital processing in the implementation of the readout layer, posing inherent limitations on their speed and capabilities. In this paper, we propose a non-feedback method that utilizes the pulse broadening effect induced by optical dispersion to implement a RL. By combining the multiplication of the modulator with the summation of the pulse temporal integration of the distributed feedback-laser diode, we successfully achieve the linear regression operation of the optoelectronic analog readout layer. Our proposed fully-analog feed-forward photonic RC (FF-PhRC) system is experimentally demonstrated to be effective in chaotic signal prediction, spoken digit recognition, and MNIST classification. Additionally, using wavelength-division multiplexing, our system manages to complete parallel tasks and improve processing capability up to 10 GHz per wavelength. The present work highlights the potential of FF-PhRC as a high-performance, high-speed computing tool for real-time neuromorphic computing.

Real-time signal acquisition based on optical excitable response by pre-designing single-tone preamble waveform

Real-time acquisition of target signals is preferred for mobile communication systems. However, under the requirement of ultra-low latency for next-generation communication, traditional acquisition methods need to temporally locate the target signal from a large amount of raw data with correlation-based computing, introducing extra latency. We propose a real-time signal acquisition method based on an optical excitable response (OER) by pre-designing a single-tone preamble waveform. The preamble waveform is designed to be within the amplitude and bandwidth of the target signal, so no extra transceiver is required. The OER generates a corresponding pulse to the preamble waveform in the analog domain, which simultaneously triggers an analog-to-digital converter (ADC) to acquire target signals. The dependence of OER pulse on the preamble waveform parameter is studied, leading to a pre-design of the preamble waveform for an optimal OER. In the experiment, we demonstrate a millimeter-wave (26.5-GHz) transceiver system with target signals of orthogonal frequency division multiplexing (OFDM) format. Experimental results show that the response time is less than 4 ns, which is far lower than the ms-level response time of traditional all-digital time-synchronous acquisition methods.

Stochastic photonic spiking neuron for Bayesian inference with unsupervised learning

Stochasticity is an inherent feature of biological neural activities. We propose a noise-injection scheme to implement a GHz-rate stochastic photonic spiking neuron (S-PSN). The firing-probability encoding is experimentally demonstrated and exploited for Bayesian inference with unsupervised learning. In a breast diagnosis task, the stochastic photonic spiking neural network (S-PSNN) can not only achieve a classification accuracy of 96.6%, but can also evaluate the diagnosis uncertainty with prediction entropies. As a result, the misdiagnosis rate is reduced by 80% compared to that of a conventional deterministic photonic spiking neural network (D-PSNN) for the same task. The GHz-rate S-PSN endows the neuromorphic photonics with high-speed Bayesian inference for reliable information processing in error-critical scenarios.