神经形态工程学
密码系统
计算机科学
加密
密码学
可扩展性
计算机体系结构
翻译(生物学)
材料科学
嵌入式系统
计算机硬件
电子工程
卷积神经网络
微处理器
人工智能
计算机工程
波形
人工神经网络
炸薯条
尖峰神经网络
逻辑门
非线性系统
理论计算机科学
干扰(通信)
作者
Riping Liu,Jiayao Duan,Yifei He,Yiming Wang,Juntao Tan,Jiaqi He,Zhuang Xie,Wan Yue
摘要
Cryptographic bionics, which seeks to emulate biological principles for secure information processing, remains largely unexplored at the hardware level. Here, we report an optically encrypted neuromorphic cryptosystem that utilizes an n-type organic optoelectrochemical synapse (n-OOECS), enabling brain-inspired encryption logic and adaptive information protection for end-to-end secure communication. This all-solid-state synaptic device, based on a single-component fused oligomeric organic mixed ionic-electronic conductor, exhibits multimodal reconfigurable plasticity, featuring record-high 10-bit conductance states, reprogrammable temporal pulse-waveforms, and tunable volatile-to-nonvolatile switching. Optoelectrochemical co-modulation enables bidirectional compensation and robust dark-to-light switching. Wafer-scale integration on 7-inch substrates demonstrates excellent device uniformity. By leveraging the principles of programmable-pulse engineering and nonlinear optoelectronic dynamics, the system simultaneously performs Morse pulse identification, optically encrypted transmission, and neural machine translation via convolutional neural networks. Notably, it visually demonstrates the entire process, from the input of pulse-encoded waveforms representing "SUN YAT SEN UNIVERSITY" to the final accurate Chinese translation output. This work establishes an OOECS platform that integrates neuromorphic computing with cryptographic functionality, offering a scalable materials framework for secure, adaptive, and bioinspired hardware encryption.
科研通智能强力驱动
Strongly Powered by AbleSci AI