In this paper, an artificial-intelligence-based fiber communication receiver model is put forward. With the multi-head attention mechanism it contains, this model can extract crucial patterns and map the transmitted signals into the bit stream. Once appropriately trained, it can obtain the ability to restore the information from the signals whose transmission distances range from 0 to 100 km, signal-to-noise ratios range from 0 to 20 dB, modulation formats range from OOK to PAM4, and symbol rates range from 10 to 40 GBaud. The validity of the model is numerically demonstrated via MATLAB and Pytorch scenarios and compared with traditional communication receivers.

集成、宽带、大色散延时的器件在微波光子滤波、真延时相控阵天线等领域有着重要的应用，可以有效地降低系统尺寸和功耗。文中提出并实现了一种基于硅基光子集成的宽带大色散延时芯片，通过采用超低损耗波导结构和侧壁法向量调制结构实现了片上集成大色散波导光栅，色散值超过250 ps/nm， 最大群延时达到2440 ps，带宽大于9.4 nm，该芯片有望用于微波光子学、高速光纤通信系统等领域。

Compressive sampling (CS) has attracted considerable attention in microwave and radio frequency (RF) fields in recent years. It enables the acquisition of high-frequency signals at a rate much smaller than their Nyquist rates. Combined with photonics technology, traditional CS systems can significantly enlarge their operating bandwidth, which offers great potential for spectrum sensing in cognitive radios. In this Letter, we review our recent work on photonic CS systems for wideband spectrum sensing. First, a proof-of-concept photonics-assisted CS system is demonstrated; it is capable of acquiring numerous radar pulses in an instantaneous bandwidth spanning from 500 MHz to 5 GHz with a 500-MHz analog-to-digital converter (ADC). To further reduce the acquisition bandwidth, multi-channel photonics-assisted CS systems are proposed for the first time, enabling the acquisition of multi-tone signals with frequencies up to 5 GHz by using 120-MHz ADCs. In addition, the system bandwidth is increased from 5 to 20 GHz by employing time-interleaved optical sampling.