Coherent optical communication technology stands as a thriving research domain in modern communication. In pursuit of higher spectral efficiency, extensive research has been conducted on high-order modulation and broadband transmission technologies. However, these advancements are highly sensitive to transceiver in-phase/quadrature (IQ) skew impairments, posing a significant task of estimation and compensation for transceiver IQ skew. In this paper, an effective approach based on a specially designed training sequence (TS) to simultaneously estimate both transmitter-side and receiver-side IQ skew (Tx/Rx-skew) for coherent orthogonal frequency division multiplexing (OFDM) systems is proposed, in which the TS is highly reusable and holds the potential for reducing the system overhead. Furthermore, the designed TS is also robust to the sampling frequency offset (SFO) and has the capacity to estimate the SFO. The effectiveness of the proposed joint transceiver IQ skew and SFO estimation scheme is verified by simulation in a dual-polarization coherent OFDM system. The results indicate that the proposed scheme can offer the joint estimation for SFO and transceiver IQ skew with the estimation errors below 2ppm and 0.2ps, for the range of [-500, 500] ppm and [-1.8T, 1.8T] accordingly.

Evaluating text-to-vision content hinges on two crucial aspects: visual quality and alignment. While significant progress has been made in developing objective models to assess these dimensions, the performance of such models heavily relies on the scale and quality of human annotations. According to Scaling Law, increasing the number of human-labeled instances follows a predictable pattern that enhances the performance of evaluation models. Therefore, we introduce a comprehensive dataset designed to Evaluate Visual quality and Alignment Level for text-to-vision content (Q-EVAL-100K), featuring the largest collection of human-labeled Mean Opinion Scores (MOS) for the mentioned two aspects. The Q-EVAL-100K dataset encompasses both text-to-image and text-to-video models, with 960K human annotations specifically focused on visual quality and alignment for 100K instances (60K images and 40K videos). Leveraging this dataset with context prompt, we propose Q-Eval-Score, a unified model capable of evaluating both visual quality and alignment with special improvements for handling long-text prompt alignment. Experimental results indicate that the proposed Q-Eval-Score achieves superior performance on both visual quality and alignment, with strong generalization capabilities across other benchmarks. These findings highlight the significant value of the Q-EVAL-100K dataset. Data and codes will be available at https://github.com/zzc-1998/Q-Eval.

Digital-to-analog converter (DAC) plays an important role in a communication system, which determines the quality of the generated signal. A high-resolution DAC is necessary to guarantee the system performance. However, the application of a high-resolution DAC results in high system cost and power consumption, which is incompatible with the future low-cost data centers. To be cost-effective and energy-efficient, a system with low-resolution DAC and negligible quantization noise induced performance degradation is highly desirable. Noise shaping (NS) technique is proposed to effectively suppress the quantization noise induced by low-resolution DAC. In this paper, we review the principle of the NS technique, and its effectiveness is investigated experimentally in 100 Gb/s intensity modulation and direct detection (IM/DD) PAM-4 system with low-resolution DAC. The experimental results show that the system with 4-bit DAC coupled with NS technique shows almost the same receiver sensitivity as the system with 8-bit DAC without NS technique over 2-km and 10-km single mode fiber (SMF) transmission in O-band. It indicates that significant performance improvement can be achieved by using NS technique in low-resolution DAC systems, and it is a promising solution for future low-cost data centers.