Search

Intensity-modulated direct-detection (IM/DD) optical systems are most widely employed in short-reach optical interconnects due to their simple structure and cost-effectiveness. However, IM/DD systems face mixed linear and nonlinear channel impairments, mainly induced by the combination of square-law detection and chromatic dispersion, as well as the utilization of low-cost non-ideal transceivers. To solve this issue, recent years have witnessed a growing trend of introducing machine learning technologies such as neural networks (NNs) into IM/DD systems for channel equalization. NNs usually present better system performance than traditional approaches, and various types of NNs have been investigated. Despite the excellent system performance, the associated high computational complexity is a major drawback that hinders the practical application of NN-based equalizers. This paper focuses on the performance and complexity trade-off of NNs employed in IM/DD systems, presenting a systematic review of the current status of NN-based equalizers as well as a number of effective complexity reduction approaches. The future trends of leveraging advanced NN in IM/DD links are also discussed. [ABSTRACT FROM AUTHOR]

Purpose: The siphon effect in the health service market is notably pronounced in many countries. How to measure and identify the determinants contributing to the siphon effect presents a substantial challenge. This study aimed to analyse the effect of two different social medical insurances, the Basic Medical Insurance System for Urban Employees (BMISUE), and the Basic Medical Insurance System for Urban and Rural Residents (BMISURR), on the siphon effect in the health services market. Methods: The data used in this study were from the 2021 Health Life Satisfaction Survey of Yangtze River Delta (HLSSYRD) conducted by Shanghai Jiao Tong University. The logistic model was used to evaluate the association between social medical insurances and individual choices of medical institutions, and the Propensity Score Matching method (PSM) was used to check the robustness of basic results. Results: Residents covered by BMISUE were more likely to choose a general hospital when they first sought medical treatment (OR = 5.377, 95% CI: 4.887, 5.915) relative to those insured by BMISURR. Further analysis showed that BMISUE would accelerate the siphon effect of general hospitals, people insured by BMISUE were still more likely to choose general hospitals despite being close to primary hospitals compared to those insured by BMISURR (OR = 3.240, 95% CI: 2.945, 3.565). Heterogeneity analysis indicated BMISUE had a greater impact on residents aged 15– 59 years and those with high income compared to older people and individuals with low income. Conclusion: Different social medical insurances can substantially affect residents' first choice of medical institutions. BMISUE with higher benefits level could exacerbate the siphon effect in the health service market. More equitable medical security system should be strengthened to bridge the benefits gap between BMISUE and BMISURR. [ABSTRACT FROM AUTHOR]

The GaoFen6 (GF-6) satellite, equipped with a wide full-swath (WFV) sensor, offers high spatial resolution and extensive coverage, making it widely utilized in agricultural and forestry classification, land resource monitoring, and other fields. Accurate on-orbit radiometric calibration of GF-6/WFV is crucial for these quantitative applications. Currently, the absolute radiometric calibration of GF-6/WFV relies primarily on vicarious calibration conducted by the China Center for Resources Satellite Data and Application (CRESDA). However, annual vicarious calibration may not adequately capture the radiometric performance of GF-6/WFV due to performance degradation. Therefore, increasing the frequency of on-orbit radiometric calibration throughout the lifetime of GF-6/WFV is essential. This study proposes a method for conducting long-term cross-radiometric calibrations of GF-6/WFV by taking the multispectral imager (MSI) onboard the Sentinel-2 satellite as a reliable reference sensor and the sites from RadCalNet as reference ground targets. Firstly, we conducted 62 on-orbit cross-radiometric calibrations of GF-6/WFV since its launch by tracking with the Sentinel-2/MSI sensor after correcting the discrepancy spectrum and solar zenith angle. Then, validation of cross-radiometric calibration results against RadCalNet products indicated an average absolute relative error between 3.55% and 4.64%. Cross-validation with additional reference sensors, including Landsat-8/OLI and MODIS, confirmed the reliability of calibration, demonstrating relative differences from GF-6/WFV of less than 5%. Furthermore, the overall uncertainty of the cross-radiometric calibration was estimated to be from 4.08% to 4.89%. Finally, trend analysis of the time-series radiometric performance was also conducted and revealed an annual degradation rate ranging from 0.57% to 2.31%. This degradation affects surface reflectance retrieval, introducing a bias of approximately 0.0073 to 0.0084. Our findings highlight the operational effectiveness of the proposed method in achieving long-time-series on-orbit radiometric calibration and degradation monitoring of GF-6/WFV. The study also demonstrates that the radiometric performance of GF-6/WFV is relatively stable and suitable for further quantitative applications, especially for long-term monitoring applications. [ABSTRACT FROM AUTHOR]