Search

Security of shipborne automatic identification system (AIS) sensor equipment in maritime traffic is not only essential for navigational safety but also crucial for global economic and social development. Sensor individual recognition (SIR) is a popular method for maritime sensor authentication, which involves extracting radio frequency fingerprint (RFF) information that is difficult to tamper with to identify sensor equipment, providing an additional layer of security for maritime sensor networks. Existing deep learning (DL) SIR methods often rely on a large amount of high-quality labeled data to guide the model in learning RFF features, thus achieving high-performance models. However, under noncooperative communication conditions, it is difficult to obtain a large amount of high-quality labeled sensor signal data, and manually annotating sensor signals is extremely time consuming. Therefore, for the realistic scenario with limited labeled data, this article introduces semi-supervised learning (Semi-SL) theory, combines encoder with perturbation consistency regularization semi-supervised methods, uses unlabeled data to reduce the model’s dependence on labeled data, and employs virtual adversarial training (VAT) to approximate the real label distribution of unlabeled data, thereby improving the model’s recognition performance in scenarios with few labeled samples. Experimental results show that the proposed method achieves an identification accuracy of over 85% on a dataset containing signals from 50 classes of shipboard AIS sensors, even with only ten labeled samples per class. This performance surpasses that of pseudolabeling and the teacher-student structured Semi-SL methods.

Graphical Abstract:

ABSTRACTSeasonal influenza remains a significant public health concern globally, with annual vaccinations as the most effectively preventive measure. This study examines influenza vaccination coverage rates across different age groups in Pudong New Area, Shanghai, China, from 2013 to 2023. This study extracted influenza vaccination data from the Shanghai Immunization Planning Information System (SIPIS) of the Center for Disease Control and Prevention Shanghai’s Pudong New Area from 2013 to 2023. The analysis utilized weighted linear regression to compare vaccination rates over the study period. From 2013 to 2023, a cumulative total of 1,421,295 influenza vaccinations were administered in Pudong New Area, with the quadrivalent inactivated influenza vaccines (IIV4) and trivalent inactivated influenza vaccines (IIV3) comprising 56.8% and 42.9%, respectively. The overall average influenza vaccine coverage rate was 2.27% (95% CI: 2.26, 2.28). The 0–4 years group exhibited the highest average annual coverage rate of 22.52% (95% CI: 22.47, 22.57) among different age groups, in stark contrast to that of the 20–24 years age group, which had the lowest at 0.32% (95% CI: 0.31, 0.33). In terms of repeat vaccinations, a significant majority (86.87%) of recipients received only 1–2 doses, while just 13.13% received 3 or more doses. Although influenza vaccination coverage among preschool children in Pudong New Area is relatively high, it falls significantly short of WHO recommendations. Enhance the level of awareness of influenza vaccine among adults and provide a free influenza vaccination strategy for specific groups such as doctors, which is helpful to increase influenza vaccination rates among populations.

ABSTRACTChildren with hematological malignancies are at increased risk of hepatitis B virus infection. This study assessed the immunogenicity and safety profile of HBV vaccination in pediatric hemato-oncological children. A nonrandomized interventional study was conducted from January 2017 to February 2020 in Shanghai, China. Seventy-three pediatric hemato-oncological children with hepatitis B surface antibody (anti-HBs) titers <10 mIU/ml were recruited. The participants received three doses of recombinant HBV vaccine according to the 0-, 1-, and 6- month immunization schedule. Adverse events following immunization and anti-HBs titers (at baseline, 1 month, and 6 months after inoculation) were recorded. Forty-three males and thirty females with median ages of 9.12 and 9.60 years, respectively, were included. The mean anti-HBs titer was 4.88 ± 2.61 mIU/ml, 893.12 ± 274.12 mIU/ml, and 711.45 ± 337.88 mIU/ml at baseline, one month, and six months after inoculation, respectively (P< .001). A total of fourteen adverse events following immunization were reported, and among them, 5 (6.85%), 5 (6.85%), and 4 (5.48%) events were reported after the first, second, and third inoculation, respectively (P= .927). In conclusions, the HBV vaccine is immunogenic and safe in children with hematological malignancies. It is worth noting that the anti-HBs titer was decreased at the 6-month follow-up, and periodic monitoring of the anti-HBs titer accompanied by timely booster vaccination should be carefully considered.Abbreviations:AEFI: Adverse events following immunization; HBV: Hepatitis B virus; Anti-HBs: Antibody against hepatitis B surface antigen; HBsAg: Hepatitis B surface antigen; APC: Antigen-presenting cell; HSCT: Hemopoietic stem cell transplantation; COVID-19: Corona Virus Disease 2019

Damping coatings with low temperature resistance on high-speed rail trains had been designed through the establishment of a polymer cross-linked network and the introduction of dimethylolpropionic acid and sodium 2-[(2-aminoethyl) amino] ethanesulfonate as a hydrophilic chain extender. The phase separation behavior and hydrogen-bonded interactions in composite materials effectively enhanced the hydrophobicity, film-forming properties, and damping performance. The phase separation behavior plays a dominant role. The damping peaks of the samples are saddle-shaped, and the peaks of the saddle-shaped saddle bottom also meet the requirements of vibration and noise reduction engineering technology. The effective damping temperature range (tan δ > 0.3) was broadened to 114.7 °C. The prepared waterborne polyurethane–acrylate composite emulsions can be applied to the damping of complex structural surfaces and are environmentally friendly. The strategy of regulating the phase separation behavior of materials to enhance the damping performance of materials also offers a new method for the synthesis of high-performance damping coatings.

The United Nations Framework Convention on Climate Change (UNFCCC) acknowledges that global cooperation is paramount to mitigate climate change and further warming. The global community is committed to renewable energy and natural materials to tackle this challenge for all humankind. The widespread use of natural materials is embraced as one such action to reach net-zero carbon emissions. Given the hierarchical framework and earth abundance, cellulose-based materials extend their negative carbon benefits to our daily products and accelerate our pace toward carbon neutrality. Here, we present an overview of recent developments of cellulose-based materials in upsurging applications in radiative cooling, thermal insulation, nanofluidics, and wearable devices. We also highlight various modifications and functionalized processes that transform massive amounts of cellulose into green products. The prosperous development of functionalized cellulose materials aligns with a circular economy. Expedited interdisciplinary fundamental investigations are expected to make fibrillated cellulose penetrate more into carbon downdraw at speed and scale.

ABSTRACTEmergency vaccination (EV) is used as effective postexposure prophylaxis (PEP) to control varicella outbreaks within 3–5 days. However, the advantages of a second dose of varicella vaccine (VarV) in students who had received one dose before an outbreak and the potential benefits of EV at more than 5 days after exposure have not been fully evaluated. This study evaluated the vaccine effectiveness (VE) of EV in preventing disease development during a varicella outbreak in Shanghai, China, in 2020. Questionnaires were used to obtain student demographic information, clinical manifestations, varicella history, vaccination status, and willingness to receive EV. The VE of EV was calculated as [1-relative risk (RR)] ×100%. Among the 1455 students included in this study, 31 cases were identified, resulting in an overall attack rate of 2.13%. There were 6 cases in unvaccinated students and 25 cases in one-dose-vaccinated students. A total of 788 students received one EV dose. The attack rates were 6.38% (6/94), 4.26% (19/446), 2.82% (2/71), and 0.56% (4/717) among unvaccinated students, students who received 1 dose of VarV, and students who received EV with the 1stand 2nddose of VarV, respectively. Compared to that in unvaccinated students, the VE of EV with the 2nddose of VarV was 88% (95% CI 49% to 97%). EV should be performed as soon as possible after exposure. Nevertheless, vaccination is still recommended at more than 5 days post exposure to control varicella outbreaks.

This paper focuses on the first-order and second-order performance of dual-hop underlay cognitive radio systems with opportunistic relaying (UCR-OR) over independent and non-identically distributed (i.ni.d) Rayleigh fading channels. For the UCR-OR systems, the tolerable maximum interference power (TMIP) Q at primary users (PUs) and the allowable maximum transmission power limit (AMTP) Pmax at secondary users (SUs) are considered, simultaneously. We first obtain the closed-form solutions to the first-order performance such as outage probability, average symbol error ratio (SER), and ergodic capacity (EC). Secondly, we investigate the second-order statistical performance, i.e., average outage rate (AOR) and average outage duration (AOD). With the consideration that in practice implementation the receiver performance is primarily influenced by the signal-to-noise ratio (SNR) (not the signal envelope), the second-order statistical performance is investigated based on the equivalent instantaneous end-to-end SNR. Finally, we present the detailed performance comparison analysis of UCR-OR systems by defining a random variable μ=Pmax/Q. The results show that the effect of μ on the first-order and second-order performances is different greatly. For the first-order performance, the performance gap is negligible when the value of μ is relatively large. However, for the second-order one, the gap is distinct.

To improve the prediction accuracy of chaotic time series, deep extraction of the system evolutionary patterns is a key problem in modeling. In this paper, we propose a deep learning model of automatic multi-scale feature extraction for chaotic time series prediction. A hybrid deep neural network named deep temporal-inception module and gated recurrent unit network (DTIGNet) is designed. The improved temporal-inception module stacks dilated causal convolution of different depth to increase the network adaptability to different scales and improve the network nonlinear characterization ability, and an optional 1 × 1 convolutional kernel as shared residual connection. The model is applied to the Mackey-Glass system, Rossler system, Lorenz system and sunspots time series to verify the applicability and effectiveness in chaotic time series prediction. The results show that the DTIGNet proposed has higher accuracy and better performance compared with other methods according to the 6 prediction evaluation metrics adopted. • We propose a deep learning model (DTIGNet) of automatic multi-scale feature extraction for chaotic time series prediction. • The improved temporal-inception module proposed extracts data features at different scales. • The results show that the proposed model has lower prediction errors and superior performance than existing models. [ABSTRACT FROM AUTHOR]