No abstract

No abstract

A high-speed multiple input multiple output (MIMO) wireless communication system is proposed for metro transportation. In order to better mitigate the inter-antenna interference (IAI) which degrades the quality of the MIMO receiver, we propose an efficient hybrid multiuser detection (MUD) technique. This approach consists of a two-stage procedure to achieve the optimum multiuser detector (OMD) by an acceptable computational complexity. The first stage performs interference cancellation by using sorted QR decomposition (SQRD), and the second stage performs the genetic algorithm (GA). It has two significant advantages: 1) The SQRD scheme provides "good initial setting knowledge" to improve the fitness of the population for GA. 2) The effect of fitness calculation is obtained from the QR decomposition (QRD) of a MIMO channel. Simulation results demonstrate that the two-stage procedure obtains a gain of 3 dB to 25 dB than other well-known MUD schemes. The computational complexity of the two-stage procedure can be reduced by 30% with QRD than other fitness calculation scheme in GA-MUD.Index Terms -multiple input multiple output, inter-antenna interference, multiuser detection, QR decomposition, genetic algorithm, evolutionary technique, optimization.

Obtaining a well-trained model involves expensive data collection and training procedures, therefore the model is a valuable intellectual property. Recent studies revealed that adversaries can `steal' deployed models even when they have no training samples and can not get access to the model parameters or structures. Currently, there were some defense methods to alleviate this threat, mostly by increasing the cost of model stealing. In this paper, we explore the defense from another angle by verifying whether a suspicious model contains the knowledge of defender-specified external features. Specifically, we embed the external features by tempering a few training samples with style transfer. We then train a meta-classifier to determine whether a model is stolen from the victim. This approach is inspired by the understanding that the stolen models should contain the knowledge of features learned by the victim model. We examine our method on both CIFAR-10 and ImageNet datasets. Experimental results demonstrate that our method is effective in detecting different types of model stealing simultaneously, even if the stolen model is obtained via a multi-stage stealing process. The codes for reproducing main results are available at Github (https://github.com/zlh-thu/StealingVerification).

No abstract

Background Continuous positive airway pressure (CPAP) is the most effective treatment for moderate to severe obstructive sleep apnea (OSA). Acceptance of and adherence to CPAP are crucial for optimal treatment outcomes. The aim of this study was to investigate the factors influencing patients’ acceptance of and adherence to CPAP treatment. Methods One hundred eighty‐eight patients with moderate to severe OSA who had received CPAP titration from October 2017 to September 2018 were recruited. They were interviewed at 2 weeks and at 6 months to assess CPAP use and barriers to acceptance and adherence. Results One hundred fourteen patients (60.6%) accepted CPAP treatment. Disease severity, assessed by apnea‐hypopnea index (AHI) (odds ratio [OR], 1.05; 95% confidence interval [CI], 1.01‐1.08), subjective satisfaction of titration (OR, 12.83; 95% CI, 3.83‐42.99), initial intention of CPAP therapy (OR, 3.33; 95% CI, 1.05‐10.51) and short‐term home CPAP trial (OR, 9.40; 95% CI, 2.85‐31.08) were associated with acceptance of CPAP treatment. Two‐third of the 98 CPAP acceptors reported good CPAP adherence at 6 months follow‐up. Average hours of CPAP use per day for the first 2 weeks (OR, 1.88; 95% CI, 1.28‐3.04) and the global problems associated with CPAP use (OR, 0.82; 95% CI, 0.73‐0.91) were independent predictors of the six‐month CPAP adherence. Conclusions Nearly 40% of patients with moderate to severe OSA did not accept CPAP treatment, and one‐third of those CPAP acceptors had poor adherence to CPAP treatment. Improvement in disease awareness, comfortable titration experience, short‐term home CPAP trial may be of help to increase CPAP acceptance and early experience with CPAP is important for long‐term adherence. The differences in predicting factors for CPAP acceptance and adherence highlight the importance of focusing on specific aspects during the whole process management of OSA.

This paper investigates the asymmetric dual–hop multiple input multiple output (MIMO) mixed radio frequency (RF)/free space optical (FSO) decode–and–forward (DF) relaying system. This kind of system can utilize two different fading characteristic channels to reduce the possibility of the system falling into deep fading. In addition, each link of the system adopts MIMO technology to mitigate the disadvantages of fading. In this paper, the closed form expressions of the outage probability, bit error rate (BER) and average ergodic capacity are derived. The approximate expression of the system outage probability considering the pointing error is also derived. Additionally, asymptotic performance for diversity order and diversity–multiplexing tradeoff (DMT) of the system is analyzed and discussed, which provides direct theoretical basis for practical engineering design.

No abstract