Heterogeneous Computing Systems (HCS) have received widespread attention due to their powerful computing power, low cost, and high scalability. In HCS ranging from small-embedded devices to large data centers, energy consumption is one of the crucial design constraints. Meanwhile, the schedule length (response time) of parallel applications directly affects their Quality of Service (QoS) experience. In this study, we address the problem of minimizing the schedule length of energy-constrained parallel applications (MSLEC) on heterogeneous computing systems. Firstly, we define the concept of task energy demand rate and energy allocation factor to reasonably allocate the allocatable energy. Secondly, We propose a two-stage hybrid energy allocation (HEA) strategy and divide the allocatable energy into two parts according to the energy allocation factor, namely static pre-allocate energy (SAE) and dynamic pre-allocate energy (DAE). In the first stage, we pre-allocate SAE for each task based on the minimum energy demand and energy demand rate before task scheduling. In the second stage, we dynamically allocate DAE to each task during the operation of the scheduling algorithm. Thirdly, We conduct a rigorous mathematical proof of the feasibility of the proposed strategy. Finally, according to the proposed strategy, we design a novel HEA-based parallel application scheduling (HEA-PAS) algorithm, which aims to solve the MSLEC problem. Experiments on realworld and randomly generated parallel applications show that the proposed HEA-PAS algorithm outperforms the state-of-the-art methods in terms of effectiveness.

The ambient population has been regarded as an important indicator for analyzing or predicting thefts. However, the literature has taken it as a homogenous group and seldom explored the varied impacts of different kinds of ambient populations on thefts. To fill this gap, supported by mobile phone trajectory data, this research investigated the relationship between ambient populations of different social groups and theft in a major city in China. With the control variables of motivated offenders and guardianship, spatial-lag negative binominal models were built to explore the effects of the ambient populations of different social groups on the distribution of theft. The results found that the influences of ambient populations of different social groups on the spatial distribution of theft are different. Accounting for the difference in the “risk–benefit” characteristics among different activity groups to the offenders, individuals from the migrant population are the most likely to be potential victims, followed by suburban and middle-income groups, while college, affluent, and affordable housing populations are the least likely. The local elderly population had no significant impact. This research has further enriched the studies of time geography and deepened routine activity theory. It suggests that the focus of crime prevention and control strategies developed by police departments should shift from the residential space to the activity space.

Crimes of various types are occurring in different areas of each country, almost every day. Hence, observing, predicting and preventing crimes is a crucial issue to ensure a peaceful and safe environment. Although several systems have been designed for analyzing crime related data, to our best knowledge, there is no system designed for travelers. This paper addresses this issue by proposing a decision support system named CPD-DSST (Crime Pattern Discovery Decision Support System for Travelers) that allows users to learn about crime occurrences in specific areas and provide suggestions to ensure travel safety. To discover and locate crimes, the system applies an efficient algorithm named Crime Classifying Discovery and Location (CCDL) based on multinomial logistic regression, and a Crime Rate Evaluation (CRE) algorithm, on spatio-temporal crime data. Experiments show that the proposed system can perform accurate predictions. Moreover, preliminary feedback indicates that the system is appreciated by users.

Nowadays, there are more and more criminal behaviors experiencing around the world, and crime spiking has become one of the most critical security and social issues in almost every country. It is critical to seek effective ways to discover these criminal behaviors and patterns and to carry out the prevention for the target place. In this paper, we formulate the problem of criminal behaviors and propose a Criminal Activity Clustering (CAC) algorithm. We introduce the fuzzy clustering method to detect potential criminal patterns in large-scale spatiotemporal datasets. In addition, for the improvement of the the proposed CAC algorithm performance, we implement a parallel solution for the algorithm in the Apache Spark cloud computing platform. The results of the experiment show that the proposed CAC algorithm can effectively detect accurate criminal patterns from large-scale spatiotemporal data.

As one of the great advances in modern technology, the microarray is widely used in many fields, including biomedical research, clinical diagnosis, and so on. Evidently, in order to extract the intensity of fluorescence bio-probes accurately, we need to pay special attention to the gridding of microarray at first. To solve the poor effect of the traditional Otsu method for microarray gridding, an innovative algorithm of Otsu optimized by multilevel thresholds is proposed to improve the accuracy and effectiveness of the microarray image gridding and segmentation. The experimental results indicate that considering the physical information carried by microarrays, the improved algorithm of Otsu optimized by multilevel thresholds achieves high-quality gridding and establishes the bio-spot coordinates more precisely. Compared with the traditional Otsu method, its gridding error is reduced to zero, and the integrated relative error of bio-spot coordinates is decreased from 2.89% to 1.05%. This optimization of Otsu combined with physical information of spot-matrix will greatly improve the performance of segmentation so as to make the contribution to extracting the fluorescence intensity of microarray accurately.

Networks interact with and depend on each other in modern society, operating and functioning as interdependent networks. In this paper, the cascading process of the interdependent networks against the attack on edges is studied. Then the robustness of the heterogeneous coupled networks and homogeneous coupled networks under random attack and targeted attack is fully explored. It is found that for heterogeneous and homogeneous coupled networks, no particular type always performs better than any other. However, in general, homogeneous coupled networks exhibit better robustness than the heterogeneous ones in most cases. It is also found that under targeted attack, no matter what the type of the interdependent network, it gets more vulnerable as the value of parameter alpha increases, while the similar feature cannot be observed in the case of random attack.

Understanding how collective actions evolve into violence is critical for guiding public safety decisions. While real world observations of collective violence are difficult and rare, simulation models can help us to explore its evolution process. We propose an Agent-Based Emotion Contagion (ABEC) model that simulates the spread of group violence when the mechanism of contagious grievance is at work. The model is motivated by related social psychological theories of group behavior and is implemented by incorporating the epidemiological emotion contagion mechanism with crowd's game-theoretic behaviors with an agent-based approach. Our simulation model generates some crowd patterns, including local outbursts of collective violence with grievance contagion, dynamic spatial clustering of violent civilians and the nonlinear evolution of collective violence. We also explore the variations of violence evolution by varying some parameters of our model. Results show that the high density crowds and the widespread of grievance promote violence outburst. These results suggest opportunities to curb collective violence through dispersing crowds and allaying the grievance.

The researches on robustness of interconnected networks have received more and more attention. However, previous studies of robustness are based on two extreme attack strategies, i.e. random attack and targeted attack. In the real world, the attack information is usually ambiguous. In this paper, the gray-information-based attack model is adopted to evaluate the robustness of the interdependent networks against the attack on edges. Through extensive simulation, we find that the robustness of the interdependent networks under gray-information-based attack gets stronger with the increase of redundancy parameter β. We also get the conclusion that the robustness of the interdependent network in the scenarios of gray-information-based attack is stronger than that of targeted attack and weaker than that of random attack. Finally, by comparison of three types of coupled networks, NW-NW and WS-WS coupled network are found to be the most robust type under gray-information attack.

It is crucial to provide compatible treatment schemes for a disease according to various symptoms at different stages. However, most classification methods might be ineffective in accurately classifying a disease that holds the characteristics of multiple treatment stages, various symptoms, and multi-pathogenesis. Moreover, there are limited exchanges and cooperative actions in disease diagnoses and treatments between different departments and hospitals. Thus, when new diseases occur with atypical symptoms, inexperienced doctors might have difficulty in identifying them promptly and accurately. Therefore, to maximize the utilization of the advanced medical technology of developed hospitals and the rich medical knowledge of experienced doctors, a Disease Diagnosis and Treatment Recommendation System (DDTRS) is proposed in this paper. First, to effectively identify disease symptoms more accurately, a Density-Peaked Clustering Analysis (DPCA) algorithm is introduced for disease-symptom clustering. In addition, association analyses on Disease-Diagnosis (D-D) rules and Disease-Treatment (D-T) rules are conducted by the Apriori algorithm separately. The appropriate diagnosis and treatment schemes are recommended for patients and inexperienced doctors, even if they are in a limited therapeutic environment. Moreover, to reach the goals of high performance and low latency response, we implement a parallel solution for DDTRS using the Apache Spark cloud platform. Extensive experimental results demonstrate that the proposed DDTRS realizes disease-symptom clustering effectively and derives disease treatment recommendations intelligently and accurately.

With the emergence of the big data age, the issue of how to obtain valuable knowledge from a dataset efficiently and accurately has attracted increasingly attention from both academia and industry. This paper presents a Parallel Random Forest (PRF) algorithm for big data on the Apache Spark platform. The PRF algorithm is optimized based on a hybrid approach combining data-parallel and task-parallel optimization. From the perspective of data-parallel optimization, a vertical data-partitioning method is performed to reduce the data communication cost effectively, and a data-multiplexing method is performed is performed to allow the training dataset to be reused and diminish the volume of data. From the perspective of task-parallel optimization, a dual parallel approach is carried out in the training process of RF, and a task Directed Acyclic Graph (DAG) is created according to the parallel training process of PRF and the dependence of the Resilient Distributed Datasets (RDD) objects. Then, different task schedulers are invoked for the tasks in the DAG. Moreover, to improve the algorithm's accuracy for large, high-dimensional, and noisy data, we perform a dimension-reduction approach in the training process and a weighted voting approach in the prediction process prior to parallelization. Extensive experimental results indicate the superiority and notable advantages of the PRF algorithm over the relevant algorithms implemented by Spark MLlib and other studies in terms of the classification accuracy, performance, and scalability. With the expansion of the scale of the random forest model and the Spark cluster, the advantage of the PRF algorithm is more obvious.

To distinguish lateral modes of broad-stripe diode laser array (LDA) is difficult, because the structure of lateral modes is complex. By using multiple beams coherent combined beam theory, the expression of far field pat-tern is got. Based on the expression, the lateral mode of emitter decides the envelope line of far field pattern. From the experimental records, with an external cavity, broad-stripe semiconductor diode laser array had been phase locked. The phase-locking between one order lateral modes (fundamental lateral mode was most easily phase locked) of emitters of the LDA is realized, the other order lateral modes was inhibited. A comparison between theory model and experimentation was also made, satisfying result had been obtained. From these results, it can be concluded that using lateral mode of LDA, the quality of LDA beams can be improved.

Under small signal situation, the analytical expression for describing the output signal emitted from an Yb3+-doped fiber amplifier with a Gaussian input pulse signal is deduced, by full considering both fiber guided modes’ propagation constant and filling factor, which depend on the signal frequency. Pumped by continue wave (CW) source, the analytical expressions of the small signal gain and each guided mode's output power are obtained. Based on these analytical expressions, the various intensity distributions of output pulses amplified by the fiber amplifier with different Gaussian input pulse widths are analyzed.

The nonlinear effect and material damage existing in high power fiber lasers restrict their output power.A large mode area fiber,which could reduce the power density and improve the nonlinear threshold,is one of the solutions to the prob-lem.A photonic crystal fiber with air holes at wavelength scales was designed based on the effective index model and finite ele-ment analysis.The single-mode properties as well as the influence of structural parameters on the mode area and dispersion were investigated.Considering practical applications,we designed a large mode area photonic crystal fiber,operating in the range of 0.40~1.55μm,mode area between 112.74~258.87μm2 ,centering at 1.27μm with dispersion compensation ability.Our in-vestigations may provide new references to the optimization and fabrication of large mode area fiber for high power fiber lasers.

A kind of Bi2O3-B2O3-ZnO Pb-free glass frits for Al paste was prepared,using Bi2O3,B2O3 and ZnO as main starting material.The as-prepared glass powder was characterized by X-ray diffraction(XRD),differential thermal analysis(DTA)and scanning electron microscope(SEM).Results showed that the glass powder was amorphous in majority,the softening point was about 500℃ and the mean grain size was less than 5μm.A series of solar cells were manufactured using Al paste by varying the content of glass powder in the Al paste,and their performance were tested.The results showed that when the content of glass powder in Al paste was 1%,the aluminum coating resistivity was lowest,and the boiling water resistance was improved with the decrease of the content of glass powder in Al paste.

Graphene is a new type of two-dimensional carbon nanomaterial,which has been discovered and synthesized in recent years.Graphene has great potential in terms of improving the thermal,mechanical and electrical properties of its composites,which is also a new hot research area of nanocomposites,due to its unique structure and novel physical and chemical properties.In this article,advances in preparation and application of graphene nanocomposites were reviewed and future development of graphene nanocomposites was also proposed.

In this paper, a bidirectional chaos secret communication system, based on mutually coupled semiconductor lasers (MCSLs) with asymmetrical bias currents, is proposed, and the synchronization characteristics and the communication performances of such a system are numerically investigated. The results show that the stable leader-laggard chaos synchronization can be achieved under relatively large asymmetrical bias current levels. Meantime, the influence of the intrinsic parameter variations of the laser on the synchronization quality is also considered, and the simulation reveals that this system still possesses good robustness to the parameter variations. Moreover, the influences of delay time and mutually coupling strength between the two lasers on chaos communication performance have also been discussed. Finally, unidirectional and bidirectional secret communication performances of such a system are examined under the chaos masking (CMS) encryption scheme, and the security of this system is also discussed.

针对脉冲周期在ms量级、脉宽为ns量级的高功率激光系统输出的短脉冲串,根据其周期性特征及放大过程的特点,在忽略损耗系数和自发辐射的情况下,将1个周期近似分为泵浦和放大2个阶段.运用端面泵浦瞬态速率方程组,导出了2阶段变换时刻所对应的上能级粒子数之间的半解析关系式.据此,在模拟光纤长度对增益影响的基础上,研究了2阶段腔内平均上能级粒子数密度随时间的周期性变化关系,定量分析了脉冲前沿消耗增益对脉冲后沿形状的影响.结果表明:光脉冲前沿增益可达29.22 dB,其后沿增益低至0.82 dB,光脉冲后沿波形必然存在畸变,验证了所得结论的合理性.

A scanning spectral filter method to improve the signal-noise-ratio in the femtosecond chains is proposed by using the characteristic that the instantaneous frequency varies with the time approximately linearly for the chirped pulse in the time-frequency domain. The scanning spectral filtering for reduing the amplification of spontaneous emission (ASE) intensity is analyzed in the time-frequency domain by using the Short-Time Fourier Transform method. The results show that the pulse contrast can be improved by two orders, and the transmission efficiency of the chirped pulse can exceed 90% when the synchronizing time jitter ranges from -2ps to 2ps and the chirp rate p from 0.9C/T-2 to 1.10C/T-2. Adopting the cascaded scanning filter to improve the pulse contrast is investigated too, which can improve the pulse contrast effectively. The great advantage of this novel nonlinear spectral filter technology is high energy and high peak intensity femtosecond chains for it filters out the ASE in the near field in temporal domain.

Fluorinated multi-walled carbon nanotubes(F-MWCNTs) were prepared by using gaseous products from the thermal decomposition of polytetrafluoroethylene(PTFE) as fluorination agent for multi-walled carbon nanotubes(MWCNTs).The specimens were analyzed by using FTIR,SEM,XRD and contact angle detection.The results showed that C-F bonds were created on the surface of MWCNTs,the structure was intact after fluorination and the hydrophobicity of the specimen was enhanced.Compared with the traditional method by using gaseous fluorine as fluorination agent,the means used in this experiment was easier,safer and cheaper,which could be a promising method in industrial production.