Search

The prevalence of smartphones has promoted the popularity of mobile apps in recent years. In this paper, we study how mobile app usage evolves over a long-term period. We first introduce an app usage collection platform named carat, from which we have gathered app usage records of 1,465 users from 2012 to 2017. We then conduct the first study on the long-term evolution processes on a macro-level, i.e., app-category, and micro-level, i.e., individual app. We discover that, on both levels, there is a growth stage enabled by the introduction of new technologies. Then there is a plateau stage caused by high correlations between app categories and a Pareto effect in individual app usage, respectively. Additionally, the evolution of individual app usage undergoes an elimination stage due to fierce intra-category competition. The inter-diversity of app-category and individual app usage exhibits opposing trends: app-category usage assimilates while individual app usage diversifies. Nevertheless, the intra-diversity of both app-category and app usage declines over time. Also, we demonstrate the country barriers of app category usage. We further investigate how different demographics affect the evolutionary processes of app usage. Our study provides useful implications for app developers, market intermediaries, and service providers

Urban anomalies may result in loss of life or property if not handled properly. Automatically alerting anomalies in their early stage or even predicting anomalies before happening are of great value for populations. Recently, data-driven urban anomaly analysis frameworks have been forming, which utilize urban big data and machine learning algorithms to detect and predict urban anomalies automatically. In this survey, we make a comprehensive review of the state-of-the-art research on urban anomaly analytics. We first give an overview of four main types of urban anomalies, traffic anomaly, unexpected crowds, environment anomaly, and individual anomaly. Next, we summarize various types of urban datasets obtained from diverse devices, i.e., trajectory, trip records, CDRs, urban sensors, event records, environment data, social media and surveillance cameras. Subsequently, a comprehensive survey of issues on detecting and predicting techniques for urban anomalies is presented. Finally, research challenges and open problems as discussed., Comment: Accepted by IEEE Transactions on Big Data

With the prevalence of smartphones, people have left abundant behavior records in cyberspace. Discovering and understanding individuals' cyber activities can provide useful implications for policymakers, service providers, and app developers. In this paper, we propose a framework to discover daily cyber activity patterns across people's mobile app usage. We first segment app usage traces into small time windows and then design a probabilistic topic model to infer users' cyber activities of each window. By exploring the coherence of users' activity sequences, the daily patterns of individuals are identified. Next, we recognize the common patterns across diverse groups of individuals using a hierarchical clustering algorithm. We then apply our framework on a large-scale and real-world dataset, consisting of 653,092 users with 971,818,946 usage records of 2,000 popular mobile apps. Our analysis shows that people usually obey yesterday's activity patterns, but the patterns tend to deviate as the time-lapse increases. We also discover five common daily cyber activity patterns, including afternoon reading, nightly entertainment, pervasive socializing, commuting, and nightly socializing. Our findings have profound implications on identifying the demographics of users and their lifestyles, habits, service requirements, and further detecting other disrupting trends such as working overtime and addiction to the game and social media.

To introduce blockchain technology to the mobile internet and Internet of Things, offloading proof-of-work tasks to edge computing nodes is considered an effective solution. Among these solutions, pricing edge computing resources through auctions has been widely accepted as a simple and efficient method. However, in these auctions, it is difficult to identify bid-rigging behavior by users, and the lack of technical means to avoid such behavior can results in a loss of revenue for edge service providers (ESPs). In this study, we first model the optimal bid-rigging mechanism in a mobile blockchain edge computing resource auction. Then, we introduce an auction method that an ESP can use to prevent the bid-rigging strategy in resource auctions. Finally, we propose a solution for the optimal reserve price of the SA (simulated annealing) algorithm, which calculates the optimal reserve price according to bidders’ willingness to pay and the number of bid-rigging participants to improve the revenue of the ESP. Through experiments, it is verified that the proposed method meets the requirements of IR (individual rationality) and IC (incentive compatibility) in resource allocation, and effectively impacts of the number of bidders who adopt a bid-rigging strategy in auctions on ESP revenue. Furthermore, the efficiency and effectiveness of the algorithm are analyzed using data. When the cooling coefficient is less than 0.95, the time complexity of the algorithm is O ( l o g N ) . Using this algorithm to conduct auctions can increase revenue by approximately 2%, requiring the ESP to pay only a small computational cost.

Unmanned aerial vehicle (UAV) multi-hop communication networks are foreseen to be widely employed in both military and civilian scenarios. However, in ultra-dense scenarios with swarm UAVs, nodes are highly dynamic mobile, ultra-dense deployment and non-centralized distribution. These characteristics make the centralized resource management policy not apply. Meanwhile, existing routing protocols can't meet the performance challenges of high dynamic, topology and link frequency changes of ultra-dense scenarios with swarm UAVs. To solve the above challenges of resource management and routing protocol, a cross-layer optimization method is presented with a novel mean field game (MFG) in this paper. It is based on the cross-layer design method of the MFG theory and jointly considers the power resources in the physical layer, frequency resources in the medium access control (MAC) layer, and routing resources in the network layer. By dividing into subproblems, the original problem is solved. Meanwhile, the optimal data transmission path can be selected through the management and allocation of frequency resources and power resources. A crosslayer resource management dynamic source routing (CLRMDSR) protocol is designed based on that which adds link quality measurement. The simulation results show that the presented CLRM-DSR with the proposed resource management scheme can improve the data packet transmission rate, reduce end-toend delay, and lower routing overhead for the multi-hop swarm UAV communication network.

Hypersharpening (namely, hyperspectral (HS) and multispectral (MS) image fusion) aims at enhancing the spatial resolution of HS image via an auxiliary higher resolution MS image. Currently, numerous hypersharpening methods are proposed successively, among which the unmixing-based approaches have been widely researched and demonstrated their effectiveness in the spectral fidelity aspect. However, existing unmixing-based fusion methods substantially employ mathematical techniques to solve the spectral mixture model, without taking full advantage of the collaborative spatial-spectral information that is usually helpful for abundance estimation improvement. To overcome this drawback, in this article, a novel unmixing-based HS and MS image fusion method, via a convolutional neural network (CNN), is proposed to promote spectral fidelity. The main idea of this work is to use CNN to fully explore the spatial information and the spectral information of both HS and MS images simultaneously, thereby enhancing the accuracy of estimating the abundance maps. Experiments on four simulated and real remote sensing data sets demonstrate that the proposed method is beneficial to the spectral fidelity of the fused images compared with some state-of-the-art algorithms. Meanwhile, it is also easy to implement and has a certain advantage in running time.

Since big data becomes a main impetus to the next generation of IT industry, data privacy has received considerable attention in recent years. To deal with the privacy challenges, differential privacy has been widely discussed and related private mechanisms are proposed as privacy-enhancing techniques. However, with today's differential privacy techniques, it is difficult to generate a sanitized dataset that can suit every machine learning task. In order to adapt to various tasks and budgets, different kinds of privacy mechanisms have to be implemented, which inevitably incur enormous costs for computation and interaction. To this end, in this paper, we propose two novel schemes for outsourcing differential privacy. The first scheme efficiently achieves outsourcing differential privacy by using our preprocessing method and secure building blocks. To support the queries from multiple evaluators, we give the second scheme that employs a trusted execution environment to aggregately implement privacy mechanisms on multiple queries. During data publishing, our proposed schemes allow providers to go off-line after uploading their datasets, so that they achieve a low communication cost which is one of the critical requirements for a practical system. Finally, we report an experimental evaluation on UCI datasets, which confirms the effectiveness of our schemes.

The shared nature of the wireless medium induces contention between data transport and backward signaling, such as acknowledgment. The current way of TCP acknowledgment induces control overhead which is counter-productive for TCP performance especially in wireless local area network (WLAN) scenarios. In this paper, we present a new acknowledgment called TACK (``Tame ACK''), as well as its TCP implementation TCP-TACK. TACK seeks to minimize ACK frequency, which is exactly what is required by transport. TCP-TACK works on top of commodity WLAN, delivering high wireless transport goodput with minimal control overhead in the form of ACKs, without any hardware modification. Evaluation results show that TCP-TACK achieves significant advantages over legacy TCP in WLAN scenarios due to less contention between data packets and ACKs. Specifically, TCP-TACK reduces over 90% of ACKs and also obtains an improvement of up to 28% on goodput. A TACK-based protocol is a good replacement of the legacy TCP to compensate for scenarios where the acknowledgment overhead is non-negligible.

Inertial measurement units (IMUs)-based sensing systems have received extensive attention for wearable gait measurement due to their small size, low price, and high accuracy. However, reconstructing walking dynamics usually requires a large number of sensor units attached to major body segments, which is generally not suitable for daily-life scenarios. Previous studies showed that two shank-mounted IMUs can detect gait events and estimate many important gait parameters. However, it is unclear whether this simple two-IMU system can be used to reconstruct walking dynamics. In this paper, we propose a two-stage method for predicting walking dynamics from measurements of two shank-mounted IMUs. Displacements of ankle joints and shank angles are first estimated from the IMU outputs. Then a model-based whole-step optimization approach is used to solve the gait dynamics by tracking the estimated shank motion. The proposed method is validated with both normal and asymmetric walking data, achieving a root mean square error of 5.3, 6.1, and 6.8 degrees in the hip, knee, and ankle joint angle estimation and 3.9% and 12.2% body weight in the fore-aft and vertical ground reaction force estimation. Comparing with the reported results in literature, the number of IMUs is significantly reduced with similar accuracy. This implies that gait dynamic information may be estimated from very limited measurements and the inherent gait characteristics can be used to reconstruct gait dynamics from the motion of a small number of segments.

An aptamer is a short oligonucleotide chain that can specifically recognize targeting analytes. Due to its high specificity, low cost, and good biocompatibility, aptamers as the targeting elements of biosensors have been applied widely in non‐invasive tumor imaging and treatment in situ to replace traditional methods. In this review, we will summarize recent advances in using aptamer‐based biosensors in tumor diagnosis. After a brief introduction of the advantage of aptamers compared with enzyme sensors and immune sensors, the different sensing designs and mechanisms based on 3 signal transduction modes will be reviewed to cover different kinds of analytical methods, including: electrochemistry analysis, colorimetry analysis, and fluorescence analysis. Finally, the prospective advantages of aptamer‐based biosensors in tumor theranostics and post‐treatment monitoring are also evaluated in this review., This review introduces the advantages of aptamer sensors in disease diagnosis compared with traditional biosensors. It also shows the design strategies of the 3 most studied aptamer sensors: fluorescent aptasensors, electrochemical aptasensors, and colorimetric aptasensors, and their application in early tumor diagnosis. Moreover, taking the fluorescent aptasensor as an example, the design of aptasensors combined with nanomaterials can not only be used for tumor imaging, but also has important significance in tumor treatment and efficacy evaluation. The design provides a new idea for the integration of tumor theranostics.

With the rapid development of society, video surveillance has progressively expanded into different areas of life, such as transportation, security inspection, banks. There are a large number of replaced and newly deployed cameras in fields such as safe cities, smart campuses and smart buildings, which leads to a huge amount of video data, slow retrieval speed in video examining, and low efficiency in understanding complete picture of videos. In this paper, we propose SurVizor, a visual analysis system to understand the key content of surveillance videos. We integrate multiple image features and employ time series analysis methods to explore key temporal patterns in the feature. We integrate multiple visualization views from three levels of video, feature, and frame to promote exploration, analysis and understanding of video content. We evaluate the proposed system through a case study based on real-world surveillance videos from multi-camera and a user study. The results demonstrate the usability and effectiveness of our system in analyzing and understanding the key content of surveillance videos.

An electronic toll collection (ETC) system is one of the most important parts of an intelligent transportation system (ITS), but existing ETC systems are not efficient and have vehicle fee evasion complications. Owing to the traceability and tamper resistance of a blockchain, the combination of a blockchain and ETC system is a feasible way to solve the above problems. However, traditional blockchain (e.g., Bitcoin and Ethereum) has high power consumption and low efficiency. Hence, it cannot be used in ETC systems with high throughput and is not suitable for power-constrained Internet of Things (IoT) devices. In this study, therefore, we propose a blockchain architecture for the ETC system. To curb vehicle fee evasion behavior, thereby reducing the economic loss caused, we propose a vehicle behavior management mechanism based on credit value. Furthermore, to reduce the burden of storage and build a chain of evidence for auditing purposes, we propose an evidence chain framework. To protect the data security in the transaction process, we designed a data protection method to encrypt the transaction data. Additionally, this system is based on an open-source alliance blockchain framework called Hyperledger Fabric, which is more in line with the current application scenarios of ETC systems. We implemented the system on a Raspberry Pi and carried out simulations. The comprehensive evaluation results and analysis show that the system effectively reduces the number of illegal acts, completes the evidence inspection, and improves the security of the data.

Background: Event log data generated in the software development process contains historical information and future trends in software development activities. The mining and analysis of event log data contribute to identify and discover software development activities and provide effective support for software development process mining and modeling. Methods: Firstly, a deep learning model (Word2vec) was used for feature extraction and vectorization of software development process event logs. Then, the K-means clustering algorithm and measure of silhouette coefficient and intra-cluster SSE were used for clustering and clustering effect evaluation of vectorized software development process event logs. Results: This paper obtained the mapping relationship between software development activities and events, and realized the identification and discovery of software development activities. Conclusion: Two practical software development projects (jEdit and Argouml) are given to prove the feasibility, rationality and effectiveness of our proposed method. This work provides effective support for software development process mining and software development behavior guidance.

This paper proposes a new non-cascade fast nonsingular terminal sliding mode control algorithm based on disturbance observers for the speed control of permanent magnet synchronous motor (PMSM). To ensure the high precision control and fast finite-time convergence, a fast nonsingular terminal sliding mode controller is designed for PMSM. The novel controller adopts the single-loop non-cascade control structure. Then, to improve the anti-disturbance ability, a nonlinear disturbance observer and a dual disturbance observer are respectively designed to estimate the disturbances during motor operation, which are used for the feed-forward compensation control. The composite control method can effectively alleviate the larger chattering caused by the high sliding mode gain. Finally, the stability is proved based on Lyapunov stability theory. Simulation and experimental results show that the proposed method has the advantages of fast convergence and strong robustness against parameter variations and external disturbances.

Deep learning-based Automatic Modulation Recognition (AMR) can improve the recognition rate compared with traditional AMR methods. However, in practical applications, as training samples and real scenario signal samples have different distributions in practical applications, the recognition rate for target domain samples can deteriorate significantly. This paper proposed an unsupervised domain adaptation based AMR method, which can enhance the recognition performance by adopting labeled samples from the source domain and unlabeled samples from the target domain. The proposed method is validated through signal samples generated from the open-sourced Software Defined Radio (SDR) GNU Radio. The training dataset is composed of labeled samples in the source domain and unlabeled samples in the target domain. In the testing dataset, the samples are from the target domain to simulate the real scenario. Through the experiment, the proposed method has a recognition rate increase of about 88% under the CNN network structure and 91% under the ResNet network structure.

In the field of artificial intelligence, classification algorithms tend to be biased toward the majority class samples when encountering imbalanced data , resulting in low recognition rates for minority class samples. Undersampling techniques address this issue by decreasing the number of majority class samples to balance the original data distribution before the dataset is learned. However, current clustering-based undersampling methods have limitations that directly affect the original imbalanced dataset and the final classification performance. To address these problems, we propose a novel three-stage undersampling framework with denoising, fuzzy c-means clustering, and representative sample selection (UFFDFR). This framework improves the classification performance on imbalanced data by removing noise and unrepresentative samples from the majority class. Experiments on 15 different imbalanced datasets demonstrate that UFFDFR effectively removed noise and unrepresentative majority class samples and improved classification performance. Furthermore, UFFDFR outperformed three classic and three state-of-the-art clustering-based undersampling methods in terms F-measure, G-mean, and AUC for five classification algorithms, which was confirmed by the Friedman and Nemenyi post-hoc statistical tests.

Ancient Chinese military castles were equipped with rapid transportation routes for mutual aid, and this was an essential indicator of regional defense capability. However, since the sites of these transportation routes have mostly disappeared, it is not easy to examine the actual distribution of these routes. It is necessary to speculate the trend and position of military routes on the basis of the castle locations. In this study, the geographic features of each castle location were extracted as factors affecting the efficiency of the intercastle transportation system using the ArcGIS cost path function. By analyzing the fit of each factor for screening and weight assignment, a time cost path was established, and a model was generated for calculating the efficiency of this transportation system. The Weihai area, a typical representation of sea defense during the Chinese Ming Dynasty, was taken as an example for simulation. Overall, five ancient military transportation routes were restored. The establishment of the Ming Dynasty Wendengying transformed the linear defense layout of the Weihai region into a longitudinal network layout, and its site selection was of great benefit to the overall defense of the coastal citadel of Weihai. This model breaks the traditional limitations of relying on subjective speculation for ancient road restoration and dramatically improves its accuracy and credibility. Moreover, it makes a significant contribution to judging the road systems of ancient cities in different regions and provides a new idea to quantify the efficiency of ancient castle defenses.

Urban rail transit stations are the key nodes of urban rail transit network. Identifying and analyzing land use characteristics around urban rail transit stations can significantly contribute to urban rail transportation operation and management. Therefore, a visualization method of land use characteristics around urban rail transit stations based on POI is proposed in this paper. In the proposed method, first, the Voronoi diagram is used to determine coverage of urban rail transit stations and each POI is put in a coverage area based on their physical location. Then, topic-oriented hierarchical POIs of each urban rail transit station are extracted based on skyline idea. Finally, the land use characteristics around an urban rail transit station are visualized based on the extracted hierarchical POIs. We carried out two case studies and a quality evaluation. By using realistic data from Beijing rail transit in order to validate the method proposed in this paper. Results show that our method can clarify various situations of land use of urban rail transit stations and may provide support for the application of transportation model technology.

Table-to-Text generation aims to generate descriptions which can be viewed as a set of field-value records for factual tables. Despite the significant progress, the state-of-the-art models suffer from two major issues: Nonfluency and Divergence. Nonfluency means descriptions generated by models are not as fluent as those generated by humans, and thus can be distinguished easily. Divergence refers to the fact that the generated sentences contain information which can not be concluded from factual tables. This could be attributed to that most neural models are trained with the Maximum Likelihood Estimation (MLE) loss and use divergence-contained references as the ground truth, which forces the models to learn what cannot be inferred from the source to some extent. Motivated by the limitations of current models, we propose a novel GAN-based model with adversarial learning mechanism, which simultaneously trains a generative model G and a discriminative model D, to address Nonfluency and Divergence issues in Table-to-Text generation. Specifically, we build the generator G as an agent of reinforcement learning with a sequence-to-sequence architecture, which takes the raw data as input and predicts the generated sentences. Meanwhile, we build the discriminator D with a Convolutional Neural Network (CNN) to calculate rewards to measure the fluency of generations. To judge the fidelity of generations with regard to the original table more accurately, we also calculate the rewards from BLEU-Table. With the fusion rewards from CNN and BLEU-Table, our methods outperform the baselines by a large margin on the WikiBio and Wiki3C benchmarks evaluated with BLEU, ROURGE, and PARENT. Specifically, our models achieve 49.0 (BLEU-4), 37.8 (ROUGE-4) and 45.4 (PARENT) on WikiBio, as well as 12.9 (BLEU-4) and 6.9 (ROUGE-4) on Wiki3C. More importantly, we construct a new Wiki3C dataset that improves the insufficiency of datasets and promote the progress in Table-to-Text generation.

The range query on encrypted databases is usually implemented using the order-preserving encryption (OPE) technique which preserves the order of plaintexts. Since the frequency leakage of plaintexts makes OPE vulnerable to frequency-analyzing attacks, some frequency-hiding order-preserving encryption (FH-OPE) schemes are proposed. However, existing FH-OPE schemes require either the large client storage of size O ( n ) or O (log n ) rounds of interactions for each query, where n is the total number of plaintexts. To this end, we propose a FH-OPE scheme that achieves the small client storage without additional client-server interactions. In detail, our scheme achieves O ( N ) client storage and 1 interaction per query, where N is the number of distinct plaintexts and N ≤ n . Especially, our scheme has a remarkable performance when N ≪ n . Moreover, we design a new coding tree for producing the order-preserving encoding which indicates the order of each ciphertext in the database. The coding strategy of our coding tree ensures that encodings update in the low frequency when inserting new ciphertexts. Experimental results show that the single round interaction and low-frequency encoding updates make our scheme more efficient than previous FH-OPE schemes.

Software is rapidly evolving and operates in a changing environment; therefore, in addition to software design and testing, it is essential to observe and understand the software execution behavior by modeling data recorded during the execution of the software to improve its reliability. The nested call relationship between methods during the execution of software is common, but most process-mining methods are unable to discover them, only generating flat models with low fitness. Meanwhile, it is easy to generate “spaghetti-like” models with low comprehensibility when dealing with complex software execution data. This paper proposes a component-based hierarchical software behavior model discovery method that can discover hierarchical nested call structures during software runtime, improving the fitness of the model; meanwhile, the proposed method partitions the discovery model into several parts by component information to improve the comprehensibility of the model, which can also reflect the interaction behavior within and between components. The proposed approach was implemented in a process mining toolkit. Using real-life software event logs and public datasets, we demonstrated that compared with other advanced process mining techniques, our approach can visualize actual software execution behavior in a more accurate and easy-to-understand way while balancing time performance.

In this paper, the control problem for a class of stochastic nonlinear systems with incomplete measurements is investigated based on Luenberger-like nonlinear state observer. The systems are descri...

Point of interest (POI) recommendation is a fundamental task in location-based social networks (LBSN). The increasing proliferation of LBSNs brings about considerable amounts of user-generated check-in data. Such data can significantly contribute to understanding user behaviors, based on which personalized recommendations can be efficiently derived. Spatial and temporal effects are crucial factors in the user’s decision-making for choosing a POI to visit. Most existing methods treat them as two independent features and cannot accurately capture users’ interests. We argue that spatial and temporal effects should be analyzed simultaneously in POI recommendations. To this end, we propose a S patioT emporal heterogeneous information Network (HIN)-based PO I RE commendation model (STORE) to model various heterogeneous context features, e.g., the joint spatiotemporal effects, types of POI, and social relations. Specifically, we defined the spatiotemporal effects entity (St) in HIN to model the joint spatiotemporal effects. Instead of modeling the traditional two-way interaction , we further design a four-way neural interaction model . In this way, our model can effectively mine and extract useful information from the meta-path-based context and spatiotemporal effects, thereby improving recommendation performance. We conduct extensive experiments on two real-world datasets, and the results demonstrate that the STORE model outperforms the best baseline by about 12% in NDCG@5 and 11% in Rec@5.

Crowd counting aims to count the number of pedestrians in an image or a video. Currently, scale variations in crowd counting are inevitable and challenging in practice. Besides, for the crowd counting task, there is only a small amount of annotated data available. It can be seen from previous methods of crowd counting and object detection that the two tasks have similar attention areas. However, existing methods in crowd counting generally ignore the similarities and specialties between the crowd counting task and the object detection task. In this paper, in order to solve the above challenges, we propose an adversarial scale-adaptive neural network (ASANet), consisting of three branches. First, a private branch for the crowd counting task concentrates on generating high-quality density maps. Second, another private branch for the object detection task aims to correctly detect and recognize objects. Third, we design a common branch to learn the similar attention area of the two tasks and assist crowd counting. Experimental results demonstrate an outstanding performance of the ASANet over state-of-the-art methods on three public datasets (ShanghaiTech, UCF_CC_50, and UCF_QNRF).

With the improvement of public security awareness, video anomaly detection has become an indispensable demand in surveillance videos. To improve the accuracy of video anomaly detection, this paper proposes a novel two-stream spatial-temporal architecture called Two-Stream Deep Spatial-Temporal Auto-Encoder (Two-Stream DSTAE), which is composed of a spatial stream DSTAE and a temporal stream DSTAE. Firstly, the spatial stream extracts appearance characteristics whereas the temporal stream extracts the motion patterns, respectively. Then, based on the novel policy joint reconstruction error, this model fuses the spatial stream and the temporal stream to extract spatial-temporal characteristics to detect anomalies. Furthermore, since the optical flow is invariant to appearances such as color or light, we introduce optical flow to enhance the capability of extracting continuity between adjacent frames and inter-frame motion information. We demonstrate the accuracy of the proposed method on the publicly available standard datasets: UCSD, Avenue and UMN datasets. Our experiments demonstrate high accuracy, which is superior to the state-of-the-art methods.

In the human-vehicle-road system of collisions, the human is the most important factor, and digital human models (DHMs) are developed with the aim of preventing or at least reducing human injury. Because most of the relevant literature is focused mainly on collisions in traffic accidents (TAs), only some of the literature reviewed in this paper involves research results on other aspects of collisions. In this review, based on the background of DHMs and the application of DHMs regarding human injury biomechanics in collisions field, research results regarding the development of DHMs are described, the methods for verifying such models are introduced, and the application of the research results is discussed based on the aspect of human injury biomechanics. From the research literature, the development and validation of DHMs and their application in human injury biomechanics are summarized, and future research trends are proposed and discussed.

Always-on keyword spotting (KWS) that detects wake-up words has been the indispensable module in the voice interaction system. However, the ultra-low-power embedded devices put forward strict requirements on energy consumption, latency, and recognition accuracy of KWS. In this work, we propose a near-sensor processing architecture of feature-configurable distributed network (NS-FDN) for always-on KWS applications. The proposed distributed network adapts to the flexible keywords demands in the actual scene by splitting the conventional single network into distributed sub-networks. We design a channel-independent training framework to improve the recognition accuracy of distributed networks. The speech features are evaluated and the redundancy is reduced in NS-FDN, which can also configure the speech features to further reduce the computing complexity and improve processing speed. For deeper optimization, we implement a 65nm-process prototype chip with near-sensor mixed-signal processing architecture avoiding energy-consuming analog-to-digital converter. By improving the system, algorithm, and hardware designs of the KWS, our co-optimized architecture eliminates the energy consumption bottleneck long-standing in conventional KWS systems and achieves state-of-the-art system performance. The experiment results show that NS-FDN achieves 31.6% energy consumption savings, 1.6 times memory savings, 57 times speedup, and 3.4% higher recognition accuracy compared with the state of the art.

As one of the most critical blocks of the CMOS image sensor (CIS), the accuracy and power consumption of the analogue‐to‐digital converter (ADC) play an important role in its performance. However, due to the mutually restrictive relationship between power consumption and accuracy in the single‐slope ADC (SS ADC), it is difficult to improve these performance indexes meantime without an efficient optimal design method. Here, an optimal design methodology based on an improved artificial fish swarm optimization algorithm (IAFSOA) is proposed for a 10‐bit SS ADC in CIS. First, a voltage storage structure and offset calibration circuit are developed for promoting the performance of the comparator. Then, IAFSOA is developed for the circuit parameters optimization with faster convergence rate and higher computational accuracy. The proposed multi‐objectives optimization algorithm can obtain the optimal parameters of the kernel circuit components and lead to an enhancement of both power consumption and accuracy. The optimized SS ADC is designed based on 0.18 µm process to verify the effectiveness of the proposed method. Compared to the earlier reported work, the IAFSOA‐based optimal design method can promote the performance of the SS ADC by an improvement of effective number of bits and a reduction of power consumption.

Wireless communication systems assisted by unmanned aerial vehicles (UAVs) can well extend the coverage of conventional communication networks and quickly provide service wherein the communication infrastructures are destroyed. However, radio resources are always limited and coupled with complex interference in the orthogonal multiple access framework. Thus, non-orthogonal multiple access (NOMA), which can enhance spectrum efficiency and inherently has capability to mitigate interference, is integrated with UAV communication networks in this article. We first propose a cooperative UAV NOMA network (CUNN) architecture and analyze its technical challenges. In addition, with the goal of cooperatively or competitively utilizing the limited resources effectively in a distributed manner, game theory shows its potential in addressing various technical challenges in wireless networks. Thus, we further survey game theoretical approaches for NOMA networks and UAV-based relay networks, and provide insights on how these game theoretical approaches can be adopted to solve technical challenges in CUNNs. Finally, aiming to achieve high spectrum efficiency and ensure fairness among user equipments, we propose a UAV-iUE (isolated user equipment) association, UAVs' position optimization, and power allocation algorithm in a dynamic cooperative game-theoretic framework. In the numerical results and discussions Section, the effectiveness of the proposed dynamic cooperative game-theoretic framework is shown through multiple simulations.

Next-generation communication networks can provide high capacity, low latency, and massive connections; however, they introduce novel challenges of management complexity, and traditional mathematical methods cannot well characterize the rational behavior of users. In this article, we pay attention to the methods of artificial intelligence (AI) and game theory. We first review the applications of machine learning (ML) and game theory models in wireless communications and summarize their advantages and disadvantages. After surveying the state of the art, in this article we propose a novel framework combining ML and game theory, which explores and exploits the benefits of the two disciplines. Finally, we apply our novel framework to solve the network selection problem in a 5G ultra-dense and heterogeneous network. Simulation results confirm the advantage of our presented framework on reducing the average delay of users.

Mobile devices connected to the Internet are a ubiquitous platform that can easily record a large amount of data describing human behavior. Specifically, the data collected from mobile devices --- referred to as mobile big data reveal important social and economic information. Therefore, analyzing mobile big data is valuable for several stakeholders, ranging from smartphone manufacturers to network operators and app developers. This thesis aims to discover and understand behavioral patterns from mobile big data based on large real-world datasets. Specifically, this thesis reveals patterns from three domains: people, time, and location. First, we explore mobile big data from the people domain and propose a framework to discover users' daily activity patterns from their mobile app usage. By applying the framework to a real-world dataset consisting of 653,092 users, we successfully extract five common patterns among millions of people, including commuting, pervasive socializing, nightly entertainment, afternoon reading, and nightly socializing. Second, still from the people domain, we derive group health conditions by using their smartphone usage data. In particular, we collect mobile usage records of 452 users in North America. We then demonstrate the potential for inferring group health conditions (i.e., COVID-19 outbreak stages) by leveraging less privacy-sensitive smartphone data, including CPU usage, memory usage, and network connections. Third, we mine the behavior patterns from the time domain. We reveal the evolution of mobile app usage by conducting a longitudinal study on 1,465 users from 2012 to 2017. The results show that users' app usage significantly changes over time. However, the evolution in app-category usage and individual app usage are different in terms of popularity distribution, usage diversity, and correlations. Last, with respect to the location domain, we leverage city-scale spatiotemporal mobile app usage data to reveal urban land usage patterns. We prove the strong correlation between mobile usage behavior and location features, which brings a new angle to urban analytics. Internetiin kytketyt mobiililaitteet ovat kaikkialla läsnä oleva alusta, joka voi helposti tallentaa suuren määrän tietoja, jotka kuvaavat ihmisen käyttäytymistä. Erityisesti mobiililaitteista kerätyt tiedot, joita kutsutaan mobiiliksi massadataksi (big data), paljastavat tärkeitä sosiaalisia ja taloudellisia tietoja. Siksi mobiilin massadatan analysointi on arvokasta useille sidosryhmille älypuhelinvalmistajista verkko-operaattoreihin ja sovelluskehittäjiin. Tämän väitöskirjan tavoitteena on löytää ja ymmärtää käyttäytymismalleja mobiilista massadatasta, joka perustuu suuriin reaalimaailman tietojoukkoihin. Erityisesti tämä väitöskirja tuottaa malleja kolmelta eri alueelta: ihmisiin, aikaan ja sijaintiin liittyen. Ensinnäkin tutkimme mobiilia massadataa ihmisiin liittyen ja ehdotamme viitekehystä, jonka avulla voidaan löytää käyttäjien päivittäisiä toimintamalleja heidän mobiilisovellustensa käytön perusteella. Soveltamalla tätä viitekehystä tosielämän tietojoukkoon, joka koostuu 653 092 käyttäjästä, löysimme onnistuneesti viisi yleistä mallia miljoonien ihmisten tiedoista, joihin kuuluivat mm. tiedot työmatkoista, sosiaalisista kontakteista, yöllisestä viihteestä, iltapäivän lukemisesta ja yöllisestä seurustelusta. Toiseksi, edelleen ihmisiin liittyen, johdamme tietoja ryhmien terveysolosuhteista käyttämällä heidän älypuhelintensa käyttötietoja. Keräsimme erityisesti 452 käyttäjän mobiilikäyttötietoja Pohjois-Amerikassa. Sitten osoitamme, että on mahdollista päätellä ryhmän terveysolosuhteet (eli COVID-19-epidemiavaiheet) hyödyntämällä vähemmän yksityisyyden kannalta arkoja älypuhelintietoja, mukaan lukien suorittimen käyttö, muistin käyttö ja verkkoyhteydet. Kolmanneksi louhimme käyttäytymismalleja aikaan liittyen. Paljastamme mobiilisovellusten käytön kehityksen tekemällä pitkittäistutkimuksen 1 465 käyttäjälle vuosina 2012–2017. Tulokset osoittavat, että käyttäjien sovellusten käyttö muuttuu merkittävästi ajan myötä. Sovellusluokan käytön ja yksittäisten sovellusten käytön kehitys on kuitenkin erilainen niiden suosion jakautumisen, käytön moninaisuuden ja korrelaatioiden suhteen. Lopuksi liittyen sijaintitietoihin hyödynnämme spatiotemporaalisten mobiilisovellusten käyttötietoja suurkaupunkitasolla paljastaaksemme kaupunkien maankäyttömallit. Todistamme vahvan korrelaation mobiililaitteiden käyttöön liittyvän käyttäytymisen ja sijaintiominaisuuksien välillä, mikä tuottaa uuden näkökulman kaupunkianalytiikkaan.

In this paper, we investigate the feasibility of using mobility patterns and demographic data to predict hospital visits. We collect mobility traces from two thousand users for around two months. We extract 16 mobility features from these passively collected mobility traces and train an XGBoost model to predict users' hospital visits. We demonstrate that the designed mobility features can significantly improve prediction accuracy (p < 0.01, AUC = 0.79). We further analyze how these mobility features affect the prediction results and measure their importance by using Shapley additive explanation values. We discover that users with less mobility activity, less visit diversity, and few sports facilities, bountiful entertainment around their visited locations are more likely to visit hospitals. Moreover, we conduct predictions on the populations with different demographic features, which achieves meaningful and insightful results, i.e. maintaining a high mobility activity is crucial for older people's health, while fast food store more substantially affects younger people's health; visit patterns can indicate females' health, while the neighborhood environment is more indicative of males, etc. These results shed light on how to use and understand large scale mobility data in health monitoring and other health-related applications in practice.

Radar cross section (RCS) reduction in the working band of antennas, likely leading to radiation performance change or deterioration, is quite difficult for both antenna elements and arrays. Aimed at this problem, this article presents a novel design strategy of microstrip patch antennas with low in-band RCS. This strategy combines metasurface with antenna from the perspectives of both structure and electrical performance. To be specific, anisotropic metasurface is proposed to replace the conventional patch and behave as the radiating structure of the antenna directly, which does not increase the original antenna size. Through proper connection of the metasurface units, the novel antenna has the same radiation performance with conventional microstrip antenna but low in-band RCS for any polarized incidence. The radiation and scattering mechanism of the novel antenna is carefully elaborated, from which a frequency reconfigurable antenna is also derived by loading varactors. Furthermore, different design strategies of low in-band-RCS antenna element and array are clarified. On the basis of which, a $4 \times 4$ array is developed to achieve in-band-RCS reduction with radiation performance kept. Both numerical and experimental results are demonstrated, and the reasonably good agreements verify the effectiveness of the proposed strategy.

The dependence on frequent acknowledgments (ACKs) is an artifact of current transport protocol designs rather than a fundamental requirement. Frequent ACKs waste resources when the overhead of ACKs is nonnegligible. However, reducing the number of ACKs may hurt transport performance. “Tame ACK” is an on-demand ACK mechanism that seeks to minimize ACK frequency, which is exactly what is required by transport.

Bridging the gap between academia and industry is an important issue to promote the practicality of i* framework. Researchers have been dealing with this issue from various perspectives, such as simplifying the meta-models or modeling processes of i* framework. In this paper, we exclusively focus on the scalability issue in laying out large-scale i* models and propose a two-level layout approach to automatically lay out i* models in an efficient and comprehensible manner, contributing to the adoption of i* framework in the industry. The proposed approach is designed by considering the semantics of i* constructs and layout conventions of i* models in order to produce meaningful layouts and can appropriately handle both the SD (Strategic Dependency) view and the SR (Strategic Rationale) view of i* models. We have implemented our approach in an open-access prototype tool, which is able to be integrated with existing iStarML-compatible modeling tools. We have conducted a controlled experiment, a case study, and performance testing to empirically and comprehensively evaluate the utility of our approach, the results of which show that our proposal can efficiently produce meaningful layouts that are as comprehensible as manually laid out models in most cases.

Sarcasm is common in social media, and people use it to express their opinions with stronger emotions indirectly. Although it belongs to a branch of sentiment analysis, traditional sentiment analysis methods cannot identify the rhetoric of irony as it requires a significant amount of background knowledge. Existing sarcasm detection approaches mainly focus on analyzing the text content of sarcasm using various natural language processing techniques. It is argued herein that the essential issue for detecting sarcasm is examining its context, including sentiments of texts that reply to the target text and user’s expression habit. A dual-channel convolutional neural network is proposed that analyzes not only the semantics of the target text, but also its sentimental context. In addition, SenticNet is used to add common sense to the long short-term memory (LSTM) model. The attention mechanism is then applied to take the user’s expression habits into account. A series of experiments were carried out on several public datasets, the results of which show that the proposed approach can significantly improve the performance of sarcasm detection tasks.

For CVaR (conditional value-at-risk) portfolio nonsmooth optimization problem, we propose an infeasible incremental bundle method on the basis of the improvement function and the main idea of incremental method for solving convex finite min-max problems. The presented algorithm only employs the information of the objective function and one component function of constraint functions to form the approximate model for improvement function. By introducing the aggregate technique, we keep the information of previous iterate points that may be deleted from bundle to overcome the difficulty of numerical computation and storage. Our algorithm does not enforce the feasibility of iterate points and the monotonicity of objective function, and the global convergence of the algorithm is established under mild conditions. Compared with the available results, our method loosens the requirements of computing the whole constraint function, which makes the algorithm easier to implement.

In recent years, semisupervised spectral–spatial feature extraction (FE) methods for hyperspectral image (HSI) classification have shown promising performance by combining spectral–spatial information and label information. A problem that has not been addressed satisfactorily is that how to effectively and collaboratively use this abundant information contained in the hyperspectral data to exhibit better HSI classification performance. In this article, a novel FE method based on joint spectral–spatial information is proposed for HSI classification, which consists of the following steps. First, an effective re-expression for the original data is constructed by incorporating texture features extracted by extended multiattribute profiles with the original HSI. Thus, every pixel can be described by diverse and complementary information in the spectral–spatial domain. Then, the improved neighborhood preserving embedding (NPE) is proposed to establish a relatively accurate reconstruction model and mine high-reliable neighborhood structure from a global perspective by a new distance metric, which incorporates spectral bands, texture features, and geographical information simultaneously. Finally, the low-dimensional and high-discriminative features for HSI classification are obtained by combining the scatter matrices of local fisher discriminant analysis based on labeled samples and the improved NPE based on the whole data. Experimental results on three real-world HSI datasets show that the proposed method can effectively utilize both the label information and the spectral–spatial information, and hence, achieve much better classification performance compared to the conventional FE methods and some state-of-the-art spectral–spatial classification methods.

Recently, crowd counting has drawn widespread attention in computer vision, but it is extremely challenging because of the varying scales and densities. Many existing methods focus on improving the multi-scale representation by utilizing multi-column or multi-branch architectures with different kernel sizes. However, such networks cannot extract the feature maps with large receptive fields due to limitation of depth. In addition, the importance of utilizing the multi-level feature information in a deep network is ignored. In this paper, we propose a multi-scale and multi-level features aggregation network (MFANet) for accurate and efficient crowd counting, and it can be trained by end-to-end. A vital component of the network is the scale and level aggregation module (SLAM), which can extract multi-scale features and make full use of multi-level feature information for more accurate estimation. When six SLAMs are stacked together and applied to our network, our method can achieve the best performance. Furthermore, we introduce a new loss function called normalized Euclidean loss (NEL) to balance the contribution of all samples to network training. To demonstrate the performance of the proposed method, extensive experiments are conducted on four benchmark crowd counting datasets, including ShanghaiTec Part A/B, UCF-CC-50, Mall, and UCF-QNRF. Experimental results show that our MFANet achieves state-of-the-art performance in crowd counting and crowd localization.

Structural health monitoring (SHM) systems have been extensively used to ensure the operational safety of long-span bridges. Large-scale bridge structural response and loading data observed from various sensors show obvious big data characteristics. However, serious “data island” problems, which exist in the conventional SHM solutions, inevitably limit the effects of sensory data analysis and information sharing. A unified bridge SHM semantic representation model is much in demand. By taking the advantages of Semantic Web technologies, this article presents a novel model, called the bridge structure and health monitoring ontology, to achieve fine-grained modeling of bridge structures, SHM systems, sensors, and sensory data from multiple perspectives. A bridge SHM big data platform is used to demonstrate the usefulness. Several representative data accessing and rule-based reasoning scenarios are employed as to illustrate the advantages of the proposed manner.

As an extensive prospect in communication technology, multi-hop unmanned aerial vehicle (UAV) faces some challenges as well. How to ensure the availability of time slots in high dynamic scenarios and the rationality and robustness of power allocation in massive scenarios remains to be resolved. In this paper, we propose a robust mean field game (MFG) framework to manage time slot resources and power resources. On the one hand, this framework can predict the potential conflicts of time slots and avoid them in time. On the other hand, it can well adapt to the massive scenarios with complex interference, in order to realize power control. The simulation results show that the robust MFG-based resource management scheme can not only reduce the probability of packet collision and packet loss, but also improve communication energy efficiency (EE) with better robustness.

Aiming at the speed regulation problem of permanent magnet synchronous motor (PMSM) drives, a novel backstepping nonsingular terminal sliding mode control (BNTSMC) method with finite-time disturbance observer (FTDO) is proposed. In order to ensure excellent tracking and anti-disturbance performance, the backstepping nonsingular terminal sliding mode speed controller based on equivalent motor model and the recursive principle is designed for PMSM. The benefits of this approach are that the controller has asymptotic stability based on backstepping design and strong robustness for the matched disturbance based on sliding mode control. Then, to further improve the anti-disturbance performance, a finite-time disturbance observer is designed, and it is used to estimate the mismatched external disturbance in motor drive system. The stability of the system is also proved with Lyapunov criterion. Finally, the experimental results show that the proposed method has the advantages of fast convergence and strong robustness against parameter variations and external disturbance.

The advent of noisy intermediate-scale quantum (NISQ) devices provide crucial promise for the development of quantum algorithms. Variational quantum algorithms have emerged as one of the best hopes to utilize NISQ devices. Among these is the famous variational quantum eigensolver (VQE), where one trains a parameterized and fixed quantum circuit (or an ansatz) to accomplish the task. However, VQE also suffers from some serious challenges, which are training difficulty and accuracy reduction due to deep quantum circuit and hardware noise. Motivated by these issues, we propose a runtime and resource-efficient scheme, Monto Carlo tree (MCT) search-based quantum circuit architecture optimization, where the ansatz is built in the variable form. Our approach first models the search space with a MCT and regards it as a supernet, where we make use of layers dependence to reduce the size of the search space. Second, a two-stage scheme is proposed for the search space training, where weight sharing and warm-up strategies are employed to avoid huge computation cost. Training results are stored in nodes of the MCT for future decisions, and hierarchical node selection is presented to obtain an optimal ansatz. As a proof of principle, we carry out a series of numerical experiments for condensed matter and quantum chemistry in a quantum simulator with and without noise. Consequently, our scheme can be efficient to mitigate trainability and accuracy issues by minimizing the ansatz depth and the number of entanglement gates.

The simple frigate shape (SFS) as defined by The Technical Co-operative Program (TTCP), is a simplified model of the frigate, which helps to investigate the basic flow fields of a frigate. In this paper, the flow fields of the different modified SFS models, consisting of a bluff body superstructure and the deck, were numerically studied. A parametric study was conducted by varying both the superstructure length L and width B to investigate the recirculation zone behind the hangar. The size and the position of the recirculation zones were compared between different models. The numerical simulation results show that the size and the location of the recirculation zone are significantly affected by the superstructure length and width. The results obtained by Reynolds-averaged Navier-Stokes method were also compared well with both the time averaged Improved Delayed Detached-Eddy Simulation results and the experimental data. In addition, by varying the model size and inflow velocity, various flow fields were numerically studied, which indicated that the changing of Reynolds number has tiny effect on the variation of the dimensionless size of the recirculation zone. The results in this study have certain reference value for the design of the frigate superstructure.

Distributed and pervasive web services have become a major platform for sharing information. However, the hypertext transfer protocol secure (HTTPS), which is a crucial web encryption technology for protecting the information security of users, creates a supervisory burden for network management (e.g., quality-of-service guarantees and traffic engineering). Identifying various types of encrypted traffic is crucial for cyber security and network management. In this paper, we propose a novel deep learning model called BGRUA to identify the web services running on HTTPS connections accurately. BGRUA utilizes a bidirectional gated recurrent unit (GRU) and attention mechanism to improve the accuracy of HTTPS traffic classification. The bidirectional GRU is used to extract the forward and backward features of the byte sequences in a session. The attention mechanism is adopted to assign weights to features according to their contributions to classification. Additionally, we investigate the effects of different hyperparameters on the performance of BGRUA and present a set of optimal values that can serve as a basis for future relevant studies. Comparisons to existing methods based on three typical datasets demonstrate that BGRUA outperforms state-of-the-art encrypted traffic classification approaches in terms of accuracy, precision, recall, and F1-score.