Assessing data quality in multivariate time-series datasets is crucial for reliable analysis, particularly when dealing with missing values, inconsistent feature availability, and massive records in large-scale edge computing and IoT clusters. Existing methods often fall short of capturing intricate patterns of missingness and co-coverage, restricting the capacity to make well-informed decisions regarding the usability of the data. In order to systematically extract reliable data segments, this paper presents a comprehensive framework that combines a heuristic model with temporal coverage, period-specific missingness, and co-coverage metrics. By integrating these metrics with visualizations such as temporal coverage heatmaps and parallel coordinates plots, the framework reveals complex patterns of missingness while supporting human involvement in validating data subsets. Our approach effectively balances automation with expert judgment, enhancing the interpretability of data quality assessments. The findings show that the proposed methods satisfy the design specifications for revealing patterns, quantifying missingness impact, measuring feature availability, guiding feature selection, and facilitating scalable, multi-scale data summarization. The framework offers a solid way to improve the quality of data in multivariate time-series analysis, opening the door to more precise and trustworthy insights for assessing data gathered from edge computing infrastructures and large-scale, heterogeneous IoT deployments, where data consistency and completeness are frequently very variable.

Cloud-based large language models (LLMs) and their variants have significantly influenced real-world applications. Deploying smaller models (i.e., small language models (SLMs)) on edge devices offers additional advantages, such as reduced latency and independence from network connectivity. However, edge devices’ limited computing resources and constrained energy budgets challenge efficient deployment. This study evaluates the power efficiency of five representative SLMs — Llama 3.2, Phi-3 Mini, TinyLlama, and Gemma 2 on Raspberry Pi 5, Jetson Nano, and Jetson Orin Nano (CPU and GPU configurations). Results show that Jetson Orin Nano with GPU acceleration achieves the highest energy-to-performance ratio, significantly outperforming CPU-based setups. Llama 3.2 provides the best balance of accuracy and power efficiency, while TinyLlama is well-suited for low-power environments at the cost of reduced accuracy. In contrast, Phi-3 Mini consumes the most energy despite its high accuracy. In addition, GPU acceleration, memory bandwidth, and model architecture are key in optimizing inference energy efficiency. Our empirical analysis offers practical insights for AI, smart systems, and mobile ad-hoc platforms to leverage tradeoffs from accuracy, inference latency, and power efficiency in energy-constrained environments.

Attention deficit disorder is a frequently observed symptom in individuals with autism spectrum disorder (ASD). This condition can present significant obstacles for those affected, manifesting in challenges such as sustained focus, task completion, and the management of distractions. These issues can impede learning, social interactions, and daily functioning. This complexity of symptoms underscores the need for tailored approaches in both educational and therapeutic settings to support individuals with ASD effectively. In this study, we have expanded upon our initial virtual reality (VR) prototype, originally created for attention therapy, to conduct a detailed statistical analysis. Our objective was to precisely identify and measure any significant differences in attention-related outcomes between sessions and groups. Our study found that heart rate (HR) and electrodermal activity (EDA) were more responsive to attention shifts than temperature. The ‘Noise’ and ‘Score’ strategies significantly affected eye openness, with the ASD group showing more responsiveness. The control group had smaller pupil sizes, and the ASD group’s pupil size increased notably when switching strategies in Session 1. Distraction log data showed that both ‘Noise’ and ‘Object Opacity’ strategies influenced attention patterns, with the ‘Red Vignette’ strategy showing a significant effect only in the ASD group. The responsiveness of HR and EDA to attention shifts and the changes in pupil size could serve as valuable physiological markers to monitor and guide these interventions. These findings further support evidence that VR has positive implications for helping those with ASD, allowing for more tailored personalized interventions with meaningful impact.

Machine learning methods have made huge improvements to medical imaging, particularly in the utilization of CT scans for intricate trauma cases. This paper initiates an exploratory journey, employing a hybrid quantum machine learning (HQML) approach to enhance the detection accuracy of abdominal trauma. Our study meticulously curates an extensive dataset from abdominal CT scans, navigating the inherent challenges presented by the voluminous nature and complex features of such medical data. We use quantum transfer learning techniques in a new way by combining the complex details of these scans into a computing framework that combines the powerful pattern recognition of quantum computing with the stability of classical machine learning. Our approach positions itself at the lead of medical innovation, poised to refine diagnostic precision and accelerate therapeutic protocols through an advanced analytical perspective. By looking at how well classical ResNet architectures and quantum models work, we find small differences in how well they work. This shows that quantum algorithms could decode medical images with a level of good accuracy. Our findings do not merely highlight the transformative potential of quantum computing in medical diagnostics but also pave the way for ensuing explorations in the domain of hybrid quantum-classical machine learning solutions.

This research delves into quantum machine learning (QML) in the context of computer vision analysis by exploring the progress made in quantum computing and its impact on machine learning applications such as managing datasets and improving large-scale data processing efficiency through QML techniques specialised for tasks like image segmentation and classification in computer vision projects, along with findings, from trials conducted using the EMNIST benchmark dataset.Our tests reached an accuracy level above 90% successfully categorising tasks, with precision. This study explores the uses of quantum machine learning (QML) in areas like identification medical scans and distant monitoring. It also delves into the existing constraints and hurdles linked to quantum computer technologies.

In recent years, significant advancements in computing power, data richness, algorithmic development, and the growing demand for applications have catalyzed the rapid emergence and proliferation of large language models (LLMs) across various scenarios. Concurrently, factors such as computing resource limitations, cost considerations, real-time application requirements, task-specific customization, and privacy concerns have also driven the development and deployment of small language models (SLMs). Unlike extensively researched and widely deployed LLMs in the cloud, the performance of SLM workloads and their resource impact on edge environments remain poorly understood. More detailed studies will have to be carried out to understand the advantages, constraints, performances, and resource consumption in different settings of the edge.This paper addresses this gap by comprehensively analyzing representative SLMs on edge platforms. Initially, we provide a summary of contemporary edge hardware and popular SLMs. Subsequently, we quantitatively evaluate several widely used SLMs, including TinyLlama, Phi-3, Llama-3, etc., on popular edge platforms such as Raspberry Pi, Nvidia Jetson Orin, and Mac mini. Our findings reveal that the interaction between different hardware and SLMs can significantly impact edge AI workloads while introducing non-negligible overhead. Our experiments demonstrate that variations in performance and resource usage might constrain the workload capabilities of specific models and their feasibility on edge platforms. Therefore, users must judiciously match appropriate hardware and models based on the requirements and characteristics of the edge environment to avoid performance bottlenecks and optimize the utility of edge computing capabilities.

Quantum Machine Learning (QML) promises the transformative potential in computer vision by utilizing quantum computing to facilitate faster high-dimensional data processing. In this paper, we will go through some of the recent works that employ QML for computer vision problems such as Image Segmentation, Classification, and Generation. Demonstrations aimed at showing where QML methods beat the state of art techniques in particular applications like facial recognition, medical imaging, and satellite imagery. QML aspires to make pathbreaking changes in a field limited by current hardware capabilities. This poster abstract summarizes the important studies, methodologies and findings to inform further research in this developing field.

Recently, with the advent of the Internet of everything and 5G network, the amount of data generated by various edge scenarios such as autonomous vehicles, smart industry, 4K/8K, virtual reality (VR), augmented reality (AR), etc., has greatly exploded. All these trends significantly brought real-time, hardware dependence, low power consumption, and security requirements to the facilities, and rapidly popularized edge computing. Meanwhile, artificial intelligence (AI) workloads also changed the computing paradigm from cloud services to mobile applications dramatically. Different from wide deployment and sufficient study of AI in the cloud or mobile platforms, AI workload performance and their resource impact on edges have not been well understood yet. There lacks an in-depth analysis and comparison of their advantages, limitations, performance, and resource consumptions in an edge environment. In this paper, we perform a comprehensive study of representative AI workloads on edge platforms. We first conduct a summary of modern edge hardware and popular AI workloads. Then we quantitatively evaluate three categories (i.e., classification, image-to-image, and segmentation) of the most popular and widely used AI applications in realistic edge environments based on Raspberry Pi, Nvidia TX2, etc. We find that interaction between hardware and neural network models incurs non-negligible impact and overhead on AI workloads at edges. Our experiments show that performance variation and difference in resource footprint limit availability of certain types of workloads and their algorithms for edge platforms, and users need to select appropriate workload, model, and algorithm based on requirements and characteristics of edge environments.

The secure authenticated key establishment between nodes in Wireless Sensor Networks (WSNs) has not been fully solved in the existing schemes. It’s a good idea to apply the Diffie-Hellman protocol to address it perfectly, but the existing authenticated Diffie-Hellman (ADH) protocols are not perfect because their authentication are partial or delayed. In this paper, we first present a concept of full authentication and propose a new fully authenticated Diffie-Hellman (FADH) prototype with light-certificate-based authentication. And then based on the theory of elliptic curve cryptography, we construct the TinyADH (Tiny Authenticated Diffie-Hellman) protocol with applying the FADH in WSNs. Compared with the existing similar solutions, TinyADH has lower communication overload, is easier to implement into existing standards, and more secure under equivalent computational complexity. The experimental results show that using this scheme for a successful key agreement between two nodes averagely takes about 54 seconds on TelosB. Moreover, the simulation results indicate that repeated key agreement can improve the secure connectivity rate. However, considering the cost performance ratio, it is advisable to take 2 runs of the negotiation.

Interstitial lung disease (ILD) causes pulmonary fibrosis. The correct classification of ILD plays a crucial role in the diagnosis and treatment process. In this research work, we propose a lung nodules recognition method based on a deep convolutional neural network (DCNN) and global features, which can be used for computer-aided diagnosis (CAD) of global features of lung nodules. Firstly, a DCNN is constructed based on the characteristics and complexity of lung computerized tomography (CT) images. Then we discussed the effects of different iterations on the recognition results and influence of different model structures on the global features of lung nodules. We also incorporated the improvement of convolution kernel size, feature dimension, and network depth. Thirdly, the effects of different pooling methods, activation functions and training algorithms we proposed has been analyzed to demonstrate the advantages of the new strategy. Finally, the experimental results verify the feasibility of the proposed DCNN for CAD of global features of lung nodules, and the evaluation shown that our proposed method could achieve an outstanding results compare to state-of-arts.

In this project, we present a real-time Internet of Things (loT)-based system to convert handwritten text into editable format by implementing Hand Gesture Recognition (HGR) with Raspberry Pi for classroom education. HGR is implemented using edge detection algorithm and it is used to reduce computational complexity and improve the efficiency of the system. Raspberry Pi is used to retrieve and perceive HGR and to build a smart classroom via loT technologies. Handwritten images are converted into editable format by using OpenCV and machine learning algorithms. In the text conversion, recognition of uppercase and lowercase alphabets, numbers, special characters, mathematical symbols, equations, and figures are included with recognition of word, lines, blocks, and paragraphs. With the help of Raspberry Pi and IoT technologies, students can access the editable format of lecture notes via a desktop application which helps students to edit and share notes and images according to their necessity. Implementation details and comprehensive evaluations of the system are summarized in the paper.

What students are supposed to learn from computing education is a fundamental question for curriculum design. Based on 1,462 articles we extracted from Web of Science, Eric, and Jstor, we performed a keyword analysis and case analysis so as to answer this question. The findings indicate that problem-solving, collaboration skills, and creativity are the common goals identified in computing education, while students' improvements of abstraction, inclusiveness, and self-efficacy are rarely studied by educators. By comparing this result to the learning objectives defined by curriculum standards, we identified a gap between the current common practices in school and the visions provided by the standards. We also summarized some best practices of improving students' multiple competences through computing education.

This research to practice full paper reports the authors' experience in teaching five computer science courses that were developed specifically for non-CS major students. These courses were taught in a span of four years in a public research institution in the South-eastern region of the United States. The investigators collected data from both the students and the instructors in the hope to seek empirical answers to the question of what non-CS major students are looking for in computer science classes. The answers to this question help shed the light on what educators should provide to the students in these courses. Based on the analysis of the data gathered , the investigators proposed a list of recommendations to help guide fellow educators in the development of non-CS major computer science courses.

In this Innovative Practice Full Paper, we propose to establish a community-based learning initiative, which integrates recruitment, advisement, and retention programs with hands-on activities that supplement core curriculum for cohorts of students in the STEM fields - from recruiting through graduation and placement into the workforce. This initiative is designed in three phases: the pilot phase, the alliance phase, and the backbone phase. This project is aimed at the pilot phase; however, its focus does not end when the piloting is over, but rather, it is intended to integrate with other pilot programs into building regional alliances. Through the regional alliances, this proposed pilot initiative is planned to become part of the national backbone program, which aims at producing well-prepared, underrepresented graduates into the STEM-related workforce. Because the percentage of women and minorities (African-American, Hispanics, and Native Americans) in STEM-related workforce is so small, the COESIND program is designed to broaden the participation of these groups. To do so, a community of the eight campuses of Chattahoochee Technical College, the five campuses of Georgia-Perimeter College, the Girls, Inc. of Greater Atlanta, Georgia, Young Women's Christian Association (YWCA) of Greater Atlanta, and three area high schools in Cobb County, all in the metropolitan area of Atlanta, Georgia, will participate in the COESIND program. A Kennesaw State University (KSU) CS mentorship program, The Girls, Inc.'s summer camps have consistently attracted several girls over the years (Google IgniteCS). These institutions, collectively, have a significantly large population of African-American (A-A), Hispanics, and Girls/Women. Faculty from the College of Computing and Software Engineering (CCSE) and College of Engineering at KSU, located in Kennesaw, Georgia, will lead the proposed program. KSU, which recently merged with Southern Polytechnic State University (SPSU), located in Marietta, Georgia, now has a student population of 5,000 in these two Colleges. Moreover, KSU was recently classified as an R2 institution. The University also offers an array of STEM programs and has been ranked 7th among institutions that offer online courses and, at the time the merger, SPSU was ranked among the top three producers of African-American engineering technology graduates in the nation. With its prior implementations of two NSF-LSAMP programs, among other undergraduate-level student training, KSU is qualified to lead such an initiative. Our focus on Computing and Engineering sub-disciplines, among the many STEM-fields, is intentional: To focus on programs and activities where computing and engineering technologies are deployed as tools to solving real-world problems. In this paper, we will describe the motivation of the project and the detailed implementation plan of the pilot phase. The assessment plan along with the measurable outcomes, are summarized in the paper as well.

In this poster, the investigators report their experience in teaching five computer science courses that were developed specifically for non-CS major students. These courses were taught in a four-year period in a public research institution in the South-eastern region of the United States. The whole study demonstrates an emerging demand for computing education from non-CS major students. The investigators collected data from both the students and the instructors in the hope to seek empirical answers to the questions of why there are such emerging demands for computing education, where are the demands coming from, and what non-CS major students are looking for in computer science classes. The answers to these questions can help shed the light on what educators should provide to the students in these courses. Based on the analysis of the data gathered, the investigators proposed a list of four recommendations to help guide fellow educators in the development of non-CS major computer science courses: (a) focus on high-level programming and modeling with low syntactic overhead; (b) broaden the breath of application demonstration to help non-CS major students to apply the knowledge to their specific domain; (c) introduction of computational tools; (d) hands on projects that match students' background and interests.

This paper studies feasibility and scalable computing processes for visualizing big high dimensional data in a 3 dimensional space by using dimension reduction techniques. More specifically, we propose an unsupervised approach to compute a measure that is called visualizability in a 3 dimensional space for a high dimensional data. This measure of visualizability is computed based on the comparison of the clustering structures of the data before and after dimension reduction. The computation of visualizability requires finding an optimal clustering structure for the given data sets. Therefore, we further implement a scalable approach based on K-Means algorithm for finding an optimal clustering structure for the given big data. Then we can reduce the volume of a given big data for dimension reduction and visualization by sampling the big data based on the discovered clustering structure of the data.

Internet of Things (IoT) is rapidly emerging as the next generation of communication infrastructure, where myriad of multi-scale sensors and devices are seamlessly blended for ubiquitous computing and communication. The rapid growth of IoT applications has increased the demand for experienced professionals in the area. Since few, if any, dedicated IoT courses are currently offered, most Science, Technology, Engineering, and Mathematics (STEM) students will have limited or no exposure to IoT development until after graduation and entrance into the workforce. Moreover, there is a little room for adding additional courses into existing STEM curriculum. Therefore, we propose to transform STEM core courses by integrating IoT-based learning framework into their corresponding lab projects. The design challenges of the new learning framework is summarized in the paper. Subsequently, we propose the effective learning approaches to address those challenges. Moreover, in this paper, we present a case study by incorporating IoT-based learning framework into a Software Engineering (SWE) embedded system analysis & design course. Specifically, we introduce a lab development kit composed of Raspberry Pi/Arduino boards and a set of sensors with Zigbee supporting to provide wireless communication in the class lab section. We adopt module design method to design the course labware. Well-developed modules are presented and one sample module is illustrated in the paper. The labware is evaluated through survey questions. The majority of the students provided positive feedback and enjoyed the IoT-based lab development kit.

In order to remove resource barriers and smooth the learning curve for education on big data analytics in STEM disciplines, we develop an portable open source labware that is called STEM-BD for promoting education on big data analytics. STEM-BD integrates the following four critical components, big data platform, big data sets, data analytics algorithms and hands-on lab exercises in a multi-dimensional and customizable way. In this paper, we provide a detailed description of the design goal of STEM-BD, its prototype, preliminary evaluation results, and future development.

Internet of Things (IoT) involves connecting physical objects to the Internet, to provide opportunities to build smart systems or applications by leveraging Radio-Frequency Identification (RFID), Near Field Communication (NFC), Wireless Sensor Network (WSN), and universal mobile accessibility advanced technologies. In this paper, we propose an Internet-of-Things Based Smart Resource Management System. To further prove the concept, we implement a case study of Intelligent Chair system, where the chairs are connected to the internet. To be specific, the Intelligent Chair system includes an assembled Arduino system, which is responsible for scanning user ID, obtaining chair occupancy status, and then sending that information to the cloud server. The collected data stored on the cloud can be retrieved at anytime anywhere, and can be displayed on an Android application with authorized user. Additionally, the analyzed data can be used in various commercial/educational systems such as students attendance checking, tutor time tracking management, and dynamic ticketing system. Finally, the energy consumption of the Intelligent Chair system is tested and analyzed in this work.

When people utilize social applications and services, their privacy suffers a potential serious threat. In this article, we present a novel, robust, and effective de-anonymization attack to mobility trace data and social data. First, we design a Unified Similarity (US) measurement, which takes account of local and global structural characteristics of data, information obtained from auxiliary data, and knowledge inherited from ongoing de-anonymization results. By analyzing the measurement on real datasets, we find that some data can potentially be de-anonymized accurately and the other can be de-anonymized in a coarse granularity. Utilizing this property, we present a US-based De-Anonymization (DA) framework, which iteratively de-anonymizes data with accuracy guarantee. Then, to de-anonymize large-scale data without knowledge of the overlap size between the anonymized data and the auxiliary data, we generalize DA to an Adaptive De-Anonymization (ADA) framework. By smartly working on two core matching subgraphs , ADA achieves high de-anonymization accuracy and reduces computational overhead. Finally, we examine the presented de-anonymization attack on three well-known mobility traces: St Andrews, Infocom06, and Smallblue, and three social datasets: ArnetMiner, Google+, and Facebook. The experimental results demonstrate that the presented de-anonymization framework is very effective and robust to noise. The source code and employed datasets are now publicly available at SecGraph [2015].

As an objective reality, the social behavior pattern of Primary Users (PUs) has significant impacts on the design and management of the secondary network. However, most of the existing works overlook this fact by simplifying the spectrum whitespace assumption. In this paper, we study the joint routing and time-domain scheduling problem for Cognitive Radio Networks (CRNs) by considering the social behaviors of PUs. Our main contributions consist of four aspects. First, we analyze the social pattern of PUs based on two practical data traces. According to the obtained social pattern, the available spectrum whitespace is derived for SUs. Subsequently, in terms of previous analysis, we propose a centralized joint routing and time-domain scheduling framework with global provable ε-optimality (ε ∊ [0,1]) by employing the branch-and-bound technique, where ε indicates the expected closeness of our solution to the optimum solution. The solution of this centralized algorithm can serve as a theoretical benchmark for developing future routing and scheduling algorithms for CRNs. Third, we design a distributed primary behavior-aware routing and scheduling algorithm with local performance guarantee, where the routing and scheduling fairness, the available bandwidth, the potential interference, etc. are taken into account. Finally, simulation results confirm our assertion that primary behaviors have significant impacts on the spectrum whitespace, and demonstrate that primary-behavior-aware joint routing and scheduling design can utilize spectrum whitespace efficiently.

Social networks are important mediums for communication,information dissemination, and influence spreading. Most of existing works focus on understanding the characteristics of social networks or spreading information through the 'word of mouth' effect of social networks. However, motivated by applications of alleviating social problems, such as drinking, smoking, addicting to gaming, and influence spreading problems, such as promoting new products, we propose a new optimization problem named the Minimum-sized Influential Node Set (MINS) selection problem, which is to identify the minimum-sized set of influential nodes, such that every node in the network could be influenced by these selected nodes no less than a threshold. Our contributions are threefold. First, we prove that, under the independent cascade model, MINS is NP-hard. Subsequently, we present a greedy approximation algorithm to address the MINS selection problem. Moreover, the performance ratio of the greedy algorithm is analyzed. Finally, to validate the proposed greedy algorithm, extensive experiments and simulations are conducted both on real world coauthor data sets and random graphs.

As a promising communication paradigm, Cognitive Radio Networks (CRNs) have paved a road for Secondary Users (SUs) to opportunistically exploit unused licensed spectrum without causing unacceptable interference to Primary Users (PUs). In this paper, we study the distributed data collection problem for asynchronous CRNs, which has not been addressed before. We study the Proper Carrier-sensing Range (PCR) for SUs. By working with this PCR, an SU can successfully conduct data transmission without disturbing the activities of PUs and other SUs. Subsequently, based on the PCR, we propose an Asynchronous Distributed Data Collection (ADDC) algorithm with fairness consideration for CRNs. ADDC collects a snapshot of data to the base station in a distributed manner without the time synchronization requirement. The algorithm is scalable and more practical compared with centralized and synchronized algorithms. Through comprehensive theoretical analysis, we show that ADDC is order-optimal in terms of delay and capacity, as long as an SU has a positive probability to access the spectrum. Furthermore, we extend ADDC to deal with the continuous data collection issue, and analyze the delay and capacity performances of ADDC for continuous data collection, which are also proven to be order-optimal. Finally, extensive simulation results indicate that ADDC can effectively accomplish a data collection task and significantly reduce data collection delay.

Data Gathering is a fundamental task in Wireless Sensor Networks (WSNs). Data gathering trees capable of performing aggregation operations are also referred to as Data Aggregation Trees (DATs). Currently, most of the existing works focus on constructing DATs according to different user requirements under the Deterministic Network Model (DNM). However, due to the existence of many probabilistic lossy links in WSNs, it is more practical to obtain a DAT under the realistic Probabilistic Network Model (PNM). Moreover, the load-balance factor is neglected when constructing DATs in current literatures. Therefore, in this paper, we focus on constructing a Load-Balanced Data Aggregation Tree (LBDAT) under the PNM. More specifically, three problems are investigated, namely, the Load-Balanced Maximal Independent Set (LBMIS) problem, the Connected Maximal Independent Set (CMIS) problem, and the LBDAT construction problem. LBMIS and CMIS are well-known NP-hard problems and LBDAT is an NP-complete problem. Consequently, approximation algorithms and comprehensive theoretical analysis of the approximation factors are presented in the paper. Finally, our simulation results show that the proposed algorithms outperform the existing state-of-the-art approaches significantly.

ABSTRACT Inspired by the backbone concept in wired networks, a virtual backbone is expected to bring substantial benefits to routing in wireless sensor networks (WSNs). A connected dominating set (CDS) is used as a virtual backbone for efficient routing and broadcasting in WSNs. Most existing works focus on constructing a minimum CDS, ak‐connectm‐dominating CDS, a minimum routing cost CDS, or a bounded‐diameter CDS. However, theload‐balancefactor is not considered for CDSs in WSNs. In this paper, a greedy‐based approximation algorithm is proposed to construct load‐balanced CDS in a WSN. More importantly, we propose a new problem: the Load‐balanced Allocate Dominatee problem. Consequently, we propose an optimal centralized algorithm and an efficient probability‐based distributed algorithm to solve the Load‐balanced Allocate Dominatee problem. For a given CDS, the upper and lower bounds of the performance ratio of the distributed algorithm are analyzed in the paper. Through extensive simulations, we demonstrate that our proposed methods extend network lifetime by up to 80% compared with the most recently published CDS construction algorithm. Copyright © 2012 John Wiley & Sons, Ltd.

ABSTRACT Query scheduling as one of the most important technologies used in query processing has been widely studied recently. In this paper, we investigate the Minimum Latency Multi‐Regional Query Scheduling (ML‐MRQS) problem in wireless Sensor Networks (WSNs), which aims to generate a scheduling plan with minimum latency under a more practical query model called Multi‐Regional Query (MRQ). An MRQ targets at interested data from multiple regions of a WSN, where each region is a subarea. Because the ML‐MRQS problem is NP‐hard, we propose a heuristic scheduling algorithm Multi‐Regional Query Scheduling Algorithm (MRQSA) to solve this problem. Theoretical analysis shows that the latency of MRQSA is upper bounded by 23A + B + Cfor an MRQ withmquery regions , where is the maximum latency for non‐overlapped regions, is the maximum latency for overlapped regions, and is the accumulated latency for data transmission from the accessing nodes to the sink. Simulation results show that MRQSA reduces latency by 42.7%to 51.63%with respect to different number of query regions, network density, region size, and interference/transmission range compared with C‐DCQS, while guaranteeing energy efficiency. Copyright © 2012 John Wiley & Sons, Ltd.

A Connected Dominating Set (CDS) is used as a Virtual Backbone (VB) for efficient routing and broadcasting in Wireless Sensor Networks (WSNs). Currently, almost all existing works focus on constructing Minimum-sized CDS under the Deterministic Network Model (DNM). However, due to the existence of many probabilistic lossy links in WSNs, it is more practical to obtain a VB under the realistic Probabilistic Network Model (PNM). Moreover, load-balance factor cannot be neglected when constructing a VB to prolong network lifetime. Hence, in this paper, we propose a Multi-Objective Genetic Algorithm (MOGA) to construct a Load-Balanced Virtual Backbone under PNM (LBVBP). Through simulations, we demonstrate that our proposed methods extend network lifetime by 65% on average compared with the existing state-of-the-art approaches.

Social networks are important mediums for spreading information, ideas, and influences among individuals. Most of existing research work focus on understanding the characteristics of social networks, investigating spreading information through the “word of mouth” effect of social networks, or exploring social influences among individuals and groups. However, most of existing work ignore negative influences among individuals or groups. Motivated by alleviating social problems, such as drinking, smoking, gambling, and influence spreading problems (e.g., promoting new products), we take both positive and negative influences into consideration and propose a new optimization problem, named the Minimumsized Positive Influential Node Set (MPINS) selection problem, to identify the minimum set of influential nodes, such that every node in the network can be positively influenced by these selected nodes no less than a threshold θ. Our contributions are threefold. First, we propose a new optimization problem MPINS, which is investigated under the independent cascade model considering both positive and negative influences. Moreover, we claim that MPIMS is NP-hard. Subsequently, we present a greedy approximation algorithm to address the MPINS selection problem. Finally, to validate the proposed greedy algorithm, extensive simulations are conducted on random Graphs representing small and large size networks.

Approximate string matching using the k-difference technique has been widely applied to many fields such as pattern recognition and computational biology. Data dependency exists in the traditional sequential algorithm. Therefore, it is hard to design a parallel algorithm for approximate string matching with k differences. This paper presents a technique to eliminate data dependency. Based on this technique, this paper also presents a parallel algorithm which can calculate the elements in the same row of the edit distance matrix in parallel by eliminating data dependency. The algorithm has high parallelism, but requires synchronization. To validate the proposed algorithm, it is implemented on GPU and multiple-core CPUs. Moreover, the CUDA optimization techniques are also presented in the paper. Finally, experimental results show that, compared with the traditional sequential algorithm on CPU with twenty-four cores, the proposed parallel algorithm achieves speedup of 7-42 on GPU.

Data collection is a common operation of wireless sensor networks (WSNs). The performance of data collection can be measured by its achievable network capacity. However, most existing works focus on the network capacity of unicast, multicast or/and broadcast. In this article, we study the snapshot/continuous data collection (SDC/CDC) problem under the physical interference model for randomly deployed dense WSNs. For SDC, we propose a Cell-Based Path Scheduling (CBPS) algorithm based on network partitioning. Theoretical analysis shows that its achievable network capacity is order-optimal. For CDC, a novel Segment-Based Pipeline Scheduling (SBPS) algorithm is proposed which combines the pipeline technique and the compressive data gathering technique. Theoretical analysis shows that SBPS significantly speeds up the CDC process and achieves a high network capacity.

As a promising communication paradigm, Cognitive Radio Networks (CRNs) have paved a road for Secondary Users (SUs) to opportunistically exploit unused licensed spectrum without causing unacceptable interference to Primary Users (PUs). In this paper, we study the distributed data collection problem for asynchronous CRNs, which has not been addressed before. First, we study the Proper Carrier-sensing Range (PCR) for SUs. By working with this PCR, an SU can successfully conduct data transmission without disturbing the activities of PUs and other SUs. Subsequently, based on the PCR, we propose an Asynchronous Distributed Data Collection (ADDC) algorithm with fairness consideration for CRNs. ADDC collects data of a snapshot to the base station in a distributed manner without any time synchronization requirement. The algorithm is scalable and more practical compared with centralized and synchronized algorithms. Through comprehensive theoretical analysis, we show that ADDC is order-optimal in terms of delay and capacity, as long as an SU has a positive probability to access the spectrum. Finally, extensive simulation results indicate that ADDC can effectively finish a data collection task and significantly reduce data collection delay.

Solving the eigenvalues of matrices is an open problem which is often related to scientific computation. With the increasing of the order of matrices, traditional sequential algorithms are unable to meet the needs for the calculation time. Although people can use cluster systems in a short time to solve the eigenvalues of large-scale matrices, it will bring an increase in equipment costs and power consumption. This paper proposes a parallel algorithm named Jacobi on gpu which is implemented by CUDA (Computer Unified Device Architecture) on GPU (Graphic Process Unit) to solve the eigenvalues of symmetric matrices. In our experimental environment, we have Intel Core i5-760 quad-core CPU, NVIDIA GeForce GTX460 card, and Win7 64-bit operating system. When the size of matrix is 10240×10240, the number of iterations is 10000 times, the speedup ratio is 13.71. As the size of matrices increase, the speedup ratio increases correspondingly. Moreover, as the number of iterations increases, the speedup ratio is very stable. When the size of matrix is 8192×8192, the number of iterations are 1000, 2000, 4000, 8000 and 16000 respectively, the standard deviation of the speedup ratio is 0.1161. The experimental results show that the Jacobi on gpu algorithm can save more running time than traditional sequential algorithms and the speedup ratio is 3.02~13.71. Therefore, the computing time of traditional sequential algorithms to solve the eigenvalues of matrices is reduced significantly.

Query scheduling as one of the most important technologies used in query processing has been widely studied recently. Unfortunately, to the best of our knowledge, no previous work focuses on the Minimum Latency Multi-Regional Query Scheduling (ML-MRQS) problem. In this paper, we investigate the ML-MRQS problem in Wireless Sensor Networks (WSNs), which aims to generate a scheduling plan with minimum latency for a more practical query model called Multi-Regional Query (MRQ). A MRQ targets at user interested data from multiple region-sofa WSN, where each region is a subarea of the WSN. We claim that the ML-MRQS problem is NP-hard. Therefore, we propose a heuristic scheduling algorithm Multi-Regional Query Scheduling Algorithm (MRQSA) to solve this problem. Theoretical analysis shows that the latency of MRQSA is upper bounded by 23A + B + C for a MRQ with m query regions R1, R2…, Rm, where A = maxi=1mDileft, B = maxi=1m{(23Di+5Δ+21)ki}, C = Σi=1mHi+5Δ−m+17, m is the number of regions, Δ is the maximum node degree in the WSN, A is the diameter of Ri, kiis the maximum overlapped degree of sensor nodes in TU, Hi represents the distance of Riwith respect to the sink, and Dileftis the diameter of the non-overlapped part of Ri. Extensive simulations are conducted to verify the performance of our algorithm, which show that MRQSA significantly reduces the query latency when compared with the most recently published multi-query scheduling algorithm.

Spatial Audio Coding (SAC) is an emerging technology with a distinguishing feature of delivering good even excellent audio quality at monotonic or stereo bitrate of conventional perceptual transform coders. By a systematic exploitation of spatial hearing, Binaural Cue Coding illustrates the power and potentials of SAC in the future for intelligent multimedia services. MPEG Surround, receiving cumulative efforts from industry and academy, strives to build a SAC system with great versatility and high quality. The initial test results of MPEG Surround show its performance advantage over conventional state-of-the-art coders in a wide range of coding configurations.

Recently, control and coordination of a set of autonomous mobile robots has been paid a lot of attentions, because the cooperation of simple robots offers several advantages, such as redundancy and flexibility, and allows performing hard tasks that could be impossible for one single robot. There are a lot of interesting applications of multiple robots, such as satellite exploration and surveillance missions. So far, there are many papers working on the coordination of mobile robots. The characteristic of simplicity of mobile robots brings potential wide applications; however this characteristic also lead to crash with higher probability during cooperation, especially in harsh environment. Surprisingly, only few researches consider the fault tolerance of mobile robots, especially for dynamic coordination application - robot flocking. In this paper, we summarize the existed flocking algorithms and discuss their characteristics. Then we briefly described our fault tolerant flocking algorithms in different models. Finally we proposed the potential future research directions for dynamic flocking of a group of mobile robots. In all, this work can provide a good reference for the researchers working on dynamic cooperation of agents in distributed system.

With the ever-increasing wireless multicast data applications recently, considerable efforts have focused on the large scale heterogeneous wireless multicast, especially those with large propagation delays, which means the feedbacks arriving at the source node are somewhat outdated and harmful to the control actions. To attack the above problem, this paper describes a novel, autonomous, and predictive wireless multicast flow control scheme, the so-called proportional, integrative plus neural network (PINN) predictive technique, which includes two components: the PI flow controller located at the wireless multicast source has explicit rate algorithm to regulate the transmission rate; and the neural network part located at the middle branch node predicts the available buffer occupancy for those longer delay receivers. The ultimate sending rate of the multicast source is the expected receiving rates computed by PI controller based on the consolidated feedback information, and it can be accommodated by its participating branches. This network-assisted property is different from the existing control schemes in that neural network controller can predict the buffer occupancy caused by those long delay receivers, which probably cause irresponsiveness of a wireless multicast flow. This active scheme makes the control more responsive to the network status, therefore, the rate adaptation can be in a timely manner for the sender to react to network congestion quick. We analyze the theoretical aspects of the proposed algorithm, show how the control mechanism can be used to design a controller to support wireless multi-rate multicast transmission based on feedback of explicit rates.