International Journal of Sensor Networks (41 papers in press)

Regular Issues

• Recent Advances in Wireless Sensor Networks with Environmental Energy Harvesting
  by Lei Shu, Wanjiun Liao, Jaime Lloret, Lei Wang 
  Keywords: .
  
  
• Collaborative Filtering Recommendation Algorithm Based on Deep Neural Network Fusion  
  by Juan Fang, Baocai Li, Mingxia Gao 
  Abstract: In order to accurately obtain potential features and improve the recommendation performance of the collaborative ?ltering algorithm, this paper puts forward a Collaborative Filtering Recommendation Algorithm Based on Deep Neural Network Fusion(CF-DNNF). CF-DNNF makes the best of the implicit attributes of data, where the text attributes and the other attributes are extracted from the data through the LSTM network and the deep neural network, respectively, so as to obtain the feature matrix that contains the user and item attribute information. DBN uses the feature matrix and outputs the probability. Besides, this paper initially discusses an Interpretable Collaborative Filtering Recommendation Algorithm Based on Deep Neural Network Fusion(ICF-DNNF). The paper compares the CF-DNNF algorithm with PMF, SVD, and RBM-CF algorithms on the MovieLens dataset and the Amazon product dataset. Results indicate that the RMSE of CF-DNNF is improved by 2.015%, and the MAE is improved by 2.222%.
  Keywords: recommendation; algorithm; feature; interpretable; fusion; neural network; collaborative ?ltering; deep learning; MovieLens; RBM; CFDNNF.
  
  
• A Data Dissemination Scheme for a Wireless Nanosensor Network towards IoNT  
  by Mohamed Mostafa A. Azim 
  Abstract: Nanotechnology has gained significant importance in many fields, especially biological, industrial, military, and environmental sectors. Rapid advancements in nanotechnology have led to the development of wireless nanosensor networks (WNSNs). Wireless nanosensor networks consist of nano-sized communication devices, including nano-nodes, nanorouters, nano-micro interfaces, and gateways. One of the most important goals for wireless nanosensor networks is data collection and delivery. However, data collection and delivery face two main challenges: resource limitation and dynamic channel state. Nanoscale devices have low limited capacity in terms of storage, computation, energy, and communication. The dynamic channel state occurs due to nanorouter links that are sensitive to molecular absorption. Most researchers focus on addressing resource limitation constraints and ignore the dynamic channel state. Therefore, in this paper, we design a fine-grained, lightweight, and energy-efficient end-to-end data dissemination scheme for WNSN that takes into consideration the dynamic channel state problem and energy efficiency. The proposed scheme divides the data dissemination process into two phases: a setup phase and a data collection phase. The setup phase aims to discover the nanonetwork topology and assign a forwarder for both the nanorouter and nanosensor nodes, while the data collection phase focuses on aggregating, disseminating data, and adopting a dynamic channel state. To validate the performance of the proposed scheme, a network simulator was developed using MATLAB to compare our proposed scheme with one of the benchmarking schemes known as the TEForward scheme. The TEForward scheme is considered to be the first solution that addresses the dynamic channel state as well as resource constraints. Simulation results showed that the proposed scheme outperforms TEForward in terms of energy consumption and packet delivery ratio. Moreover, we investigate the effect of increasing the number of gateways in the network on energy consumption. Our simulations indicate that a small-sized network (a networ with a small number of nodes) using one gateway consumes less energy than those using more than one. For large-sized networks (networks with a large number of nodes) using more than one gateway is more appropriate from the energy consumption point of view than depending on only one gateway. Hence, we conclude that the number of gateways in the network has an effect on energy consumption.
  Keywords: Wireless Nanosensor Network; WNSN; Internet of Things; Internet of NanornThings; IoT; IoNT; gateway; nanosensor; nanorouter.
  
  
• Anisotropic Diffusion Based on FermiDirac Distribution Function and its Application in the ShackHartman Wavefront Sensor  
  by Yanyan Zhang, Chengsheng Pan, Luyao Wang, Suting Chen 
  Abstract: In this study, the anisotropic diffusion technique is applied to estimate spots in the noise signals of the ShackHartmann wavefront sensor. Based on the analysis of the classical anisotropic diffusion function and on an improved algorithm, a diffusion function is proposed based on the FermiDirac distribution. It is proved mathematically that the new function has a higher convergence speed and a better performance. Monte Carlo simulations are used to verify the applicability of the new function subject to the noise limit and signal level. The simulation and experimental results show that the anisotropic diffusion algorithm can effectively filter out the noise. The integrity of the spots can be maintained, and the centroid detection accuracy and signal-to-noise ratio are also improved.
  Keywords: Shack–Hartmann wavefront sensor? noise? anisotropy? diffusion function.
  
  
• Distributed Mobile Wireless Sensor Node Localization using RSSI-aided Monte Carlo Method  
  by Timoteo Cayetano-Antonio, M. Mauricio Lara, Aldo G. Orozco-Lugo 
  Abstract: Localization, also known as positioning, is a key issue in mobile wireless sensor networks. There are different positioning algorithms for low-cost sensor nodes in the literature; but most of them are focused on the basic idealized scenario of the free-space radio propagation model. In this paper, a new algorithm is proposed based on Monte Carlo localization for positioning mobile wireless sensor nodes in the more challenging scenario of the shadowing radio path loss propagation model. The received signal strength indicator (RSSI) is integrated into the Monte Carlo algorithm as an undemanding method of distance estimation. Besides, multilateration based on the concept of radical axes and the use of Least Squares is also proposed to increase the number of localized nodes. The key difference with previous works comes from an extension of the concept of neighborhood of nodes which is more suitable for shadowing channels. The proposed algorithms show an improvement in the localization precision compared with other works in the literature.
  Keywords: Mobile Wireless Sensor Node Localization; Shadowing; Monte Carlo Localization; RSSI Localization; Positioning.
  
  
• IoTtalk Experience on Building Commercial IoT/AI Applications  
  by Yi-Bing Lin, Tai-Hsiang Yen 
  Abstract: Many smart applications have been developed using the Internet of Things (IoT) technology. Unfortunately, some of them are not sustainable and cannot be commercialized. This paper describes our observation on mistakes made in IoT application development, and introduces an IoT application development platform called IoTtalk that supports sustainable applications. Specifically, to achieve the above goal, we introduce two powerful mechanisms of IoTtalk: MapTalk and the IoTtalk control board. In the summary, we show how a 2019 Novel Coronavirus (2019-nCoV) monitoring system with privacy can be conveniently built in IoTtalk.
  Keywords: 2019 Novel Coronavirus (2019-nCoV); Internet of Things; Smart City; sustainable applications.
  
  
• Wind Turbine Blades Icing Failure Prognosis Based on Balanced Data and Improved Entropy  
  by Cheng Peng, Qing Chen, Xiaohong Zhou, Songsong Wang, Zhaohui Tang 
  Abstract: To improve the accuracy of icing failure prediction, which is often limited due to unbalanced condition data, a novel balancing algorithm based on boundary division synthetic minority oversampling technology (BD-SMOTE) and a method for predicting the icing failure of wind turbine blades in the short term based on multiple neural network combination are presented. First, the original data set obtained by sensors is balanced by BD-SMOTE. Then, the key features are extracted by multivariate and multiscale entropy based on a continuous smooth coarse (CSMMSE) algorithm, and the values of three kinds of features in the near future are predicted by the Elman neural network (ENN). Finally, a back-propagation (BP) neural network is adopted to predict the icing failure of wind turbine blades. Compared with the results of other methods, the prediction deviation of the ENN is smaller; the prediction results demonstrated the effectiveness and superiority of the proposed method.
  Keywords: sensors; icing failure; BD-SMOTE; boundary division synthetic minority oversampling technology; CSMMSE; multivariate and multiscale entropy based on a continuous smooth coarse; ENN; Elman neural network; BP; back propagation.
  
  
• Automated Real-Time Anomaly Detection of Temperature Sensors through Machine-Learning  
  by Debanjana Nayak, Harry Perros 
  Abstract: Fast identification of faulty sensors is necessary for guaranteeing their robust functions in diverse applications ranging from extreme weather prediction to energy saving to healthcare. We present an automated machine-learning based framework that can detect anomalies of temperature sensor data in real-time. We adopted a purely temporal approach that utilizes a univariate time-series (UTS) generated by a single sensor. The framework divides the UTS into subsequences, models each subsequence stochastically as an autoregressive function, and finally mines the function parameters with a One-Class Support Vector Machine (OC-SVM) that classifies any outlier as an anomaly. Extensive experimentation showed that the framework identifies both normal and anomalous data correctly with high degrees of accuracy.
  Keywords: Univariate Time-Series; Anomaly Detection; Temperature Sensors; One-Class Support Vector Machines; Autoregression.
  
  
• Throughput and delay optimization for NOMA systems with energy harvesting  
  by Nadhir Ben Halima, Hatem Boujemaa 
  Abstract: In this paper, we analyze and optimize the total throughput and average delay in Non Orthogonal Multiple Access (NOMA) with energy harvesting. The source harvest energy from Radio Frequency (RF) signals received from another node $A$. Node A can be any node transmitting RF signals. The frame with duration $T$ contains two time slots. The first slot has duration $\\delta T$ and is dedicated to energy harvesting where $0< \\delta
  Keywords: NOMA; energy harvesting; Rayleigh channel; optimal power allocation; optimal harvesting duration; packets\' waiting time; delay analysis.
  DOI: 10.1504/IJSNET.2020.10030152
   
  
• Improving Activity Recognition for Multiple-Node Wireless Sensor Network System Based on Compressed Sensing  
  by Duo Yang, Jiangtao Huangfu 
  Abstract: The article proposes a multiple-node activity recognition wireless sensor network system based on compressed sensing (CS), where effective use of transmission resources, the real-time performance, low data recovery errors, and high activity recognition accuracy are achieved. A WSN system model for data compression-transmission-recovery (CTR-WSN) in the case of data loss is built. It performs different degrees of compression according to different types of activities characteristics. With consideration of data loss, transmission resources are effectively used, and the most time-saving data recovery methods are evaluated based on SP/SAMP algorithms, and the error range of different activities are all basically controlled within
  Keywords: activity recognition; compressed sensing; data loss; multiple-node; wireless sensor network.
  
  
• A Dynamic Data Driven Indoor Localization Framework based on Ultra High Frequency Passive RFID System  
  by Sara Masoud, Young-Jun Son, Chieri Kubota, Russell Tronstad 
  Abstract: Better monitoring of workers and the materials flow within a production system can potentially enhance any facilitys productivity and efficiency. This paper proposes a data driven framework to affordably localize indoor workers and materials using a passive Radio Frequency Identification (RFID) system in large scale. Here, indoor wireless sensor networks are developed via passive Ultra-High Frequency (UHF) tags, where Received Signal Strength Indicator (RSSI) is measured by different access points (APs) to generate a fingerprinting database. Then, this database not only translates the signal strength reported by APs to distance through regression models but also helps to localize each tag utilizing our proposed k-nearest neighbors (KNN) algorithm. Our improved KNN algorithm dynamically defines different neighbourhoods, in terms of size and topology considering environment status. Results from multiple experiments under different scenarios reveal that our proposed methods can detect and localize objects with an error as low as 0.36 m.
  Keywords: RFID; radio frequency identification; Passive UHF RFID; WSNs; wireless sensor networks; RFID–sensor networks; distance estimation; localization; KNN; Dynamic Data Driven; Machine learning; Statistical Modeling.
  
  
• A Contactless Sensing System for Indoor Fall Recognition Based on Channel State Information  
  by Wei Zhuang, Yixian Shen, Jiefeng Zhang, Chunming Gao, Yi Chen, Dong Dai 
  Abstract: This paper introduces the designs and the implementation of a non-invasive indoor fall recognition system based on Channel State Information (CSI) in the Wi-Fi physical layer. We use a wireless router and a laptop computer equipped with an Intel Wi-Fi Link 5300 network card (802.11n) to setup a hardware platform. The platform receives and stores CSI data under various circumstances when a person in the Wi-Fi covered area stands up, sits down, walks, and falls. The CSI data are then processed and analyzed using MATLAB tools. Feature variables such as signal offset strength, period of motion, normalized standard deviation, median absolute deviation, interquartile range, and signal entropy are examined and best feature variables are chosen. Finally, cross validation algorithm and support vector machine are used to establish the pattern recognition model. We tested the system in a laboratory environment and the experimental results showed that the fall incidents were effectively differentiated from other movements.
  Keywords: CSI; movement recognition; fall detection; support vector machine.
  
  
• A Brain-Computer Interface System for Smart Home Control Based on Single Trial Motor Imagery EEG  
  by Wei Zhuang, Yixian Shen, Lu Li, Chunming Gao, Dong Dai 
  Abstract: With the development of science and technology in recent years, many researches on brain signal recognition and brain-computer interface control have made great progress. By analysing electroencephalogram (EEG), a specific brain activity can be detected and the signal can be used to control smart devices and help people to complete many difficult and complicated tasks, especially for people with disabilities. This paper presents the design and implementation of a novel brain-computer interface system for smart home control using single trial motor imagery EEG. The system adopts STM32 MCU (Microcontroller Unit) and TGAM (ThinkGear Asic Module) modules to realize the acquisition and recognition of EEG signals. It can transfer the signals to portable devices through low power consumption Bluetooth modules. Three main EEG features including Alpha, Beta, and Gamma waves are discussed. It is tested during simple actions such as blinking in various situations. The experimental results show that the implemented system is suitable for extracting specific EEG signals to control smart home devices.
  Keywords: EEG recognition; STM32 embedded control; TGAM; smart home control.
  
  
• Triplet Erasing-Based Data Augmentation for Person Re-identification  
  by Wei Sun, Xu Zhang, Xiaorui Zhang, Guoce Zhang, Nannan Ge 
  Abstract: Occlusion is a fundamental yet challenging problem in person re-identification task. Previous work like Random Erasing randomly selects a rectangle region in an image and erases its pixels with random values, which does not consider the correlation between images when triplet loss is employed. To address the problem, we propose an end-to-end approach called Triplet Erasing-based data augmentation for person re-identification. We apply this approach to train a CNN network with two branches. Local distance branch determines the location of the part that needs to be erased in the image, then Triplet Erasing branch erases a rectangle region in the determined part. By generating a variety of occlusion samples, Triplet Erasing improves the robustness of the model to occlusion. Triplet Erasing can increase the distance between the positive sample pairs and decrease the distance between the negative sample pairs, which improves the generalization ability of the network.
  Keywords: deep learning; person re-identification; triplet loss; occlusion.
  
  
• A multi-objective emergency vehicle scheduling optimization model  
  by Jiao Yao 
  Abstract: Various types of emergencies occur frequently, which may cause casualties and huge economic losses to certain extent. How to optimize the dispatch of emergency rescue vehicles and improve the rescue efficiency quickly and effectively after the occurrence of emergency incidents is of great significance. Based on the analysis of the influencing factors of emergency vehicle scheduling, to overcome its unreasonable problems, the emergency vehicle scheduling optimization model was firstly established, oriented to achieve three objects, minimizing total travel time, total cost of dispatching travel, and maximizing the path reliability. Moreover? model was solved by genetic algorithm based on non-dominated sorting with elite strategy (NSGA-II algorithm). Finally, the local road network in Huangpu District of Shanghai was taken as an example to verify the model in this study. The results show that the optimal solution of the model can satisfy the optimal total travel time, total travel cost of the scheduling scheme and the path reliability can also achieve requirements, which means good validity and practicability.
  Keywords: Emergency vehicle; multi-objective planning; NSGA-II algorithm; path reliability; effectiveness coefficient comprehensive evaluation method.
  
  
• Road Traffic Optimization based on a Learning Approach  
  by Ahmed Mejdoubi, Hacene Fouchal, Ouadoudi Zytoune, Mohamed Ouadou 
  Abstract: Road traffic management is an important issue for authorities, road operators and road users. Researchers from optimisation area have worked for many years in order to propose smart solutions. Many of them have been adopted by road operators and applied on road signaling or on traffic light management. During the last decade, Cooperative Intelligent Transport Systems (C-ITS) have emerged and are being deployed. They permit on one hand to increase the number of connected vehicles and on another hand to deploy RSU (Road Side Units) along roads and unctions. RSUs together with central road operator servers compose the road infrastructure used for Vehicle to Infrastructure (V2I) or Infrastructure to Vehicle (I2V). RSUs aim is to disseminate continuously messages about road safety and trafic flows to vehicles. C-ITS have a high dissemination potential about road traffic informations. Vehicles and RSUs are important actors in C-ITS. They are able to measure, to compute, and to disseminate. In this study, we add to these actors the ability to learn mainly in the purpose to provide the optimal path in terms of travel time for vehicles seeking to reach their destination The solution is based on a reinforcement learning technique, in particular Q-learning, that is used to learn the best action to take into account according to various situations. The transit delay from one location to another is used to determinate the rewards. The simulation results confirm that the proposed Q-learning approach outperforms the existing greedy algorithm with better performances in terms of transit delay.
  Keywords: C-ITS ; VANETs ; Reinforcement learning ; Distributed traffic management ; Travel time.
  
  
• LAVR: Link failure Avoidance and Void Recovery Routing Protocol for Underwater Sensor Networks  
  by Zinal Patel, Keyur Rana 
  Abstract: Recently, Underwater Wireless Sensor Network (UWSN) has paved the way for a new era of exploring ocean realm. The researchers are attracted towards UWSN due to the envisioned landscape of underwater applications including oceanographic exploration, assisted navigation, disaster prevention, underwater pollution monitoring, tactical surveillance, mine reconnaissance, etc. These applications can be made viable through underwater communication by using acoustic waves as a communication medium since radio waves propagate at extra low frequencies in water and optical waves suffer heavily due to scattering. The change in communication medium gives rise to several design challenges like limited bandwidth, high bit error rate, large and variable propagation delay, frequent loss of connectivity, multi-path scattering and fading, etc. which leads to the need of designing efficient and reliable routing protocols for UWSN. Reliable data delivery in UWSN is always a tempting issue to be addressed, which is jeopardized by link failures and communication void problem. Under this context, Link failure Avoidance and Void Recovery (LAVR) routing protocol is proposed to achieve reliability. Link quality is computed using Triangle Metric and routing is done on the basis of link quality to avoid link failure. A void node recovery procedure is proposed to displace the void node to a new depth to increase the packet delivery ratio. Simulations are performed in Aqua-Sim simulator and the performance of LAVR is evaluated by comparing it with existing protocols. It is observed that the proposed protocol outperforms other protocols in terms of packet delivery ratio and network lifetime.
  Keywords: Communication Void; Link Failure; Routing Protocol; Underwater Wireless Sensor Networks; Void Recovery.
  
  
• Optimized Joint Resource Allocation for NOMA MIMO-basedWireless Powered Sensor Networks  
  by Qiang Wang, Hai-Lin Liu 
  Abstract: Wireless powered sensor network has captured a lot of attentions due to its higher lifetime compared with the battery powered sensor network. The multiple-input multiple-output(MIMO) can be used to improve the wireless energy transfer efficiency by beamforming. In this paper, we investigate the resource allocation in the non-orthogonal multiple access (NOMA)MIMO-based wireless powered sensor networks in terms of the wireless energy transfer time allocated and power allocation to maximize the system throughput. The joint resource allocation is\r\nformulated as a nonconvex optimization problem which is difficult to solve due to its high computational complexity. To reduce the complexity, the original optimization problem is decomposed into two optimization subproblems, namely the time allocation subproblem and power allocation subproblem. Specifically, the particle swarm optimisation (PSO) is adopted to solve the time allocation subproblem. When the time allocation is fixed, the power allocation subproblem is still a nonconvex optimization problem. The DC (difference of two convex functions) programming method is adopted to solve it. We first transform the objective function as a difference of two convex functions and then the objective function is approximated as a convex function by the first order Taylor expansion. The Lagrangian dual method is used to solve the\r\napproximated convex optimization problem. Simulation results illustrate that the proposed joint resource allocation scheme can significantly improve the system throughput.
  Keywords: Resource allocation; MIMO; Wireless Sensor Networks; Evolutionary algorithm.
  
  
• Energy Balanced, Delay Aware Multi-Path Routing using Particle Swarm Optimization in Wireless Sensor Networks  
  by Priti Maratha, Kapil Gupta, Pratyay Kuila 
  Abstract: World-wide use of wireless sensor networks has urged the need for energy-efficient, distributed routing algorithms. The interest of researchers from the past decade is in energy-efficient routing. In this paper, sequential quadratic programming (SQP) based multi-path routing formulation focusing on improving lifetime and delay is represented, namely EBDA-DEFL. This SQP based formulation is solved using the optimization tool after that same formulation is solved using particle swarm optimization (PSO). Also, a quota strategy for traffic load distribution is also introduced to mitigate the negative effects of multi-path routing. The proposed work is experimented and compared with existing algorithms to analyze its quality over previous work. Comparison has been done in terms of first node death, half node death, last node death, delay, and time consumed by Fminimax and PSO. Simulation results confirm the supremacy of proposed work over the existing ones.
  Keywords: Wireless sensor networks; residual energy; network lifetime; delay; load; particle swarm optimization.
  
  
• ST-IFC: Efficient Spatial-Temporal Inception Fully Connected Network for Citywide Crowd Flow Prediction  
  by Yan Kang, Bing Yang, Hao Li, Lan Zhang, Tie Chen 
  Abstract: Traffic flow prediction is important to urban management for the development of smart cities as well as further contribution to public safety. In the era of big data, large amounts of data related to traffic flow have been exponentially produced every day. However, vehicle streaming data may also record the mobility of human traffic. It could reflect urban traffic conditions to a certain extent. This paper analysed the spatial and temporal characteristics of human traffic in depth and proposed an efficient ST-IFC (Spatial-Temporal Inception Fully Connected) network for citywide traffic prediction. An IFC (Inception Fully Connected) unit was proposed to directly capture the spatial dependence and multi-scale characteristics of the entire data set of the urban traffic. In addition, this paper also proposes a multi-level feature fusion strategy to effectively combine the flow features of low-level surface and high-level abstract to avoid feature loss. Therefore, the proposed strategy greatly enhances the utilization of computing resources while ensuring the significant improvements of the prediction results for the proposed model. The simulations were carried out using the trajectory data of Beijing taxis and New York City bicycles. The experimental results show the advantages of our model in predicting the accuracy in addition to operating at a higher speed.
  Keywords: Deep Learning; Urban Computing; Spatio-temporal Data; Traffic Forecast.
  
  
• Compressive Sensing Multi-target Diffusive Source Localization using Sparse Recovery Algorithms in Sensor Networks  
  by Zhang Yong, Zhi Yan, Qi Chen, Teng Fei, Liyi Zhang 
  Abstract: According to the multi-target diffusive source localization in sensor networks, a compressive sensing sparse recovery algorithm was proposed for the mismatching problem of the target sources sparsity and the high-dimensional redundant sampling signals. Firstly, the compressive sensing system model and the related terms were given and explained. Then, the joint optimal estimation of the sparse diffusive source vector and the diffusion distribution state were realized with the variational Bayesian expectation maximization algorithm (VB-EM). In which, the dynamic compressive sensing dictionary model of the real target source sparse representation was designed and adjusted with the grid division parameters optimization for the dictionary mismatch problem solving. Finally, the simulation results show that the proposed compressive sensing method with VB-EM algorithm could effectively achieve the diffusive source parameters estimation and its diffusion distribution state prediction. Compared with the traditional compressive sensing sparse recovery algorithms, it could obtain higher robustness performance for the rapid and accurate localization in complex environment.
  Keywords: sensor network; compressive sensing; variational Bayesian expectation maximization.
  
  
• MPI hardware framework for many-core based embedded systems  
  by Rodrigo Vinicius Mendonça Pereira, Laio Oriel Seman, Marcelo Daniel Berejuck, Douglas Rossi De Melo, Analucia Schiaffino Morales, Eduardo Augusto Bezerra 
  Abstract: Multiprocessor System-on-Chip (MPSoCs) designs interconnected by high-speed networks has a crucial role at embedded systems, leading to next level sensors applications and interfaces' services. This paper presents the results regarding an investigation and evaluation of the services and infrastructure performance of a software and hardware implementation subset of the Message Passing Interface (MPI) standard. The proposal for an efficient MPI Hardware (MPIHW) and MPI Software (MPISW) models, along with the presentation and evaluation of its queuing model, aims at giving the system design a framework to assist. Comparative results are presented between MPI in hardware and software such as silicon consumption, processing time and transfer rate of the system related to the size of buffers. Also, tests in an environment consisting of an MPSoC model integrated on a Network-on-Chip (NoC) were performed, including classical algorithms such as pi calculation and the Dining Philosophers problem, to evaluate the proposed model functionality. Experimental results demonstrated the effectiveness of the proposed approach and the precision of the obtained implementation, although this comes at the cost of increased use of silicon in hardware implementation. This trade-off must be taken into account by the system designer of the sensors network.
  Keywords: MPI Hardware; MPSoC; Queue Model; Analytical Model; Sensor Network.
  
  
• Detection and Recognition of Text Traffic Signs above the Road  
  by Wei Sun, Yangtao Du, Xu Zhang, Guoce Zhang 
  Abstract: Based on the similarity between traffic sign images in the source and target domains, we use the parameters migrated from the source domain as the initial parameters of the Faster-R-CNN network, which is trained for detecting text traffic sign, then fine-tune the network parameters based on the samples in target domain to obtain the final network parameters. Moreover, we convert the traffic sign images from RGB color space to HSV color space and use the converted images in HSV color space as the training samples of the network, thereby overcoming the under-learning problem of model caused by less training samples. We tailor the traditional EAST text detection network model and propose a new recognition model based on the extreme learning machine (ELM) classifier to identify and classify the detected text traffic signs above the road. Experimental results in the natural scene demonstrate the effectiveness of the proposed method.
  Keywords: Text traffic sign; Computer vision; Convolutional neural network; Text detection.
  
  
• An Identity Authentication Method for Ubiquitous Electric Power Internet of Things Based on Dynamic Gesture Recognition  
  by Pingping Yu, Jincan Yin, Yi Sun, Zheng Du, Ning Cao 
  Abstract: This paper presents a novel algorithm for gesture recognition and identity authentica-tion based on continuous hidden Markov model (CHMM) and optical flow method. This study aims to solve the information se-curity problems about ubiquitous electric power Internet of Things. In this system, the optical flow method is used to segment and extract the features of the preprocessed dy-namic gesture information to obtain the fea-tures of the dynamic gesture motion track, and the CHMM is chosen to establish a valid user dynamic gesture model, which leads to ensuring the dynamic gestures are accurately recognized. The proposed method is test on accurately recognize the dynamic gestures and the result is compared with the Dynamic Time Warpring (DTW) algorithm and Practi-cal Swarm Optimisation-Radial Basis Func-tion Network (PSO-RBFN) algorithm. The result of the comparisons illuminates the su-periority of the proposed method in terms of accuracy of identity authentication.
  Keywords: Dynamic gesture recognition; Identity authentication; Ubiquitous electric power Internet of Things; Information safety; Continuous hidden Markov.
  
  
• A Secure and Privacy-Preserving Key Agreement and Mutual Authentication Scheme  
  by Hui Li, Tao Jing, Jin Qian 
  Abstract: The development of manufacturing, communication, and computing technologies has promoted convergence of different network architectures, aiming to make full use of all resources to create valuable information. While heterogeneous integrated networks play a key role in providing highly effective and efficient solutions for medical users, they also raise various security and privacy issues. Strengthen the security and privacy protection and efficiency before transmitting sensitive data is a primary challenge to be addressed. Traditional authentication methods fail to differentiate the privacy sensitivity difference between data collected by distinct devices. Instead of treating all devices just as one ``tool'', context privacy-aware access has been prompted to solve this issue. We propose a secure and efficient key management scheme based on a key-exposure resistant chameleon hash function. Through adopting the devices' privacy sensitivity and users' attribute credential as two factors to establish trust between different entities, we proposed a secure and anonymous authentication protocol. The security analyses indicate that the proposed protocol can satisfy security requirements.
  Keywords: Mutual-authentication; Chameleon Hash Function; Fog Computing; Key Management.
  
  
• WQMS - Water Quality Monitoring Station for IoT  
  by Sergio Diaz, Andres Molano, Christian Erazo, Juan C. Monroy 
  Abstract: Water pollution threatens public health with infectious and transmissible diseases. The traditional approach of water quality monitoring consists in collecting samples of water and transporting them to specialized laboratories, which is a waste of manpower. Nowadays, IoT devices are capable of monitoring the water quality and reporting the data to a cloud server without any human intervention. However, related work does not present a comprehensive solution that covers the main design aspects; thus, we designed and built a Water Quality Monitoring Station (WQMS) that includes power management, data measurement, data transmission, and Internet of Things (IoT). We deployed the WQMS in a remote fishpond with tilapias and measured the water quality for seven days in a row. The results show that the pH and dissolved oxygen can be expressed in terms of temperature as a quadratic function; besides, our solar energy harvesting module is a sustainable source of energy as long as there are no failures in the system.
  Keywords: Water Quality; Water Monitoring; Internet of Things.
  
  
• Task Scheduling for Mobile Edge Computing Enabled Crowd Sensing Applications  
  by Jingya Zhou, Jianxi Fan, Jin Wang 
  Abstract: Crowd sensing has emerged as a new promising application paradigm that collects real-time information about the physical world from individuals via their own devices. It effectively solves the dilemma of massive data collection faced by most data-driven applications such as traffic control, air quality prediction, disaster relief, etc. {Most of those applications are latency-sensitive, while current cloud-based crowd sensing systems cannot fully guarantee the response latency due to the restricted bandwidth of the backbone}. In recent years, mobile edge computing (MEC) is proposed to extend the frontier of cloud to the network edge so that it is quite suitable to integrate MEC with current crowd sensing systems. In this paper, we focus on the basic problem of task scheduling in a MEC-enabled crowd sensing system. Though many efforts have been devoted to the task scheduling research, the problem discussed here has some unique challenges, e.g., {edge devices are not dedicated to perform sensing tasks, task scheduling on edge devices and edge servers are highly coupled, and it is hard to achieve long-term objectives}. To address these challenges, we first present a system workflow framework that captures the unique execution logic of sensing tasks. Based on the framework, we propose a staged scheme to decouple the original scheduling problem and divide it into two sub-problems, i.e., task offloading problem and task shifting problem. Moreover, we leverage Lyapunov optimization technique to design a multi-period optimization for long-term objective achievement. The extensive experiments show that our proposed algorithm can effectively reduce cost with the guaranteed latency constraint.
  Keywords: Task scheduling; Crowd sensing; Mobile edge computing; Lyapunov optimization.
  
  
• Instant Messaging User Geolocating Method Based on Multi-Source Information Association  
  by Jiadong Guo, Rui Xu, Wenqi Shi, Pei Zhou, Xiangyang Luo 
  Abstract: The increased intermingling of instant messaging (IM) and the Internet of Things puts forward higher requirements for the security, the research on IM user geolocation technology is particularly important in maintaining the security of IoT. Most of the existing IM user geolocating methods analyze the relationship between the reported and real distance of the user to geolocate, but the number of geolocating users is limited. Therefore, we propose the instant messaging user geolocating method based on multi-source information association. This method converts the geolocating problem of a single user into the association of multi-source IM user information, and solves the problem of automatic and accurate acquisition of user information. The geolocating experiments are carried out for WeChat, Momo, and QQ. The results show that the method can achieve reliable acquisition for user information, and the number of geolocating users can be much greater than the existing geolocating method.
  Keywords: instant messaging; user information acquisition; user geolocating; user associating.
  
  
• Review of Single Image Defogging  
  by Baowei Wang, Bin Niu, Peng Zhao, Neal Xiong 
  Abstract: With the great advance of computer vision technology, the application of images in daily production work is more and more extensive. However, fairly substantial images collected in foggy weather show significant degradation. Image defogging technology is developed to solve this problem. After the defogging operation, the visual effect will be obviously improved, and it will also bring convenience to subsequent processing. This paper discusses the research background and current status of single image defogging strategies, and discusses the advantages and disadvantages of some classical algorithms. At the same time, combined with the analysis of these algorithms, some expectations are proposed.
  Keywords: Image defogging;Image enhancement;Image restoration; Fusion strategy;Retinex theory;Atmospheric scattering model.
  
  
• Towards Green Computing: Intelligent Bio-Inspired Agent for IoT-enabled Wireless Sensor Networks  
  by Sheetal Ghorpade, Marco Zennaro, Bharat Chaudhari 
  Abstract: With the emergence of the Internet of Things and machine to machine communications, massive growth in the IoT-enabled wireless sensor node deployment is expected in the near future. The critical challenges for the sensor network include energy efficiency, optimum route calculation, and the overall transmission cost. Although several kinds of research have been carried out to address such challenges, most of the reported bio-inspired optimizations are based on a single objective function considering other objectives as constraints. Some have considered a multi-objective optimization approach; however, the updating of particle positions was slower and challenging. To avoid the bias toward one of the objectives and also to facilitate ease of position updating, we propose a novel multi-objective optimization agent based on particle swarm gray wolf optimization (PSGWO) and inverse fuzzy ranking. We initially developed an enhanced PSGWO model, and then it is utilized for the development of population and multi-criteria based soft computing algorithm, called fuzzy PSGWO. The inverse fuzzy ranking guides the optimizer in updating the positions of particles better and faster way. The performance of the proposed algorithm is validated and compared with the well-known techniques. The results show that for the proposed algorithm, residual energy of the nodes is much higher than that of other algorithms, and save up to 48% energy along with smaller variation in the standard deviation. The results also demonstrate the smaller average values of fitness function and computationally efficient capabilities of the proposed algorithm.
  Keywords: Bio-inspired optimization; Energy efficiency; Inverse fuzzy ranking; Wireless sensor networksrnrn.
  
  
• Lightning Location Method Based on Improved Fuzzy C-Means Clustering Algorithm  
  by Tao Li 
  Abstract: Location accuracy is an important index for the evaluation of location networks and the localisation algorithm related to the accuracy of the results. Classical location algorithms scarcely correct errors accurately. Moreover, they have poor resistance to error interference and low location accuracy. To achieve error-resistant lightning localisation, weighted rough-fuzzy C-means (WRFCM) was introduced in location calculation. The performance of this localisation algorithm was analysed using a lightning accident case through regional simulation. Results show that the lightning localisation algorithm based on WFCM overcomes the disadvantage in which traditional location algorithms easily diverge; the algorithm also has an improved ability to resist error interference and can solve the lightning points steadily and accurately.
  Keywords: lightning localisation; time difference of arrival; cluster analysis; fuzzy C-means.
  
  
• SemicNet: A semicircular network for the segmentation of the liver and its lesions  
  by Zhihua Zheng 
  Abstract: The traditional neural network used for medical image segmentation was not clear on the network depth. In view of these problems, we propose a convenient and efficient liver and lesion segmentation system, which uses a double-layer codec semi-circular network to combine the deep and shallow semantic information through dense jump connection, which is easier to be processed by the optimizer; The transition zone between liver and lesion segmentation is designed so that the result of liver segmentation can be effectively transmitted to lesion segmentation; We believe that the selection of complementary loss function combination for in-depth supervision can effectively receive the anti-propagation gradient signal and obtain more regularization effects. Finally, in terms of liver segmentation, in addition to the model with lower accuracy than multiple loss functions for joint decision-making, all other evaluation indexes, including lesions, exceeded the fusion results of multiple models.
  Keywords: Liver and lesion segmentation; Codec network; Dense connections; Transition zone; The depth of the supervision.
  
  
• A computationally efficient authentication and key agreement scheme for multi-server switching in WBAN  
  by Zisang Xu, Cheng Xu, Jianbo Xu, Xiangwei Meng 
  Abstract: Wireless Body Area Network (WBAN) is mainly used in the medical field. The wearable device in WBAN can monitor the physiological information of the patient and send information to a server. The doctor can remotely diagnose the patient by accessing the server in the hospital. As the patient's physiological information is sensitive, transmitting the data in the WBAN may reveal patient's privacy. Hence, WBAN needs a reliable authentication and key agreement scheme. In addition, each hospital or health care provider usually has a server that is independent of each other. Once the patient needs to change hospitals or health care providers to receive medical services, he/she needs to transfer his/her historical data to the new server, the process which is called multi-server switching. Most existing authentication schemes for WBAN either use a single-server model or do not consider multi-server switching issues. Therefore, we propose a computationally efficient mutual authentication and key agreement scheme for multi-server switching in WBAN. Our scheme ensures that patients in WBAN can implement secure switching servers at any time in a multi-server environment, and it is also lightweight enough because only hash function operations, XOR operations, and symmetric encryptions/decryptions are employed. Our scheme proves to be secure under the Real-Or-Random (ROR) model and ProVerif. In addition, compared with related schemes, our scheme solves the server switching problem while reducing the computational cost.
  Keywords: authentication; cryptography; key agreement; WBAN.
  
  
• Novel greedy grid-voting algorithm for optimization Placement of multi-camera  
  by Hocine Chebi 
  Abstract: The optimal placement of surveillance cameras can be used for other purposes such as network performance on information detection, in addition to improving the efficiency of surveillance in public places as well as the cost of the installation. Determining the configuration of exposure routes to ensure optimal coverage is essentially a combinatorial optimization problem. In this paper, we proposed two approaches to identify and locate the location of cameras in an area of interest with planning already done at the site. The major contribution of the paper is to propose the modification on the greedy grid voting algorithm, which can control the overlap between each cameras coverage based on the application requirements. The proposed method gives preference to cover unique regions first. Unlike other existing greedy methods, the greedy grid-voting algorithm doesnt follow any execution order for optimization purposes. We also modified the traditional global greedy algorithm and it produced better coverage of the area of interest compared to the greedy approaches used to calculate the maximum coverage. The proposed methods of optimization methods are evaluated in an examination hall. The results obtained tell us that the approach ensures total coverage with a minimum of sensors than other techniques exist in the literature.
  Keywords: Placement of multi-camera; video surveillance; combinatorial optimization; sensors network.
  
  
• Application of convolutional neural networks and image processing algorithms based on traffic video in vehicle taillight detection  
  by Ning Cao, Wei Huo, Gangshan Wu 
  Abstract: With the sharp increase in car ownership, frequent traffic accidents have caused huge losses to the national economy and people's lives. How to take effective measures to assist the safe driving of vehicles has become a hot issue in today's traffic safety research. The headlight of a vehicle is an important way to exchange information with surrounding vehicles while the vehicle is running. In the process of assisted driving, accurately understanding the linguistic information transmitted by surrounding vehicles is the prerequisite for making correct driving decisions. In this paper, the neural network is partly used for vehicle detection. The recognition of the front vehicle taillights is based on the taillight recognition mechanism and image processing technology. The taillights are then positioned by using their colour and shape characteristics. The RGB and CMY colour spaces are used to establish a taillight recognition mechanism to detect the taillight status of the front car, thereby the driving intention of the front car is understood. The experimental results show that the method can accurately identify the state of the front taillights during the day.
  Keywords: driving assistance; vehicle detection; taillight detection; signal recognition.
  
  
• Parallel Cuckoo Search for Cognitive Wireless Sensor Networks  
  by Tong Bang Jiang, Jeng-Shyang Pan, Yu Mo Gu, Shu-Chuan Chu 
  Abstract: In cognitive wireless sensor networks(CWSNs), although each sensor node has the characteristics of cognitive radio, the limited energy of the sensor node is still the core defect that restricts its comprehensive network performance. This paper proposes a novel medoids generation mode named Parallel Cuckoo Search medoids (PCS-medoids) algorithm with a new communication strategy in Cuckoo Search(CS) to manage the energy consumption in CWSNs efficiently. The PCS-medoids can match cluster heads dynamically and get optimal cluster heads in CWSNs. Firstly, in order to speed up the convergence of CS, an improved Parallel Cuckoo Search algorithm(PCS) is proposed, then, the PCS is applied to k-medoids to get cluster heads quickly. Finally, the PCS-medoids is presented to manage the consumption of each sensor node. First experiment results illustrate that PCS tends to get optimal solutions quickly and accurately compared to CS and PSO. The other experiment results demonstrate that PCS-medoids has advantage over energy management in CWSNs compared to low-energy adaptive clustering hierarchy, LEACH-centralized, and hybrid energy-efficient distributed clustering. In addition, the advantage is more obvious with the increase of sensor nodes in CWSNs.
  Keywords: Cognitive Wireless Sensor Networks (CWSNs); Parallel Cuckoo Search medoids (PCS-medoids); Energy management.
  
  
• A Dynamic Resource Assignment Scheme with Aggregation Node Selection and Power Conservation for WBAN Based Applications  
  by Mahfuzulhoq Chowdhury 
  Abstract: Wireless body area network (WBAN) is regarded as one of the crucial\r\ntechnology to monitor patient health status by collecting real-time health-related data with the help of interconnected human body sensors. For real-time health status monitoring and treatment, the development of an efficient resource management scheme is an important research issue for WBAN based medical applications. To cope with the different requirements of WBAN based applications, the coordination of cluster head\r\nselection, dynamic medium access control mechanism (MAC) considering both critical and normal data transfer, end-to-end connectivity, data transfer to multiple processing and destination device, intermediate node selection, and proper path selection, is mandatory.\r\nWBAN sensor nodes energy power conservation and network lifetime maximization is another major research challenge for WBAN based medical applications. The existing resource management scheme suffers from higher delay and energy power consumption due to a lack of coordination between MAC and routing protocol along with the cluster head nodes huge working load. In this paper, a dynamic resource assignment scheme\r\nis proposed that assigns bandwidth and medical server resources to WBAN users for\r\ndata transfer without any backoff delay and incorporates a power conservation approach\r\nfor WBAN members. To minimize the cluster head node working load, the proposed\r\ndynamic resource management scheme selects both the cluster head node and aggregation\r\nsensor node for the aid of WBAN sensors. This paper compares the proposed dynamic\r\nresource assignment scheme against a random resource assignment scheme in terms of\r\ntotal delay overhead, energy overhead, transmission latency, goodput, number of nonalive nodes, packet delivery, remaining energy, network stable, and unstable lifetime. The\r\nevaluation results show that the proposed dynamic resource assignment scheme performs\r\nmuch better results than the compared random resource assignment scheme in terms of\r\nthe delay overhead, energy overhead, goodput, and the network lifetime.
  Keywords: Wireless body area network (WBAN); Resource Assignment Scheme; Power Conservation; Cluster Head Selection; Goodput; Network Lifetime.
  
  
• Realtime Soldier's Health Monitoring system Incorporating low power LoRa communication  
  by Bhargav Jethwa, Milit Panchasara, Abhi Zanzarukiya, Rutu Parekh 
  Abstract: In this paper, we have implemented a low-cost embedded system developed for soldiers assistance. The system consists of interconnected body sensor networks for real-time health monitoring and environmental analysis of the soldier and the communication is done with the base station using LoRa module. The data collected from each sensor is processed using a robust and steady algorithm to decide the health and environmental conditions of the soldier. The health status is further classified into healthy, wounded or dead. All the collected data along with the obtained results about the soldier's health condition are then encrypted and transmitted to the base station securely. The system uses only 3.2 Wh energy making it energy efficient to extend the operational time. If we use a commercially available 50,000 mAh battery, operating at 5 V, with the worst-case current drawing capacity of 90\%, the system could remain functional for about 75 hrs. The information can be transmitted up to 700 meters at 2400 baud rate and could be increased further by reducing the baud rate.
  Keywords: LoRa; BSN; ECG; IMU; NMEA; WBASN; AES encryption; Signal processing.
  
  
• Dual Hop Relaying using CDMA and Reconfigurable Intelligent Surfaces (RIS)  
  by Sami Touati, Rachid Sammouda, Musaed A. Alhussein 
  Abstract: In this paper, we suggest the use of Reconfigurable Intelligent Surfaces (RIS) in dual hop relaying. There are two hops, the first one uses Code Division Multiple Access (CDMA) and the second one uses RIS as transmitter or reflector. We show that the incorporation of RIS allows to increase the throughput of conventional CDMA systems. The throughput of dual hop relaying using a combination of CDMA and RIS is larger than that of dual hop relaying based entirely on CDMA.
  Keywords: Code Division Multiple Access (CDMA); Reconfigurable Intelligent Surfaces (RIS); throughput analysis; Block error probability.rn.
  
  
• A Novel Offloading Strategy for High Speed Mobile and Data Intensive Intelligent Sensor  
  by Changming Zhao, Mingdong Li, Tiejun Wang, Hao Yang 
  Abstract: In the paper, it proposes a novel offloading computing strategy for the mobile intelligent sensors in the high speed mobile scene. At present, the local computing resources are not able to afford the computing demand for the data intensive sensors. The current offloading computing mechanism are not capable of approaches for offloading computing in high speed scene that it may cause extremely serious delay by the handover delay and the wireless channel fading. In the paper, it proposes a novel strategy named Asynchronous Parallel Offloading Computing Strategy (APOCS). It makes use of one type of spatial computing diversity method to restrain the wireless channel fading.. The simulation results show that it is able to reduce the offloading delay caused by wireless channel fading in the high speed moving scene effectively.
  Keywords: Parallel Computing; Edge Computing; Offloading; fast fading.
  
  
• Low-Delay Fair-Reliability Scheduling in Multi-hop IEEE802.15.4e Time Synchronized Channel Hopping Networks  
  by Junhua Zhang, Yuanyi Wang, Zhenqian Wu 
  Abstract: IEEE802.15.4e is the latest generation of high-reliability, low-power medium access control protocols. As part of IEEE802.15.4e, Time Synchronized Channel Hopping (TSCH) provides support for multi-hop and multi-channel communications. IEEE802.15.4e only defines when a schedule is executed; it does not specify a concrete resource schedule strategy. In this study, an IEEE802.15.4e TSCH scheduling algorithm called Fair Reliability Scheduling Algorithm (FRESA) is proposed to focus on both less delay and fair reliability. Through leaf node prior and multiple nodes sharing the same resource, there is less delay to transmit messages. Balancing the assignment of different qualities of channel offset guarantees fair reliability in message transmission. We design two strategies to ensure fair reliability, one based on per Link and the other based on per Path, and then accordingly divide our algorithm into FRESA-L and FRESA-P. Simulation results demonstrate our algorithms performance. We also observe that FRESA-P can achieve higher reliability than FRESA-L.
  Keywords: IEEE802.15.4e; time synchronized channel hopping; wireless sensor network; industrial Internet of Things; fair reliability.