Forthcoming and Online First Articles

International Journal of Wireless and Mobile Computing

International Journal of Wireless and Mobile Computing (IJWMC)

Forthcoming articles have been peer-reviewed and accepted for publication but are pending final changes, are not yet published and may not appear here in their final order of publication until they are assigned to issues. Therefore, the content conforms to our standards but the presentation (e.g. typesetting and proof-reading) is not necessarily up to the Inderscience standard. Additionally, titles, authors, abstracts and keywords may change before publication. Articles will not be published until the final proofs are validated by their authors.

Forthcoming articles must be purchased for the purposes of research, teaching and private study only. These articles can be cited using the expression "in press". For example: Smith, J. (in press). Article Title. Journal Title.

Articles marked with this shopping trolley icon are available for purchase - click on the icon to send an email request to purchase.

Online First articles are published online here, before they appear in a journal issue. Online First articles are fully citeable, complete with a DOI. They can be cited, read, and downloaded. Online First articles are published as Open Access (OA) articles to make the latest research available as early as possible.

Open AccessArticles marked with this Open Access icon are Online First articles. They are freely available and openly accessible to all without any restriction except the ones stated in their respective CC licenses.

Register for our alerting service, which notifies you by email when new issues are published online.

We also offer which provide timely updates of tables of contents, newly published articles and calls for papers.

International Journal of Wireless and Mobile Computing (81 papers in press)

Regular Issues

  • Fast retrieval of similar images of pulmonary nodules based on deep multi-index hashing
    by Rui Hao, Yaxue Qin, Yan Qiang 
    Abstract: CT image retrieval of pulmonary nodules mainly uses deep neural network embedded in hash layer to extract hash codes, which are directly as the index address for linear search. With the increasing number of clinical lung CT images and the complexity of image expression, the traditional retrieval methods are inefficient. We propose a multi-index hash retrieval algorithm based on deep hash features. First, a hash layer is added to the convolutional neural network (CNN) which can simultaneously learn the high-level semantic features of images and the corresponding hash function expression. Secondly, the hash codes extracted are effectively divided and multi-index tables are constructed. The query algorithm is designed based on the drawer principle. Finally, the complexity analysis of the whole index algorithm is carried out and experimental results show that the proposed algorithm can effectively reduce the retrieval cost while maintaining the accuracy.
    Keywords: pulmonary nodule; deep learning; hash feature; image retrieval; multi-index hashing.

  • Multicast stable path routing protocol for wireless ad-hoc networks   Order a copy of this article
    by K.S. Saravanan, N. Rajendran 
    Abstract: Wireless Ad-Hoc Networks (WANETs) enable steady communication between moving nodes through multi-hop wireless routing path. The problem identified is how to improve the lifetime of the route and reduce the need for route maintenance. This helps to save bandwidth and reduce the congestion control available in the network. This paper aims to focus on redesign and development of multicast stable path routing protocol with special features that determine long-living routes in these networks. An extensive ns-2 simulation based performance has been analysed of three widely recognised stability oriented wireless ad-hoc network routing protocols, namely are Associativity Based Routing (ABR) protocol, Flow Oriented Routing Protocol (FORP) and Lifetime Route Assessment Based Routing Protocol (LRABP). The order of ranking of the protocols in terms of packet delivery ratio, average hop count per route, end-to end delay per packet and the number of route transitions is presented.
    Keywords: wireless ad-hoc networks; multicast routing protocol; wireless communication; routing protocol.

  • Study on carbon footprint model and its parameter optimisation of wave soldering process based on response surface method   Order a copy of this article
    by Renwang Li, Haixia Liu, Jiaqi Li, Jinyu Song, Rong Jie 
    Abstract: In order to respond to low carbon manufacturing, from energy, materials and process carbon emissions, etc., this paper constructs a carbon footprint model for the wave soldering process in the module workshop of H Company. Based on this model, the carbon footprint value of the wave soldering process is calculated. On the basis of selecting appropriate parameter factors, a parameter optimisation model of the wave soldering process life-cycle carbon footprint is constructed, and the optimum parameters combination is analysed by the response surface method, which contains surface area, clip velocity, clipping angle, flux flux, purity of solder, temperature of tin furnace, height of wave peak, etc. The the response values obtained are verified. The experimental results show that the optimised parameters are used to process and manufacture the wave soldering process, and the carbon emissions produced by wave wave welding can be controlled from above 15 kg CO2 to 12 kg CO2.
    Keywords: wave soldering; carbon footprint; calculation model; response surface method; parameter optimisation.

  • New transformation method in continuous particle swarm optimisation for feature selection   Order a copy of this article
    by Kangshun Li, Dunmin Chen, Zhaolian Zeng, Guang Chen, James Tin-Yau Kwok 
    Abstract: Feature selection is a very important task in many real-world problems. Because of its powerful search ability, particle swarm optimisation (PSO) is widely applied to feature selection. However, PSO was originally designed for continuous problems, and therefore, the transformation between continuous particles and binary solutions is needed. This paper proposes a new transformation methods-based PSO (PSOS) in which the related feature subset of a particle is decided by a sine function rather than comparing with a single threshold. To further upgrade the performance of the proposed method, an extra increment generated by the Gaussian distribution is added to the marginal positions (PSOSI). The experimental results show that PSOS and PSOSI can select smaller feature subsets with higher classification accuracy than all the other algorithms compared in this paper. Furthermore, in most cases, the performance of the second method is better than the first one.
    Keywords: particle swarm optimisation; feature selection; classification; sine function; Gaussian distribution; transformation method.

  • Research of small fabric defects detection method based on deep learning network   Order a copy of this article
    by Siqing You, Kexin Fu, Peiran Peng, Ying Wang 
    Abstract: For quality improvement of textile products, fabric defects detection is significant. In this paper, the detection capacity of SSD for small defects was studied. The loss of feature information was reduced through the reduction of layers of SSD network; then the size of the default box was adjusted based on the K-means clustering algorithm, and the adaptive histogram equalisation algorithm was applied to enhance the defect features and effectively improve the detection accuracy. The improved SSD network model was tested to verify the fabric defects dataset, which further improved the accuracy of detection. In addition, the two-stage algorithm was compared to find the optimal algorithm for small object detection. According to the test results, the subsequent improvement method for small object detection with SSD was proposed.
    Keywords: fabric defects detection; default box; feature enhancement; SSD; faster RCNN.

  • Field theory trusted measurement model for IoT transactions   Order a copy of this article
    by Meng Xu, Bei Gong, Wei Wang 
    Abstract: The Internet of Things (IoT) allows the concept of connecting billions of tiny devices to retrieve and share information regarding numerous applications, such as healthcare, environment, and industries. Trusted measurement technology is crucial for the security of the sensing layer of the IoT, especially the trusted measurement technology oriented to transaction IoT nodes. In the traditional trust management system, historical behaviour data are considered to predict the trust value of the network entity, while the nodes' trust between network entities is rarely considered. This paper proposes a novel fi eld theory trusted measurement model of the sensing layer network, which can well adapt to the transaction scenarios of the IoT.
    Keywords: field theory; internet of things; trust measurement; transaction scenario.

  • Dynamic time warping-based evolutionary robotic vision for gesture recognition in physical exercises   Order a copy of this article
    by Quan Wei, Kubota Naoyuki, Ahmad Lotfi 
    Abstract: In this paper, we propose a three-dimensional posture evaluating system from two-dimensional images, which can be implemented in physical exercises for elderly people. In this system, two-dimensional coordinates of human joints are first captured and calculated, then our proposed Dynamic Time Warping Steady State Genetic algorithm (DTW-based SSGA) is used for the evaluation of three-dimensional rotational variables from RGB images for the human arm. Finally, these predicted rotational variables would be compared with the template of sample posture by Dynamic Time Warping (DTW) to check the complement of physical exercises. The experimental result shows that our proposed DTW-based SSGA performs with higher accuracy than other evolutionary algorithms, such as standard Steady State Genetic Algorithm (SSGA) and Particle Swarm Optimisation (PSO) when evaluating human joint variables with templates, especially in the physical exercises for rehabilitation.
    Keywords: gesture recognition; forward kinematics; evolutionary computing; dynamic time warping.

  • Research on trusted SDN network construction technology   Order a copy of this article
    by Fazhi Qi, Zhihui Sun, Yongli Yang 
    Abstract: In this paper, we combine trusted computing with SDN. By active measurement of the SDN controller when it is starting and running, we can guarantee the trust of the SDN controller. By actively measuring the behaviour of the SDN data transponder in the domain, we can guarantee trust of the SDN data transponder. When the cross-domain data interaction is involved, by trusted network connection mechanism, we can guarantee the trust of the transmission of data in different domains so as to build a trusted SDN network as a whole.
    Keywords: trusted computing; SDN; active measurement.

  • A method of spatial place representation based on visual place cell firing   Order a copy of this article
    by Naigong Yu, Hui Feng 
    Abstract: Constructing a model of visual place cells (VPCs), which produce sensitive firing to visual information, is of great significance for studying bionic positioning and bionic navigation. Based on the physiological research of place cells and the analysis of existing VPC generation models, a firing model of VPCs based on the distance perception of landmarks by the agent is proposed in the paper. Based on the firing activity of VPCs, a spatial place representation method is proposed. The method mainly includes exploring the environment and detecting landmarks, calculating the firing rate of VPCs, adding VPCs and constructing the map of VPCs. Through simulation experiments, the reliability of the positioning performance of the proposed method is verified, and the influence of various parameters in the model on the accuracy of spatial representation of the VPCs map is analysed.
    Keywords: visual place cell; spatial representation; bionic positioning; bionic navigation.

  • Point cloud registration algorithm based on 3D-NDT algorithm and ICP algorithm   Order a copy of this article
    by Jiangge Huang, Bo Tao, Fei Zeng 
    Abstract: The purpose of point cloud registration is to minimise the difference of spatial position between point clouds. In addition, the point cloud registration process needs to be performed with high efficiency and accuracy. This paper combines the high efficiency of the 3D normal distribution transformation (3D-NDT) algorithm with the high precision of the iterative nearest point (ICP) algorithm, and proposes a fusion registration algorithm. At the same time, the fusion algorithm can still keep high efficiency and high precision registration. First, the 3D-NDT algorithm is used to select appropriate parameters, so that the point cloud to be registered is closer to the target. It provides an excellent initial position for the ICP algorithm to complete coarse registration. Secondly, in order to improve the efficiency of solving transformation matrix in ICP algorithm, kd-tree is introduced for acceleration. The experimental results show that the fusion point cloud registration algorithm proposed in this paper is better than the 3D-NDT algorithm and the ICP algorithm in efficiency and accuracy. The method proposed in this paper has more obvious advantages in dealing with larger point clouds.
    Keywords: 3D-NDT algorithm; ICP algorithm; point cloud registration; point cloud search.

  • Research on leak detection and location of urban gas pipeline network based on RSSI algorithm   Order a copy of this article
    by Liming Wei 
    Abstract: To solve the leakage problem of urban gas pipelines, this paper presents a method of detecting and locating leakages based on the RSSI algorithm. This technique can analyse and calculate the signal strength received between ZigBee nodes when a pipeline leaks and ultimately obtain the location of the leak. Firstly, the algorithm model is established by using the RSSI signal strength values between the leak target point and each receiving point. Secondly, the distance between the leak point and each receiving point is obtained by the model. Lastly, the approximate coordinates of the leak point are obtained by the least squares method. The simulation results show that the proposed algorithm has high positioning accuracy and wide application prospects.
    Keywords: gas pipeline network; fire early warning; least squares method; RSSI algorithm; ZigBee technology.

  • Multi-objective workflow scheduling in the cloud environment based on NSGA-II   Order a copy of this article
    by Tingting Dong, Chuangbai Xiao 
    Abstract: The emergence of cloud computing offers a novel perspective to solve large-scale computing problems. Workflow scheduling is a major problem in the cloud environment, and parallelism and dependency are two important characteristics of tasks in a workflow, which increases the complexity of problem. Workflow scheduling is also a multi-objective scheduling problem, and task execution time and cost are the two extremely significant goals for users and providers in the cloud environment. Existing heuristic algorithms are popular, but they lack of robustness and need to be revised when the problem statement changes. Evolutionary algorithms have a complete algorithm system, which is widely used in the multi-objective scheduling problem. In this paper, Nondominated Sorting Genetic Algorithm-II (NSGA-II) is utilized to solve the workflow scheduling problem aiming at minimizing the task execution time and cost. Some real-world workflows are used to make simulation expriments, and comparative simulations with genetic algorithm are given. Results show that NSGA-II is effective for the workflow scheduling.
    Keywords: cloud computing; workflow scheduling; non-dominated sorting genetic algorithm-II; multi-objective scheduling.

  • Application of improved deep reinforcement learning algorithm in traffic signal control   Order a copy of this article
    by Wang Qiang, Song Shuaidi, Zhang Tengyun, Wang Zelin 
    Abstract: For a regional road network, the signal control system lights at multiple intersections belong to the core technology of intelligent innovation. The relevant personnel need to integrate and analyse the relevant status information of the intersection area based on the DQN method and strategy, and strengthen the control of the signal light system effect, to achieve fast and effective detection. In this paper, we propose a reinforcement learning DQN+ algorithm by using the improved DQN reward and punishment function. Experiments show that DQN+ has obvious advantages in terms of average queue length (AQL), average speed (AS) and average waiting time (AWT) at four intersections.
    Keywords: intelligent transportation; traffic signal control; reinforcement learning; deep reinforcement learning.

  • Enhancing artificial bee colony algorithm with depth-first search and direction information   Order a copy of this article
    by Xinyu Zhou, Hao Tang, Shuixiu Wu, Mingwen Wang 
    Abstract: In recent years, artificial bee colony (ABC) algorithm has been criticized for its solution search equation, which makes the search capability bias towards exploration at the expense of exploitation. To solve the defect, many improved ABC variants have been proposed aiming to use the elite individuals. Although these related works have shown effectiveness, they rarely take the factor of search direction into account. In fact, the search direction has an important role in determining the performance of ABC. Thus, in this work, we are motivated to investigate how to combine the idea of using the elite individuals with the search direction, and a new ABC variant, called DDABC, is designed. In the DDABC, the depth-first search (DFS) mechanism and direction information learning (DIL) mechanism are introduced, and the former mechanism is to allocate more computation resources to the elite individuals, while the latter mechanism aims to adapt the search to the promising directions. To verify the effectiveness of the DDABC, experiments are carried out on 22 classic test functions, and three relative ABC variants are included as the competitors. The comparison results show the competitive performance of our approach.
    Keywords: artificial bee colony; exploration and exploitation; depth-first search; direction information learning.

  • Performance evaluation of AODV, DSDV and DSR protocols in wireless sensor networks   Order a copy of this article
    by Rohin Rakheja, Sonam Khera, Neelam Turk 
    Abstract: In wireless sensor networks, the routing protocol and path selection are two of the most important factors when designing the network. Owing to the severe limitations on the resources available, the selected protocol should provide high energy efficiency without compromising in terms of data delivery rate, security, integrity. Hence the analysis of the characteristics of these protocols is the major step before selecting them for real-world applications. In this paper, AODV (Ad-hoc On-demand Distance Vector), DSDV (Destination Sequenced Distance Vector) and DSR (Dynamic Source Routing) protocols have been simulated using Network Simulator 2 (NS2.35) software package. Performance parameters, such as instantaneous throughput, total energy consumption of the network, residual energy of each node, packet delivery rate and average throughput, have been calculated over multiple networks of 100, 150 and 200 nodes. The simulation runs for 65 seconds and each network has two static client and sink nodes. The data traffic starts at t = 1.0 and stops at t = 61.0, where (t) is time in seconds. Graphical representation has been done with the help of data extraction and manipulation based on trace files and drawing software. Experiments reveal the characteristics and behaviour of these protocols to substantiate the conclusions.
    Keywords: wireless sensor networks; OSPF; shortest paths; energy consumption.

  • PeneVector Marine multifunctional geological sampling/testing integrated equipment   Order a copy of this article
    by Wei Zhang, Qi Chen, Linqi Xia, Tao Li 
    Abstract: An integrated seabed geological sampling/testing system is developed for marine geological survey. The system has the functions of static pressure penetration sampling, vibration penetration sampling and in-situ test. The static pressure penetration sampling function adopts the fuzzy control algorithm to control the speed and torque of the friction wheel to ensure the synchronism of the two friction wheels. The double friction wheel holds the sampling pipe to penetrate into the stratum for sampling, which has high fidelity. In the sand layer, the sand can be liquefied and sampled by the function of vibration penetration sampling. For in situ testing, multi-functional static cone penetration is used to obtain the mechanical properties, resistivity and geothermal gradient of the seabed directly.
    Keywords: geological survey; seabed type; static pressure sampling; vibration sampling; static cone penetration.

  • Network lifetime maximisation with low power consumption by the use of an ANFIS-based technique in wireless sensor networks   Order a copy of this article
    by Nune SrinivasRao, K.V.S.N. Rama Rao 
    Abstract: The routing in a wireless sensor network (WSN) is important for improving the network's functioning, because inappropriate routing methods and routing degrade the sensor network energy, impacting the network lifetime. Clustering strategies for reducing the energy consumption and extending the network life have been employed widely. The clustering mechanism can extend the network's service life and network failure. In this study, the IoT has contributed in improving network performance with a new energy efficient ANFIS-based routing approach for WSN. A new distributed cluster creation methodology that enables the self-organisation of local nodes, a novel method for the adjustment of clusters and the turning of the cluster head centre location to distribute the energy burden equally through all sensing nodes incorporates the suggested ANFIS-based routing. The simulation result shows that the proposed scheme outperforms conventional methods with an improvement of 80% in network lifetime and 27% in throughput.
    Keywords: energy-efficient routing protocol; base station; cluster head; network lifetime; wireless sensor network.

  • A blind receiver for OFDM communications   Order a copy of this article
    by Min Lu, Min Zhang, Guangxue Yue, Bolin Ma, Wei Li 
    Abstract: Owing to channel fading and noise interference in different environments, how to accurately restore the transmitted bit stream at the receiving end has become a key issue of the orthogonal frequency division multiplexing (OFDM) systems. We propose a dual-path mixed deep learning (DMDL) framework for the blind OFDM receiver, which combines the densely connected convolutional networks (DenseNets) and the residual networks (ResNets). The DMDL receiver can solve the problem of gradient explosion and feature disappearance in the network training, and it does not require pilots for the channel estimation. The experimental results show that on the additive white Gaussian noise (AWGN) channel, the performance of the DMDL receiver can be improved by 1.62 dB over the traditional receiver. On the Rayleigh fading channel, the performance improvement of the DMDL receiver can reach 1.94 dB. The DMDL model also has excellent performance in cyclic prefix-free and Doppler frequency shift environment.
    Keywords: blind receiver; deep learning; DenseNet; OFDM; ResNet; signal detection; wireless communication.

  • Research on video image face detection and recognition technology based on improved MTCNN algorithm   Order a copy of this article
    by Jinfeng Liu 
    Abstract: With the development of modern computer technology and artificial intelligence, face image processing technology has been widely used in people's lives and work. In order to realise face image detection and recognition in a dynamic video, this paper proposes a face detection and recognition technology based on the MTCNN algorithm. This algorithm includes R-Net, o-net and p-net deep network models, which can realise face image deep processing in dynamic videos. In order to train the MTCNN algorithm deeply, Selected_Face and CelebA database training sets were used to train the additional test tasks and regression key points of the model. After setting the main parameters of the MTCNN algorithm, the algorithm is simulated and analysed. Through the comparative simulation analysis of traditional algorithms, we can see that MTCNN algorithm has better performance than traditional CNN algorithms, and can meet the requirements of face image detection and recognition in dynamic videos.
    Keywords: video image; face detection; distinguish; MTCNN algorithm; sample training.

  • Machine learning-based approach for the detection of phishing websites   Order a copy of this article
    by Yaqin Wang, Jingsha He, Nafei Zhu 
    Abstract: Compared with traditional forms of crime, cyber-attacks and cyber-crimes have removed the limitation on distance and speed. With very low cost, phishing is a very effective way of launching network attacks with the purpose of obtaining sensitive information about users, such as username, password and payment voucher, through counterfeiting regular websites so as to steal users private information and personal property using the obtained information. Both the trust that internet users have and the development of the internet itself can be affected by this kind of attack, making it imperative to detect this type of attack. Many methods have been proposed for the detection of phishing websites in the literature in recent years based on techniques ranging from conventional classifiers to complex hybrid classifiers. Meanwhile, although convolutional neural networks (CNNs) can achieve very high accuracy in classification tasks, not much research has been done on the use of CNNs for the detection of phishing websites. This paper proposes a CNN-based scheme for the detection of phishing websites in which four dimensions of the features of phishing websites are defined and CNN is used to extract local features. The proposed CNN-based scheme is compared with several machine learning-based methods on the effectiveness of detecting phishing websites, which shows that the proposed scheme can achieve the accuracy rate of 97.39% and is better than the other classification methods in terms of accuracy, recall and F1-score.
    Keywords: convolutional neural network; classification; machine learning; phishing website detection.

  • A method of walking trajectory for biped robot based on Newton's interpolation   Order a copy of this article
    by Yingli Shu, Quande Yuan, Huazhong Li, Wende Ke 
    Abstract: Trajectory design of a biped robot is the premise of effective walking. By analysing the kinematics characteristics of a biped robot, Newton's interpolation is used to design the trajectory of the insertion point when planning periodic motion and realising ZMP (zero moment point) constraint. With the increase of interpolation points, the curve fitting effect is improved, and it will not lead to more expensive calculation cost, which can effectively meet the real-time requirements. Simulation results show the effectiveness of the method.
    Keywords: biped robot; interpolation; trajectory; walking.

  • Research on driving control of an electric vehicle based on fuzzy control algorithm   Order a copy of this article
    by Jinxia Yang 
    Abstract: In the development process of automobile industry in full swing, the problem of traffic safety is becoming increasingly prominent. In order to effectively reduce the incidence of traffic accidents, this experiment deeply explores the drive control of electric vehicles. By integrating the fuzzy control algorithm into the drive control system of electric vehicles, the simulation analysis under different working conditions is carried out. The simulation results show that the driving control of electric vehicle based on fuzzy control algorithm can effectively control the driving speed, acceleration and distance from the front vehicle under four working conditions. For example, when the front vehicle leaves, the electric vehicle can play a superior speed following performance. When the current vehicle accelerates away from the self vehicle, the self vehicle can quickly switch to the constant speed cruise mode; The actual distance between the two vehicles is basically consistent with the expected safety distance, which can realize effective distance control; The acceleration trend of the electric vehicle is very close to the expected acceleration, and it can maintain the constant speed cruise mode after 40s of the simulation experiment. This shows that the electric vehicle control algorithm based on fuzzy control has stable control performance.
    Keywords: fuzzy control; electric vehicle; simulation; drive control.

  • Narrowband internet of things: performance analysis of coverage enhancement in uplink transmission   Order a copy of this article
    by Rasveen Singh, Shilpy Agrawal, Khyati Chopra 
    Abstract: Narrowband Internet of Things (NB-IoT) is a wireless standard and a novel technology for the IoT devices and applications. The extended coverage, low cost, and long battery life make NB-IoT an excellent candidate for IoT applications. One of the primary goals of NB-IoT is to enhance coverage beyond the existing cellular technologies like general packet radio service and long-term evolution (GPRS and LTE). To accomplish this, the NB-IoT system utilizes a repetition technique in which the same signal is repeated several times with different subcarrier spacing in the uplink. We propose the repetition model with the optimization algorithm, i.e., moth flame optimisation (MFO). The optimisation reduces the block error rate (BLER) even in the worst channel condition, which increases the performance evaluation in a single-tone and multi-tone transmission with sufficient transmission time, eventually increasing the radio coverage. The conducted evaluation showed that signal could be recovered even in low S/N, thereby providing better coverage.
    Keywords: repetition; resource unit; single-tone; multi-tone; coverage enhancement.

  • Channel estimation for hybrid mmWave massive MIMO via low rank Hankel matrix reconstruction   Order a copy of this article
    by Yujian Pan, Zongfeng Qi, Jingke Zhang, Feng Wang 
    Abstract: The underdetermined model in the hybrid massive multiple-input and multiple-output (MIMO) brings challenges to channel estimation. This paper proposes a low rank Hankel matrix reconstruction based method for this problem. First, the channel is modelled as a superposition of finite complex exponential functions based on the millimetre wave (mmWave) channel sparsity in angular domain. Then, channel estimation is converted into seeking a low rank Hankel matrix of the channel. For the low rank matrix reconstruction, an inequality constrained nuclear norm minimisation problem is built, and an efficient alternating direction method of multipliers (ADMM) based algorithm is derived for solving this problem. The new method estimates the channel using only one pilot. It is gridless, efficient, free of path number estimation, and has no minimum angle separation requirement. Its performances are verified by simulations and compared with representative algorithms.
    Keywords: ADMM; channel estimation; hybrid massive MIMO; low rank Hankel matrix reconstruction; nuclear norm minimisation.

  • Performance evaluation of FBMC vs OFDM in tapped delay line doubly selective channels   Order a copy of this article
    by Ritesh Baranwal, B.B. Tiwari 
    Abstract: As the technology grows in wireless communication, we move towards 5G. In mobile communication, requires a greater number of users in limited bandwidth, and also the technology used is easily accessible to all. For these requirements, several types of research have been conducted to fulfill these requirements. One of the researches is to find a suitable waveform for 5G. But requirements for 5G waveform are very low out-of-band (OOB) radiation, low Peak to Average Power Ratio (PAPR), low reliability, and low latency communication, enhanced mobile broadband. This paper investigates filter Bank multicarrier (FBMC) waveform in time and frequency selective channel used Gabor Theory. Performance evaluation for the parameters like Bit Error Rate (BER), PAPR, Power Spectral Densities for FBMC and OFDM in various real-time doubly selective TDL-A, TDL-B, TDL-C, pedestrian channel, vehicular channel as suggested by the Third Generation Partnership Project (3GPP) has been done.
    Keywords: 5G; Python; FBMC; OFDM; PHYDYAS; vehicular communication; TDL channel.

  • An adaptive sub-pixel edge detection method based on improved Zernike moment   Order a copy of this article
    by JiaDi Mo, He Yan, JiHong Liu 
    Abstract: Sub-pixel edge detection is one of the most basic procedures in the field of vision measurement as an important step for high-precision measurement. For the traditional Zernike moment-based sub-pixel edge detection algorithms, it's difficult to obtain a suitable grayscale threshold for different images that largely affects the accuracy of vision measurement. This paper proposes a new sub-pixel edge detection method based on improved Zernike moment, which is an adaptive, robust and effective method for high-precision measurement. The ideal step edge is modelled in three-grey-step edge model, for the solution of edge parameters, only two Zernike moment are required. According to the characteristics of greyscale in three-grey-step edge model, the greyscale of noise and edge can be clarified into two categories to obtain a suitable threshold according to k-means clustering. Experimental results show that the proposed method can obtain an appropriate greyscale threshold, and has good performance in locating edges.
    Keywords: Zernike moment; sub-pixel edge detection; adaptive threshold; three-grey-step edge model.

  • Research on modular design and manufacturing of ship anchor winch structure under artificial intelligence optimisation   Order a copy of this article
    by Can Wu, Shuguang Wang, Jinjun Long, Qiang Liu 
    Abstract: The anchor winch is mainly used for the operation of ship anchoring and berthing. The normal operation of the anchor winch ensures the safe and stable operation of the ship. The design and manufacture of the traditional anchor winch has significant problems of high cost and low efficiency, which cannot meet the development needs of the modern shipbuilding industry. This research introduces the concept of modular design, and determines the division of different modules by combining the artificial intelligence method of fuzzy clustering. According to the specific working principle of the ship anchor winch, the parameters of each stage in the operation process of the winch are calculated, and the overall structure design of the winch is completed. Finally, the 1020 anchor winch is selected for specific application analysis, and the basic design parameters and main module design parameters of the winch are calculated by modular design method. Finally, the number of outfitting is 1600, the anchoring depth is h < 82.5 m, and the coaxial bedroom single side double drum design is adopted.
    Keywords: ship; anchor winch; modularisation; artificial intelligence; cluster analysis.

  • Cluster head replacement strategy for wireless sensor networks Connectivity repairing   Order a copy of this article
    by Xibao Wu, Hao Wu, Jia Sun, Wenbai Chen 
    Abstract: In practical applications, the connectivity repair simulation of wireless sensor networks is quite different from the actual network performance. In view of this difference, we simulate a cluster head replacement strategy. Specifically, the cluster head node collects perceptual packet sent to the aggregation node. After the cluster head node is faded, the network starts performing repair action, and the repair node moves to the position of the cluster head node to replace the cluster head. Further, the cluster wireless sensor network is modelled and the network repair process after node failure is simulated by OPNET software. Network connection repair processes in 2D plane scenarios and 3D terrain scenarios are also compared. The experimental results show that the strategy of the cluster head can complete the repair of network connectivity, and the three-dimensional terrain field centralized node receives power higher than in the two-dimensional planar scene.
    Keywords: OPNET; wireless sensor network; terrain scene; cluster head replacement strategy.

  • Mapping between baseband units and remote radio heads based on channel prediction in CRAN   Order a copy of this article
    by Lilly Odwa, Yueyun Chen 
    Abstract: In cloud radio access networks, it is difficult to efficiently use the baseband unit resources. Among the existing methods to improve resource use, a centralised baseband unit pool with traffic load prediction has not been considered. To enhance spectral and energy efficiencies, this paper proposes the Predictive Borrow and Lend method, which maps between baseband units and remote radio heads based on channel prediction. The channel state is predicted based on regularised particle filters, and multiple remote radio heads are mapped to a single baseband unit depending on baseband unit capacity. The proposed method aims at maximising the number of busy baseband units, by combining remote radio heads to a single baseband unit. When baseband unit utilization exceeds the upper limit, the remote radio heads are switched to another baseband unit with low resource usage. The simulation results further approves that the spectral and energy efficiencies of the proposed method are significantly improved, compared with the borrow and lend method. Moreover, the delay is also reduced because the borrow and lend processing in baseband units is not needed in the proposed method, which is suitable for real-time applications.
    Keywords: cloud radio access network; resource allocation; regularized particle filter; baseband unit aggregation; spectral efficiency; energy efficiency.

  • Rapid detection and identification of major vegetable pests based on machine learning   Order a copy of this article
    by Changzhen Zhang, Yaowen Ye, Deqin Xiao, Long Qi, Jianjun Yin 
    Abstract: To develop strategies for vegetable pest control, information on the types of pest and the quantity of the pests is essential. In this study, an automatic pest monitoring system has been developed by combining remote information processing technology and machine vision technology. A Vegetable Pest Counting Algorithm Based on Machine Learning (VPCA-ML) was proposed and implemented in a vegetable field to monitor four major pests: Phyllotreta striolata (F.), Frankliniella occidentalis (P.), Bemisia tabaci (G.), and Plutella xylostella (L.). Results show that a bag-of-features model in the algorithm is feasible for representing pest features, and an improved SVM model is suitable for pest classifications. Compared with the manual counts, the VPCA-ML results in an overall relative error of less than 10%. The system based on the VPCA-ML can quickly and accurately acquire the types and dynamic quantities of major vegetable pests, and have a stable operation in a field environment.
    Keywords: machine learning; real-time detection; vegetable pests; identification and count.

  • PMix: a method to improve the classification of Xray prohibited items based on probability mixing   Order a copy of this article
    by Qitong Lu, Peng Han, Jian Qiu, Kunyuan Xu, Kaiqing Luo, Dongmei Liu, Li Peng 
    Abstract: In recent years, much more attention has been paid to the field of automatic unmanned security checks. However, its difficult to collect X-ray images about different types of dangerous items, and the X-ray imaging is characterised by a perspective view with a limited observation angle, which easily result in the X-ray image blurring, information loss, and mistakes of prohibited items recognition. Therefore, there are few data sets about X-ray dangerous goods image and data imbalance exists at the same time. To solve this problem, this paper proposes a probabilistic image mixed data processing method called PMix. This method can increase the data volume of positive samples without destroying the unbalanced proportion of data sets and improve the accuracy of the model. Compared to the baselines, the experiments show that the proposed method can effectively promote the accuracy of model classification by more than 5% under the condition of unbalanced data.
    Keywords: X-ray; Mixup; CutMix; data imbalance; multilabel classification.

  • Cold-start recommendation algorithm based on user preference estimation   Order a copy of this article
    by Biao Cai, Jiahui Xin, Xu Ou 
    Abstract: In order to improve the dilemma of collaborative filtering in the face of cold start and achieve a better balance between accuracy and diversity, this paper considers the influence of user characteristics on recommendation results and proposes a Preference Estimation Network (PEN) based on maximum likelihood. PEN uses the user's characteristic information to estimate the user's preference information, and represents the user's preference vector with the item's label system. On this basis, PEN-Rec, an improved version of the traditional recommendation algorithm based on preference vector estimation and particle swarm optimisation, is proposed. Finally, the PEN-Rec algorithm is compared with the benchmark algorithm on six public evaluation indicators using open datasets, and the experimental results show that the accuracy, diversity and novelty of the PEN-Rec algorithm are all improved.
    Keywords: recommendation; feature impact; preference estimation; label vector.

  • Enhancing the latency of mobile cloud gaming system   Order a copy of this article
    by Vaithegy Doraisamy, Saleh Ali Alomari, Rasslenda Rass Rasalingam 
    Abstract: The Mobile Cloud Gaming (MCG) is a large network that comprises cloud computing and wireless network services. It is a network that provides online gaming services to mobile users. This paper proposes a model that enhances the existing MCG model by introducing Mobile Cloud Cloudlet Gaming (MCCG) system to provide better gaming service to the mobile users by reducing the transmission latency by placing a cloudlet as an intermediate between the main cloud and mobile nodes. Besides that, the mobile nodes within the transmission area of the cloudlet can form an ad hoc network to share data among the devices when the cloudlet has reached its maximum capacity. The Clustered Destination-Sequenced Distance Vector (c-DSDV) protocol is also introduced in this paper for data transmission between the ad-hoc cloudlet and mobile nodes to further reduce the transmission delay. Besides reducing the latency, the model also could reduce congestion and enable low-end mobile devices to join MCG services.
    Keywords: cloud computing; cloudlet; DSDV; mobile cloud computing; latency; mobile cloud gaming.

  • Link prediction with Fusion of DeepWalk and node structural information   Order a copy of this article
    by Xinhui Xiang, Biao Cai, Yunfen Luo 
    Abstract: The existing link prediction algorithms are mainly based on structural information or network embedding, but minimal research has been conducted on the fusion of these algorithms. It is found that the structure-based algorithms have high accuracy, but the complexity is higher owing to the introduction of high-order structural information while the network embedding algorithms have low complexity, but because the structural information of the node is not fully used, the accuracy is not as good as some structure-based algorithms. Therefore, by combining the structural attributes of nodes and the degree of convergence between node pairs, this paper proposes two new improved similarity algorithms the similarity algorithm based on edge-degree DeepWalk cosine (EDDWC) and the similarity algorithm based on preferential attachment mechanism DeepWalk cosine (PADWC). Experiments show that the performances of the proposed algorithms are greatly improved over that of the DeepWalk algorithm, and they are also better than other link prediction algorithms.
    Keywords: link prediction; DeepWalk; edge-degree; preferential attachment mechanism; cosine similarity.

  • Radio signal modulation recognition algorithm based on convolutional neural network   Order a copy of this article
    by Liuxun Xue, Wenqi Zhang, ZhiYang Lin, Hui Li 
    Abstract: In modern communications, it is often necessary to correctly identify the modulation of a signal with almost zero a priori knowledge as a non-collaborator for subsequent work, such as demodulation and analysis. However, the traditional modulation identification process requires manual extraction of signal features, which is a tedious process with large uncertainties and cannot meet the efficiency requirements of modern communication. In order to avoid the drawbacks brought by manual processing, more and more scholars have started to use deep learning methods to extract features directly from the raw data of communication signals, which ensures real-time and robustness of automatic modulation recognition of communication signals. This paper is mainly based on the convolutional neural network algorithm to study the modulation recognition algorithm of radio signals. Firstly, this paper studies the format, parameters and channel model of the radio signal dataset, and construct a dataset in the format of I/Q data. Then, this article uses Convolutional Neural Network (CNN) algorithm to identify and classify the signal. Finally, to address the defects and training time problems in the classical convolutional neural network recognition method, this paper proposes an improved convolutional neural network named X-CNN. X-CNN is optimised in terms of connection structure, enhancement mechanism, and pooling method. The experimental results show that the improved convolutional neural network in the radio signal modulation pattern recognition, the recognition rate and the training process have obtained better performance gains, and the overall fitting ability of the network is improved.
    Keywords: convolutional neural network; deep learning; modulation recognition; radio signal.

  • Hop count, ETX and energy selection based objective function for image data transmission over 6LoWPAN in IoT   Order a copy of this article
    by Archana Bhat, Geetha V 
    Abstract: Internet of things (IoT) is technology that connects millions of things to the internet for collecting data and controlling things. 6LoWPAN looks promising for future IoT networks as it works with IPv6, which is essential to address millions of things. However, as the 6LoWPAN devices are resource constrained with payload constraint at the data link layer, it needs efficient mechanisms to send packets over IEEE 802.15.4 MAC layer. The challenge increases when the sensors used in the devices are camera or audio recordings. Multimedia data transmission over 6LoWPAN is great challenge, and this paper addresses the same with respect to selection of Objective Function (OF) for multimedia data traffic. A new hop count, ETX and energy selection based OF is proposed in this work. The proposed technique is compared with existing OF, and the simulation results shows that the proposed technique provides better performance.
    Keywords: 6LoWPAN; objective function; IPv6; multimedia; RPL; IEEE 802.15.4.

  • UAV path planning based on an elite-guided orthogonal diagonalised krill herd algorithm   Order a copy of this article
    by Renxia Wan, Fangxing Zhang, Tao Zhou 
    Abstract: At present, obstacle avoidance and the shortest path to the destination are the main problems that need to be considered when the UAV (unmanned aerial vehicle) performs its mission. In this paper, an elite-guided orthogonal diagonalised krill herd algorithm (EODKH) is developed to solve the path-planning problem of a UAV in three-dimensional space with complex terrain. EODKH algorithm divides the krill population into two parts: one part is composed of elite krill, and the evolution of krill population is guided by the experience of elite krill with orthogonal diagonalisation, and the elite krill with higher rank can obtain greater 'respect' in the evolution procedure; the other part evolves according to the original krill herd method, so as to improve the convergence diversity of the whole krill herd. Experimental results show that EODKH is better than other krill herd algorithms in UAV path planning.
    Keywords: krill herd; elite krill; orthogonal diagonalization; fitness value; unmanned aerial vehicle; path planning; three-dimensional.

  • Research on application of spatial attention mechanism in the super-resolution reconstruction of single-channel greyscale image of mouse brain   Order a copy of this article
    by Yanan Wang, Yi Luo, Jia Ren 
    Abstract: High-resolution medical images are an essential basis for scientific research and clinical judgement. At this stage, saq is still limited by the hardware conditions of imaging equipment and the generalisation of imaging methods. An economical method would involve the application of Single Image Super-Resolution (SISR) technology in the acquisition of high-resolution images in the biomedical field. Based on the classic super-resolution network, the cascaded residual plate aimed at CARN is studied into in this paper. Regarding the problem of the insufficient feature learning process, the sSE module of spatial incentives was introduced, and the network structure of CARN was optimised. The mouse brain nerve image reconstructed by the CARN-L algorithm is significantly improved compared with the original image.
    Keywords: medical image; image super-resolution; deep learning; cascade residuals.

  • Power allocation to enhance non-orthogonal multiple access performance   Order a copy of this article
    by Dipa Nitin Kokane, Geeta Nijhawan, Shruti Vashist, 
    Abstract: Recently, researchers are having more focus to work in the field of non-orthogonal multiple access (NOMA) networks to address the emerging challenges of advanced wireless communication owing to its cost effectiveness to enhance spectral efficiency. This helps in increasing the capacity of advanced wireless systems without expanding the bandwidth. This paper investigates various power allocation schemes that are used in power domain non-orthogonal multiple access wireless communication and suggests an improved power allocation technique under optimum fairness constraints between users capacity. We also discuss spectral efficiency and energy efficiency tradeoff for proposed NOMA and conventional orthogonal system, and simulation results show that the proposed scheme for NOMA has improved results compare with different power allocation technique and orthogonal multiple access (OMA) system.
    Keywords: FPA; FTPA; NOMA ; OMA; power allocation.

  • Risk assessment for vehicle injury accidents in non-coal mines based on Bow-tie model   Order a copy of this article
    by Bo Wei, Yuan Li, Guixian Liu, Yi Zhao 
    Abstract: In order to reduce the incidence of vehicle injury accidents in non-coal mines, a quantitative risk assessment method based on the fuzzy bow-tie model is proposed. First, the bow-tie model of vehicle injury accidents in non-coal mines is established. Then, the fuzzy failure probability of vehicle injury accidents is calculated by using fuzzy set theory and expert evaluation method. Finally, the risk value of non-coal mines vehicle injury accident consequence is obtained based on fuzzy analytic hierarchy process. Taking a mine vehicle injury accident as an example, the results show that the probability of vehicle injury accident is 2.104
    Keywords: bow-tie model; vehicular injury accident in non-coal mines; risk assessment; fuzzy set theory; fuzzy analytic hierarchy process.

  • Research and application of spoken English evaluation system based on biological voiceprint feature recognition   Order a copy of this article
    by Ying Zhang 
    Abstract: The design of an oral English evaluation system based on biological voiceprint feature recognition technology can deal with the shortcomings of speech teaching in traditional computer-aided technology, and can improve the learning interest and learning effect of oral English learners. In this paper, the MFCC method is selected to construct the evaluation model of oral English, which can achieve good speech recognition effect. Through hierarchical recognition strategy, the algorithm mechanism of oral English is further optimised. Through simulation analysis, it can be seen that the proposed system design can objectively reflect the actual oral English level of oral English evaluators, and has good application effect and practical value.
    Keywords: biological voiceprint; voiceprint features; spoken English; MFCC evaluation.

  • A self-management mobile application system for patients with mild cognitive impairment and mild dementia   Order a copy of this article
    by Fadi Thabtah, Arun Padmavathy, Thanh Trung Thai, Daymond Goulder-Horobin 
    Abstract: The projected increase in the number of dementia cases has prompted the development of an automated solution to assist patients with self-management in everyday life. This paper proposes a new mobile application (app) called DDoMate to help users with mild cognitive impairment (MC) or mild dementia, and their caregivers, to organise tasks to improve the patients quality of life. DDoMate has been developed using the latest in research design for medical mobile apps for elderly people to ensure a carefree experience without hassle. DDoMate offers easy to navigate interfaces with large fonts and a dynamic environment that enables users to record their own voices for reminders to manage tasks. DDoMate is implemented in the Android environment and is accessible through the Google App Store. Data within the proposed app is anonymous and can be further analysed using machine learning to improve the self-management characteristics of mild dementia patients.
    Keywords: cognitive computation; data management; early dementia; digital informatics; mobile computing.

  • Fast trace ratio-based feature selection   Order a copy of this article
    by Bo Liu, Ling Ling Tao, Xi Ping He 
    Abstract: The quality of features has a great impact on machine learning tasks. Feature selection is to get a high-quality feature subset from data, which has been widely studied because of high interpretability. In this paper, we propose a novel feature selection algorithm, termed trace Ratio-Based Feature Selection (RBFS), which first defines the distance of different classes and same classes for a given sample, and then projects these distances into the subspace. The margin is defined by the trace ratio of these two distances. The objective function is formulated by maximizing the margin. To avoid a trivial solution, the orthogonal subspace and the $L_{2,1}$ norm are incorporated into the objective function, and then theoretically analyse that the rewritten objective function can get the optimal solution through alternating iteration. In addition, power iteration is introduced to reduce the computational cost. Comprehensive experiments are conducted to compare the performance of the proposed algorithm with six other state-of-the-art ones.
    Keywords: feature selection; trace ratio criteria; large margin; subspace learning.

  • Research on application of deep neural network model in College English skill training system   Order a copy of this article
    by Li Miao, Qian Zhou 
    Abstract: In view of the current situation that the College English skill training system has not yet realised the scoring of subjective questions, the proposed system adopts the deep neural network model and linear regression model to realise the automatic scoring of subjective questions of oral English expression. The results show that the automatic scoring system for college students' English skill training can effectively improve the accuracy and coverage of the six dimensions of lr-dnn fusion model. The 1-6 dimensional accuracy of lr-dnn fusion model is 93.51%, 92.45%, 9037%, 92.36%, 91.81% and 92.15%, respectively. The results are of great value to improve students' English skill training system.
    Keywords: deep neural network; language training; scoring; linear regression; intelligence.

  • The internet of things enabling communication technologies, applications and challenges: a survey   Order a copy of this article
    by Sihem Tlili, Sami Mnasri, Thierry Val 
    Abstract: Recently, the IoT has gained great importance, which results in the evolution of communication technologies to meet the needs of different IoT applications. In addition, several domains integrate the IoT technologies in several applications of our professional life and daily activities. However, the IoT still presents many challenges and issues. This article describes in detail the emerging communication technologies used in the IoT networks and enumerates the common domains of their application. It also describes the main challenges of the IoT and its use in order to exploit its advantages.
    Keywords: internet of things; communication technologies; IoT domain applications; IoT use cases.

  • Research on construction engineering safety early warning based on BP neural network   Order a copy of this article
    by Linyu Liu, Xuejie Wang, Congru Zhang, Xikun Li 
    Abstract: In the process of China's reform and opening up, the development of the construction industry has greatly promoted the development of China's economy, and the construction industry is also one of China's pillar industries. Owing to the particularity of the construction industry, construction engineering safety accidents occur frequently and there are many factors that influence such accidents. In order to break through this limitation, this research introduces an adaptive learning algorithm with additional momentum to optimise and improve the model. Through the simulation analysis of 10 groups of sample data of MATLAB, it can be seen that the proposed construction engineering safety early warning model has good convergence and accuracy, and can be applied to specific engineering practice.
    Keywords: BP neural network; building construction; project safety; early warning.

  • A novel convolution kernel based robust watermarking scheme applied in medical image   Order a copy of this article
    by Junhua Zheng, Jingbing Li, Jing Liu, Mengxing Huang, Yen-Wei Chen, Uzair Aslam Bhatti 
    Abstract: When the existing digital watermarking algorithm is applied to the medical image, the watermark embedded in the region of interest (ROI) will change the properties of the original medical image, causing misdiagnosis. To solve these problems, a robust watermarking method for medical images based on feature extraction of convolution kernels was proposed in this paper. First, the embedded watermark is encrypted by Logistic mapping to improve the security of the transmission of patients' personal information on the Internet. Then, the convolution kernels are used to perform feature extraction on the original medical image to increase the algorithm's ability to resist geometric attacks. Finally, the zero-watermarking technology is used to organically combined the watermark and the medical image to prevent the region of interest of the medical image from being damaged by the watermark. Experimental results show that the algorithm has good performance on security and robustness.
    Keywords: digital watermark;medical image;convolution kernel;zero watermark;feature extraction.

  • Energy-efficient cooperative spectrum sensing for detection of licensed users in a cognitive radio network using eigenvalue detector   Order a copy of this article
    by Samson Ojo, Zacheaus Adeyemo, Festus Ojo 
    Abstract: Spectrum Hole Detection (SHD) in a Cognitive Radio Network (CRN) is of great importance to prevent licensed users from harmful interference. However, channel impairment affects the SHD resulting in interference. The existing Cooperative Spectrum Sensing (CSS) used to solve this problem suffers from large reporting overhead, resulting in energy inefficiency. Hence, this paper proposes an Energy-Efficient CSS (EECSS) for SHD in a CRN using different Secondary Users (SUs) to carry out local sensing with eigenvalue detector. The received signals from the primary user form a square matrix to determine the ratio of maximum to minimum eigenvalue. The SUs form clusters to reduce the reporting overheads, which are combined at the Cluster Head (CH) using the majority fusion rule. The proposed technique is simulated using MATLAB software and evaluated using Probability of Detection (PD) and Sensing Time (ST). The results obtained show that EECSS gives better performance than CSS with higher PD and lower ST values.
    Keywords: eigenvalue detector; probability of detection; sensing time; secondary user; primary user; cluster.

  • Adaptive group Riemannian manifold learning for hyperspectral image classification   Order a copy of this article
    by Haoxiang Tao, Xiaofeng Xie, Rongnian Tang, Wen Feng, Jingru Li, Youlong Chen, Hou Yao, Gaodi Xu 
    Abstract: Hyperspectral image classification is an important topic in the field of remote sensing. However, the high dimensionality and high spatial-spectral correlation of hyperspectral image will easily leads to poor classification performance due to the Hughes phenomenon. In this paper, we proposed a adaptive group local Riemannian embedding, called AGLRE, to extract the spatial and spectral features from hyperspectral image. It firstly mapped original data into a Riemannian manifold by constructing region covariance matrices for each pixel of hyperspectral image. And the multiple local tangent space on Riemannian manifold were learned by adaptive neighbourhood strategy. Lastly, the local linear embedding was applied to reduce the dimensionality and merge multiple tangent space into a global coordinate. Experimental results on public hyperspectral dataset shown that the proposed method can achieve higher classification performance than other competing algorithms.
    Keywords: hyperspectral image; Riemannian manifold; classification; dimensionality reduction; adaptive selection.

  • Real-time detection system of bird nests on power transmission lines based on lightweight network   Order a copy of this article
    by Haopeng Yang, Enrang Zheng, Yichen Wang, Junge Shen 
    Abstract: In response to real-time detection requirements for bird nests and other hidden danger on power grid transmission lines, this paper proposes a lightweight real-time detection system of bird nests. In terms of bird nests on transmission towers, there are many small targets, which may lead to possible loss of data. Thereby, the algorithm detects small targets of bird nests through three scales: low, middle and high scales. At the same time, the DIoU-NMS calculation method is used to make the prediction box closer to the real box. The average accuracy of the improved algorithm is 90.05%, which is 7.38% higher than the original one. The detection speed of the detection system of bird nests in NVIDIA Xavier NX, an embedded device, is 26.3 FPS. With higher detection accuracy and real-time detection speed, the requirements of high-precision and real-time inspection of the state grid in line inspection can be met.
    Keywords: bird nests detection; lightweight network; multi-scale fusion; attention mechanism; non-maximum suppression; object detection.

  • Real-time fire detection and alarm system using edge computing and cloud IoT platform   Order a copy of this article
    by Guo Chenglin, Yong Bai, Mei Wu, You Zhou 
    Abstract: Fires often cause huge loss of personnel and property, hence it is very important to monitor fires and send alarms to users in real time. With the development of the IoT (Internet of Things) technology, intelligent edge devices can reduce delay in fire detection. This paper proposes a real-time video fire monitoring and alarm system based on edge computing and cloud IoT platform. The intelligent edge device is implemented based on Nvidia Jetson Nano with object detection network YOLOv5s deployed for fire detection, where YOLOv5s is accelerated by TensorRT and DeepStream. Then real-time message notification is performed with a local server, and fire event and metadata can be further delivered to the Azure IoT platform. The experiment demonstrates that our system is effective for real-time fire detection and message notification.
    Keywords: fire detection; IoT; DeepStream; edge computing; deep learning.

  • Targeted sentiment classification with multi-attention network   Order a copy of this article
    by Xiao Tian, Peiyu Liu, Zhenfang Zhu 
    Abstract: Targeted sentiment classification aims to recognise the sentiment polarity of specific targets. However, existing methods mainly depend on a crude attention mechanism, while neglecting the mutual effects between target and context. In order to solve this problem, this paper introduces a multi-attention network (MAN) for aspect level sentiment classification. We jointly modelled intra-level and inter-level attentional components to capture the interaction between target and context. The former attention mechanism pays attention to the context relation, whereas the latter attention mechanism considers important parts in a sentence. The experiments conducted on Laptop, Restaurant and Twitter datasets indicate that our model surpasses the baseline model.
    Keywords: attention mechanism; self-attention; targeted sentiment analysis; emotion analysis; neural network.

  • Research on heavy truck recognition algorithm based on deep learning   Order a copy of this article
    by Huan Wang, Dun Zhang, Zhikai Huang 
    Abstract: The Chinese economy has developed rapidly by taking advantage of convenient and rapid freight transport. Heavy trucks have always attracted much attention as the leading force in freight transportation. Although heavy trucks have significant advantages in freight transportation, their high emissions cause air pollution and high accident rates, which have always been criticised. Monitoring of heavy trucks has been improving in China, and this paper adopts deep learning to identify heavy trucks to strengthen their supervision. Given the multi-target recognition problem in the actual scene, this paper uses a one-stage algorithm for target detection. The representative network SSD (Single Shot Multi-Box Detector) and YOLO (You Only Live Once) are compared. YOLO adopts the YOLOv5s structure, and the SSD network is subdivided into two networks by replacing the backbone structure. One backbone structure is VGG, and the other is Mobilenetv2. The final experimental results show that the SSD network with VGG as the backbone structure achieves the best map value of 93.24%, which is 6.02% and 8.14% higher than the SSD and YOLOv5s training models with Mobilenetv2 backbone structure, respectively.
    Keywords: convolutional neural network; deep learning; target detection; heavy truck recognition.

  • Weld defect detection of power battery pack based on image segmentation   Order a copy of this article
    by Bo Tao, Fuqiang He, Quan Tang, Zhinan Guo, Hansen Long, Shidong Li, Yongcheng Cao, Guijian Ruan 
    Abstract: The safety and production efficiency are an important part of the power batteries production process and need to be considered seriously. Aiming at the welding quality of a power battery, a three-dimensional detection method based on the line laser sensor was proposed. Firstly, the depth data of the weld surface of the battery top cover is obtained by using a line laser sensor, and the defect area was segmented by using a multi thresholds segmentation method based on contour lines. Through the connected domain algorithm, the centres of defective areas are located. And the defect type is determined according to distance between the centres of the defect areas. Experimental results show that the detection rate reaches 97%, which indicates that the scheme has high detection accuracy and strong stability, and verifies the effectiveness of the method.
    Keywords: power battery; line laser sensor; threshold segmentation; connected domain; defect classification.

  • The application of improved UCT combined with neural network in Tibetan JIU chess   Order a copy of this article
    by Yajie Wang, Kai Liang, Jilin Qiao, Yanyan Xie 
    Abstract: Tibetan JIU chess is a kind of ethnic game, which is mainly popular in Tibetan areas of China. Because of the large state space, it is impossible to use the ready-made Upper Confidence Bound Apply to Tree (UCT) game model. Given the above problem, an algorithm combining neural network and UCT is proposed. First, the strategy value function is designed to assist the simulation of UCT to prevent the inaccuracy of random simulation results. Second, the multi-process method is introduced to improve the search efficiency of the game model. Third, we simplified the structure of the neural network and improved the two-dimensional Gaussian distribution matrix to facilitate the training of the network. The experimental results demonstrate that the win rate of this reinforcement learning model has been improved, which verifies the feasibility of the proposed method.
    Keywords: Tibetan JIU chess; UCT; neural network; strategy value function; multi-process.

  • Target detection and recognition method based on embedded vision   Order a copy of this article
    by Xiao Zhao, Qi Zou, Zhenjia Chen 
    Abstract: Vehicles have become an essential means of transportation in peoples daily lives. A large number of vehicles will need scientific and effective detection and management. The practical application of vehicle detection and recognition technology is imperative. The existing vehicle recognition technology is only for computer training and operation, and the application in the actual environment on the embedded platform may not achieve good results. If bad vehicle perspective or licence plate information is not detected, the effect is general. We propose a vehicle detection and recognition model for embedded platform and apply it to the actual environment. Support vector machine (SVM) uses histogram of oriented gradient (HOG) feature combined with window sliding detection to complete vehicle detection. On this basis, convolutional neural network (CNN) is used to realise licence plate recognition. Furthermore, oriented fast and rotated brief (ORB) feature extraction method is used to extract vehicle key information quickly and accurately. Licence plate information and vehicle ORB features are stored on the embedded device as the unique features of the vehicle. Moreover, ORB can be used to match the extracted information with the feature database, so as to identify the recorded vehicles. We have deployed to the embedded platform, with good timeliness, high accuracy and practical value, which can be applied to parking lot or high-speed detection port and other scenes.
    Keywords: embedded vision; vehicle recognition; feature extraction.

  • Remaining useful life prediction for lithium-ion battery using a data-driven method   Order a copy of this article
    by Zhiyang Jin, Chao Fang, Jingjin Wu, Jinsong Li, Wenqian Zeng, Xiaokang Zhao 
    Abstract: Accurate prediction of the remaining useful life (RUL) of lithium-ion batteries is one of the key technologies in the battery management system (BMS). To boost the prediction accuracy of Li-ion battery RUL, a data-driven approach is developed, through the combination of long and short-term memory (LSTM) and complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN). First and foremost, the battery capacity extracted from the National Aeronautics and Space Administration (NASA) battery data set is used as original data and the CEEMDAN is used to divide the original data into components of dissimilar frequencies. Then, the LSTM model is used to predict components of different frequencies. Finally, the CEEMDAN-LSTM prediction results are integrated to acquire the final prediction of the Li-ion battery RUL. The results show that the proposed method is superior for Li-ion battery RUL prediction.
    Keywords: Li-ion battery; RUL; LSTM; CEEMDAN.

  • Research on remote sensing image classification method using two-stream convolutional neural network   Order a copy of this article
    by Kai Peng, Juan Hu, Siyu Liu, Fang Qu, Houqun Yang, Jing Chen 
    Abstract: Owing to the lack of remote sensing image dataset and no regional pertinence in terms of characteristics for classification, we have published the remote sensing image of some areas in Haikou City, Hainan Province, and made a HN-7 dataset, which has the regional characteristics specific of Hainan Province. The HN-7 dataset consists of seven classes, of which the construction site and dirt road categories appear in the public remote sensing dataset for the first time. Owing to the limited quantity of the HN-7 dataset, we decided to train a small convolutional neural network from scratch for the classification task, by using a three-layer two-stream network for improving the accuracy of the neural network model; our model achieved 98.57% accuracy on the test set. We compared the accuracies of four common networks trained on HN-7, and the results showed that our model achieves the best performance.
    Keywords: classification method; remote sensing image; two-stream convolutional.

  • Apple's internetwork operating system and Google's Android in sub-Saharan Africa: the mobile internet services dimension   Order a copy of this article
    by Francis Osang 
    Abstract: We extended the technology acceptance model (TAM) to enable us carry out a comparative study of iOS and Android in the sub-Saharan Africa on mobile internet services for educational purpose in a natural end-user environment. To test the model, we conducted a survey of 180 students from a private university in Nigeria. We tested the exploratory factor analysis (EFA) on the latent variables to analyse test items measuring the constructs and the ordinary least squares multiple regression to analyse the results. We found a 75% and 36% predictive power of the model for iOS and Android, respectively. Additionally, perceived ease of use significantly influenced adoption decisions for both iOS and Android. Perceived cost, social norms and all other constructs except perceived enjoyment were supported for iOS and rejected for Android. Strategies that factors in reduced cost and ease of use must be considered if full penetration is to be achieved.
    Keywords: Android; iOS; decision to adopt; mobile internet; perceived enjoyment; perceived cost; social norms; i-Phone.

  • Poor and rich squirrel algorithm-based Deep Maxout network for credit card fraud detection   Order a copy of this article
    by Annu Paul, Varghese Paul 
    Abstract: This paper proposes a Poor and Rich Squirrel Algorithm (PRSA)-based Deep Maxout network to find fraud data transactions in the credit card system. Initially, input transaction data is passed to the data transformation phase, transforming data using Yeo-Johnson (YJ) transformation. Then, the feature selection procedure is done by the Fisher score for creating the unique and significant features. Next, based on the selected textures, the data augmentation mechanism is done using the oversampling model. At last, the fraud detection is carried out by the Deep Maxout network, which is trained by the proposed PRSA optimisation algorithm, derived by integrating Poor and Rich Optimisation (PRO) and Squirrel Search Algorithm (SSA). The integration of parametric features of the PRSA algorithm trained the classifier to update weights to generate the best solution by considering fitness measures. The proposed method achieved the best accuracy, sensitivity, and specificity measures of 0.96, 0.95, and 0.94, respectively.
    Keywords: credit card; deep learning; fraud detection; data augmentation; data transformation.

  • Chinese Sentiment Analysis with Multi-granularity Vector Representation and Multi-channel Network   Order a copy of this article
    by Suoyu Zhang, Yong Wang, Xinyi Lyu 
    Abstract: Aiming at the problem that normal vector representation method cannot fully represent the emotional semantic information contained in the text when dealing with Chinese text sentiment analysis task, a multi-granularity convolutional capsule neural network model is constructed. The input of sentiment analysis model is vector representation of the texts trained through the language model. As there exists the problem that a single language model is not enough to abstract text features, a multi-granularity text vector representation method based on BERT is proposed. The text vector representations with different granularities are input into the improved multi-channel convolutional capsule neural network. The capsule layer can associate the low-level and high-level text features, and it will extract information selectively through the dynamic routing algorithm to construct emotional and semantic features of the whole text. Multiple comparative experiments confirm that the method proposed in this paper will efficiently improve the accuracy of Chinese sentiment analysis.
    Keywords: multi-granularity; vector representation; multi-channel; convolutional; capsule layer.

  • A dynamic displacement map based on deep Q network to assist the rendering of stylised 3D models   Order a copy of this article
    by Hao Zheng, Houqun Yang, Mengshi Huang, Yizhen Wang 
    Abstract: In the process of rendering with NPR, there will often be a problem of reduced perception of the NPR effect caused by the fixed spatial structure of the static mesh. Therefore, this research first establishes a convolutional neural network to conduct supervised training on numerous hand-drawn target stylised images, and perform continuous angle recognition through stylized images then output its angle vector. Secondly, generating the displacement map in real time by inputting the observation angle through the fully connected neural network model. After that, sampling the real-time displacement map to dynamically generate the deformation of the model. In the end, the goal of breaking the model's sense of space and enhancing the NPR rendering effect can be achieved. Besides, the experimental results of this study verify the effectiveness of the method.
    Keywords: non-photorealistic rendering; displacement map; stylised rendering; deep reinforcement learning.

  • Robust zero-watermarking algorithm for medical images based on K-means and discrete cosine transform   Order a copy of this article
    by Wenxing Zhang, Jingbing Li, Uzair Aslam Bhatti, Mengxing Huang, Jixin Ma, Cheng Zeng 
    Abstract: To better protect patient information in medical images and improve the security of medical image transmission, this paper studies a robust watermarking algorithm for medical images based on K-means and discrete cosine transform (DCT). Firstly, the watermark is preprocessed by chaotic encryption to make it more secure. Then use the K-means clustering algorithm to classify the grey values of the pixels in the medical image to obtain the feature image after the cluster segmentation; then use the DCT to extract the feature coefficient matrix and transform it into the feature hash sequence of the image. Finally, the zero-watermark technology is used to combine the feature hash sequence with the encrypted watermark to realize the embedding and extraction of the watermark. Experiments show that the algorithm not only can resist conventional attacks, but also has good robustness against geometric attacks.
    Keywords: medical image; K-means clustering; feature vector; DCT; zero watermark.

  • Study of the monitoring system for double row steel sheet pile cofferdam engineering   Order a copy of this article
    by Jianjun Wang, Guiqin Liu 
    Abstract: The construction monitoring and control standard of the steel sheet pile cofferdam is still the standard of the foundation pit of civil structures. In this paper, by referring to the literature related to double-wall steel sheet pile cofferdam, the detection content, structure calculation method and error analysis of double-wall steel sheet pile cofferdam project are summarised systematically. To ensure the safety of the cofferdam structure, the automatic monitoring system of steel cofferdam is established to realise the organic combination of real-time monitoring and control of steel cofferdam, so as to take timely measures to ensure the smooth progress of the project and analyse the error which will be used to improve the model and algorithm. Finally, the calculation error of between the model and algorithm will shrink and be smaller, making the results more reliable.
    Keywords: cofferdam; monitoring; safety.

  • Energy efficient sink relocation using whale optimisation technique in virtual grid based wireless sensor network   Order a copy of this article
    by A. Keerthika, Victor Berlin Hency 
    Abstract: Wireless Sensor Network (WSN) is an efficient network for monitoring and recording the physical environment and transfers the monitored data into the central location using widely distributed sensor nodes. One of the main problems in WSN is the issue of developing an energy-efficient routing protocol that achieves less energy consumption and enhances the lifetime of the network. During the past decades, researcher used the mobile sink to reduce the energy problem and hotspot problems. In this work, Virtual Grid-Based Energy Efficient Sink Relocation (VGESR) is proposed to solve these issues. The grid clustering is achieved by employing K-means clustering. After clustering, the Leader Node (LN) selection is done by calculating the Acceptability Factor (AF). Acceptability factor is calculated based on the nodes residual energy, available bandwidth and Received Signal Strength (RSS). Whale Optimisation (WOA) technique is employed for the optimal sink relocation based on the fitness value of the nodes. The results obtained from the simulation prove that the proposed VGESR performs well in terms of life span and energy use. The proposed VGESR simulation is performed with the omnet++ tool.
    Keywords: acceptability factor; clustering; network lifetime; sink relocation; whale optimisation; wireless sensor networks.

  • Direction-of-arrival estimation for partially polarised signals with switch-based multi-polarised uniform linear array   Order a copy of this article
    by Yujian Pan, Jingke Zhang, Zongfeng Qi 
    Abstract: In this paper, a new switch-based multi-polarised receiver architecture and two compatible direction-of-arrival (DOA) estimation algorithms are proposed for the partially polarised signals. In the receiver, each polarised element in an antenna is connected to a common radio frequency (RF) chain via a switch, which reduces the number of RF chains. For DOA estimation, an ESPRIT-based algorithm and a joint annihilation-based algorithm are proposed. The ESPRIT-based algorithm is based on summing the covariance matrices of different polarized outputs, and the joint annihilation-based algorithm is based on annihilating different polarized outputs by a common filter. Compared with other algorithms, the ESPRIT-based algorithm, which only takes about 91 us to perform one estimation, is more efficient, and the joint annihilation-based algorithm, which can approach the Cramer-Rao lower bound (CRLB), is more accurate. It is also concluded that the tri-polarised uniform linear array (ULA) can offer more accurate estimation than the dual-polarised ULA.
    Keywords: direction-of-arrival estimation; ESPRIT; joint annihilation; multi-polarised array; partially polarised signal.

  • Empirical study on the algorithms of food cold chain logistics for multi-regional and large-scale athletic sports   Order a copy of this article
    by Zixia Chen, Hanrui Lyu, Hanmin Zhu, Jianliang Peng 
    Abstract: Distribution of food for multi-regional distributed large-scale events generally involves the optimisation of multi-input and multi-output logistics system of the secondary distribution network. Based on the location distribution of 43 venues and the layout of 17 front-end warehouses determined by hierarchical clustering analysis algorithm, the innovative time-distance saving algorithm, considering the influence of distance and time at the same time, is designed to optimise the design and empirical research on the real delivery lines of the inner-ring and outer-ring from the food cold chain logistics centre to the front-end warehouse. Using the OMAP+Autonavi navigation tool and comprehensively considering the weighting factors of delivery time and delivery mileage, the satisfactory solution of the distribution of multiple actual delivery lines was obtained. The simulation results showed that the total delivery mileage and time savings of the optimised food cold chain logistics network are significantly better than those of traditional saving algorithms.
    Keywords: hierarchical clustering analysis algorithm; time-distance saving algorithm; cold chain of food materials; optimisation of delivery route; empirical analysis.

  • Computation offloading using K nearest neighbour time critical optimisation algorithm in fog computing   Order a copy of this article
    by Ashwini Kumar Jha, Minal Patel, Tanmay Pawar 
    Abstract: The wide range of IoT devices and wireless devices used in health care, hospitals, and the enterprise generates a large volume of digital data that must be processed, analysed, and stored. Owing to the small processing capacity of these devices, the data generated cannot be processed on-board. Therefore, we suggest offloading this data to an efficient server. Time-critical applications cannot rely on the availability of cloud servers since they are in a remote location. The paper examines algorithms such as Deep Reinforcement Learning for Online Computation Offloading (DROO), Coordinate Descent, Adaptive boosting, and then implements the K nearest neighbour time critical optimization algorithm as a fog offloading network topology. The offloading decision is based on the cost function, which includes latency, memory consumption, and model accuracy. The topology implementing KNN can be trained quickly and offers almost 99% accuracy when it comes to data offloading. Based on the comparative analysis, it excels over other machine learning approaches.
    Keywords: fog computing; edge computing; computation offloading; cloud computing; K-nearest neighbour.

  • A study on dual-sense broadband circularly polarised monopole antenna for UWB applications   Order a copy of this article
    by Umesh Singh, Kalyan Mondal, Rajesh Mishra 
    Abstract: The proposed work is designed with the embedded of stubs and Parasitic Strips (PSs) under the radiator. An FR4 substrate is used to design the antenna (?_r= 4.4, h = 1.6 mm). The overall size of the antenna is 0.8?_0
    Keywords: monopole; CP; dual-sense; stubs; ARBW; satellite;.

  • Pattern recognition of surface electromyography based on multi-scale convolutional neural network with attention mechanism   Order a copy of this article
    by Beibei Wang, Hui Zheng, Jing Jie, Miao Zhang, Yintao Ke, Yang Liu 
    Abstract: Natural control methods based on surface electromyography (sEMG) pattern recognition have been widely applied in the field of hand prostheses. However, the control robustness and accuracy are difficult to meet many real-life applications. This paper proposes a multi-scale convolutional neural network (MSCNN) model based on the attention mechanism, which can automatically learn gesture features through convolution. The model generates features through convolution kernels of different sizes to achieve the fusion of features of different degrees firstly. After that, the attention mechanism is used to calculate the weights of different scales, and then the fused comprehensive features are obtained. The proposed model has been verified on the SIA_delsys_16_movement and NinaPro datasets. The experimental results showed that the proposed model has better classification accuracy, and the attention mechanism can validly improve the classification performance of the convolutional neural network.
    Keywords: surface electromyography; convolutional neural network; gesture recognition; machine learning; attention mechanism.

  • Systematic analysis and review of trust management schemes for IoT security   Order a copy of this article
    by Shilpa Vijayrao Shankhpal, Brahmananda S H 
    Abstract: The advancements in Internet of Things (IoT) gained a massive interest amongst researchers by providing interesting services when huge number of entities connects together to solve complicated scenarios. In addition, the IoT network connects all the devices using internet connectivities. However, the IoT network faces many issues and one of them is the trust management. Several algorithms had been devised for trust management in IoT, but the problems persist in enhancing the IoT network. In this research, the review article provides the detailed review of 50 research papers presenting the suggested trust-based IoT security methodologies. Moreover, an elaborative analysis and discussion are made by concerning the year of publication, employed methodology, evaluation metrics, trust parameters, and the implementation tool. Eventually, the research gaps and issues of various trust-based IoT security schemes are presented for extending the researchers towards a better contribution of significant trust-based IoT network.
    Keywords: IoT; trust management; authentication; security; access control.

  • YOLO-C: a new surface defect detection model of steel plate based on YOLO optimisation   Order a copy of this article
    by Jia Chongliu, Wang Lu, Zhang Juan, Jian Yuemou 
    Abstract: It is necessary to detect steel defects in the manufacturing industry. Proper quality control can reduce problems arising from steel defects implemented by traditional digital image processing or sight inspection in the past. However, it cannot meet the requirements of accuracy and real-time for intelligent manufacturing lines. With the extensional combination of IT and manufacturing, related scholars used deep learning to detect the steel defects, despite the limitations of data set and application scenarios. We propose a detection model called YOLO-C, an optimisation model that combines convolutional neural networks (CNN) with attention mechanisms. The proposed model adds several attention mechanism modules in the CNN structure, and the attention mechanism module is used to enhance the performance of feature extraction. The experimental results show that the YOLO-C model proposed is superior to other models, and the precision reaches 86%.
    Keywords: YOLO; attention mechanism; defect detection; convolutional neural network.

  • Hardware implementation of approximate multipliers for signal processing applications   Order a copy of this article
    by Elango Konguvel, I. Hariharan, R. Sujatha, M. Kannan 
    Abstract: Multiplication is a complex and substantial arithmetic task involved in signal processing applications. The hardware complexity of the multiplier is always high when compared with any other arithmetic operation. Approximate multiplication is a common operation used in many signal processing applications for improved performance and low power computation. The proposed approximate multiplier design is based on the approximate 4-2 compressor and self-error recovery technique. A small modification of the truth table entries in the approximate 42 compressor shows performance improvement at a cost of small accuracy. The designed multiplier promises to have improved performance when compared with the earlier approximate designs. The computational errors arising because of this multiplication approximation can be considered as trade-off for the significant gains in power and area.
    Keywords: approximate computing; adders; multipliers; hardware; error analysis; VLSI design.

  • Fuzzy Borda combined model in small town sewage treatment process Alternative Selection   Order a copy of this article
    by Fang He, Bo Wu 
    Abstract: With the increasing pressure of sewage treatment in small towns, it is necessary to establish a set of systematic and objective evaluation system to select the most suitable sewage treatment technology in small towns. In this paper, according to sewage characteristics, sewage disposal difficulties, treatment requirements and development of small towns in our country, seven processes were selected as follows: A/O, oxidation ditch, CASS, SBR, biological contact oxidation, BAF and artificial wetland. The evaluation index system was set up through literature research and consulting experts. Comprehensive index method, analytic hierarchy process and weighted arithmetic average method were used to evaluate the priority of each process. The consistency was tested by Spearman rank correlation coefficient. Based on three single evaluation results, fuzzy Borda combination evaluation model was established to evaluate the priority of sewage treatment process. Finally, an example was introduced to prove the feasibility of the combination evaluation model.
    Keywords: small towns; sewage treatment process; combination evaluation; priority; fuzzy Borda method.

  • Smart water grid technology based on deep learning: a review   Order a copy of this article
    by Huan Wu, Lin Peng, Feng Jiang, Shuiping Cheng, Jie Chen, Linda Yang 
    Abstract: In recent years, the development of deep learning technology has made breakthroughs in computer vision, natural language processing, and other fields. The Smart Water Grid (SWG) technology based on deep learning has also been a hot area of research in recent years. It has achieved better performance in the related detection and prediction of urban pipe networks. Therefore, this survey paper presents an extensive review of the application of deep learning to several different issues related to the SWG. This paper emphasizes feasibility studies and summarizes the state-of-the-art development in this field from a technical point of view, which consists of pipeline leakage and burst detection, contamination source identification, and water demand forecasting. Furthermore, this paper also proposes challenges and future directions in these key research areas, demonstrating that deep learning based SWG technology is still an emerging and encouraging research field.
    Keywords: smart water grid; pipeline leakage and burst detection; contamination source identification; water demand forecasting; deep learning.

  • Segmentation of lung parenchyma based on a new U-NET network   Order a copy of this article
    by Cheng Liying, Jiang Longtao, Wang Xiaowei, Liu Zuchen, Zhao Shuai 
    Abstract: In this paper, the U-NET network was selected as the basic segmentation model. It was found in the experiment that the segmentation accuracy of U-NET for upper lung and lower lung parenchyma was low. In view of this phenomenon, a new network model, New U-NET, was proposed. It adds input images of the same depth and corresponding input images of different depths as additional information and directly adds them to the result of deconvolution, so that the network can obtain more feature information in the decoding process, and the original information will be retained completely. Experimental data show that the proposed New U-NET network model solves the problem of low segmentation accuracy of the original U-NET network segmentation model at both ends of lung, improves the segmentation accuracy of lung parenchyma on the whole, and verifies that the New U-NET network model is more suitable for parenchyma segmentation.
    Keywords: New U-NET; lung parenchymal segmentation; CT images of lung; deep learning.

  • DstNet: deep spatial-temporal network for real-time action recognition and localisation in untrimmed video   Order a copy of this article
    by Zhi Liu, Junting Li, Xian Wang 
    Abstract: Action recognition is a hot research direction of computer vision. How to deal with human action in untrimmed video in real time is a very significant challenge. It can be widely used in fields such as real-time monitoring. In this paper, we propose an end-to-end Deep Spatial-Temporal Network (DstNet) for action recognition and localization. First of all, the untrimmed video is clipped into segments with fixed length. Then the Convolutional 3 Dimension (C3D) network is used to extract highly dimensional features for each segment. Finally, the extracted feature sequences of several continual segments are input into Long Short-Term Memory (LSTM) network to find the intrinsic relationship among clipped segments to take action recognition and localization simultaneously in the untrimmed video. While maintaining good accuracy, our network has the function of real-time video processing, and has achieved good results in the standard evaluation performance of THUMOS14.
    Keywords: action recognition; action localisation; LSTM; C3D; untrimmed video.

  • Robust zero-watermarking algorithm for medical images based on Hadamard-DWT-DCT   Order a copy of this article
    by Mingshuai Sheng 
    Abstract: As the IT industry grows rapidly, information security is particularly important. Digital development is a double-edged sword for the medical field, which not only brings convenience to patients and doctors, but also has hidden dangers for information security. We propose a Hadamard-DWT-DCT-based zero watermark algorithm for medical images for the problem of privacy information leakage when medical images are transmitted on the Internet. First, the raw medical image is chunked using the Hadamard transformation and produces a coefficient matrix, which is then transformed into a wavelet coefficient matrix, which can be efficiently compressed and stored. The wavelet coefficient matrix is then DCT-transformed to finally obtain the eigenvector. Experimental results show that the proposed algorithm is not only highly robust, but the watermark can still be effectively extracted against different degrees of geometric and conventional attack interference, but also can encrypt the patient privacy information contained in the medical image.
    Keywords: Hadamard-DWT-DCT; invisibility; medical images; robustness; zero watermark.

  • Sub-optimal antenna selection technique over Weibull-Gamma fading channel for MIMO communication systems   Order a copy of this article
    by Selvam Paranche Damodaran, Vijayakumar Perumal, Ganesan Verappan 
    Abstract: In this paper, the orthogonal space-time block code (OSTBC) of the Multi-Input Multi-Output (MIMO) system is considered. The advantage of the MIMO system is higher multiplexing gain and diversity gain. Owing to the increased complexity of the MIMO systems, it is difficult to use the MIMO system for practical applications. In this paper, we considered the antenna selection technique to reduce the cost and complexity to achieve the desired gain. The multipath and shadowing degrades the system performance in wireless communication. To model the multipath and shadowing effects, the composite WeibullGamma fading (WGF) channel is considered in this paper. The performance of a MIMO communication system is assessed in this paper using antenna selection techniques (AST) over the WGF channel. The channel state information on the transmitter side (CSIT) can be used to improve the system capacity and error rate at the same time with reduced complexity of the hardware. The CSIT in AST is used for orthogonal space-time block code (OSTBC) of the MIMO system over the WGF channel to improve the bit-error rate (BER) of the system. The SNR performance improvement is discussed over Weibull and gamma fading parameters. In this paper, both optimal and suboptimal AST analysis is derived for various numbers of antennas and capacity improvement of the channel is analysed. The simulations are done for the proposed sub-optimal algorithm to show the performance in high SNR region with fewer antennas selected for transmitting. The suboptimal AST shows the desired capacity improvement with less complexity.
    Keywords: MIMO communication; antenna selection techniques; Weibull—Gamma fading; channel state information on the transmitter side.

  • Split and rotated microstrip patch antenna with improved performance   Order a copy of this article
    by Josephin Pon Gloria Jeyaraj, Anand Swaminathan 
    Abstract: A novel low profile split and rotated microstrip patch antenna (SRMPA) with improved antenna performance is presented. The proposed antenna overcomes the limitations of the conventional microstrip patch antenna (CMPA) such as low gain, less front-to-back ratio (FBR), and more spurious radiation by the ground plane by a simple alteration in the antenna geometry. Its performance is mainly based on the opening angle (D) between the two arms of the antenna. Because the near-field interactions between the two arms are stronger at smaller opening angles, the current and scattering field amplitude are expected to be maximum. Therefore, for D<90
    Keywords: SRMPA; surface current distribution; opening angles; antenna parameters.

  • A novel trust-based approach for intrusion detection architecture in wireless sensor networks   Order a copy of this article
    by Mr. Jeelani, Kishan Pal Singh, Aasim Zafar 
    Abstract: Wireless sensor networks (WSNs) is a new technology that can be used to monitor the environment. Because sensor nodes in wireless sensor networks are installed in an open environment, they are more vulnerable to attacks. The sensor network lifetime improvement is dependent on minimum energy use. Protection is also a major concern when it comes to designing protocols for multi-hop secure routing. The results based on trust have proven to be more effective in addressing malicious node attacks. In this article, we propose a novel trust-based approach for intrusion detection architecture (IDA) in a wireless sensor network that is called the trust-based approach for varying nodes with energy (TBNE) model. TBNE finds the misbehaving nodes in the network. The structure is based on the trust model for secure communication in WSN and improves the performance of nodes. The simulation has been done with QualNet 5.0 simulator.
    Keywords: wireless sensor network; throughput; packet delivery ratio.