International Journal of Nanotechnology (39 papers in press)

Regular Issues

• A simulation-based study on the disc brake temperature distribution for optimizing hole geometry
  by Shyam Sunder Sharma, Hariharan Raju, Pranay Singh Tomar, Rajesh Jangid, Rahul Khatri 
  Abstract: Disc brakes used in automotive are responsible for braking to ensure a smooth and safe ride. This study deals with the thermal analysis of a disc brake rotor under various geometry of holes cut on the disc rotor surface. The friction on the disc escapes in the form of heat from the surface of the disc rotor. The temperature observed on the surface of the rotor, because of the friction developed between the brake pads and the rotor is analysed using ANSYS 18.1. The rotor is designed by assuming appropriate parameters in SOLIDWORKS 17. The temperature distribution and total heat flux were observed using ANSYS 18.1. The analysis was carried out on different hole geometries i.e. circular, square, 3/4th circular, straight slots, and rotor without holes. The dissipation of heat was found better in disc rotor with holes as compared to rotor without holes. The simulation study shows that the slotted holes on the disc rotor has surface temperature i.e. 89.356
  Keywords: Automotive disc brake; Simulation; Hole geometry; Heat dissipation.
  
  

Special Issue on: Smart Bio-Signal Acquisition System

• Analysing behavioral and academic attributes of students using educational data mining  
  by Muhammad Umer, Saima Sadiq, Arif Mehmood, Imran Ashraf, Gyu Sang Choi, Sadia Din 
  Abstract: Educational data mining has attracted significant consideration over the last few years. Information that is stored online using educational systems is increasing tremendously. The online learning environment can be improved by analysing and mining this information to extract representative features about students' behaviour and academic skills. Various classifiers are investigated to analyse the prediction of students' academic performance based on the attributes from the Kalboard360 learning management system (LMS). The selection of significant features can substantially improve the prediction, hence, ANOVA and chi-square filter approaches and forward feature selection and backward feature elimination wrapper approaches are examined for their efficacy. Results reveal that an extra tree classifier can achieve an accuracy of 0.8755 when trained on backward feature elimination selected features. Wrapper approaches prove to be effective in determining the most significant attributes for student performance prediction. 'Visited resources', 'raised hands', 'relation', 'parent answering survey', and 'student absent days' are regarded as the most significant attributes to determine student performance. Education specialists and institutions can leverage these findings to improve the student learning process and enhance their academic performance.
  Keywords: metadata; educational data mining; feature extraction; ANOVA; chi-square filter; forward feature selection; backward feature elimination; extra tree classifier.
  
  
• Multi to binary class size based imbalance handling technique in wireless sensor networks  
  by Neha Singh, Deepali Virmani, Gaurav Dhiman, S. Vimal 
  Abstract: Wireless sensor networks are used various disciplines including healthcare, banking, transportation, ocean and wildlife monitoring, earthquake monitoring, and numerous military applications. Now-a-days, there is escalation in size of data which makes it unfeasible to analyze it with accuracy. There are numerous problems that are faced when detecting patterns between structured and unstructured data that are unworkable by humans, so to make computation fast, easy and accurate, Machine Learning came in existence. Machine learning is extensively used in Wireless sensor networks. To make a machine learn, a training dataset is required and output is predicted by testing the dataset. A dataset in wireless sensor network has multi-class in its dependent variable. This multi-class classification causes class imbalance problem. This paper proposes MBSCIH (multi to binary class size based imbalance handling) technique in wireless sensor networks to solve the class imbalance problem in multi-class classification. MBSCIH converts multi-class classification into binary-class classification. MBSCIH is applied on WSN-DS, NSL-KDD and KDD-Cup 99 datasets and is tested with five major machine learning algorithm: Naive Bayes, Random Forest, Decision Tree, Support Vector Machine (SVM) and k-Nearest Neighbour (KNN). The test method used for testing is 10-fold cross validation. Results indicate that the proposed method increases the existing efficiency by 15.13%, 0.28%, 0.01%, 0.01%, 0.12% for Na
  Keywords: wireless sensor network; multiclass classification; binary classification; intrusion detection; WSN-DS dataset.
  
  
• Optimising pharmacokinetics via ADMET, bioactivity of Zr substituted samarium-doped ceria nanomaterials  
  by Bradha Madhavan, Suvitha A, Nagaraj Balakrishnan, Ananth Steephen, Ajay P 
  Abstract: The properties of ADMEs, bioactivity and restricted molecular medicinal chemistry were analysed via Swiss ADME in order to facilitate drug development. The important activities of Absorption, Distribution, Metabolism, Excretion and Toxicity (ADMET) in optimising pharmacokinetics and assessing the Ce0.8-xZrxSm0.2O2?? compounds were estimated. Six physico-chemical characteristics, bioactivity radar, donor acceptors and molar refractivity are being tested. In search of significant improvement in CeO2 fluorite structure, a promising doping, co doping approach was adopted. By varying the mole ratios of samarium dopant and Zr substitution, the Ce0.8-x ZrxSm0.2O2 nanomaterials were synthesised through simple chemical processes. Basic electronic structure calculation, bonding graph and crystallographic representation were analysed via a materials project tool. From the bonding graph, material design, synthesis glitches and stability assessments from the energies were calculated. ADMET Predictor is a product tool that rapidly and reliably forecasts more than 40 properties, including dissolvability, log P, CYP digestion positions and mutagenicity of Ames. Hence, these efforts are promising in monitoring pharmacokinetics via bio-electronics sensors and interfaces.
  Keywords: bio sensors; bio-potential ceria; bio-electronics; bioactivity; pharmacokinetics.
  
  
• Predictive protein module based on PPI network and double clustering algorithm  
  by Sicong Huo, Quansheng Liu, Tao Lu 
  Abstract: Protein-protein interaction (PPI) is a kind of biomolecular network which plays an important role in biological activities. In order to improve the accuracy of protein function module prediction, obtain the protein function module and run timely, this paper proposes a predictive protein module based on PPI network and double clustering algorithm (ISCC), which considers the characteristics of PPI network and considers nodes as two-dimensional data points. First of all, improved Density-Based Spatial Clustering of Applications with Noise (IDBSCAN) determines the central cluster, and then uses Spectral Clustering (SC) to redivide the weight; secondly, CFSFDP and chameleon algorithm are used to filter the similarity of the central cluster, and Support Vector Machine (SVM) is used to get the final clustering result. Finally, the experiments are compared with CDUN, EA, MCL and MCODE in terms of accuracy, sensitivity, F value and the number of protein functional modules. The experimental results show that the F value of ISCCD is 70% higher than that of EA, the number of recognition modules is 257 higher than that of CDUN, and the running time is 494 s faster than that of MCODE
  Keywords: PPI network; protein function module prediction; clustering; SVM; IDBSCAN.
  
  
• Experimental analysis of boost converter performance with non-ideals for sustainable energy applications  
  by Chegireddy Naga Kota Reddy, Choppavarapu Sai Babu 
  Abstract: In sustainable energy applications, boost converters are preferred owing to the low voltage profile of sustainable energy sources. Because of their light weight, high reliability, and small size, boost converters play a vital role in biomedical portable devices. In this paper, boost converter performance is examined using nonidealities of semiconductor devices. The impact of such nonidealities on the different parameters is investigated. To compensate for the damage incurred by nonidealities, expressions for output voltage and duty cycle are derived. The expression for voltage gain has been modified as a side effect of these nonidealities. The design equation for the inductor is modified based on these nonidealities. Further, nonidealities based expressions for the average value and ripple of current flowing through the inductor have also been established. Ultimately, these theoretical findings are validated with the results of simulation in MATLAB and hardware setup. In the hardware, the desired output voltage of 24 V is obtained from a 12 V input source by enhancing the duty cycle from 0.5 to 0.5174.
  Keywords: sustainable energy sources; ideal boost converter; non-ideal Boost converter; output voltage; duty cycle; voltage gain; critical inductance; design of inductor; inductor current ripples.
  
  
• Feature extraction based on motor imagery EEG  
  by Qingjun Wang, Zhihan Lv 
  Abstract: In order to study the entire process of EEG signals, the feature extraction and classification of motion imaging EEG signals becomes very meaningful. This article aims to study the feature extraction based on moving image EEG. Today, more mature motor imaging EEG signals are mostly two-dimensional signals, and the recognition results for 3D or 4D are low. In this paper, the wavelet transform is used as the feature extraction method, and support vector machine is used as the classifier to study the recognition of the four imaginary movements of left and right and tongue and leg. The simulation results of 20 subjects show that the four EEG signals have significant time-frequency domain differences and the recognition result is above 80%, with good recognition effect.
  Keywords: motion imaging; electroencephalogram; wavelet transform; support vector machine; feature distribution.
  
  
• Visible property enhancement techniques of IoT cameras using machine learning techniques  
  by Subbiah Narayanan, G.H. Sastry, Shobha Aswal, Venkatadri Marriboyina, R. Sankaranarayanan, Varsha Ph.D 
  Abstract: Visual perception in low light is challenging owing to low signal to noise ratio and photon counts. Deep learning is a kind of machine learning that is revolutionising picture identification and computer perception. In this study, deep learning will be used to enhance low-light picture filtering. To do this, a literature review will be performed to gather inspiration for methods and features that may be applied to the final networks. A fully functioning deep learning picture-filtering system will then be created, allowing networks to be trained using guided learning and the filtered resulting images to be recorded to files. With its output pictures plainly showing it was filtering low-light shots, the network functioned effectively. To maximise the network's potential, it must be run for a longer length of time.
  Keywords: IoT cameras; visual improvement; signal to noise ratio; machine learning.
  
  
• Real time crop field monitoring system using agriculture IoT system  
  by Pankaj Agarwal, Deepthi Gorijavolu, Hanumat Sastry G, Venkatadri Marriboyina, D. Vijendra Babu, Kishore C K 
  Abstract: Modern agriculture systems use features up to the maximum extent. In this paper, a propose wireless smart automation (WSA) with IoT system that interfaces with users' mobiles for real time crop field monitoring and control through the internet anytime from anywhere in the world. It is mainly focused on adapting smart methodology for farming, field monitoring and enhancing crop production. It provides a low cost and reliable monitoring towards field crops through IP connectivity for accessing and controlling devices from a smart mobile app. In addition to this, proper security authentication is given to this system through the Arduino UNO microcontroller. The user can operate drones for different control purposes, such as pesticide spraying, crop size monitoring, and field water level. The major advantage of the proposed system is adapting automation technology for continuous monitoring of crop fields with an IoT system.
  Keywords: wireless smart automation; IoT system; Arduino; UNO microcontroller.
  
  
• Implementation of intrusion detection system and improvement using genetic algorithm  
  by Ke Huang, Bichuan Sun, Xianming Sun, Mohammad Shabaz, Rijwan Khan 
  Abstract: The rapid evolution of computer technology due to the vast services and applications has made people dependent on computer technology. As a result, there could prevailing threats that need to be addressed, while dealing with the networking background. Therefore, we require the security to assist the networking technology in revealing the vulnerability abuses against the uses of a computer or its applications. In todays interacting environment, an Intrusion Detection System (IDS) is one of the major security components. It uses the security tools in the traditional way, and firewalls are also a significant method. The IDS is a security system that provides effective methods for computer network safety. This paper addresses the detection rate maximisation and false rate minimisation that are a major problem owing to its inability to discover a particular attack. This problem is tackled by the Genetic Algorithm (GA) approach presented in this paper, using the fuzzy methods for the IDS development. As a robust technology, it is most commonly used for IDS design and is based on machine learning. It is a search algorithm based on natural selection and genetics principles. For the GA specific problem solution, the fittest survival principle is used by GA functions for better approximation generation. In our approach, two datasets are used to perform the experiments. In the first dataset, 137 attacks and 840 normal connections, 977 connections in total, are kept in dataset 1; and in dataset-2, 234 attacks and 744 normal connections, 978 connections in total, are included. For both experiments, the presented technique manages a high detection rate, high accuracy, and a low false alarm. Therefore, the proposed technique outperforms the existing techniques with 96.57% detection rate and 3.12% false alarm.
  Keywords: intrusion detection system; genetic algorithm; detection rate; false alarm; security system; intrusion prevention system; firewall; network-based intrusion. Detection(NIDS),Knowledge Discovery in Databases (KDD).
  
  
• Analysis of resting state functional magnetic resonance images for evaluating the changes in brain function depression  
  by Hao Yu, Ye Yuan, Ashutosh Sharma, Abolfazl Mehbodniya, Mohammad Shabaz 
  Abstract: Continuous emotion of sadness is habitually considered as Major Depressive Disorder (MDD) that has parallel signs like other mental illnesses. These parallel indicative features can frequently lead to suffering of depression and other psychological conditions, and therefore involve experts to predict such symptoms and the timely treatment of MDD in order to evade the adverse effects. Magnetic Resonance Imaging (MRI) has a vital role in deducing the pathologies related to MDD. This paper deals with the application of data collection for the characteristics of spontaneous brain activity in the basic state of depression in patients, using resting state functional magnetic resonance images (fMRI), and discusses the changes in the brain function during a depression stage. In this paper, 16 patients with depression underwent 5 minutes and 12 seconds of brain functional MRI scan, and the Hamilton depression scale was used to evaluate the severity of the condition. The ReHo software was used to examine local brain regions on the image data. It is revealed that the resting brain fMRI-ReHo method found that the abnormal brain function area of patients with depression included: left thalamus, left temporal lobe, left cerebellum, occipital lobe, and the spontaneous activity consistency of patients in these areas was reduced. This work is done by SVM approach that uses AUC value of 0.885 for prediction, and it outperforms the state-of-the-art methods in a brain abnormality prediction by a maximum improvement of 22.24% and minimum improvement of 13.75%.
  Keywords: data collection; magnetic resonance imaging; functional magnetic resonance images; resting state function; depression; brain function; Hamilton depression scale; default mode network.
  
  
• Lightweight and secure IoMT edge device architecture through computing base minimization and computing process optimisation  
  by Prateek Mishra, Sanjay Kumar Yadav, Amit Kishor, Ravi Kumar Sachdeva 
  Abstract: Internet of Medical Things (IoMT) edge devices are resource-limited medical devices in terms of battery, footprint and processing power. Owing to limited resources, security and performance in presence of bulky hardware and software resources are challenging. Bulky resources are insecure owing to higher attack surface area. Performance optimisation of IoMT edge devices needs optimum resource surface area, known as trusted computing base. Minimisation of the trusted computing base in terms of number and size decreases visibility to the unauthorised world and increases lightweight and security of the architecture. This paper presents an Arduino IDE and ESP32 micro controller based IoMT edge device architecture to minimise the computing base in terms of least number and size of resources for lightweight and secure architecture. Computing process optimisation is proposed using lightest secure IoT algorithm with minimum size key. Comprehensive and direct comparisons between existing and proposed architectures result in a tightly secure and lightweight IoMT edge device architecture. Owing to the resource-constrained nature of the IoMT edge device, the aim of this work is to propose a lightweight and secure IoMT edge device architecture using minimum computing base and computing process optimisation.
  Keywords: computing base minimisation; computing process optimisation; IoMT edge device; lightweight; secure IoT algorithm.
  
  
• A novel hybrid model for automatic diabetic retinopathy grading and multi-lesion recognition method based on SRCNN and YOLOv3  
  by Prasanna Lakshmi Akella, Kumar Rajagopal, Fadi Al-Turjman 
  Abstract: Automatic grading and lesion identification of Diabetic Retinopathy (DR) is important for researchers because it is the leading cause of diabetes. Due to diabetes, the tiny blood vessels within the fundus are damaged and multiple lesions such as microaneurysms, haemorrhages, hard exudates, and soft exudates appear in the retina and cause multiple vision-related complications, which can drive to total vision loss without early examination and treatment. For clinical screening and diagnosis of DR, retinal fundus images are commonly used. Fundus images were taken by operators with different levels of experience, however, have a broad variance in quality. Low-resolution images of the fundus raise the risk of misdiagnosis and makes it more difficult to observe clinically. In order to avoid low resolution fundus images and to be able to diagnose DR carefully, authors developed a new hybrid structure in our proposed system to ensure that DR detection and classification processes become much more precise and faster compared with existing models. In the image pre-processing stage, the proposed model adopts a Super-Resolution Convolutional-Neural-Network to enhance the pixel density of low-quality fundus images. In the next step, to identify the DR grade, an advanced deep-learning model called You-Only-Look-Once Version 3 is used. Finally, another You-Only-Look-Once Version 3 network stage is applied using a bounding box to recognize the multiple lesions in the fundus images. The proposed system is evaluated on an openly accessible MESSIDOR dataset, and the results show that the system achieves 96.89% overall accuracy for DR grading and 97.6% accuracy for lesion detection with a high detection speed of 5.6 seconds.
  Keywords: diabetic retinopathy; multiple lesions; diabetes; blood vessels; deep-learning; SR-CNN- YOLOv3.
  
  
• Gas chromatography-mass spectrometry determination of polycyclic aromatic hydrocarbons in oil fried quail meat vs rabbit meat  
  by Rabia Siddique, Amna Sarfraz, Ameer Fawad Zahoor, Shazia Naheed, Muhammad Faisal Manzoor 
  Abstract: Processing improves the microbiological profile of food, but also introduces carcinogenic compounds, such as polycyclic aromatic hydrocarbons (PAHs) in food items. It has been noticed that generation of these carcinogenic compounds can be reduced by the marination process. There is a difficulty to check PAH levels and make comparison of results when many variables are involved. The study analysed the concentration of PAH standards in black and brown quail meat, checked the effect of different recipes (in which different spices are used), and also compared the recipe-wise PAH concentrations. This study also focused to compare the PAH values in both quails (brown and black) as well as their PAH results with a previous study of rabbits. PAHs were analysed by gas chromatography mass spectrometry from 42 samples of black and brown quail meat samples. The maximum levels of naphthalene were noticed in recipe V (country fried kebab) (1.71
  Keywords: polycyclic aromatic hydrocarbons; quail meat; GC-MS; frying recipes; naphthalene; Tukey test.
  
  
• Comparative approach for discovery of cancerous skin using deep structured learning  
  by Varun Kumar, T. Sucharitha, R. Priyadarshini, N. Rajendran 
  Abstract: Skin cancer incidences have increased recently owing to ozone layer deterioration. UV rays immediately enter the human skin, causing skin cancer. Thus, a novel approach for early skin cancer detection using digital data and image processing is required. Skin cancer detection research has been quite active since 2016. To increase patient detection accuracy and early diagnosis, we use both machine learning and deep learning algorithms to identify skin cancer. In this model, we employ naive Bayes, decision trees, and KNN. A convolutional neural network (CNN) is a deep learning technology that may be used to automate skin cancer screening. In this work, we compare 93.54% model accuracy.
  Keywords: decision tree algorithm; KNN; skin cancer; deep structured learning; image processing.
  
  
• Design and analysis of power-efficient hybrid full adder using static CMOS and transmission gates  
  by Priyank Sharma, Sanjay Sharma 
  Abstract: Energy-efficiency and high performance are the key requirements for any designer to achieve while making a device. For any controller or processor, the arithmetic logic unit is of utmost importance. The addition is the basic and necessary operation on which device performance depends. This paper presents the analysis of a low power hybrid full adder with the two commonly used logic techniques, Static CMOS adder and transmission gate adder circuit. This hybrid full adder circuit has been implemented by both the static CMOS and the transmission gate. This is a unique design of a hybrid full adder for the application of low power VLSI circuits. Using these two logics, the transistor count is reduced. The goal of this research paper is to influence the power, noise, and delay of the proposed full adder with the two commonly used TG and standard CMOS adders. All the three full adder circuits are simulated in the cadence virtuoso with GPDK at 45 nm.
  Keywords: transmission gate; CMOS; noise; low power; full adder; delay.
  
  
• A novel SVM and LOF-based outlier detection routing algorithm for improving the stability period and overall network lifetime of WSN  
  by Tripti Sharma, Amar Kumar Mohapatra, Geetam Tomar 
  Abstract: Wireless sensor network data are frequently erroneous owing to inevitable environmental factors such as intrusion attacks, signal weakness, and noise, which may vary depending on the situation. Outlier detection, often known as anomaly detection, is a technique for detecting anomalies and recognising noisy data in the aforementioned scenarios. In the proposed work, efforts have been made to design a routing algorithm that can detect anomalies based on LOF and SVM and is more energy-efficient. The primary objective of the proposed algorithm is to design an energy-efficient routing algorithm that is capable of detecting anomalies present in the environment with improved stability period and overall network lifetime. The sensor dataset provided by the Intel Berkeley Research Lab was simulated to assess the suggested approachs efficiency and competency. The simulation results reveal that this identification of anomalous nodes leads to the development of a more energy-efficient routing algorithm with a better stable region and a higher network lifetime. The proposed algorithm gives the best result with LOF. However, SVM with a gamma of 0.0005 could be used successfully in densely deployed wireless sensor networks. The LOF gives a 98% accuracy in finding anomalies present in the dataset chosen for the simulation.
  Keywords: support vector machine; local outlier factor; clustering; wireless sensor network; routing; energy efficiency; anomalies; K-means clustering; outliers; data transmission.
  
  
• Deep learning technique in CT image reconstruction and segmentation: a systematic literature review  
  by Shailendra Tiwari, Manju Devi, Sukhdip Singh 
  Abstract: Deep Learning (DL) in Computed Tomography (CT) is an important research area in computer vision and it provides fast advancement in the field of medical imaging. DL enables automated extraction of features and real-time estimation, whereas the traditional image reconstruction methods approximate the inverse function based on historian parameters to maintain reconstruction efficiency. This systematic literature review is based on last five-year data with the help of digital libraries (IEEE, ACM, Springer, Wiley, ScienceDirect) to find the research articles. The final work includes a systematic mapping report of the selected 88 research articles after applying the inclusion exclusion technique. This paper describes the basis of nine research questions, which include deep learning methods, framework, parameters etc., used in this literature. It concludes by highlighting the challenges of DL in the area of medical imaging, particularly in application of reconstruction and segmentation, and potential future development in the area.
  Keywords: medical imaging; image reconstruction; image segmentation; deep learning.
  
  
• Health information transmission system with less error rate based on wireless network  
  by Jing Zhang, Yanfang Liu 
  Abstract: Aiming at the problem of high bit error rate in traditional health information transmission systems, a health information transmission system of less error rate specialty based on a wireless network is designed. In terms of hardware, the accompanying cable, CI slot wireless network card, and information server are designed. At the same time, we obtain the information transmission time series of the final medical database, set the information transmission standard as IEEE 802.11b based on the wireless network, calculate the information transmission entry parameters, design the transmission status code, and then realise the information transmission of the final health data, and complete the system design. The experimental results show that the designed system has a lower transmission error rate and can solve the problem of high transmission error rate in traditional health information transmission system
  Keywords: wireless network; healthcare and literature; final examination wireless information; information transmission system.
  
  
• Design of IOT-aided prevention and control platform for major public health emergencies  
  by Yanfang Ma, Chunmeng Lu, Cunhong Li 
  Abstract: In view of the low traceability rate of traditional major public health emergency prevention and control platform, a new type of major public health emergency prevention and control platform based on the internet of things is designed. The information of major public health emergencies is collected, and the data is transmitted through the internet of things. The federal learning neural network is used to calculate the risk of major public health emergencies, analyse the data, and visually process the data to determine the type of prevention and control, so as to realise the auxiliary prevention and control of major public health emergencies through the internet of things. The experimental results show that the traceability rate of the experimental group is significantly higher than that of the control group, which can solve the problem of low traceability rate of traditional prevention and control platform.
  Keywords: auxiliary prevention and control platform; internet of things; public health emergencies;.
  
  
• Design of health system based on collaborative filtering algorithm  
  by Zheng Yi 
  Abstract: In view of the imperfect health system in Chinese society, this paper puts forward the design of a health management system based on collaborative filtering algorithm. In the original hardware system structure, we add automatic reset circuit, detect circuit state, ensure charge stability, avoid resource information outflow, reduce the interdependence between modules, use collaborative filtering algorithm, improve the system management function structure, use fitness function, refine the curriculum management scheme, and thus complete the design of the health management system. The test results show that the original system cannot meet the needs of users for the management of health and physical fitness test, and this system makes up for this defect and provides convenience for society to work and study.
  Keywords: collaborative filtering; electronic engineering; physical education; health management; teaching management.
  
  
• Interactive e-health care design system based on artificial intelligence technology  
  by Caibo Wang, Huanhuan Ge, Yaxing Lu 
  Abstract: This paper designs an interactive system of e-healthcare based on artificial intelligence technology, and optimises the hardware structure and software function of the system. In addition, through the introduction of roaming logic, navigation logic and other control logic, users are allowed to roam and navigate in the virtual scene. The system has the characteristics of a strong sense of reality, friendly interface and interactivity, which can meet the needs of an e-health system in external publicity, internal guidance, e-health planning and information management. Virtual reality technology and artificial intelligence technology are used to build a three-dimensional model and a virtual scene, with the introduction of the corresponding control logic to build a realistic and interactive, virtual city environment landscape that supports roaming, navigation and other operations, which can be deployed on different platforms to provide services for e-healthcare publicity, information management and other aspects.
  Keywords: artificial intelligence; e-health environment; landscape design.
  
  
• Optimisation of cache replacement policy using extreme learning machine  
  by Swapnita Srivastava, P.K. Singh 
  Abstract: In multiprocessors, all the cores ordinarily share the Last Level Cache (LLC). The memory systems of multi-core CPUs are often affected by irregular memory access patterns. The gap between the memory system and LLC prompts the research for an effective Cache Replacement Policy (CRP). Current processors use a variant of the Least Recently Used (LRU) policy to identify which should replace victims. However, there is a significant gap between the LRU policy and Belady's MIN policy, which is the ideal CRP in all the scenarios. Since Belady's algorithm needs future knowledge, it is optimal but not practically possible. This paper shows how CRP can be trained from recent cache accesses to guide future replacement decisions. Recent research on anticipating the reuse of cache blocks has enabled substantial improvement in cache speed and efficiency. This paper presents the ELM-SSO policy that uses Salp Swarm Optimisation (SSO) to optimise the weights coefficients of Extreme Learning Machine (ELM) to perform cache replacement classification. Furthermore, the use of SSO in optimising the ELM is examined to increase system accuracy and overcome the drawback of traditional ELM. The findings demonstrate that the proposed ELM-SSO policy outperforms the traditional cache replacement policy in terms of improvement rate, cache hit rate and cache miss rate. The proposed ELM-SSO policy improves the system performance by 36.66%, 6.25%, 11.71%, 11.35%, 10.32% and 10.99% over Optimal (OPT), Least Recently Used (LRU), Least Frequently Used (LFU), Logistic Regression (LR), K-Nearest Neighbour (K-NN) and Neural Network (NN), respectively.
  Keywords: computer architecture; cache memory; latency; eviction set; hit ratio; Belady's replacement policy; least recently used; Salp swarm optimisation; extreme learning machine.
  
  
• A hybrid WSN-based two-stage model for data collection and forecasting water consumption in metropolitan areas  
  by Mohammad Faiz, A.K. Daniel 
  Abstract: The improper distribution of in-house water consumption in the metropolitan regions of several Indian states has raised severe issues during the last few decades. Owing to increased human population and inefficient water usage, the average volume of water in the country's aquifers has begun to decline. Traditional water distribution and monitoring systems are unable to address this serious issue. The water crisis in the metropolitan area needs more efficient and reliable solutions to overcome this water distribution problem. The working of the proposed model is as follows. In the first stage, the data collection technique is proposed for water distribution in metropolitan areas through Energy Efficient Two-Phase Routing Protocol (EE-TPRP) using cloud-assisted wireless sensors. In the second stage, an efficient water demand prediction model (EWDM) using Backpropagation Feed-forward Neural Network (BP-FNN) is used for the prediction of water consumption for optimal distribution to users. The EE-TPRP protocol is compared to LEACH, MOD-LEACH, and DEEC protocols, where it has reduced overhead and enhanced network lifetime. The BP-FNN is compared to the regression model, fuzzy model, and ARIMA model, where it has improved the prediction efficiency of the water distribution in metropolitan areas.
  Keywords: water distribution; WSN; artificial neural network; cloud; feed-forward; gateway; sensor node.
  
  
• Detection of brain tumour using machine learning based framework by classifying MRI images  
  by P. Nancy, Murugesan G, Abu Sarwar Zamani, Karthikeyan Kaliyaperumal, Malik Jawarneh, Surendra Kumar Shukla, Samrat Ray, Abhishek Raghuvanshi 
  Abstract: The fatality rate has risen in recent years owing to an increase in the number of encephaloma tumours in each age group. Because of the complicated structure of tumours and the involution of noise in magnetic resonance (MR) imaging data, physical identification of tumours becomes a difficult and time-consuming operation for medical practitioners. As a result, recognizing and locating the tumour's location at an early stage is crucial. Cancer tumour areas at various levels may be followed and prognosticated using medical scans, which can be utilized in concert with segmentation and relegation techniques to provide a correct diagnosis at an early time. This article aims to develop image processing and machine learning based framework for early and accurate detection of brain tumours. This framework includes image preprocessing, image segmentation, feature extraction, and classification using the SVM, KNN, and Nave Bayes algorithms. Image preprocessing is performed using Gaussian Elimination, image enhancement using histogram Equalization, image segmentation using k Means and feature extraction performed using PCA algorithm. For performance comparison, parameters like- Accuracy, sensitivity and specificity are used. Experimental results have shown that the KNN is getting better accuracy for classification of brain tumour related images. KNN is performing admirably in terms of accuracy. In terms of specificity, SVM and KNN perform similarly well. KNN outperforms other algorithms in terms of sensitivity. Accuracy of KNN classifier is around 98 percent in brain tumour image classification.
  Keywords: brain tumour detection; MRI images; machine Learning; Gaussian elimination; K means; KNN; SVM; image segmentation; feature extraction; image classification.
  
  
• Credence-Net: a semi-supervised deep learning approach for medical images  
  by Pawan Kumar Mall, Pradeep Kumar Singh 
  Abstract: Deep learning uses a large-scale labelled dataset to ensure a high degree of accuracy. This technology is increasingly data-driven in medicine and biology imaging, and labelled data is more difficult and expensive to retrieve. Various studies are being conducted on semi-supervised deep learning models (SSDLM) and self-supervised deep learning. In order to increase the quantity of labelled data necessary for deep learning, researchers are increasingly looking at SSDLM and its applications. The motivation for the proposed Credence-Net is similar to how physicians handle uncertain or questionable instances in reality, based on their colleague's or senior's consultation. Proposed model Credence-Net has attained the best accuracy and specificity, sensitivity, precision, Matthews correlation coefficient, false discovery rate, false-positive rate, f1 score, negative predictive value, and false-negative rate 91.834%, 85.268%, 97.008%, 89.356%, 83.648%, 10.644%, 14.732%, 93.016%, 95.696%, and 2.992% for unseen dataset respectively. This research work leads to a more accurate and efficient semi-supervised deep learning model.
  Keywords: deep learning; semi-supervised learning; shoulder fracture; X-ray; medical images.
  
  
• Health evaluation system for hospitals based on big data and deep learning model  
  by Yuchen Xie 
  Abstract: Owing to the influence of redundant quality evaluation data, the existing college students academic quality evaluation system has too few hardware structure test points, which leads to the large power consumption of the evaluation system in actual operation. To solve this problem, an academic quality evaluation system for college students based on big data and deep learning model is designed. In the hardware part, microprocessor is used to build a CPU platform supporting big data. According to the discharge size of the hardware circuit, the test site position is set. In the software part, the deep learning model is used to construct the academic quality evaluation algorithm. The traditional evaluation system and the evaluation system designed in this paper are used to carry out experiments. The results show that the power consumption of the evaluation system designed in this paper is the minimum.
  Keywords: big data; deep learning model; academic quality; hardware data parameters.
  
  
• AI reconstruction method of health planning using IoT  
  by Y.A.O. Hang, IANG Yan, YANG Rongging 
  Abstract: With the increasing complexity of human settlements and the deterioration of the ecological environment, health systems are shouldering arduous responsibilities. However, owing to various reasons, the digital process of landscape planning and design has lagged behind the times. This paper discusses the current difficulties of AI reconfiguration design in landscape architecture, and points out the necessity of popularising this. The paper defines the definition, characteristics, and concept pedigree of landscape architecture AI reconfiguration design, analyses the main digital design methods and application software platform in the concept pedigree, and summarises the flow chart of landscape architecture AI reconfiguration design. At the end of this paper, the limitations and misunderstandings in the current AI reconfiguration design are analysed.
  Keywords: VR technology; health; AI reconstruction; internet of things.
  
  
• Detection and classification of brain abnormality by a novel hybrid Efficinentnet-Deep Autoencoder CNN model from MRI brain images in smart health diagnosis  
  by Dillip Ranjan Nayak, Neelamadhab Padhy, Ashish Singh, Pradeep Kumar Mallick 
  Abstract: This paper presents the novel smart hybrid EfficientNet-Deep Autoencoder (EF-DA) deep neural network model to classify brain images. This is the successor of modified EfficientNetB0 with a deep autoencoder to detect tumours. Initially, the feature extraction is done by modified EfficientNet, and then classification is done by the proposed smart deep autoencoder. The images are filtered, cropped by morphological operations, and augmented to train a deep hybrid EF-DA model in the first stage. In the second stage, a modified deep autoencoder is used for classification. The statistical result analysis of the hybrid model is assessed using seven types of degree metrics including F-score, precision, recall, specificity, kappa score, accuracy, and area under the ROC curve (AUC) score. It is compared with three types of pre-trained model, MobileNet, MobileNetV2, and ResNet50 for analysis. The EF-DA model has achieved an overall accuracy of 99.34% and an AUC score of 99.95%.
  Keywords: hybrid; EfficientNet; data augmentation; deep autoencoder; deep neural network; AUC score; overfitting; recall; precision; F-score.
  
  
• Intelligent overlay algorithm for medical data management based on wireless communication technology and feature fusion  
  by Changrong Peng, Xiaodong Zhang, Qian Liu, Xiaofang Zhao, Chenyang Dai 
  Abstract: Medical data management through wireless communication system become essential to make data available at all time. To address the problem of poor quality of management, a intelligent overlay algorithm based on wireless communication technology and feature fusion is proposed. The algorithm first uses sensors remote sensing equipment to collect patient data and transmit them by wireless communication, followed by image and data filtering, then feature extraction and feature fusion, and finally seamless overlaying by projection model. The results show that the spatial frequency and average gradient of the superimposed patient data management meets the requirements, indicating that the resultant data after the application of the sensing data is superimposition algorithm based on wireless communication technology and feature fusion retain the detail components of the patient data more realistically, with good clarity, and the image information is better maintained.
  Keywords: wireless communication technology; feature fusion; medical data; sensors; intelligent overlay algorithm.
  
  
• Research on health service technology based on multimedia imagery training method  
  by Wanchun Kang, Jian Wang, Hongseol Kim, Xinchao Du 
  Abstract: Health serving technique is one of the basic technical movements in serving patients. The traditional health technique pays attention only to the training of service technique, but ignores the patients' psychological state, which greatly limits the learning effect. The application of multimedia imagery training teaching method in hospital health service technology teaching not only conforms to the characteristics of modern health and mind learning, but also increases the stability and accuracy of patients health, helping patients quickly establish the power of correct action, so as to quickly and effectively master the action of service technology. This paper analyses the problems existing in the teaching of service technology in hospitals and medical colleges, and finally explores the specific application strategies of multimedia imagery training method in the teaching of nursing service technology in colleges and hospitals.
  Keywords: multimedia imagery; health data; detection; knowledge data.
  
  
• Evaluation method of resource fusion using artificial intelligence technology  
  by Juan Li 
  Abstract: Aiming at the problems of low application performance and long retrieval time of traditional resource database construction methods, this paper proposes a teaching resource database construction method based on neural network, which combines electronic engineering education with physical education. Firstly, the inventory structure of teaching resources is determined according to the type of hardware equipment. In the environment of a teaching platform andthe internet, the initial teaching resources are collected and preprocessed from two aspects of electronic engineering education and physical education. The neural network iterative algorithm is used to extract the characteristics of resources, realise the division of resource types, and fuse the same type of resource data. Finally, through the unified storage format, the teaching resources are substituted into the storage structure to realise the construction of the teaching resource database. It is found that the retrieval time of resources is reduced by 0.21 s, which improves the retrieval speed.
  Keywords: neural network; electronic engineering education; physical education; teaching resources; database construction.
  
  
• Evaluation method for colour matching using artificial intelligence technology  
  by Lijuan Yao, Ling Tang 
  Abstract: The existing colour matching evaluation methods have the problem of fuzzy colour attributes, which leads to high image distortion. This paper designs an evaluation method of public space indoor landscape colour matching based on artificial intelligence technology. The method quantifies the colour layout of the public space, determines the main colour of the space, identifies the colour attributes of the indoor landscape, deploys the combined colour phase ring, uses artificial intelligence technology to extract the colour matching features, calculates the colour distance combined with the transition colour frequency information, and adopts the colour quantisation algorithm to set the evaluation model. The experiment results show that the average distortions of the evaluation method and the other two evaluation methods is 30.12, 38.96, are 38.87, respectively, which proves that the colour matching evaluation method combined with artificial intelligence technology has higher use value.
  Keywords: artificial intelligence technology; colour layout; colour matching; evaluation method; public space; interior landscape;.
  
  
• Stereoscopic display of architectural design images based on virtual reality technology  
  by Ling Tang, Lijuan Yao 
  Abstract: The current image stereoscopic display method mainly displays images stereoscopically from the perspective of human left and right eye visual imaging, which not only displays images with distortion and missing details, but also makes it difficult to realise interaction for complex image stereoscopic display. This paper proposes a stereoscopic display method of architectural design images based on virtual reality technology. The images are drawn using DIBR technology and the depth images are processed using Gaussian filtering and so on. After designing the virtual interaction of the image stereoscopic display scene, EON is used to analyse the lighting of the building exterior and realise the stereoscopic display of the image. The simulation experimental data of the stereoscopic display method show that the proposed image stereoscopic display method relatively improves the display effect by about 66.7% and has good adaptability for different grey value images.
  Keywords: architectural exterior; design images; image presentation; stereoscopic presentation; virtual interaction; virtual reality technology.
  
  
• Studying the impact of anti-oxidant extracts of different vegetables on the formation of PAHs in rabbit meat  
  by Rabia Siddique, Ameer Fawad Zahoor, Sajjad Ahmad, Hamad Ahmad, Abid Hussain 
  Abstract: Polycyclic aromatic hydrocarbons (PAHs) are powerful noxious compounds which are produced in well-done processed meat products. Incomplete burning of fuels produces PAHs and soot, by the combination of combustion and pyrolysis. The PAH yields are connected with diverse factors, such as type of meat, heating temperature, cooking time, processing technique, additives and storage time. This research paper examines the formation of PAHs in cooked European rabbit meat. Additionally, extraction process and quantification procedure to evaluate the PAHs in these compounds are reported. Finally, an overview is represented on the impact of the anti-oxidant properties of vegetable extracts to decrease the generation of PAHs in cooked rabbit meat.
  Keywords: rabbit meat; vegetables extracts; anti-oxidant; GC-MS; PAHs.
  
  
• Deep learning-based feature extraction coupled with multi-class SVM for COVID-19 detection in the IoT era  
  by Auwalu Mubarak, Sertan Serte, Fadi Al-Turjman, Rabiu Aliyu, Zubaida Said, Mehmet Ozsoz 
  Abstract: The deadly coronavirus virus (COVID-19) was confirmed as a pandemic by the World Health Organisation (WHO) in December 2019. Prompt and early identification of suspected patients is necessary to monitor the transmission of the disease, increase the effectiveness of medical treatment and as a result, decrease the mortality rate. The adopted method to identify COVID-19 is the Reverse-Transcription Polymerase Chain Reaction (RT-PCR), the method is affected by the shortage of RT-PCR kits and complexity. Medical imaging using deep learning has proved to be one of the most efficient methods of detecting respiratory diseases, but efficient deep learning architecture and low data are affecting the performance of the deep learning models. To detect COVID-19 efficiently, a deep learning model based feature extraction coupled with Support Vector Machine (SVM) was employed in this study, Seven pre-trained models were employed as feature extractors and the extracted features are classified by multi-class SVM classifier to classify CT scan images from COVID-19, common pneumonia and healthy individuals. To improve the performance of the models and prevent overfitting, training was also carried out on augmented images. To generalise the model's performance and robustness, three datasets were merged in the study. The model with the best performance is the VGG19 which was trained with augmented images: it achieved an accuracy of 96%, a sensitivity of 0.936, a specificity 0f 0.967, an F1 score of 0.935, a precision of 0.934, a Yonden Index of 0.903 and AUC of 0.952. The best model shows that COVID-19 can be detected efficiently on CT scan images.
  Keywords: artificial intelligence; COVID-19; SVM; feature extraction.
  DOI: 10.1504/IJNT.2021.10040115
   
  
• Power optimisation of wireless sensing network through quantum deep learning  
  by Zhongzhen Yan, Kewei Kewei Zhou, Feng Guo, Na Hou, Jiangyi Du 
  Abstract: In current research, a secure wireless communication system with high data rate obtained by quantum computing (QC) is used. To speed up the data rate of transfer, the QC has been combined with machine learning algorithms. Then the large volume of data that are transferred, stored and processed in wireless systems leads high energy of the system. This article proposes a power minimisation approach with QC-based approach called improved sequential parametric convex approximation (ISPCA) trained by deep learning algorithm called Graph Convolutional Neural Network (GCNN). The proposed approach has been used to minimise the power and enhance the energy efficiency of wireless communication systems, such as LTE/5G. Evaluated results show that our proposed approach consumes only 44.82J energy for computing 1000 samples. The proposed technique outperforms existing ones by consuming much less energy.
  Keywords: wireless communication system; 5G; SPCA; convolution network; deep learning; energy efficiency; power optimisation.
  
  
• Integrated agriculture IoT-based farm monitoring and management systems  
  by Satish Sampatrao Salunkhe, Aarti Amod Agarkar, Mandar Karyakarte, Jaison Mulerikkal, Prathiba Jonnala, Divyendu Kumar Mishra 
  Abstract: The internet of things (IoT) is a cutting-edge phenomenon in the digital realm, wherein objects connect, and processes are autonomous and managed via the internet. In this research, an architecture for integrating the IoT with crop production is designed, and various measurements and techniques for crop monitoring using cloud computing have been employed. The method enables real-time evaluation of data gathered from sensors put in crops that delivers a response for the farmer for crop production monitoring and saves the farmer's time as well as energy. The data acquired in the fields are saved in the cloud and analysed to allow for automation using IoT devices. The experimental results show the specifics of soil moisture, temperature, humidity, and water use in the field through a monitoring system, as well as managing the farmer's input.
  Keywords: internet of things; agriculture; farming; monitoring system; management system.
  
  

Special Issue on: Eco-Friendly and Sustainable Cognitive Green Nano-Technologies for the Mitigation of Emerging Environmental Pollutants

• Preparation of titanium dioxide composite nanomaterials using copper catalysis and their dynamic adsorption and photocatalytic performance in water treatment  
  by Ye Tian 
  Abstract: The aim is to investigate the dynamic adsorption performance of titanium dioxide (TiO2) nanocomposite materials in water treatment, providing direction for water purification. The copper-catalysed living free-radical polymerization method polymerizes the prepared TiO2 particles with tertiary amine polymer to manufacture the TiO2 polymer nanocomposite materials. The prepared TiO2 nanocomposite materials are then modified to obtain the quaternised TiO2 polymer nanocomposite materials (quaternised TiO2@poly(DEAEMA)), which are characterized and analysed. Finally, the water treatment performance of quaternised TiO2@poly(DEAEMA) is judged through photocatalysis and adsorption experiments, while the antibacterial performance of the prepared materials is judged using the common Escherichia coli and Staphylococcus aureus. Results demonstrate that the quaternised TiO2@poly(DEAEMA) polymer nanocomposite materials are completely and tightly wrapped, presenting a flower-like appearance, with a significantly-increased diameter and an average size of about 600nm, which can be utilized as the pollutant adsorbent. Water treatment simulation reveals the fastest adsorption rate and the highest adsorption capacity of quaternised TiO2@poly(DEAEMA), reaching 265 mg/g given the same reaction time. The catalytic removal rate in ultraviolet and visible light reaches 94%, and the photocatalysis of visible light reaches 69%. Until the reaction lasts for 45 minutes, its antibacterial activity is optimal, and the diameter of the inhibition zone against Escherichia coli and Staphylococcus aureus exceeds 16 mm. Therefore, the prepared TiO2 nanomaterials have high adsorption properties, good photocatalysis performance, and excellent antibacterial properties, which can provide an experimental basis for the treatment and purification of water resources in the industry.
  Keywords: titanium dioxide; water treatment; dynamic adsorption; photocatalysis; nanocomposite material.