Growing plants in the gulf region can be challenging as it is mostly desert, and the climate is dry. A few species of plants have the capability to grow in such a climate. However, those plants are not suitable as a food source. The aim of this work is to design and construct an indoor automatic vertical hydroponic system that does not depend on the outside climate. The designed system is capable to grow common type of crops that can be used as a food source inside homes without the need of large space. The design of the system was made after studying different types of vertical hydroponic systems in terms of price, power consumption and suitability to be built as an indoor automated system. A microcontroller was working as a brain of the system, which communicates with different types of sensors to control all the system parameters and to minimize the human intervention. An open internet of things (IoT) platform was used to store and display the system parameters and graphical interface for remote access. The designed system is capable of maintaining healthy growing parameters for the plants with minimal input from the user. The functionality of the overall system was confirmed by evaluating the response from individual system components and monitoring them in the IoT platform. The system was consuming 120.59 and 230.59 kWh respectively without and with air conditioning control during peak summer, which is equivalent to the system running cost of 13.26 and 25.36 Qatari Riyal (QAR) respectively. This system was circulating around 104 k gallons of nutrient solution monthly however, only 8-10 L water was consumed by the system. This system offers real-time notifications to alert the hydroponic system user when the conditions are not favorable. So, the user can monitor several parameters without using laboratory instruments, which will allow to control the entire system remotely. Moreover, the system also provides a wide range of information, which could be essential for plant researchers and provides a greater understanding of how the key parameters of hydroponic system correlate with plant growth. The proposed platform can be used both for quantitatively optimizing the setup of the indoor farming and for automating some of the most labor-intensive maintenance activities. Moreover, such a monitoring system can also potentially be used for high-level decision making, once enough data will be collected. This work presents significant opportunities for the people who live in the gulf region to produce food as per their requirements.
One of the major causes of death all over the world is heart disease or cardiac dysfunction. These diseases could be identified easily with the variations in the sound produced due to the heart activity. These sophisticated auscultations need important clinical experience and concentrated listening skills. Therefore, there is an unmet need for a portable system for the early detection of cardiac illnesses. This paper proposes a prototype model of a smart digital-stethoscope system to monitor patient's heart sounds and diagnose any abnormality in a real-time manner. This system consists of two subsystems that communicate wirelessly using Bluetooth low energy technology: A portable digital stethoscope subsystem, and a computer-based decision-making subsystem. The portable subsystem captures the heart sounds of the patient, filters and digitizes, and sends the captured heart sounds to a personal computer wirelessly to visualize the heart sounds and for further processing to make a decision if the heart sounds are normal or abnormal. Twenty-seven t-domain, f-domain, and Mel frequency cepstral coefficients (MFCC) features were used to train a public database to identify the best-performing algorithm for classifying abnormal and normal heart sound (HS). The hyper parameter optimization, along with and without a feature reduction method, was tested to improve accuracy. The cost-adjusted optimized ensemble algorithm can produce 97% and 88% accuracy of classifying abnormal and normal HS, respectively.
Gait analysis is a systematic study of human locomotion, which can be utilized in variousapplications, such as rehabilitation, clinical diagnostics and sports activities. The various limitationssuch as cost, non-portability, long setup time, post-processing time etc., of the current gait analysistechniques have made them unfeasible for individual use. This led to an increase in research interestin developing smart insoles where wearable sensors can be employed to detect vertical groundreaction forces (vGRF) and other gait variables. Smart insoles are flexible, portable and comfortablefor gait analysis, and can monitor plantar pressure frequently through embedded sensors thatconvert the applied pressure to an electrical signal that can be displayed and analyzed further.Several research teams are still working to improve the insoles' features such as size, sensitivity ofinsoles sensors, durability, and the intelligence of insoles to monitor and control subjects' gait bydetecting various complications providing recommendation to enhance walking performance. Eventhough systematic sensor calibration approaches have been followed by different teams to calibrateinsoles' sensor, expensive calibration devices were used for calibration such as universal testingmachines or infrared motion capture cameras equipped in motion analysis labs. This paper providesa systematic design and characterization procedure for three different pressure sensors: forcesensitiveresistors (FSRs), ceramic piezoelectric sensors, and flexible piezoelectric sensors that canbe used for detecting vGRF using a smart insole. A simple calibration method based on a load cellis presented as an alternative to the expensive calibration techniques. In addition, to evaluate theperformance of the different sensors as a component for the smart insole, the acquired vGRF fromdifferent insoles were used to compare them. The results showed that the FSR is the most effectivesensor among the three sensors for smart insole applications, whereas the piezoelectric sensors canbe utilized in detecting the start and end of the gait cycle. This study will be useful for any researchgroup in replicating the design of a customized smart insole for gait analysis.
Numeral recognition is considered an essential preliminary step for optical character recognition, document understanding, and others. Although several handwritten numeral recognition algorithms have been proposed so far, achieving adequate recognition accuracy and execution time remain challenging to date. In particular, recognition accuracy depends on the features extraction mechanism. As such, a fast and robust numeral recognition method is essential, which meets the desired accuracy by extracting the features efficiently while maintaining fast implementation time. Furthermore, to date most of the existing studies are focused on evaluating their methods based on clean environments, thus limiting understanding of their potential application in more realistic noise environments. Therefore, finding a feasible and accurate handwritten numeral recognition method that is accurate in the more practical noisy environment is crucial. To this end, this paper proposes a new scheme for handwritten numeral recognition using Hybrid orthogonal polynomials. Gradient and smoothed features are extracted using the hybrid orthogonal polynomial. To reduce the complexity of feature extraction, the embedded image kernel technique has been adopted. In addition, support vector machine is used to classify the extracted features for the different numerals. The proposed scheme is evaluated under three different numeral recognition datasets: Roman, Arabic, and Devanagari. We compare the accuracy of the proposed numeral recognition method with the accuracy achieved by the state-of-the-art recognition methods. In addition, we compare the proposed method with the most updated method of a convolutional neural network. The results show that the proposed method achieves almost the highest recognition accuracy in comparison with the existing recognition methods in all the scenarios considered. Importantly, the results demonstrate that the proposed method is robust against the noise distortion and outperforms the convolutional neural network considerably, which signifies the feasibility and the effectiveness of the proposed approach in comparison to the state-of-the-art recognition methods under both clean noise and more realistic noise environments.
In this article, we utilize Digital Twins (DT) with edge networks using blockchain technology for reliable real-time data processing and provide a secure, scalable solution to bridge the gap between physical edge networks and digital systems. Then, we suggest a Federated Learning (FL) framework for collaborative computing that runs on a blockchain and is powered by the DT edge network. This framework increases data privacy while enhancing system security and reliability. The provision of sustainable Resource Allocation (RA) and ensure real-time data-processing interaction between Internet of Things (IoT) devices and edge servers depends on a balance between system latency and Energy Consumption (EC) based on the proposed DT-empowered Deep Reinforcement Learning (Deep-RL) agent. The Deep-RL agent evaluates the performance action based on RA actions in DT to distribute its bandwidth resources to IoT devices based on iteration and the actions taken to generate the best policy and enhance learning efficiency at every step. The simulation results show that the proposed Deep-RL-agent-based DT is able to exploit the best policy, select 47.5% of computing activities that are to be carried out locally with 1 MHz bandwidth and minimize the weighted cost of the transmission policy of edge-computing strategies.
Recently, green computing has received significant attention for Internet of Things (IoT) environments due to the growing computing demands under tiny sensor enabled smart services. The related literature on green computing majorly focuses on a cover set approach that works efficiently for target coverage, but it is not applicable in case of area coverage. In this paper, we present a new variant of a cover set approach called a grouping and sponsoring aware IoT framework (GS-IoT) that is suitable for area coverage. We achieve non-overlapping coverage for an entire sensing region employing sectorial sensing. Non-overlapping coverage not only guarantees a sufficiently good coverage in case of large number of sensors deployed randomly, but also maximizes the life span of the whole network with appropriate scheduling of sensors. A deployment model for distribution of sensors is developed to ensure a minimum threshold density of sensors in the sensing region. In particular, a fast converging grouping (FCG) algorithm is developed to group sensors in order to ensure minimal overlapping. A sponsoring aware sectorial coverage (SSC) algorithm is developed to set off redundant sensors and to balance the overall network energy consumption. GS-IoT framework effectively combines both the algorithms for smart services. The simulation experimental results attest to the benefit of the proposed framework as compared to the state-of-the-art techniques in terms of various metrics for smart IoT environments including rate of overlapping, response time, coverage, active sensors, and life span of the overall network.
In this study, porous silicon (PSi) was prepared and tested as an extended gate field-effect transistor (EGFET) for pH sensing. The prepared PSi has pore sizes in the range of 500 to 750 nm with a depth of approximately 42 µm. The results of testing PSi for hydrogen ion sensing in different pH buffer solutions reveal that the PSi has a sensitivity value of 66 mV/pH that is considered a super Nernstian value. The sensor considers stability to be in the pH range of 2 to 12. The hysteresis values of the prepared PSi sensor were approximately 8.2 and 10.5 mV in the low and high pH loop, respectively. The result of this study reveals a promising application of PSi in the field for detecting hydrogen ions in different solutions.
This study presents a novel feature-engineered-natural gradient descent ensemble-boosting (NGBoost) machine-learning framework for detecting fraud in power consumption data. The proposed framework was sequentially executed in three stages: data pre-processing, feature engineering, and model evaluation. It utilized the random forest algorithm-based imputation technique initially to impute the missing data entries in the acquired smart meter dataset. In the second phase, the majority weighted minority oversampling technique (MWMOTE) algorithm was used to avoid an unequal distribution of data samples among different classes. The time-series feature-extraction library and whale optimization algorithm were utilized to extract and select the most relevant features from the kWh reading of consumers. Once the most relevant features were acquired, the model training and testing process was initiated by using the NGBoost algorithm to classify the consumers into two distinct categories ("Healthy" and "Theft"). Finally, each input feature's impact (positive or negative) in predicting the target variable was recognized with the tree SHAP additive-explanations algorithm. The proposed framework achieved an accuracy of 93%, recall of 91%, and precision of 95%, which was greater than all the competing models, and thus validated its efficacy and significance in the studied field of research.
: Concrete failure will lead to serious safety concerns in the performance of a building structure. It is one of the biggest challenges for engineers to inspect and maintain the quality of concrete throughout the service years in order to prevent structural deterioration. To date, a lot of research is ongoing to develop different instruments to inspect concrete quality. Detection of moisture ingress is important in the structural monitoring of concrete. This paper presents a novel sensing technique using a smart antenna for the non-destructive evaluation of moisture content and deterioration inspection in concrete blocks. Two different standard concrete samples (United Kingdom and Malaysia) were investigated in this research. An electromagnetic (EM) sensor was designed and embedded inside the concrete to detect the moisture content within the structure. In addition, CST microwave studio was used to validate the theoretical model of the EM sensor against the test data. The results demonstrated that the EM sensor at 2.45 GHz is capable of detecting the moisture content in the concrete with linear regression of R² = 0.9752. Furthermore, identification of different mix ratios of concrete were successfully demonstrated in this paper. In conclusion, the EM sensor is capable of detecting moisture content non-destructively and could be a potential technique for maintenance and quality control of the building performance.
Exposure to mental stress for long period leads to serious accidents and health problems. To avoid negative consequences on health and safety, it is very important to detect mental stress at its early stages, i.e., when it is still limited to acute or episodic stress. In this study, we developed an experimental protocol to induce two different levels of stress by utilizing a mental arithmetic task with time pressure and negative feedback as the stressors. We assessed the levels of stress on 22 healthy subjects using frontal electroencephalogram (EEG) signals, salivary alpha-amylase level (AAL), and multiple machine learning (ML) classifiers. The EEG signals were analyzed using a fusion of functional connectivity networks estimated by the Phase Locking Value (PLV) and temporal and spectral domain features. A total of 210 different features were extracted from all domains. Only the optimum multi-domain features were used for classification. We then quantified stress levels using statistical analysis and seven ML classifiers. Our result showed that the AAL level was significantly increased (p < 0.01) under stress condition in all subjects. Likewise, the functional connectivity network demonstrated a significant decrease under stress, p < 0.05. Moreover, we achieved the highest stress classification accuracy of 93.2% using the Support Vector Machine (SVM) classifier. Other classifiers produced relatively similar results.
In real-life applications, electroencephalogram (EEG) signals for mental stress recognition require a conventional wearable device. This, in turn, requires an efficient number of EEG channels and an optimal feature set. This study aims to identify an optimal feature subset that can discriminate mental stress states while enhancing the overall classification performance. We extracted multi-domain features within the time domain, frequency domain, time-frequency domain, and network connectivity features to form a prominent feature vector space for stress. We then proposed a hybrid feature selection (FS) method using minimum redundancy maximum relevance with particle swarm optimization and support vector machines (mRMR-PSO-SVM) to select the optimal feature subset. The performance of the proposed method is evaluated and verified using four datasets, namely EDMSS, DEAP, SEED, and EDPMSC. To further consolidate, the effectiveness of the proposed method is compared with that of the state-of-the-art metaheuristic methods. The proposed model significantly reduced the features vector space by an average of 70% compared with the state-of-the-art methods while significantly increasing overall detection performance.
With continuous advancements in Internet technology and the increased use of cryptographic techniques, the cloud has become the obvious choice for data sharing. Generally, the data are outsourced to cloud storage servers in encrypted form. Access control methods can be used on encrypted outsourced data to facilitate and regulate access. Multi-authority attribute-based encryption is a propitious technique to control who can access encrypted data in inter-domain applications such as sharing data between organizations, sharing data in healthcare, etc. The data owner may require the flexibility to share the data with known and unknown users. The known or closed-domain users may be internal employees of the organization, and unknown or open-domain users may be outside agencies, third-party users, etc. In the case of closed-domain users, the data owner becomes the key issuing authority, and in the case of open-domain users, various established attribute authorities perform the task of key issuance. Privacy preservation is also a crucial requirement in cloud-based data-sharing systems. This work proposes the SP-MAACS scheme, a secure and privacy-preserving multi-authority access control system for cloud-based healthcare data sharing. Both open and closed domain users are considered, and policy privacy is ensured by only disclosing the names of policy attributes. The values of the attributes are kept hidden. Characteristic comparison with similar existing schemes shows that our scheme simultaneously provides features such as multi-authority setting, expressive and flexible access policy structure, privacy preservation, and scalability. The performance analysis carried out by us shows that the decryption cost is reasonable enough. Furthermore, the scheme is demonstrated to be adaptively secure under the standard model.
The Internet of Things (IoT) has transformed our interaction with technology and introduced security challenges. The growing number of IoT attacks poses a significant threat to organizations and individuals. This paper proposes an approach for detecting attacks on IoT networks using ensemble feature selection and deep learning models. Ensemble feature selection combines filter techniques such as variance threshold, mutual information, Chi-square, ANOVA, and L1-based methods. By leveraging the strengths of each technique, the ensemble is formed by the union of selected features. However, this union operation may overlook redundancy and irrelevance, potentially leading to a larger feature set. To address this, a wrapper algorithm called Recursive Feature Elimination (RFE) is applied to refine the feature selection. The impact of the selected feature set on the performance of Deep Learning (DL) models (CNN, RNN, GRU, and LSTM) is evaluated using the IoT-Botnet 2020 dataset, considering detection accuracy, precision, recall, F1-measure, and False Positive Rate (FPR). All DL models achieved the highest detection accuracy, precision, recall, and F1 measure values, ranging from 97.05% to 97.87%, 96.99% to 97.95%, 99.80% to 99.95%, and 98.45% to 98.87%, respectively.
The significant surge in Internet of Things (IoT) devices presents substantial challenges to network security. Hackers are afforded a larger attack surface to exploit as more devices become interconnected. Furthermore, the sheer volume of data these devices generate can overwhelm conventional security systems, compromising their detection capabilities. To address these challenges posed by the increasing number of interconnected IoT devices and the data overload they generate, this paper presents an approach based on meta-learning principles to identify attacks within IoT networks. The proposed approach constructs a meta-learner model by stacking the predictions of three Deep-Learning (DL) models: RNN, LSTM, and CNN. Subsequently, the identification by the meta-learner relies on various methods, namely Logistic Regression (LR), Multilayer Perceptron (MLP), Support Vector Machine (SVM), and Extreme Gradient Boosting (XGBoost). To assess the effectiveness of this approach, extensive evaluations are conducted using the IoT dataset from 2020. The XGBoost model showcased outstanding performance, achieving the highest accuracy (98.75%), precision (98.30%), F1-measure (98.53%), and AUC-ROC (98.75%). On the other hand, the SVM model exhibited the highest recall (98.90%), representing a slight improvement of 0.14% over the performance achieved by XGBoost.
The fabrication of lightweight, ultra-thin, low power and intelligent body-borne sensors leads to novel advances in wireless body area networks (WBANs). Depending on the placement of the nodes, it is characterized as in/on body WBAN; thus, the channel is largely affected by body posture, clothing, muscle movement, body temperature and climatic conditions. The energy resources are limited and it is not feasible to replace the sensor's battery frequently. In order to keep the sensor in working condition, the channel resources should be reserved. The lifetime of the sensor is very crucial and it highly depends on transmission among sensor nodes and energy consumption. The reliability and energy efficiency in WBAN applications play a vital role. In this paper, the analytical expressions for energy efficiency (EE) and packet error rate (PER) are formulated for two-way relay cooperative communication. The results depict better reliability and efficiency compared to direct and one-way relay communication. The effective performance range of direct vs. cooperative communication is separated by a threshold distance. Based on EE calculations, an optimal packet size is observed that provides maximum efficiency over a certain link length. A smart and energy efficient system is articulated that utilizes all three communication modes, namely direct, one-way relay and two-way relay, as the direct link performs better for a certain range, but the cooperative communication gives better results for increased distance in terms of EE. The efficacy of the proposed hybrid scheme is also demonstrated over a practical quasi-static channel. Furthermore, link length extension and diversity is achieved by joint network-channel (JNC) coding the cooperative link.
In Australia, droughts are recurring events that tremendously affect environmental, agricultural and socio-economic activities. Southern Queensland is one of the most drought-prone regions in Australia. Consequently, a comprehensive drought vulnerability mapping is essential to generate a drought vulnerability map that can help develop and implement drought mitigation strategies. The study aimed to prepare a comprehensive drought vulnerability map that combines drought categories using geospatial techniques and to assess the spatial extent of the vulnerability of droughts in southern Queensland. A total of 14 drought-influencing criteria were selected for three drought categories, specifically, meteorological, hydrological and agricultural. The specific criteria spatial layers were prepared and weighted using the fuzzy analytical hierarchy process. Individual categories of drought vulnerability maps were prepared from their specific indices. Finally, the overall drought vulnerability map was generated by combining the indices using spatial analysis. Results revealed that approximately 79.60% of the southern Queensland region is moderately to extremely vulnerable to drought. The findings of this study were validated successfully through the receiver operating characteristics curve (ROC) and the area under the curve (AUC) approach using previous historical drought records. Results can be helpful for decision makers to develop and apply proactive drought mitigation strategies.
Lung cancer is the leading cause of cancer death and morbidity worldwide. Many studies have shown machine learning models to be effective in detecting lung nodules from chest X-ray images. However, these techniques have yet to be embraced by the medical community due to several practical, ethical, and regulatory constraints stemming from the "black-box" nature of deep learning models. Additionally, most lung nodules visible on chest X-rays are benign; therefore, the narrow task of computer vision-based lung nodule detection cannot be equated to automated lung cancer detection. Addressing both concerns, this study introduces a novel hybrid deep learning and decision tree-based computer vision model, which presents lung cancer malignancy predictions as interpretable decision trees. The deep learning component of this process is trained using a large publicly available dataset on pathological biomarkers associated with lung cancer. These models are then used to inference biomarker scores for chest X-ray images from two independent data sets, for which malignancy metadata is available. Next, multi-variate predictive models were mined by fitting shallow decision trees to the malignancy stratified datasets and interrogating a range of metrics to determine the best model. The best decision tree model achieved sensitivity and specificity of 86.7% and 80.0%, respectively, with a positive predictive value of 92.9%. Decision trees mined using this method may be considered as a starting point for refinement into clinically useful multi-variate lung cancer malignancy models for implementation as a workflow augmentation tool to improve the efficiency of human radiologists.
The intelligent transportation system, especially autonomous vehicles, has seen a lot of interest among researchers owing to the tremendous work in modern artificial intelligence (AI) techniques, especially deep neural learning. As a result of increased road accidents over the last few decades, significant industries are moving to design and develop autonomous vehicles. Understanding the surrounding environment is essential for understanding the behavior of nearby vehicles to enable the safe navigation of autonomous vehicles in crowded traffic environments. Several datasets are available for autonomous vehicles focusing only on structured driving environments. To develop an intelligent vehicle that drives in real-world traffic environments, which are unstructured by nature, there should be an availability of a dataset for an autonomous vehicle that focuses on unstructured traffic environments. Indian Driving Lite dataset (IDD-Lite), focused on an unstructured driving environment, was released as an online competition in NCPPRIPG 2019. This study proposed an explainable inception-based U-Net model with Grad-CAM visualization for semantic segmentation that combines an inception-based module as an encoder for automatic extraction of features and passes to a decoder for the reconstruction of the segmentation feature map. The black-box nature of deep neural networks failed to build trust within consumers. Grad-CAM is used to interpret the deep-learning-based inception U-Net model to increase consumer trust. The proposed inception U-net with Grad-CAM model achieves 0.622 intersection over union (IoU) on the Indian Driving Dataset (IDD-Lite), outperforming the state-of-the-art (SOTA) deep neural-network-based segmentation models.
Iris biometric detection provides contactless authentication, preventing the spread of COVID-19-like contagious diseases. However, these systems are prone to spoofing attacks attempted with the help of contact lenses, replayed video, and print attacks, making them vulnerable and unsafe. This paper proposes the iris liveness detection (ILD) method to mitigate spoofing attacks, taking global-level features of Thepade's sorted block truncation coding (TSBTC) and local-level features of the gray-level co-occurrence matrix (GLCM) of the iris image. Thepade's SBTC extracts global color texture content as features, and GLCM extracts local fine-texture details. The fusion of global and local content presentation may help distinguish between live and non-live iris samples. The fusion of Thepade's SBTC with GLCM features is considered in experimental validations of the proposed method. The features are used to train nine assorted machine learning classifiers, including naïve Bayes (NB), decision tree (J48), support vector machine (SVM), random forest (RF), multilayer perceptron (MLP), and ensembles (SVM + RF + NB, SVM + RF + RT, RF + SVM + MLP, J48 + RF + MLP) for ILD. Accuracy, precision, recall, and F-measure are used to evaluate the performance of the projected ILD variants. The experimentation was carried out on four standard benchmark datasets, and our proposed model showed improved results with the feature fusion approach. The proposed fusion approach gave 99.68% accuracy using the RF + J48 + MLP ensemble of classifiers, immediately followed by the RF algorithm, which gave 95.57%. The better capability of iris liveness detection will improve human-computer interaction and security in the cyber-physical space by improving person validation.
Postquantum cryptography for elevating security against attacks by quantum computers in the Internet of Everything (IoE) is still in its infancy. Most postquantum based cryptosystems have longer keys and signature sizes and require more computations that span several orders of magnitude in energy consumption and computation time, hence the sizes of the keys and signature are considered as another aspect of security by green design. To address these issues, the security solutions should migrate to the advanced and potent methods for protection against quantum attacks and offer energy efficient and faster cryptocomputations. In this context, a novel security framework Lightweight Postquantum ID-based Signature (LPQS) for secure communication in the IoE environment is presented. The proposed LPQS framework incorporates a supersingular isogeny curve to present a digital signature with small key sizes which is quantum-resistant. To reduce the size of the keys, compressed curves are used and the validation of the signature depends on the commutative property of the curves. The unforgeability of LPQS under an adaptively chosen message attack is proved. Security analysis and the experimental validation of LPQS are performed under a realistic software simulation environment to assess its lightweight performance considering embedded nodes. It is evident that the size of keys and the signature of LPQS is smaller than that of existing signature-based postquantum security techniques for IoE. It is robust in the postquantum environment and efficient in terms of energy and computations.