Recent Trends in Communication, Computing, and Electronics

This paper presents the design of improved MEMS shunt switch for RF communication applications. The switch was designed to provide a better performance in 10–100 GHz range. The switch was optimized in terms of width of the beam and air gap between the fixed type beam and dielectric layer to improve the isolation, insertion, and return loss. This study concludes that materials with high k-dielectrics and high Young’s modulus are desirable for better performance in high-frequency range. The isolation, insertion, and return loss for the designed switch are obtained as –12 dB, –0.05 dB, and –45 dB, respectively.

The comparative response evaluation of M-ary phase-shift keying and M-ary quadrature amplitude modulation schemes is presented in this paper. The constellation diagrams for 16-PSK and 16-QAM are plotted to show the exact location of symbols, and the expressions for error probability are derived for both signaling schemes. The performance is compared by way of bit error probability and bandwidth efficiency. The simulated bit error rate graph plotted for different value of M are presented and compared with theoretical bit error rate plots separately. The combined BER graphs are also plotted for M = 16 and M = 64 for both signaling schemes and combined bandwidth efficiency graph plotted for M-ary PSK and M-ary QAM.

This paper depicts a modified return-to-zero differential quadrature phase shift keying (RZ-DQPSK) modulation scheme. A simulative analysis has been performed for an optimization of the proposed scheme. The analysis has also been extended for two different types of commercial fibers; standard single-mode fiber (SSMF) and true wave-reduced slope (TW-RS) fiber to report the similar qualitative performance characteristics. For the sake of better understanding of the proposed modified carving signal RZ-DQPSK scheme, optical spectrum and eye pattern plots are also investigated.

The main goal of this article is to design an optimized optical logic gates which are a key point to improve and enhance the speed of next era photonic integrated circuit. These kinds of integrated circuits are much more capable for high-speed transmission of information and very low energy loss. The energy dissipation by any integrated circuit in form of heat reduces the life span of device. This kind of design protects it and improves the life expectancy of the integrated circuit. In this article, opto-electronic conversion and vice versa are taken out.

Photonics crystal fiber-based sensors have small size, high sensitivity, flexibility, robustness, and ability of remote sensing. They can be used in unfavorable environmental conditions such as strong electromagnetic field, nuclear radiation, noise, and high voltages for explosive or corrosive media at high temperature. Fabrication of these sensors is so simple, and altogether makes them as a very efficient sensing solution for medical and industrial applications. A simple configuration of hollow-core photonic crystal fiber (HC-PCF) is presented for application as a micro-strain sensor to exhibit better sensitivity then the typical fiber Bragg grating (FBG)-based fiber optic strain sensors. Also, cross-sensitivity to changes in surrounding refractive index is avoided. The performance of the designed sensor is investigated for different strain levels ranging from 0 to 2000 µɛ at the wavelength, 1550 nm. Additionally, due to the air-hole structure of HC-PCF, the sensitivity of strain measurement remains unaffected of the changes in surrounding refractive index (SRI). Sensitivity of HC-PCF also depends on the fiber parameters like pitch and air-filling fraction.

The electron beam trajectories were simulated in a single-stage depressed collector for a millimeter-wave gyrotron. This collector was designed to handle the spent beam obtained after beam-wave interaction in the 100 kW gyrotron (accelerating voltage 55 kV and beam current 5 A). Similar to other high-power gyrotrons, this collector has larger volume considered at ground potential. The cathode and the beam-wave interaction cavity were considered at −40 kV and +15 kV, respectively. The collector sees the depression of 15 kV. The geometry of the collector is considered as three sections: (i) the open entrance conical section, (ii) the smooth cylindrical section, and (iii) the closed conical section. In order to simulate the beam trajectory from nonlinear taper to collector, the electron trajectories and spent beam power distribution data obtained from large-signal analysis have been fed at the entrance of mode converter of the gyrotron with required potentials applied. The collector geometry and the magnetic field are profiled to ensure the landing of the gyrating electrons to the wider smooth cylindrical section for better thermal management. The sensitivity of the magnetic field profile is studied and observed that for ±5% variation in the magnetic field profile would not shift the electron beam landing to conical sections, and the spent electron beam has no interception. The power dissipation on the collector is found to be 80.55 kW. The collector efficiency is calculated as ~48% for 120 kW RF output. The maximum thermal loading on collector inner surface is estimated as 0.38 kW/cm2.

A novel and compact inverted V-shaped slot loaded patch antenna is presented. The proposed antenna comprises of an inverted V-shaped slot, two rectangular slots, and one vertical notch which are connected to each other on the patch, two via hole for shorting ground and patch. Intermediate and final designs are simulated for comparison. The performance analysis of these antennas is compared and analyzed in terms of return loss and gain. The proposed antenna resonates at three frequency bands at resonating frequency of 12.5 GHz, 25.8 GHz, and 28.8 GHz having impedance bandwidth of 6%, 19.45%, and 20.61% with peak gain of 4 dBi, 8 dBi, and 9 dBi, respectively. The first resonating band is useful for Ku band, second is useful for K band, and third is useful for Ka band and 5G applications.

This paper presents the analysis of plus shape slot loaded patch antenna. It is observed that wideband property of antenna is found by suitably selecting the positions and dimensions of the slot on the patch. The operating frequency band of the proposed antenna is achieved from 6.60 to 8.15 GHz with bandwidth of 21.01% (simulated) and 6.50–8.33 GHz with 24.67% bandwidth (theoretical). The characteristics of the antenna are also observed by varying parameters like slot width, substrate height, and dielectric constant. The gain of the antenna is about 6.5 dBi in the operating frequency range. The simulated and theoretical results are compared which are in good agreement. The proposed antenna can be used in C-band applications.

This proposed article presents multiband circular ring loop patch antenna (CLPA). The design is simulated using HFSS tool and studied by varying the ground patch size (area). The substrate material FR-4 is used and studied for the frequency range from 1 to 30 GHz in the proposed article. After analysing the effect of change in ground (GND), it is found that multiband characteristics slightly change in terms of S11 (dB) parameter below 20 GHz and above 20 GHz bandwidth increases. The CLPA shows eight to ten resonating frequency bands. Simulated data show −40.98 (dB) reflection coefficient at the frequency 15.88 GHz in design 1; similarly, design 2, design 3, design 4, design 5, and design 6 obtain maximum reflection coefficient −33.76 (dB) at 10.70 GHz, −37.27 (dB) at 27.65 GHz, −36.43 (dB) at 10.70 GHz, −37.08 (dB) at 21.33 GHz, and −29.11 (dB) at 15.35 GHz. The proposed antenna can be applied in numerous wireless applications by selecting different ground size.

In this paper, W-slot microstrip antenna is explored to obtain multiband resonance. The proposed design shows three resonance frequencies which pivot on the structure of W-slot in the radiating patch. The frequency range of the proposed antenna is between 5 and 10 GHz. The operating bands are at 5.1, 8.3, 9.5 GHz and is suitable for C and X bands. The results are analysed on HFSS simulator.

This paper describes a patch antenna with slotted ground to operate at public domain frequency bands. Proposed design has a ground and a patch on the lower and upper planes of the PCB, respectively. FR4 material is used as substrate for PCB having thickness 0.8 mm. Proposed antenna offers a wide frequency ranges from (2.17–2.53) GHz, (3.0–3.72) GHz, and (5.05–6.5) GHz for Bluetooth, WiFi, WiMAX, and WLAN applications. Ansoft HFSS software is used for simulation process of the proposed design.

A compact two notch-loaded patch antenna with multiple shorting pins for X and Ku band applications is presented. Two different substrates FR-4 and RT Duroid 5880 are used to compare the antenna characteristics. Multiple shorting pins are used to enhance the bandwidth and gain of antenna. Volume of the proposed antenna is 785 mm3, and it resonates at 10.5 GHz and 14.65 GHz with impedance bandwidth of 10.2% and 6.49% and gain of 7.42 dBi and 12 dBi, respectively. This antenna is useful for X band and Ku band applications.

Wireless sensor network (WSN) has become a prominent technology in order to access data from the remote or non-remote areas, e.g., forests, battlefields, hospitals, homes. As WSN has energy constraints, energy-efficient routing protocols are required to prolong the network lifetime. Delay is one of the important parameters for WSN because of it data losses its importance of data and creates congestion in the network also. In this paper, a delay-aware energy-efficient routing protocol is proposed. Delay is minimized with the help of mobile base station, and optimized numbers of hops improve the energy efficiency of proposed routing protocol. Simulation results show the improvement in performance over the existing routing protocol. Extensive simulation study is carried out to evaluate the performance of the proposed protocol with respect to delay, throughput, average residual energy, and network lifetime.

LTE is a widespread technology for high-speed wireless communication that is used in our mobile networks and various data terminals. With the increase in number of users for broadband communication, the requirement for data rate has increased significantly due to the evolution of LTE took place. LTE is the evolvement of high-speed packet access (HSPA) that was assimilated by third-generation partnership project (3GPP) release 8, in order to accomplish widening demand for a very high-speed and proficient access of data. LTE promoted such a high-speed data by enabling larger bandwidth. This paper focuses on the performance of LTE networks at changing spectral frequency bands. In our work, we have tried to simulate LTE networks using NS-2 simulator. Our primary focus is on specifications that greatly affect the behavior of LTE networks. The key specifications that we have used in our work are throughput, average throughput, and jitter. After simulation of our LTE networks, we have concluded that with changing frequency bands, throughput is not affected until and unless bandwidth and modulation type are not varied. We have observed same behavior in case of average throughput but it is smaller in comparison to throughput as bandwidth is shared among multiple broadband users; hence, it can be noted that average throughput falls off with the rise in number of nodes. Jitter does not show any distinct behavior with changing spectral frequency band. It may rise for some instance of time and diminish for other instants.

In recent years, location-based services like emergency, rescue, and response, and location-based marketing are gaining popularity. In all such services, location of involved entity should be located accurately and timely. GSM system is the most popular communication network around the world. Hence, such location detection techniques which require very less or no modification in the existing infrastructure are required. This paper talks about two enhanced ToA-based localization methods using two and three base transreceiver stations. The comparative study between standard ToA techniques and proposed techniques shows encouraging results in various regions like urban area, suburban area, hilly area, and open area.

This paper investigates the use of cooperative multiple-input-multiple-output orthogonal frequency-division multiplexing (MIMO-OFDM) technique to limit energy consumption used to set up communications among distant hubs in a remote wireless sensor network (WSN). As energy exhaustion is a pertinent issue in WSN field, various methods aim to safeguard such asset, particularly by means for harvesting energy amidst communication among sensor hubs. One such widely utilized strategy is multi-hop communication to reduce the energy required by a single hub to transmit a given message, giving a homogeneous utilization of the energy assets among the hubs in the system. The case of multi-hop communication is not continuously more effective than single-hop. In multi-hop communication, energy efficiency will depend upon the distance between transmit and receive clusters. If the distance is large, then the energy expended will be less than single-hop communication. In this paper, an agreeable MIMO-OFDM transmission strategy for WSN is exhibited, which is contrasted with single-hop. The cooperation among adjacent nodes is analyzed, highlighting its points of interest in connection with both. The inference drawn from performance improvement manifests itself in the utility of applying the proposed strategy for energy-sparing purposes.

Wireless sensor and actuator network is the extension of wireless sensor network. Wireless sensor and actuator network is made of sensor and actuator nodes. To overcome energy constraints of sensor networks, some energy-rich nodes (actuator nodes) are deployed in surveillance field. Due to energy heterogeneity of deployed nodes, sensor and actuator nodes have different responsibility. Sensor nodes gather data from surveillance field, and actuator nodes are responsible for transmission of collected data from sink. In this paper, an energy-balanced cluster-based routing protocol for wireless sensor and actuator networks is proposed. The network is divided into clusters, and the most energy-rich node among sensor nodes is selected as cluster head. Cluster head selection and actuator node selection metrics balance the energy consumption of network. Simulation results show the improvements in the performance parameters over the existing routing protocol. Simulation is carried out on network simulator (ns-2.35).

Vehicular ad hoc network deals with traffic congestions. It also ensures the safety and convenience of drivers and passengers. Safety is one of the most important aspects in VANETs, and to make it safe and reliable, its performance needs to be enhanced. To enhance the performance, there is a need to control the congestion. In the high-density region, congestion control becomes a challenging task and special characteristics of VANETs (e.g. high rate of topology changes, high mobility) make it more challenging. In this paper, collective scheduling strategy is proposed. In this strategy, priority of the message is defined which mainly depends on three factors, i.e. size of messages, type of messages and state of networks. Based on these factors, messages priorities are calculated and then these messages are scheduled. Collective scheduling uses the concept of clustering while assigning priority of information. Simulation is carried out to demonstrate improvement in comparison with collective scheduling with tabu search scheduling.

Mobile ad hoc networks are established and deployed spontaneously without any infrastructure in geographical area. The performance of network is satisfied only when all the member nodes have intensity to work in cooperative manner. But due to lack of any centralized unit, it is vulnerable to various attacks of malicious nodes. To overcome these types of attacks, the network has to be enhanced to provide secure delivery services. Our proposed Trusted-Differential Evolution algorithm deals with malicious node and inhibits them to become a member of data transmission path. It has two components: one to find the fittest path and other to deal with fluctuating credibility of nodes through trust. The dynamic of trust is handled by new trust-updation scheme along with punishment factor for malicious node. The proposed algorithm is compared with DSR and genetic algorithm.

Internet of things (IoT) has emerged as a natural evolution of environmental sensing systems such as wireless sensor networks (WSNs). Wireless sensor nodes being resource constrained in terms of limited energy supply through batteries, the communication overhead and power consumption are the most important issues for WSNs design. The sensor nodes have non-replaceable battery with limited amount of energy, which determines the sensor node’s life as well as network lifetime. A key challenge in the design as well as during the operation of WSNs is the extension of the network lifetime even in harsh environment. This paper presents a review on various data aggregation techniques to reduce redundancy and proposes an approach for redundancy elimination during data aggregation exploiting support vector machine (SVM) so that network lifetime can be prolonged during communication of information from sensor nodes to base station. The proposed approach is simulated, and the results are analyzed in terms of average packet delivery ratio, average residual energy, accuracy, and aggregation gain ratio with state-of-the-art techniques.

In the recent era, content-centric network (CCN) is emerging as a future Internet paradigm to leverage scalable content distribution. Similarly, the Internet of things (IoT) is another upcoming technology which integrates as well as manages heterogeneous connected devices over the Internet. Recent literature have shown that IoT architecture can efficiently perform if it is implemented in CCN environment. In addition, considering the openness of the Internet used in IoT communication and limited capacity of IoT devices, security becomes a serious challenge which demands attention of the research community. In this paper, our main objective is to design a secure IoT communication framework that operates in CCN. A certificateless public key infrastructure is designed for our resource-constrained IoT communication framework in CCN. We have incorporated elliptic curve cryptography (ECC), a state-of-the-art lightweight cryptosystem, to ensure security of the proposed scheme. Finally, an in-depth security analysis confirms that the proposed scheme is resistant to various relevant cryptographic attacks.

Internet of Things (IoT) is the interconnection of physical entities to be combined with embedded devices like sensors, activators connected to the Internet which can be used to communicate from human to things for the betterment of the life. Information exchanged among the entities or objects, intruders can attack and change the sensitive data. The authentication is the essential requirement for security giving them access to the system or the devices in IoT for the transmission of the messages. IoT security can be achieved by giving access to authorized and blocking the unauthorized people from the internet. When using traditional methods, it is not guaranteed to say the interaction is secure while communicating. Digital certificates are used for the identification and integrity of devices. Public key infrastructure uses certificates for making the communication between the IoT devices to secure the data. Though there are mechanisms for the authentication of the devices or the humans, it is more reliable by making the authentication mechanism from X.509 digital certificates that have a significant impact on IoT security. By using X.509 digital certificates, this authentication mechanism can enhance the security of the IoT. The digital certificates have the ability to perform hashing, encryption and then signed digital certificate can be obtained that assures the security of the IoT devices. When IoT devices are integrated with X.509 authentication mechanism, intruders or attackers will not be able to access the system, that ensures the security of the devices.

With the urbanization, exposure of mankind to the noise is increasing day by day leading to many health issues. For low-frequency noise reduction, active noise control system is widely applied in many applications. In the present paper, the effect of different secondary path lengths in FxLMS and virtual sensing technique based ANC system is studied. The virtual sensing technique based ANC is applied in the cases where it is not feasible to place the error microphone physically at the desired location. The filter coefficients of secondary path of ANC system are measured experimentally for the three different filter lengths, k (i.e., k = 64, 128, and 256) using Texas Instruments TMS320C6713 processor in the semi-anechoic chamber. The performance of ANC system with different filter lengths is analyzed in the terms of residual noise, signal-to-noise ratio, computational load, and error plots. The comparison suggests that secondary path of different filter lengths suits for FxLMS and virtual sensing technique based ANC systems.

Filtering is an immense process in spectral analysis of signals. In designing of filters, window functions are usually used. In this paper, we present the variety of window functions based on the scale-free COordinate Rotation DIgital Computer (CORDIC) algorithm for the target angle which covers the complete coordinate space. To overcome the problem of more occupied area and speed, we present a study of a different design that is scale-free CORDIC algorithm-based window function architectures. The current paper presents the simulation and synthesis results of two designs which are coded in very high speed integrated circuit hardware description language (VHDL). The Xilinx 13.1 software is used for the simulation and synthesis of coded design, and also these designs are mapped into Virtex-5(XC5VLX20T-FF323) field-programmable gate array (FPGA) device.

Medical images have made a great influence on medicine, diagnosis, and treatment. The essential element of medical image processing is segmentation for identifying the region of interest. The fundamental methods of image segmentation are unable to process the large dataset of images and require scaling to make them more interactive. In order to address this issue, an exponential entropy-based image segmentation methods are proposed which are based on boundary demarcation, contour- and learning-based approaches. To accelerate these methods on graphical processing unit, a well-defined concept of memory preallocation and vectorization are incorporated in the novel approach. Results have been investigated on 240 gold standard dermoscopy images. These results reveal that the optimized methods of segmentation are computationally benefited from GPU processing in terms of speed and accuracy for skin lesion detection.

Segmenting object from a moving camera is a challenging task due to varying background. When camera and object both are moving, then object segmentation becomes more difficult and challenging in video segmentation. In this paper, we introduce an efficient approach to segment object in moving camera scenario. In this work, first step is to stabilize the consecutive frame changes by the global camera motion and then to model the background, non-panoramic background modeling technique is used. For moving pixel identification of object, a motion-based approach is used to resolve the problem of wrong classification of motionless background pixel as foreground pixel. Motion vector has been constructed using dense flow to detect moving pixels. The quantitative performance of the proposed method has been calculated and compared with the other state-of-the-art methods using four measures, such as average difference (AD), structural content (SC), Jaccard coefficients (JC), and mean squared error (MSE).

An image quality index plays an increasingly vital role in image processing applications for dynamic monitoring and quality adjustment, optimization and parameter setting of the imaging systems, and finally benchmarking the image processing techniques. All the above goals highly require a sustainable quantitative measure of image quality. This manuscript analytically reviews the popular reference-based metrics of image quality which have been employed for the evaluation of image enhancement techniques. The efficiency and sustainability of eleven indexes are evaluated and compared in the assessment of image enhancement after the cancellation of speckle, salt and pepper, and Gaussian noises from MRI images separately by a linear filter and three varieties of morphological filters. The results indicate more clarity and sustainability of similarity-based indexes. The direction of designing a universal similarity-based index based on information content of the image is suggested as a future research direction.

Digital images are used in courtrooms as evidence. We cannot predict nativity of the image without forensic analysis. Tampering with the image is common nowadays with a lot of online and offline tools. To hide an object in an image, regions of the same image are copied and pasted on that object, and this is known as copy-move forgery. In this paper, we have introduced a technique to detect such type of forgery, known as CMFD. In this technique, the image is pre-processed by converting RGB into YCbCr and then Y channel is decomposed into four components of translation-invariant stationary wavelet transform (SWT). Its LL (approximation) component is then divided into 8 × 8 blocks. Further, from each block, we have taken six mean features which are calculated by dividing each block into four squares and two triangular blocks and put them into feature vector with block location. After sorting these feature vectors into lexicographical order, we get the location of forged regions.

Image fusion is a process of combining two or more images into one, in order to obtain more relevant data or information from it. In this paper, we have proposed dual discrete wavelet transform (DDWT)-based image fusion method in frequency domain with averaging the principal component analysis that overcomes the spatial distortion, blocking artifact, and shift variance of the fusion methods. The results of the proposed method have been promising for qualitative and quantitative evaluations that are performed on subjective and objective criteria, respectively, which are shared in the experimental results for considered test images over fusion methods.

Researches on person re-identification have become more prevalent as a core technique in visual surveillance systems. While a large volume of person re-identification (Re-ID) methods adopt experimental settings with a large patch size, relatively small ones are not much considered. Thus, this work focuses on investigating comparison of person Re-ID performance when color and shape features-based are adopted for small patch size images with low resolution. First, the deep decompositional network is used to divide the person into upper and lower body parts. Then, color and shape features are extracted. Finally, using single or combination of features, similarity-based ranked matching scores are computed. The person Re-ID performance is evaluated based on VIPeR dataset. From the experiment results, we found that color-based feature is better than shape-based features, and the combination of color and shape features-based can be meaningful.

Human detection and tracking is one of the most crucial tasks in video analysis. We can find its applications in areas like video surveillance, augmented reality, traffic supervision. KLT and CAMSHIFT are two popular algorithms for this task. In this paper, we present a comparison of their performance in different scenarios. As a result, this paper provides concrete statistics to choose an appropriate algorithm for tracking, given the nature of the objects and surrounding. Our experiments show that KLT algorithm is advantageous for crowded scenes, whereas CAMSHIFT performs better for tracking a specific target. Based on our analysis, we conclude that KLT algorithm performs more efficiently than CAMSHIFT algorithm for video object tracking.

Video-based human activity recognition has fascinated researchers of computer vision community due to its critical challenges and wide variety of applications in surveillance domain. Thus, the development of techniques related to human activity recognition has accelerated. There is now a trend towards implementing deep learning-based activity recognition systems because of performance improvement and automatic feature learning capabilities. This paper implements fusion-based dual-stream deep model for activity recognition with emphasis on minimizing amount of pre-processing required along with fine-tuning of pre-trained model. The architecture is trained and evaluated using standard video actions benchmarks of UCF101. The proposed approach not only provides results comparable with state-of-the-art methods but is also better at exploiting pre-trained model and image data.

While human detection has been significantly recognized and widely used in many areas, the importance of human detection for behavioral analysis in medical research has been rarely reported. Recently, however, efforts have been actively made to recognize behavior diseases by measuring gait variability using pattern analysis of human detection results from videos taken by cameras. For this purpose, it is very crucial to establish robust human detection algorithms. In this work, we modified deep learning models by changing multi-detection into human detection. Also, we improved the localization of human detection by adjusting the input image according to the ratio of objects in an image and improving the results of several bounding boxes by interpolation. Experimental results demonstrated that by adopting the proposals, the accuracy of human detection could be increased significantly.

Keyframe extraction of videos is useful in many application areas such as video copy detection, retrieval, indexing, summarization. In this paper, we propose a novel shot-based keyframe extraction algorithm. The proposed algorithm is capable of detecting both shots and keyframes of any video efficiently. For extraction of keyframes, frames of video are clustered into shot transitions. These shot transitions of the video are obtained using higher-order color moments and zero-normalized pixel correlation coefficients. In each shot, all the frames are scanned to detect frame with highest standard deviation in that particular shot and chosen as keyframe to that shot. The proposed method is tested on videos of personal interviews with luminaries. Performance of the proposed method is evaluated on the basis of five parameters—recall, figure of merit, detection percentage, accuracy and missing factor. The proposed method is able to detect both abrupt and gradual shot transitions with comparatively less computational complexity. The exhaustive analysis of results shows the sound performance of the proposed method over the methods used in this study.

In this paper, we solve the problem of salient object detection by using an ensemble mechanism of edge detection and active contours. Edge cues are used to provide a solution to the initialisation problem of active contour. The active contour method suffers from the initialisation problem. If the initial contour lies in a region with low probability of salient object, the final salient object detection provides inaccurate results. In this paper, the problem is addressed by generating a binary mask using Sobel edge detection method which acts as the initial contour. The binary mask makes sure that the contour lies in the region with high probability of finding a salient object. This work is an improvement upon the active contour model. The method is simple and fast and follows basic human intuition to find salient objects. The proposed work is compared against seven recent works and gives better results in terms of precision, recall and false positive rates.

An apple is one of the most cultivated and consumed fruits in the world and continuously being praised as a delicious and miracle food. It is a rich source of Vitamin A, Vitamin B1, Vitamin B2, Vitamin B6, Vitamin C, and folic acid etc, whereas the rotten fruits affect the health of human being as well as cause big economical loss in agriculture sectors and industries. Therefore, identification of rotten fruits has become a prominent research area. This paper focuses on the classification of rotten and good apple. For classification, first extract the texture features of apples such as discrete wavelet feature, histogram of oriented gradients (HOG), Law’s Texture Energy (LTE), Gray level co-occurrence matrix (GLCM) and Tamura features. After that, classify the rotten and good apples by applying various classifiers such as SVM, k-NN, logistic regression, and Linear Discriminant. The performance of proposed approach by using SVM classifier is 98.9%, which is found better with respect to the other classifiers.

The human eye is one of the most beautiful and important sense organs of human body as it allows visual perception by reacting to light and pressure. Human eyes are capable of differentiating approximately 10 million colors. It contains more than 2 million tissues and cells. Along with these entire specialties, human eyes are the most delicate and sensitive organ. If not taken proper care, it may be infected with various diseases like glaucoma, myopia, hyper-myopia, diabetic retinopathy, age-related macular disease. Therefore, early-stage detection of these diseases could help in curing them completely and prevent from complete blindness. In this paper, we propose an approach to classify the normal (healthy) and abnormal (disease-infected) retinal images by using retinal image feature-based machine learning classification approach. The performance of proposed approach by using SVM classifier is 77.3%, which is found better with respect to the other classifiers like k-NN, linear discriminant, quadratic discriminant and decision tree classifiers.

In today’s digital world due to unlimited content, product and services available online, finding an item that satisfies user requirement and taste by simply web searching is near impossible. Recommender systems are information filtering tool which provides personalized results. Movie and video recommender system is also gaining popularity due to the growth in online streaming video content Web sites and its subscriber. Accuracy and efficiency are two major aspects of a recommendation engine because it is directly related to user experience. To achieve higher accuracy, user feedback is required which can be collected either explicitly or implicitly. Explicit feedback is not always available and not always unbiased, so implicit feedback seems to be a better option for user preference collection. In this paper, a new framework is proposed which collects the implicit user feedback (along with explicit) for a movie and video recommender system. Implicit feedbacks can be converted to explicit feedback using the proposed UARCA which can be used to improve the accuracy of recommendation engine.

Large volume of data are generated in educational institutions, which are of heterogeneous and unstructured nature. However, there is a dearth of effective data mining tools and techniques which can handle these voluminous academic data and support exploration of essential knowledge. Educational data mining (EDM) is an emerging research area dedicated toward development of tools and techniques for exploring data in educational settings. In this paper, we propose a trusted EDM framework that can deliver multiple academic tasks according to the need of various stakeholders. In order to deliver such purposes, our framework utilizes data mining tools and techniques over unified data collected from institution’s databases and various knowledge sources. As an example of the concept, we utilize data provided by National Institutional Ranking Framework (NIRF) for showing how same data can be mined to fulfill different needs of various stakeholders through our proposed framework.

Machine translation is the noticeable field of the computational etymology. Computational phonetics has a place with the branch of science which bargains the dialect perspectives with the help of software engineering innovation. In this field, all handling of regular dialect is finished by the machine (PC). Calculation is done by considering all features of the language and in addition vital principal of sentence like its structure semontics and morphology. Machine ought to see all these conceivable parts of the dialect, yet past work does not deal with alternate prerequisites amid machine interpretation. Current online and work area machine interpretation frameworks disregard numerous parts of the dialects amid interpretation. Because of this issue, numerous ambiguities have emerged. Because of these ambiguities, current machine interpreter is not ready to deliver right interpretation. In this proposed work, genetic algorithm-based machine translation system is proposed for the translation of Sanskrit into Hindi language which is more efficient than the existing translation systems.

As a non-invasive optical neuroimaging technique, functional near infrared spectroscopy (fNIRS) is currently used to assess brain dynamics during the performance of complex works and everyday tasks. However, the deep learning approaches to distinguish stress levels based on the changes of hemoglobin concentrations have not yet been extensively investigated. In this paper, we evaluated the efficiencies of advanced methods differentiating the rest and task periods during stroop task experiments. First, we explored that the apparent changes of oxy-hemoglobin (HbO) and deoxy-hemoglobin (HbR) concentrations associated with two mental stages did exist across each participant. Then, a novel discrimination framework was studied. Deep learning approaches, including convolutional neural network (CNN), deep belief networks (DBN), have enabled better classification accuracies of 84.26 ± 9.10% and 65.43 ± 1.59% as our preliminary study.

This paper presents the mathematical modeling-based design and simulation of normal mode MEMS capacitive pressure sensor for blood pressure sensing application. The normal blood pressure of human being is 120/80 mmHg. But this range varies in case of any stress, hypertension and some other health issues. Analytical simulation is implemented using MATLAB®. Basically, normal mode capacitive pressure sensors have a fixed plate and a moveable diaphragm which deflects on application of pressure with the condition that it must not touch the fixed plate. Deflection depends on material as well as thickness, shape and size of diaphragm which can be of circular, elliptical, square or rectangular shape. In this paper, circular shape is chosen due to higher sensitivity compared to other diaphragm shapes. Deflection, base capacitance, change in capacitance after applying pressure and sensitivity are reported for systolic and diastolic blood pressure monitoring application, and study involves determining the optimized design for the sensor. Diaphragm deflection shows linear variation with applied pressure, which follows Hook’s law. The variation in capacitance is logarithmic function of applied pressure, which is utilized for analytical simulation.

This paper shows a small experiment for analyzing the yield of Allium cepa with the help of checkbook method. This manual method is used for maintaining the schedule of irrigation in a kitchen garden which is a closed area nearby the living place. The paper also focuses on the careful steps needed in performing the agricultural practices for cultivating any vegetable crop. The result of experiment suggests the need for automation in irrigation as various limitations are observed in following the conventional cultivation practices. Hence for improving the yield of the crop, a scientific automated irrigation system is proposed by harnessing solar energy in the urban and rural sector.

The remarkable advances in health industry have led to a significant production of data in everyday life. These data require processing to extract useful information, which can be useful for analysis, prediction, recommendations, and decision making. Data mining and machine learning techniques are used to transform the available data into valuable information. In medical science, disease prediction at the right time is the central problem for professionals for prevention and effective treatment plan. Sometimes, in the absence of accuracy this may lead to death. In this study, we investigate supervised machine learning algorithms—Naive Bayes, random forest, and decision tree algorithm—for prediction of anemia using CBC (complete blood count) data collected from pathology centers. The results show that Naive Bayes technique outperforms in terms of accuracy as compared to C4.5 and random forest.

This paper presents the learning pattern analysis of online learning management system Moodle. The experimental work was carried out in Banaras Hindu University, India on the students of three years post graduate course on Computer Application. The comparative study has done on learning patterns of the students on tradition pedagogy with online pedagogy through the learning management system. The Moodle data analytics tool was used for the purpose of reporting students’ data. The results of students learning performance were compared and analyzed using t-test, content analysis and various other mining and analytics tools.

Recognizing the epileptogenic area of a brain is done by analyzing the electroencephalogram signal. This area is responsible for the occurrence of seizure activity in a brain. In this paper, a methodology has been presented for the analysis of electroencephalogram to recognize epileptogenic area of brain. Ensemble empirical mode decomposition (EEMD) has been used for the estimation of intrinsic mode functions (IMFs), and six parameters consisting of statistical and frequency-based feature have been extracted from first ten IMFs. The ReliefF algorithm has been used to select the relevant features for the training of artificial neural network (ANN) for recognition of epileptogenic area. The methodology has been evaluated based on accuracy, specificity and sensitivity. The comparison has also been made with other methods of epileptogenic area detection where it has been observed that the proposed method outshines other.

A cyber attack on critical infrastructure differs from attack on general information and communication systems. Recent trends of cyber attacks on critical infrastructure are found to be complex cyber attacks (CCA) because they are multistage, multi-phase and multi-pace. Detection of these complex cyber attacks is yet a challenging problem because they are intractable to describe and analyze. In this paper, complex cyber attacks are analyzed and as a response to detection of an attack, a forensic investigation framework for CCA is proposed. This paper focuses on forensic investigation framework for CCA in cyber physical system, which is large and geographically distributed. A model for forensics investigation process is proposed which is based on goals and sub-goals of an attack. This helps to reconstruct the event and collect data for evidence. Since complex cyber attacks are constructed with a variety of malwares and some of them show the property of self-propagation, an epidemic analysis in forensic investigation process determines the spread of infection in large infrastructures. Addition of epidemic behavior of malware in forensic investigation process is helpful to understand the dynamics of infection in a large, heterogeneous infrastructure.

In machine learning and data mining tasks, feature selection has been used to select the relevant subset of features. Traditionally, high-dimensional datasets have so many redundant and irrelevant features, which degrade the performance of clustering. Therefore, feature selection is necessary to improve the clustering performance. In this paper, we select the optimal subset of features and perform cluster analysis simultaneously using modified-BPSO (Binary Particle Swarm Optimization) and K-means. Optimality of clusters is measured by various cluster validation indices. By comparing the overall performance of the modified-BPSO with the BPSO and BMFOA (Binary Moth Flame Optimization Algorithm) on six real datasets drawn from the UC Irvine Machine Learning Repository, the results show that the performance of the proposed method is better than other methods involved in the paper.

The silicon-on-insulator (SOI) technology provides the higher current driving capability, low power consumptions, reduced SCEs and extensive scaling of the channel length. But SOI-based MOSFETs are weak in thermal stability like self-heating. In this paper, a comparative analysis of asymmetrical double metal double gate (ADMDG) and symmetrical double metal double gate (SDMDG) at the zero-temperature-coefficient (ZTC) bias point is proposed. ZTC is the bias point where the device constraints become free of variation in temperature. ADMDG and SDMDG devices are simulated by 2-D Atlas simulator. 2D-Atlas simulation revealed the figure of merit (FOMs) such as transconductance (gm), output conductance (gd), intrinsic gain (Av), on-current (Ion), off-current (Ioff), on–off current ratio (Ion/Ioff) and cutoff frequency (fT). The simulation results give the presence of inflection point of the devices. The variation of ZTC point for transconductance (ZTCgm) and drain current (ZTCIDS) for ADMDG and SDMDG MOSFETs is compared.

The valuable treating of parallelism on an interconnection network entails optimizing inconsistent performance indices, such as the reduction of communication and scheduling overheads and also uniform distribution of load among the nodes. In this kind of a system a number of nodes process the numerous jobs concurrently. A novel dynamic scheduling scheme that supports task unbiased structure approach is proposed for a particular class of multiprocessor networks known as linearly extensible multiprocessor networks. The significance of proposed scheduling scheme is remedying the communication overhead, delay in task execution and efficient processor utilization, which ultimately improves the total execution time. The proposed algorithm is implemented on a set of processors known as nodes which are linked through certain interconnection network. In particular, the performance is evaluated for linear type of multiprocessor architectures. In addition, a comparison is also made by implementing standard scheduling algorithm on same architectures with same number of nodes. The metrics used for comparison are Load Imbalance Factor (LIF), which represents the deviation of load among processors after achieving load balancing and execution time. The comparative simulation study shows that the proposed scheme gives better performance in terms of task scheduling and execution time when implemented on various linearly extensible multiprocessor networks.

In this paper, an architecture is presented for a fused floating-point three operand adder unit. This adder executes two additions within a single unit. The purpose of this execution is to lessen total delay, die area, and power consumption in contrast with traditional addition method. Various optimization techniques including exponent comparison, alignment of significands, leading zero detection, addition, and rounding are used to diminish total delay, die area, and power consumption. In addition to this, the comparison is described of different blocks in term for die area, total delay, and power consumption. The proposed scheme is designed and implemented on Xilinx ISE Design 14.7 and synthesized on Synopsis.

Moving towards micrometre scale to nanometre scale device shrinks down emerging nanometre technology such as quantum-dot cellular automata as a nesting success. The introduced architecture is robust where the explicit design of full adder and full subtraction uses for Ex-OR design. A new architecture of Ex-OR based on one majority gate is proposed, which its most optimized architecture and its placement of cells from the novel design. The analysis based on simulation showed that the introduced Ex-OR and full adder makes only 11 and 46 cells count, respectively. In proposed Ex-OR design, first output is received with no any latency which can be a suitable design for implementation of the high-speed full adder design. In addition, power estimation results are obtained after simulation of proposed designs in QCAPro tool. Therefore, the novel designs improve the energy dissipation parameters such as mean leakage energy dissipation, mean switching energy dissipation and total energy dissipation 75, 11.28 and 82.19% in comparison with the most robust design in existing.

In this paper I have designed a low power SARADC based on charge redistribution. In the proposed design, I have designed a successive approximation register ADC using low power double tail dynamic comparator with control transistors (MC1 & MC2). The peculiar advantage of low power double tail dynamic comparator with control transistor (MC1 & MC2) is that the power consumption is reduced by three times as compared with the existing low power double tail dynamic comparator. The proposed comparator having the power consumption 58 µW can be operated at maximum frequency 16 MHz. Best efforts has been made to design an efficient SAR and control unit in the proposed design. In this 4-bit SAR ADC which can be operated at clock frequency of 16.5 MHz and having sampling frequency of 3.3 MHz is designed. The power consumption of the proposed circuit is 113 µW. In the paper quantization noise, SNR, SNDR and ENOBs are also obtained which are better in performances than existing ADCs.

The magnetic levitation (maglev) system has become a very efficient technology in the rapid mass transportation system due to its frictionless motion. It is an open-loop unstable system, so it requires a controller implementation for its stabilization and position-tracking. Since it is inherently a nonlinear system, its controller design is a challenging problem. This paper presents a 2-degree-of-freedom PID controller designed to stably levitate the object in the magnetic field as well as for the position-tracking. All the simulation works are performed under MATLAB environment, and the simulation results have been discussed at the end of this paper.

A high-gain low-power CMOS low-noise amplifier is simulated using TSMC 0.18-µm CMOS technology. The cascade topology is used to get the high-gain and low-noise figure value. The source degeneration technique is used for the wideband matching. The circuit is simulated for 3.1–10.6 GHz in ultawideband. The simulated results show the maximum gain of 21.574 dB at 6.378 GHz and positive gain maintained during the entire frequency range. The highest noise figure value is 4.311 dB at 7.662 GHz, and the lowest value is 2.477 dB at 3.1 GHz. The matching circuit at input and output terminals shows the input return loss of 22.262 dB at 10.38 GHz, while the output return loss of 30.936 dB at 4.1 GHz. The circuit is simultaed at 1.2 V which draws the power consumption of 17.734 mW. The designed circuit shows the optimum value of gain, noise figure, matching and power consumption.

In this paper, development of nano-rough Zn0.92 Fe0.08O thin film (NFZO) on glass substrate and application in the detection of methane are reported. The deposition of film used technique is high electro-spin 500–3500 rpm pattern. Doping of 8% iron in zinc oxide synthesized by solid-state route and chemical route is applied in spin coating. Study of thin film characterizions by HR-SEM, EDX and AFM, while sensor response for 100–300 ppm methane at 150 °C. There are iron ions like ferrous and ferric improved to the gas sensing properties. Overall sensitivity rapidly increases with concentrations of methane at 150 °C. The deposition of NFZO by the high rpm spin pattern is low cost and simple operation.