Search

Accurate disease categorization using endoscopic images is a significant problem in Gastroenterology. This paper describes a technique for assisting medical diagnosis procedures and identifying gastrointestinal tract disorders based on the categorization of characteristics taken from endoscopic pictures using a vision transformer and transfer learning model. Vision transformer has shown very promising results on difficult image classification tasks. In this paper, we have suggested a vision transformer based approach to detect gastrointestianl diseases from wireless capsule endoscopy (WCE) curated images of colon with an accuracy of 95.63\%. We have compared this transformer based approach with pretrained convolutional neural network (CNN) model DenseNet201 and demonstrated that vision transformer surpassed DenseNet201 in various quantitative performance evaluation metrics.

Smart meter is a unique metering system that provides users the capability to monitor power generation, transmission and consumption by enhancing the maintenance of real-time supervision and communication. By hacking and breaching the meter, faulty data can be transferred to the utility by the hacker. The power consumption data stored in a smart meter can be spoofed by two different hacking mechanisms. They are: Communication Interface Attack and Physical Memory Attack. In this paper, the malicious behaviour of a specific set of code is measured implementing fuzzy analytical hierarchy process. Analytical hierarchy process is engaged in the formation and inspecting of sophisticated issues by mathematical calculations. This scheme considers the relative priorities of factors or alternatives with a systematic and logical approach and serves the best alternative. Fuzzy theory has been incorporated into traditional AHP method for selecting and justifying alternatives in a systematic way.

Low-DropOut (LDO) regulators are one of the most essential and critical analog blocks in power management of System-on-Chip (SoC) design. In this work, we present an external capacitor-less LDO voltage regulator design implemented in 45nm technology. The proposed design eliminates the resistive feedback network with a transistor which allows more control over output voltage and improve the transient response and PSRR. The design consists of Error Amplifier (EA), one common-source (CS) stage, one buffer stage and 2 pass transistors which offers a wide range of output voltage ranging from 0.4V-1.2V with an input voltage of 1.8V. The design also ensures low power operation with 7.824uA quiescent current. Furthermore, the design incurs minimum area of 0.0149mm2and voltage overhead of 25mV(max) along with a stable output voltage with less than 1mV ripple. The schematic and layout designs are implemented and simulated in Cadence Virtuoso using 45nm bulk CMOS process.

Graphene Nano Ribbon Field Effect Transistors (GNRFETs) are being investigated for a wide range of applications due to its unique characteristics like controllable threshold voltage, high I on -I off ratio, etc. One of the emerging areas of research on GNRFET is its utilization in beyond-binary logic and memory circuits designs. This paper presents new designs of a basic ternary D latch and a positive edge-triggered Master-Slave D flip-flop. The proposed designs exploit the electrical properties of GNRFETs and appears to be better than other similar designs in terms of delay, static power, and the average power. Simulations are done using a 16nm three ribbon MOS-GNRFET HSPICE model.

Graphene Nano Ribbon Field Effect Transistors (GNRFET) have recently started drawing attractions due to its’ unique electrical properties. The possibility of controlling threshold voltage by controlling the width of the Ribbon presents us with a great advantage of using GNRFET into a Multi-Valued Logic (MVL). In this paper, different electrical aspect like threshold voltage, on-off characteristics of GNRFET has been studied. And based on these, the scope of using it into a ternary logic has been projected.

The purpose of this paper is to explore the role of Battery Energy Storage System (BESS) in improving the frequency response of a low inertia power system. For the recent years, prolific amount of renewable energy resources that include wind and photovoltaic (PV) sources have been introduced in South Australian power system. These wind generators are mostly variable speed machines, hence they neither have inertia response nor can they perform governor like response following a contingency. The demand in South Australia can be significantly catered utilizing renewable sources and considerable power import from the adjacent state Victoria. Therefore, the number of synchronous generators in the South Australian network can be small. In such situations, the frequency response following a contingency can be concerning and the network may undergo substantial under frequency load shedding. To overcome this challenge, this paper proposes a frequency responsive BESS model to improve the frequency response of a low inertia power system that resembles the South Australian network. The network is investigated under various operating conditions to explore the performance of BESS regarding frequency response enhancement.

Zinc Telluride (ZnTe) is a very promising binary semiconductor material is chosen for the ultra-thin approach due to its high absorption coefficient, wider band gap of 1.42 eV and higher thermal stability. This paper illustrates the numerical analysis of the insertion of a thin Lead Telluride (PbTe) BSF layer between the ZnTe absorber layer and the back contact in the ultra-thin ZnTe solar cell is investigated by the wxAMPS simulation software. This PbTe BSF layer offers an extra hole tunneling action that produces a quasi ohmic contact near to the back contact and BSF region. The BSF layer minimizes the recombination losses and the quasi ohmic contact increases the carrier collection that improves the cell performance of the ZnTe solar cell. The simulated result was found 17.55% efficiency for the proposed ultra-thin cell without BSF and the improved cell efficiency was gained 22.15% for only $0.9{\mu} m$ ZnTe layer with PbTe BSF layer. Besides, it was found higher thermal stability of the proposed cells where the temperature coefficient is of ${(-0.029\%/^{\circ} \mathbf{k}).}$

The target of this research work is to investigate the role of High Voltage DC (HVDC) interconnection for enhancing frequency response in a low inertia grid. In recent times, installed capacity of wind and photovoltaic (PV) sources has significantly increased in South Australian power system. Most of these wind power plants deploy variable speed wind machines, whereas PV sources are predominantly distributed at consumer end. Therefore, wind power plants and PV generators in South Australia do not participate in frequency regulation. Under prolific penetration of wind and PV and considerable power ingress from adjoining system, an insignificant number of synchronous machines may be present in the South Australian network. During such condition, a big contingency (trip of an interconnection/a major generator) may result in unacceptable frequency response. Consequently, the network might face substantial under frequency load shedding. To mitigate this challenge, a solution for improving frequency response by taking into account the frequency control support from an HVDC interconnection is explored in this paper. To this end, a low inertia network, which mimics the South Australian grid, is analysed under various operating conditions.

This paper represents the upcoming vital issue of the end-of-life management of waste generated by the photovoltaic module for Bangladesh. Due to inadequate sources of conventional fossil fuel, Bangladesh government is looking forward to producing electricity from renewable sources, especially solar energy. As the PV power generation is increasing with time, so will the volume of obsolete PV panels. Growing PV panel waste presents a new environmental challenge, but also unprecedented opportunities to create value and pursue new economic avenues. The PV waste management regulation of different regions of the world is presented here and in line of those a guideline is constructed for Bangladesh. From the available data it is found that proper PV -waste management can unlock a large stock of raw materials and other valuable components. Proper management and recycling can at a time protect our environment from being polluted by the PV-waste, also it will support considerable economic value creation by adding financial support to our economy.

Fully Depleted Silicon-on-Insulator (FDSOI) and Fin Field Effect Transistor (FinFET) are replacing the bulk MOSFET processes for lower technology nodes. Bulk MOSFETs suffer from severe short channel effects and leakage issues. Sleep transistor based power gating circuits are typically used to address leakage power. In this paper, data retention transistor along with sleep transistor is used to improve the performance. The proposed technique is implemented in 20nm FDSOI and 20nm FinFET technology. Simulations are performed in HSPICE and 2-input NAND gate is used for test purpose. It is observed that FinFET-NAND gate consumes 3.75 times less energy compared to FDSOI-NAND gate during active mode. However, FinFET-NAND gate consumes 1.05 times more energy than FDSOI-NAND gate during hold mode.

In this paper, multi-carrier code division multiple access (MC-CDMA) based wireless sensor network is proposed for underwater applications. We have also proposed Optical communication along with Acoustic communication technique to improve the performance of Underwater Wireless Sensor Network (UWSN) in the harsh underwater environment. We have demonstrated a new path loss model for the optical signal in underwater channel and derived a SNR equation for MC-CDMA. Our purpose of this work is to improve the performance of UWSN through optical communication and combat interference problem utilizing MC-CDMA technique. Our proposed model successfully removes interference problem of optical signal in UWSN due to multipath, scattering, reflection or refraction problem of underwater channel. The BER performance shows that the proposed MC-CDMA based Optical-Acoustic UWSN outperformed over traditional acoustic signalling based UWSN.

Ternary logic devices are expected to lead to an exponential increase of the information handling capability, which binary logic cannot support. Memcapacitor is an emerging device that exhibits hysteresis behavior, which can be manipulated by external parameters, such as, the applied electric field or voltage. One of the unique properties of the memcapacitor is that by using the percolation approach, we can achieve Metal-Insulator-Transition (MIT) phenomenon, which can be utilized to obtain a staggered hysteresis loop. For multivalued logic devices staggered hysteresis behavior is the critical requirement. In this paper, we propose a new conceptual design of a ternary logic device by vertically stacking dielectric material interleaved with layers of graphene nanoribbon (GNR) between the two external metal plates. The proposed device structure displays the memcapacitive behavior with the fast switching metal-to-insulator transition in picosecond scale. The device model is later extended into a vertical-cascaded version, which performs as a ternary device.