This paper presents a voice-activated health monitoring system implemented using a Raspberry Pi. The prototype consists of two nodes: a medical sensor node with the user and a processing node, which communicates via a radio link. This system relies on speech recognition as the basis of its voice-activated design. It uses Mel-frequency cepstral coefficients for feature extraction and a dynamic time-warping algorithm for keyword recognition and comparing reference patterns. The system offers a stable, secure, and cloud-independent solution, achieving a keyword spotting accuracy of 95%.

IoT networks have become widespread in different technological industries due to the development of 6G networks. In this paradigm shift, industries like healthcare and autonomous vehicle research have incorporated Large Language Models (LLMs) into their applications and services. This combination has improved the effectiveness of Internet of Thing (IoT)-driven applications requiring intelligent interactions between humans and machines, bridging these wireless services to real-time intractability. However, as the IoT network grows, scalability and security issues arise. We present a private blockchain-based edge IoT platform to address these concerns in IoT-based LLM services. We evaluated our design's feasibility by testing its responsiveness and analyzing its security contributions. The results show the potential of our platform to improve the scalability of the IoT network through edge computing and reinforce its security through the private blockchain.

Integrating smart technologies in vehicles has brought rise to connected and autonomous vehicles (CAVs), One of the services impacted by this paradigm shift is driving assistance. Most systems use learning-based approaches such as object recognition to improve transportation quality. So, these services require exchanging and sharing data along the CAV network, which raises security issues. Within this work is a federated learning (FL)-based platform as a step towards secure CAV systems. It uses FL to secure client data locally and alleviate pressure on CAV servers and services against targeted attacks. A testbed evaluates the platform's feasibility in preserving the integrity of its implemented classifier while keeping a level of security of its training data. According to experimental results, the FL-based implementation can maintain the integrity of the classifier's accuracy even after introducing its distributive scheme. Also, security evaluations present the benefits of FL reinforcing network security and improving client data privacy. Based on the results, the proposed platform proves its feasibility as an integrity-preserving and secure option for object recognition-based driving assistance services in CAVs.

In recent years, healthcare systems have included the Internet of Things (IoT) technology in their services, such as in remote patient monitoring systems. Wearable IoT devices can provide information regarding the patient's health that are accurate and time-sensitive. However, vulnerabilities are evident as more IoT devices connect to the network. For healthcare services, the security of patient data is an issue. At the same time, with real-time data transmissions, the network runs into manageability concerns. In this work, we propose a private blockchain-based Wireless Body Area Network (WBAN) platform to aid wearable IoT devices in healthcare services. We chose this blockchain technology due to its strengths in security. Then, we enable a distributive architecture using WBANs to introduce a decentralized configuration that can ensure privacy among wearable IoT devices within the network. To evaluate the feasibility of the proposed platform in terms of latency and throughput, we conducted experiments with several wearable IoT devices. The results show that integrating a WBAN to create a fog server improves the network performance with an increasing number of IoT devices and packet size. Also, the blockchain showed its ability to address security threats in healthcare services. We evaluate our proposed platform through a performance test and a STRIDE threat model, and we prove its feasibility in improving the security and manageability of wearable IoT devices in healthcare.

Several signals can be collected from wearable devices containing important physiological and psychological information. Understanding various physiological signals is significant for computers to recognize human emotional states. Electrodermal Activity (EDA), originating from the spontaneous activation of sweat glands in the skin, is closely related to mood, arousal, and attention and is the most widely used measurement in the physiological response system for emotional state detection. However, extracting valuable features from EDA signals and making accurate emotional classification predictions has always been challenging. With the continuous development of models with representation learning capabilities, the use of deep learning models to automatically learn physiological signal features and perform classification learning is promising. In order to improve the shortcomings of traditional emotion recognition methods, which require a deep understanding of physiological signals and artificial extraction of relevant features, this paper proposed a Recurrent Neural Network (RNN) -based method for automatic feature extraction from EDA's spectrograms. A Convolutional Neural Network (CNN) is used to learn the extracted features further and output the determined emotional state. The results show that the classification accuracy for arousal and valence has reached 83.4% and 81.2%, respectively, which is promising in extracting features automatically and tackling the emotional state classification problem.

A new tripler charge pump with switch bootstrapping technique is presented. The circuit is fully integrated in a standard 0.18-μm CMOS process and operates with a 1.8-V supply. Analysis and measurement results demonstrate a faster startup transient than the conventional design and, at the same time, a lower energy consumption when charging a capacitive load. In static conditions, driving capability and conversion efficiency are improved as well.

This paper presents a MOS active diode circuit that is suitable for energy harvesting applications. The proposed active diode has a minimized reverse leakage and achieves a low forward voltage. The design is implemented and simulated in TSMC 65 nm CMOS technology. Compared to the diode-connected MOS transistor, which has a forward voltage around 300 to 400 mV, the proposed diode demonstrates a near zero forward voltage with minimum supply voltage. The application of the proposed design is demonstrated through a regenerative energy harvester, and the performance is simulated and compared with measurement results from a harvester with regular diodes.

This paper introduces a model for fully-integrated dc-dc converters that takes both resistive and capacitive losses into consideration. The parasitic elements of the inductor, which become crucial when using integrated inductors, are also taken into account. Different tradeoffs are studied and highlighted. Three boost converters are designed in TSMC 65-nm technology using integrated chamfered inductors. The simulation results are in good agreement with the proposed model.

This paper studies design trade-offs of polygonal integrated inductors of different shapes for dc-dc converters. A model that relates the energy conversion efficiency to the inductance time-constant ratio is introduced. Square and octagonal spirals are compared and an optimized chamfered inductor is proposed. Two boost converters are designed and simulated in TSMC 65-nm technology. The results show an energy conversion efficiency improvement of about 5% when using the proposed shape compared to a conventional square inductor, which is in good agreement with the analysis.

This paper presents an analytical model for estimating the resistance of planar polygonal inductors of different shapes. Planar inductors contain a number of corners based on their shapes and number of turns. The resistance of corners is difficult to estimate precisely, since the current density around them is not uniform as in a straight segment and therefore alters the value o f the resistance. A conformal mapping model is used to map a bent metal segment into a rectangular strip. The current density in a bent conductor and hence the value of the resistance are calculated using Schwarz-Christoffel transformations. Our results are compared with ASITIC simulations and are in good agreement.

Electrostatic energy harvesters powered by ambient motion and integrated at the microscale are an attractive replacement for batteries in small, low-power electronic devices. This paper contains model and design considerations for electrostatic generators based on mechanically-variable capacitors. The performance of single-stage and multistage microgenerators are compared. The main design trade-offs and the improvement of conversion efficiency are illustrated through design considerations and simulation results.

A double-ladder pMOS charge pump circuit is proposed in this paper. It requires two phases (and their complements), CMOS standard process, without triple well. With this configuration an output voltage of 10.88 V can be reached with capacitive load, while every device (transistors and capacitors) sustain a maximum voltage not higher than VDD. Taking parasitic capacitances into account, the proposed structure can reach a 93% voltage boosting efficiency and a 52% power efficiency.

In this work we propose a technique for bootstrapping CMOS switches in voltage doublers and double charge pumps. The technique prevents short-circuit losses, improves driving capability, and enables efficient operation at low supply voltages. The effectiveness of our approach is verified through simulations of design examples, which also illustrate the improvements in conversion efficiency, voltage gain, and output resistance.

A monolithic step-up dc-dc converter with on-chip spiral inductors is designed and simulated to determine its feasibility for low-power portable applications. The converter is operated at a relatively high frequency of 600 MHz to reduce passive component sizes. Quality factor limitations of on-chip inductors are mitigated without increasing the area by implementing multiple spirals on different layers and exploiting their mutual inductance. The simulations for a 0.18 μm CMOS process demonstrate the viability of the proposed circuit with a peak efficiency of 74.4% at a load current of 15 mA.

Detecting and monitoring odour around livestock farms are difficult. In this paper, a dynamic neural network based model is proposed to locate odour dispersion around livestock facilities. The proposed dispersion model can dynamically represent complex or non-steady-state meteorological and topographical features in and around livestock farm areas. The proposed approach can also model odour dispersions from multiple odour sources, and from various types of sources such like point source, line source, and area source. In addition, the proposed model simulates the odour dispersion through the dynamic neural activity landscape, without explicitly additional models of the dynamic environment, odour sources, and farming activities.

This work lays out our investigation into the noninvasive, automatic measurement of blood pressure. We have investigated the use of a thermal camera to detect changes in surface temperature due to the pulse in the radial and ulner arteries and then moved on to detecting the same pulse using a low-cost thermistor. Our preliminary results presented here indicate that the pulse is indeed detectable using a simple thermistor aparatus and that it should be possible in the future to develop a calibrated system to measure blood pressure using the pulse transit time method.

In this paper, we present a circuit that protects smart cards against differential power analysis attacks. The circuit is based on a current flattening technique, is designed using a standard 0.18-µm CMOS technology, and can be integrated on the same die or in the same package with the smart card microcontroller. We evaluate the current flattening performance and the effectiveness of the protection against differential power analysis attacks. Our analysis is based on transistor-level simulations in Cadence environment using experimental current traces collected from an 8-bit microcontroller for smart cards executing DES encryptions. The proposed circuit effectively protects against differential power analysis attacks with small chip area overhead and limited increased power consumption during the encryption cycles.

This paper presents the circuit-level implementation for a current flattening system designed to control the power-supply current and to protect cryptosystems from power analysis attacks. When required, the proposed circuit dynamically controls the power consumption by injecting an extra current and by scaling the power-supply voltage. The circuit can be integrated on the same chip with a cryptographic processor. In this way the power-supply current has little meaningful information for a side-channel attack. The proposed circuit has been designed in a 0.18 mum CMOS technology and operates with a nominal 1.8-V power supply

This paper describes a power conversion and data recovery system for a microwave powered sensor platform. A patch microwave antenna, a matching filter and a rectifier make the system frontend and implement the RF-to-DC conversion of power carrier. The efficiency of the power conversion is as high as 47% with an input power level 250 µW at 2.45 GHz. Then, a 0.18 µm CMOS integrated circuit extracts the clock and the digital data. A modified pulse amplitude modulation scheme is used to modulate the data on the 2.45 GHz carrier frequency for combined data and power transmission; this scheme allows very low power consumption of the entire IC to be less than 10 µW and making the system suitable for an autonomous wireless connected sensor module.

In new-generation flash memories, issues such as disturbs and data retention become more and more critical as a consequence of reduced cell size and decreased oxide thickness. Furthermore, the progressive increase in the cell count within a single die tends to decrease device reliability. In particular, reliability issues turn out to be more critical in multilevel (ML) flash memories, due to the reduced spacing between adjacent programmed levels. It is therefore deemed that the use of on-chip error correction codes (ECCs) will gain widespread acceptance in large-capacity flash memories. ECCs for flash memories must have very fast and compact encoding/decoding circuitry so as to have a minimum impact on memory access time. The area penalty due to check cells must also be minimized. Moreover, specific codes must be developed for ML storage. This paper presents error control coding techniques and schemes for new-generation flash memories, focusing on ML devices. The basic concepts of error control coding are reviewed, and the on-chip ECC design procedure is analyzed. Dedicated codes such as polyvalent ECCs, able to correct data stored in ML memories working at a variable number of bits per cell, and bit-layer organized ECCs are described.