A non-linear Liènard-type system based on trigono-metric functions is designed and used as a highly sensitive signal detector. Based on the intermittence mechanism presented in some chaotic systems, an adaptive detection array formed by five chaotic oscillators is designed, tuned, and tested. In the array, each individual oscillator identifies changes in frequency for non-stationary weak signals and, with the array working in an adaptative mode, gives an estimate of the timing at which intermittence occurred allowing the representation of a time-frequency spectrogram even when high-level noise is present. This work also shows that the trigonometric based system has an asymptotically stable limit cycle centered at the origin through two methods: first under the fixed-point analysis and later by the Melnikov method that uses a family of periodic orbits. Once the self-sustained oscillations are produced, the system is configured to drive itself into a saddle-node bifurcation through an iterative method for searching and tuning parameters which, in turn, assures the route to chaotic intermittence. Through this mechanism, the new system can be used as a non-stationary signal detector that also provides a high resolution time-frequency representation. The computer experiments showed that the adaptive array is perfectly compatible with other second-order non-linear oscillators configured in the chaotic intermittence regime such as Duffing, Van der Pol, and Van der Pol - Duffing, achieving high detection rates under noisy conditions. Moreover, when the new Liènard-type oscillators are incorporated into the adaptive array, the detection sensitivity overcomes the SNR threshold present in all oscillators mentioned above and allows for reaching and maintaining a relative error around 1% under severe noise conditions down to about -60 dB SNR.

A document by Narrendar RaviChandran . Click on the document to view its contents.

In recent decades, with the placement of LiDAR remote sensing instruments in orbit, we now have global coverage of the bare-ground elevation on the Earth, Mars and beyond. Encoded in such planetary LiDAR data are interesting landscape features that promise to further our knowledge of planetary topography. However, discovery of such features entails 3 major challenges: 1) massive data; 2) the need for local multi-scale features; 3) sensitivity to interfering factors. To address these challenges, we propose FARMYARD, a generic pipeline for \underline{F}e\underline{a}ture Discove\underline{r}y Fro\underline{m} Planetar\underline{y} LiD\underline{AR} \underline{D}ata Data. To our knowledge, this is the first time such a pipeline has been proposed, which provides a brand new methodology for comparative studies of planetary topography. Specifically, drawing on the parallel computing power of the Graphics Processing Unit (GPU), we propose a novel pseudo-on-pass sweep (POPS) framework for fast and memory-efficient feature extraction for massive planetary LiDAR data, a two-level division scheme for local regions with support for multi-scale features, and a Domain-Shifted Partition (DSP) scheme for feature evaluation that is robust against interfering factors. To showcase the utility of our FARMYARD pipeline, we deploy it to a real-world research project, which seeks to find topographical signatures of life by discovering features that can potentially distinguish between the Earth and alien worlds with no known life activity. We also highlight the efficiency of our POPS framework with experiments on both synthetic and real data, which can be thousands of times faster than its CPU-based counterpart, including a multi-core parallel solution.

The great variety of human emotional expression as well as the differences in the ways they perceive and annotate them make Speech Emotion Recognition (SER) an ambiguous and challenging task. With the development of deep learning, long-term progress has been made in supervised SER systems. However, the existing convolutional neural networks present certain limitations, such as their inability to well capture global features, which contain important emotional information. In addition, due to the subjective nature and continuity of emotion, the instance segments in which emotional speech is typically segmented do not fully reflect the true labels and cannot describe dynamic temporal changes. Thus, accurate emotional representation cannot be learnt in the process of feature extraction. In order to overtake these limitations, we propose an end-to-end network only for speech that maps sequences of different lengths to a fixed number of chunks and strictly preserves the order of chunks by adaptively adjusting their overlap. Subsequently, it extracts log-mel spectrogram features from chunk-level segments and feeds them into the Residual Multi-Scale Convolutional Neutral Networks with Transformer(RMSCTx) model framework. Finally, by keeping the order of the chunk-level segments, a temporal domain mean layer is used to further extract utterance-level feature representations. With this method, we perform multidimensional SER, i. e., the prediction of arousal, valence, and dominance. The experimental results on three popular corpora demonstrate not only the superiority of our approach, but also the robustness of the model for SER, showing an improvement of the recognition accuracy in the newest version of the public dataset MSP-Podcast (1.9).

Two-dimensional (2D) analytical models are only approximations for 3D structures where one cross-sectional dimension is much larger than the other. This paper uses the finite-difference time-domain method (FDTD) to perform numerical experiments on fully 3D dielectric waveguide structures to compute the wave-impedance and the propagation-constant for finite-width dielectric waveguides. These data are used to determine the width required to achieve good correlation against 2D analytical models. Results show that width ≥ 10x height is the limit for good approximation.

One of the fundamental challenges in 5G and beyond technologies is to support short packet transmissions while ensuring ultra-reliable and low latency  communication. Due to the distributed nature of the network, interference is almost unavoidable in most of the emerging communication scenarios. In addition to this, data arrival at the users is bursty in nature. The performance of various interference mitigation techniques for short packet communication under bursty traffic is not well understood from the existing literature. To address these issues this work considers the problem of short packet communication over two-user Rayleigh fading Z-interference channel (Z-IC) under random arrival of data at the users. The work characterizes the stability region of the 2-user Z-IC which takes account of random arrival of data at the users. The evaluation of the stability region involves determination of probability of successful decoding for the various interference mitigation techniques such as joint decoding and successive interference cancellation. The probabilities of successful decoding are characterized for various interference mitigation techniques using the framework of finite block-length information theory. The developed results take into account the packet length, rate, underlying channel model and random arrival of data at the users. The developed results also help to explore the impact of interference on average delay and average age of information for various interference mitigation techniques.

We apply machine learning (ML) techniques to identify the modes in rectangular waveguides from images of 2D modal field patterns injected with uniform, exponential, correlated exponential, and Gaussian noise distributions. A binary classifier is used to identify either transverse electric (TE) or transverse magnetic (TM) modes, and a Multi-class classifier is used to identify the mode numbers. Signal to noise ratios of 1, 0.1, and 0.01 are used to show the effectiveness of each model. Results show accuracy scores up to 99.95%. Several examples demonstrate that noisy modal patterns (unidentifiable to human eyes) may be successfully classified by the ML model.

Domain shift refers to change of distributional characteristics between the training (source) and the testing (target) datasets of a learning task, leading to performance drop. For tasks involving medical images, domain shift may be caused because of several factors such as change in underlying imaging modalities, measuring devices and staining mechanisms. Recent approaches address this issue via generative models based on the principles of adversarial learning albeit they suffer from issues such as difficulty in training and lack of diversity. Motivated by the aforementioned observations,  we adapt an alternative class of deep generative models called the Energy Based Models (EBMs) for the task of unpaired image-to-image translation of medical images. Specifically, we propose a novel method called the Boundary Preserving Twin EBMs (BPT-EBM) which employs a pair of EBMs in the latent space of an Auto-Encoder trained on the source data. While one of the EBMs translates the source to the target domain the other does vice-versa along with a novel boundary preserving loss, ensuring translation symmetry and coupling between the domains. We theoretically analyze the proposed method and show that our design leads to better translation between the domains with reduced langevin mixing steps. We demonstrate the efficacy of our method through detailed quantitative and qualitative experiments on image segmentation tasks on three different datasets vis-a-vis state-of-the-art methods.

Recently, the decreasing cost of storage technologies and the emergence of economy-driven mechanisms for energy exchange are contributing to the spread of energy communities. In this context, this paper aims at defining innovative transactive control frameworks for energy communities equipped with independent service-oriented energy storage systems. The addressed control problem consists in optimally scheduling the energy activities of a group of prosumers, characterized by their own demand and renewable generation, and a group of energy storage service providers, able to store the prosumers’ energy surplus and, subsequently, release it upon a fee payment. We propose two novel resolution algorithms based on a game theoretical control formulation, a coordinated and an uncoordinated one, which can be alternatively used depending on the underlying communication architecture of the grid. The two proposed approaches are validated through numerical simulations on realistic scenarios. Results show that the use of a particular framework does not alter fairness, at least at the community level, i.e., no participant in the groups of prosumers or providers can strongly benefit from changing its strategy while compromising othersâ\euro™ welfare. Lastly, the approaches are compared with a centralized control method showing better computational results. This preprint has been accepted for publication in Control Engineering Practice, Elsevier. How to cite: Nicola Mignoni, Paolo Scarabaggio, Raffaele Carli, Mariagrazia Dotoli, “Control frameworks for transactive energy storage services in energy communities”, Control Engineering Practice, Volume 130, 2023, 105364, ISSN 0967-0661, https://doi.org/10.1016/j.conengprac.2022.105364. © 2022. This manuscript version is made available under the CC-BY-NC-ND 4.0 license.

According to the 6G vision, the evolution of wireless communication systems will soon lead to the possibility of supporting Tbps communications, as well as satisfying, individually or jointly, a plethora of other very stringent quality requirements related to latency, bitrate, and reliability. The achievement of these goals will naturally raise many research issues within radio communications. In this context, a promising 6G wireless communications enabler is the reconfigurable intelligent surface (RIS) hardware architecture, which has already been recognized as a game-changing way to turn any naturally passive wireless communication setting into an active one. This paper investigates RIS-aided multicast 6G communications by first modeling the system delay as a first-come-first-served (FCFS) M/D/1 queue and analyzing the behavior under different blockage conditions. Then the study of multi-beam operation scenarios, covering multicast and RIS-aided multicast communications, is conducted by leveraging an M/D/c queue model. Achieved results show that large-size RISs outperform even slightly obstructed direct BS-to-user paths. In contrast, RISs of smaller sizes require the design of sophisticated power control and sharing mechanisms to achieve better performance.

The most widely-used active upper-limb (UL) prostheses worldwide are body-powered (BP), which is a two-century-old technology. Despite their affordability and functional benefits to users, these devices are prone to poor outcomes for many patients. Additionally, BP devices have witnessed limited improvements compared to their externally-powered counterparts. Literature indicates a strong need for appropriate prosthetic solutions for children and adolescents. Our previous work introduced a first-of-its-kind breathing-powered UL prosthesis (“Airbender”) that can overcome several limitations of the current BP systems (e.g., restricted operation space, user discomfort caused by the harness to which the cables are attached). Users can regulate their breathing, and this controllable airflow is subsequently used to power a small (purpose-built and optimised) Tesla turbine that can accurately control the opening and closing of the prosthetic hand. The current work explores device usability in children and adolescents with a UL difference (n = 15). Further, we gathered feedback, suggestions, and satisfaction levels from the study participants and their parents on breathing as a modality of controlling a prosthetic device. The collected responses and study observations were subjected to qualitative and statistical analysis. This study showcases real-world testing of a breathing-powered prosthesis and proves that UL-deficient children and adolescents can indeed operate the device (i.e., volitionally open or close) with their breathing input. The perceived level of difficulty in opening or closing the device tended to be on the ‘easy’ side. We report generally favourable feedback obtained from participants and their parents. Additionally, design suggestions and satisfaction levels concerning different device attributes help us involve the key stakeholders in co-creating and proactively developing a robust product development roadmap. This work is aligned with creating a step-change in the potential BP prosthesis options for patients in the future, propelled by a need-led approach.

Active cyber defense is the most recent model that cybersecurity has adopted from its traditional military counterpart. It is considered the most modern approach to defensive design and operations that today's cybersecurity recognizes, and as such still has much to mature. This paper presents the state of the art of active cyber defense in its theoretical and practical aspects and proposes a realistic technical approach for the provision of related professional services, based on other modular services, and considering that existing proposals in the industry are still disparate and uncommon worldwide.

The Radio Access Network (RAN) architecture evolves with different generations of mobile communication technologies and forms an indispensable component of the mobile network architecture. The main component of the RAN infrastructure is the base station, which includes a Radio Frequency unit and a baseband unit. The RAN is a collection of base stations connected to the core network to provide coverage through one or more radio access technologies. The advancement towards cloudnative networks has led to centralizing the baseband processing of radio signals. There is a trade-off between the advantages of RAN centralization (energy efficiency, power cost reduction, and the cost of the fronthaul) and the complexity of carrying traffic between the data processing unit and distributed antennas. 5G networks hold high potential for adopting the centralized architecture to reduce maintenance costs while reducing deployment costs and improving resilience, reliability, and coordination. Incorporating the concept of virtualization and centralized RAN architecture enables to meet the overall requirements for both the customer and Mobile Network Operator. Functional splitting is one of the key enablers for 5G networks. It supports Centralized RAN, virtualized Radio Access Network, and the recent Open Radio Access Networks. This survey provides a comprehensive tutorial on the paradigms of the RAN architecture evolution, its key features, and implementation challenges. It provides a thorough review of the 3rd Generation Partnership Project functional splitting complemented by associated challenges and potential solutions. The survey also presents an overview of the fronthaul and its requirements and possible solutions for implementation, algorithms, and required tools whilst providinga vision of the evaluation beyond 5G second phase.

Manuscript submitted to the IEEE Transations in Power Systems in October 2022. Title: Coordination of Frequency Reserves in an Isolated Industrial Grid Equipped with Energy Storage and Dominated by Constant Power Loads Abstract: This paper examines the use of interconnected synchronous system requirements for frequency containment reserves (FCR) on isolated industrial grids that use turbogenerators as main source of energy, have high penetration of wind energy, are equipped with energy storage, and have a high level of constant power loads coupled by power electronic converters. Leveraging on the recent Nordic requirements for reserves in islanded operation (FCRI), we propose an expansion that allows prioritizing among various reserve providers under different isolated grid conditions. The study case of a complex, isolated industrial grid is selected to test this approach. The stability of this grid is evaluated via eigenvalues and participation factors considering the detrimental effects of constant power loads. It is demonstrated that, by prioritizing the reserve allocation to the faster converter-interfaced storage devices and loads, the overall stability is increased in addition to allowing the turbogenerators to operate at a more constant load. The results are supported by computer simulations of a rotating mass model, of the complex isolated grid in PowerFactory, and by laboratory power-hardware-in-the-loop tests. The computer simulation models developed for this paper are made publicly available for reproducibility purposes.

The  energy transition towards  increased  electric power production from renewable energy (RE) resources creates new challenges to ensure the  stability  of power grids. In conventional power grids voltage  fluctuations  can be controlled locally. Here, we explore whether the energy transition changes this situation. We  study systematically the  transients  of voltage amplitude,  phase and frequency deviations  due to  local contingencies in dependence on system inertia, heterogeneity and topology. The 3rd order dynamic power grid model is studied numerically and analytically and compared with  real grid simulations for the Nigerian (330 kV) power grid and other grid models, using DigSILENT PowerFactory software.  We provide a quantitative analysis of  the parametric dependence of the  velocity with which a disturbance propagates throughout the grid, and of the period of oscillations of the frequency and voltage transients. We find  beating patterns in the transients which we identify as footprints of the location of the fault bus. We confirm  that  voltage deviations remain local for  realistic ranges of parameters. However, we  find that this no longer holds true when the electrical power in the grid approaches its critical value. We furthermore consider time dependent second moments of  geodesic distance, weighted with frequency deviations and voltage deviations, respectively. We  confirm thereby  ballistic disturbance propagation  in homogeneous model grids. However, in real grid simulations, we observe  a linear time dependence of deviations indicating diffusive  propagation due to multiple scattering from the inhomogeneities in these power grids.

In this paper, we propose a robot end-effector for fabric folding of a fabric part of a garment along a straight line. In the garment production process, some of the edges of fabric parts of a garment need to be folded before sewing. Conventional automated folding system has a fixture designed for each folding fabric part and operation. The fixture is not universal and needs to be redesigned when the shape of the folded fabric part changes. In this paper, we consider performing fabric folding without using a fixture. We propose a concept of a robot end-effector for fabric folding of a fabric part along a straight fold line and develop a prototype of an end-effector referred to as“ F-FOLD ”(Free-form FOLDing). Folding of the edge of a fabric part is achieved by moving FFOLD along the desired straight fold line. Experimental results illustrate how F-FOLD works to fold a fabric part along a straight line. *The complete version of this preprint paper is published in  IEEE/ASME (AIM) International Conference on Advanced Intelligent Mechatronics . Please refer to “https://ieeexplore.ieee.org/document/10196184”.

Federated Learning (FL) is a new paradigm aiming to solve the data access problem. It is gaining an increasing interest in a variety of research fields, including the Biomedical and Financial environments, where lots of valuable data sources are available but not often directly accessible due to the regulations that protect sensitive information. FL provides a solution by moving the focus from sharing data to sharing models. The FL paradigm involves different entities (institutions) holding proprietary datasets, contributing with each other to train a global Artificial Intelligence (AI) model using their own locally available data. Although several studies propose ways to distribute the computation or aggregate results, fewer efforts have been made on how to implement it. With the ambition of helping accelerate the exploitation of FL frameworks, this paper proposes a survey of public tools that are currently available for building FL pipelines, an objective ranking based on the current state of user preferences, and the assessment of the growing trend of the tool’s popularity over a six months time window. Finally, a ranking of the maturity of the tools is derived based on keyaspects to consider when building an FL pipeline.

This paper proposes CNN-based visual servoing for simultaneous positioning and flattening of a soft fabric part placed on a table by a dual manipulator system. We propose a network for multimodal data processing of grayscale images captured by a camera and force/torque applied to force sensors. The training dataset is collected by moving the real manipulators, which enables the network to map the captured images and force/torque to the manipulatorâ\euro™s motion in Cartesian space. We apply structured lighting to emphasize the features of the surface of the fabric part since the surface shape of the non-textured fabric part is difficult to recognize by a single grayscale image. Through experiments, we show that the fabric part with unseen wrinkles can be positioned and flattened by the proposed visual servoing scheme.  *The complete version of this preprint paper is in proceedings of ICRA 2023. Please refer to “https://ieeexplore.ieee.org/document/10160635”.Â