Special Sessions

 

Machine Learning Applications in Communications and Electronics:

        Prof. Sotirios K. Goudos, Aristotle University of Thessaloniki (AUTH), Greece.

        Prof. Marco Salucci, University of Trento, Italy

        Prof. Panagiotis Sarigiannidis, University of Western Macedonia, Greece

        Prof. Shaohua Wan, Zhongnan University of Economics and Law, China

Power Management and Monitoring of Photovoltaic Cells:

        Prof. Stylianos Siskos, Aristotle University of Thessaloniki (AUTH), Greece.

        Prof. Eftichis Koutroulis, Technical University of Crete, Greece, Greece.

Smart Systems:

        Prof. Kostas Siozios, Aristotle University of Thessaloniki (AUTH), Greece.

        Prof. Yanqiu Huang, University of Twente, Netherlands.

Fifty Years of Memristors: History, State-of-the-Art, Future Opportunities and Challenges Ahead:

        Prof. Ronald Tetzlaff, Dresden University of Technology (TU Dresden), Germany.

        Dr.-Ing. habil. Alon Ascoli, Dresden University of Technology (TU Dresden), Germany.

 

Machine Learning Applications in Communications and Electronics

 

Machine learning (ML), artificial intelligence (AI) and its learning, adaption paradigms are providing an effective solution in engineering applications. ML can adapt to new conditions and to detect and estimate patterns. Machine learning (ML) has gained recent and well-deserved attention in many fields of engineering and science mostly due to the development of high-performance graphical-processing units (GPU) as well as development of deep neural networks (DNN) based algorithms. Several leading technology companies are heavily investing in AI/ML and academia is following suit to develop more powerful algorithms that utilize the new hardware. In addition to these new developments, practitioners have found new ways to utilize the many existing machine learning algorithms in their respective domains. The fields of wireless communications, electromagnetics (EM), antennas and electronics also benefit in a variety of ways from application of machine learning, deep learning and artificial intelligence. Several applications of ML to communications and electronics already exist. These, among others, evolutionary algorithms (EAs), Decision Trees, Random Forests, Support Vector Machines, Nearest Neighbors, Extreme Learning Machines, Gaussian Processes, Artificial Neural Networks (ANNs), Ensemble learning methods, and Deep Learning Networks (DNNs). The use of all of the above has an increasing impact to key enabling technologies for wireless communications, antenna design, propagation modeling and electronics. Additionally, hybrid combinations of ML techniques and other methods are also emerging. The aim of this special session is to use the ML computing paradigm to bring more awareness on applicability to the communications and electronics domain.

Potential topics include but are not limited to the following:

  • Machine learning techniques for wireless communications
  • Machine learning techniques for propagation modeling
  • Machine learning techniques for antenna design
  • Machine learning techniques for other EM problems
  • Machine learning techniques for 5G Networks and beyond
  • Machine learning techniques for VLSI design
  • Machine Learning techniques for signal processing
  • Machine Learning techniques for leakage detection problems
  • Machine Learning techniques for wired and wireless network
  • ML techniques for biomedical applications and wireless monitoring
  • Surrogate models for antenna design problems
  • Other innovative ML techniques

Special session organizers:

  • Prof. Sotirios K. Goudos, Aristotle University of Thessaloniki, Greece
  • Prof. Marco Salucci, University of Trento, Italy
  • Prof. Panagiotis Sarigiannidis, University of Western Macedonia, Greece
  • Prof. Shaohua Wan, Zhongnan University of Economics and Law, China

 

Power management and monitoring of photovoltaic cells

 

Solar energy is one of the most common energy sources between the alternative renewable energy production alternatives. Photovoltaic (PV) systems are based on the use of solar cells (PV cells) to harvest energy. Since the sunlight intensity varies significantly during a day and depends on the location of the PV system, the presence of clouds on the sky etc. the energy production of a PV cell varies significantly during operation. Power management and conditioning of the converted energy is a critical task to increase the efficiency of PV systems. Additionally, data-acquisition systems enable to monitor the operation of PV cells in order to evaluate their energy-production performance and detect possible malfunctions. With this special session, we invite researchers and engineers to present and share their latest research results on energy harvesting from solar cells, relevant to power management and monitoring of photovoltaic systems.

Potential topics include but are not limited to the following:

  • Power management of solar cells
  • High efficiency DC/DC and DC/AC converters for photovoltaic applications
  • MPPT techniques
  • Sensors, data-acquisition and control for photovoltaic systems

Special session organizers:

  • Stylianos Siskos, Aristotle University of Thessaloniki, Greece
  • Eftichis Koutroulis, Technical University of Crete, Greece

 

Smart Systems

 

Recently, the convergence of emerging embedded computing, information technology, and distributed control became a key enabler for the future technologies. Among others, a new generation of systems with integrated computational and physical capabilities that can interact with humans through many new modalities have been introduced. Furthermore, it is expected that computing and communication capabilities will soon be embedded in all types of objects and structures in the physical environment. Applications with enormous societal impact and economic benefit will be created by harnessing these capabilities across both space and time domains. Such systems that bridge the cyber world of computing and communications with the physical world are referred to as Cyber-Physical Systems (CPSs) and Internet-of-Things (IoT). Both rely on a collection of task-oriented or dedicated systems that pool their resources and capabilities together to create a new, more complex system, which offer more functionality and performance than simply the sum of the constituent sub-systems. Among others, these new design paradigms interact with, and expand the capabilities of, the physical world through monitoring, computation, communication, coordination, and decision-making mechanisms. Such an emerging multidisciplinary frontier will enable revolutionary changes in the way humans live, while it is also expected to be a key enabler for future technology developments.

Special session organizers:

  • Prof. Kostas Siozios, Aristotle University of Thessaloniki (AUTH), Greece
  • Prof. Yanqiu Huang, University of Twente, Netherlands.

 

Fifty Years of Memristors: History, State-of-the-Art, Future Opportunities and Challenges Ahead

 

As, over the recent past, the aggressive CMOS transistor size reduction rate slows down, due to the exorbitant manufacturing costs associated to the production of atom- sized devices, which lead foundries to question whether it is still profitable to follow Moore’s directions, the scientific community has intensified research on novel nanotechnologies which may allow to foster integrated circuit design progress in alternative forms. In this regard memristors stand out as one of the most promising nanotechnologies to complement or extend the functionalities of CMOS circuits beyond the Moore era. Out of an extraordinary line of reasoning, based on robust theoretical foundations, Leon Chua postulated the existence of memristors fifty years ago. Since then, and especially after Stanley Williams’ recognition of memristive fingerprints in the resistance switching transitions of thin oxide films under appropriate stimuli and initial conditions, the interest of the scientific community in the theory and applications of memristors grew exponentially over the years. With their extraordinary versatility and large variety of functionalities, depending upon constitutive materials, fabrication process, and operating modes, which enables the circuit implementation of novel signal processing paradigms, including in-memory and spike-based computing, memristors represent undoubtedly one of the most promising nanotechnologies for enabling the transition from the Moore era, where increasing the number of transistors on the available chip area was of primary importance to improve the performance of integrated circuits, to a new phase, where the time- and energy-efficiency of computing machines is rather enhanced by managing information in disruptive forms, taking inspiration, particularly, from the operating principles of the neural networks in the human brain. This special session will look back at the history of memristors, discuss the lessons learnt from the past fifty years of research in this field, and highlight future opportunities and challenges ahead. The special session represents an important scientific forum for sharing knowledge on the latest developments in the field of memristors, and is open to the whole spectrum of scientific fields, which will contribute to the establishment of these nano- technologies in circuit and system design in the years to come, including material engineering, device physics, nonlinear circuit and system theory, analogue and digital circuit design, computer architecture, without overlooking their potential for modelling biological systems, which may pave the way toward the hardware realization of bio- plausible neuromorphic networks in the future.

Special session organizers:

  • Prof. Ronald Tetzlaff, Dresden University of Technology (TU Dresden), Germany.
  • Dr.-Ing. habil. Alon Ascoli, Dresden University of Technology (TU Dresden), Germany.