Text
E-book Special Topics in Information Technology
A fundamental pillar for the research work in the Department of Electronics,Information, and Bioengineering (DEIB) is its Ph.D. program in InformationTechnology, with more than 50 doctors graduating each year. The program ischaracterized by a broad approach to information technology, in the areas ofComputer Science and Engineering, Electronics, Systems and Controls, andTelecommunications.The characteristics of the IT Ph.D. doctoral studies in the DEIB department is anemphasis on interdisciplinarity, that is becoming more and more important in therecent technological developments in which the boundaries between disciplines arebecoming increasingly fuzzy in collaborative projects and education of youngresearchers. Therefore, the focus of the program is on providing a broad doctoral-level educational approach and research environment and at the same time, a rig-orous approach to the specific research topics developed by the Ph.D. candidates indifferent aspects of Information Technology, based on both on theoreticalapproaches and on the validation of new ideas in selected application domains.Starting with the present volume, we present the ongoing research work indoctoral studies in the department, through a collection of summaries illustratingthe results of the best Ph.D. theses defended in the academic year 2018–19. In thischapter, the coordinators of each of the areas are going to introduce the four areasof the Ph.D. program in Information Technology, namely Computer Science andEngineering, Electronics, Systems and Controls, and Telecommunications, andprovide a short introduction to the papers selected for the volume. The Ph.D. focused in Electronics explores the enabling frontier scenarios of currentand future era technologies of information, communication, control, automation,energy, and mobility.Design and innovation of devices, circuits and electronic systems continue toprovide the fundamental building blocks necessary for modern life in all its per-spectives, including its most recent declinations“smart-”(smart cyberphysical-systems, smart industries, smart manufacturing, smart living, smart mobility, smartlighting, smart cities, smart aging, etc.) and“autonomous-”(autonomous driving,autonomous vehicles, autonomous-manufacturing, autonomous agents, etc.), sopervasive in modern daily human activities.For instance, the new concept of smart mobilityfinds one of the most significantdeclinations in self and highly assisted driving, where the continuous-wavefrequency-modulated (FMCW) radar is a key element. This short-range measuring adar set capable of determining distance increases reliability by providing distancemeasurement along with speed measurement, which is essential in modern auto-motive applications. Dmytro Cherniak has given a strong contribution to thefield ofintegrated FMCW radar. These innovations support the emerging market forautonomous vehicles, which will rely heavily on radar and wireless sensing toachieve reliable all-weather mobility. High performance and low-power con-sumption meet in thefirst implemented digital PLL-based FMCW modulatorprototype fabricated in 65-nm CMOS technology that demonstrates theabove-state-of-the-art performance of fast chirp synthesis, also developed andfabricated in 28-nm CMOS technology focusing on low-fractional spur operation.From a completely different perspective, the human brain is a marvelousmachine, which is capable of solving challenging problems in a very short time andwith small energy consumption. Recognizing a face, learning to walk, and taking aquick decision under a large number of stimuli are generally straightforwardfunctions for mankind. However, recreating this type of computation in silicon withthe same speed, energy consumption, and hardware resources is practicallyimpossible with the current microelectronic technology. The work by Valerio Miloaddresses these challenges, by showing that a new type of devices and architectureis needed if we want to mimic the brain computation. It is shown that synapticdevices made of inorganic materials, such as metal oxides, show the same type ofplasticity that is observed in biological synapses. By taking advantage of suchbiological plasticity rules within artificial spiking neural networks, it is possible tolearn, make associations, and take decisions, which are some of the most funda-mental cognitive functions of the human brain. Although the work is at the veryearly stages, the results are very promising for supporting the grand challenge ofbio-inspired circuits that can compute as the brain, and even illuminate about thebiological processes of learning in our brain.
Tidak tersedia versi lain