Program

Summer school on Complex Networks and Telecommunications July 5-9-2021

Ultimofavalli

Student

Abstracts of the lectures

  • George Alexandropoulos: Cell-Free Communications and Reconfigurable Intelligent Surfaces: Two Candidate 6G Technologies and Their Joint Potential.

This talk will cover the basic principles and opportunities of the beyond 5G technologies of cell-free communications and Reconfigurable Intelligent Surfaces (RISs) in two separate parts, which will be followed by a last part elaborating on their distinctive challenges together with their joint potential for  6G smart and energy efficient wireless connectivity. In the first part, emphasis will be given in cell-free massive MIMO communication systems presenting optimization formulations for the uplink and downlink operations, as well as the channel estimation problem, and identifying similarities and differences with technologies for the interference channel and distributed antenna systems. In the second part, the available hardware architectures for RISs will be introduced together with representative models for the effective response on the RIS elements. In addition, optimization formulations for representative RIS-empowered wireless communication systems will be discussed together with approaches for the major problem of channel estimation in such systems. Finally, the last part of the talk will highlight the open challenges with both technologies and present their fascinating potential synergies.

  • Sergio Barbarossa: New tools for 6G networks: From topological signal processing to goal-oriented communications.

The challenges foreseen for 6G networks call for a radical change of perspective, where the communication network needs to be designed not only to ensure the efficient and reliable transmission of symbols from source to destination, but rather on the efficient exchange of information among agents involved in taking accurate decisions about complex environments, possibly under tight time constraints and using limited resources. In this context, artificial intelligence is expected to play a key role, in terms of knowledge representation systems and machine learning. The goal of this talk is to introduce topological tools to extract relevant information from data, generalizing the findings obtained in the emerging field of graph signal processing to higher order structures. We will then show how to generalize convolutional neural networks to process input signals defined over a non-Euclidean domain, whose geometry needs to be inferred from the data themselves. Finally, we will consider an efficient way to design communication systems using a goal-oriented communication paradigm, combining the information bottleneck principle with Lyapunov optimization to derive efficient dynamic strategies operating in an environment with multiple sources of uncertainties.

  • Jean Claude Belfiore: Topos for semantic communications.

This lecture is one of the first attempts to answer the question “How can intelligent machines efficiently communicate?” which is one of the main goals of the so-called “semantic Communication”.
I will present a joint work with Daniel Bennequin which shows our progresses towards a mathematical theory of semantic communication, inspired by the foundational works of Claude Shannon and Alexander Grothendieck. To communicate efficiently we need a language. Using category theory, we can define a category transporting the semantics of a language. We will see that the notion of semantics depends on many aspects that can be found in machine learning: Sampling (the data), structures (a kind of presemantic that will be carefully defined), ..
Some important mathematical notions as Grothendieck Toposes and Stacks will be introduced through simple examples and we will see how neural networks can be modelled this way). Finally, after showing how a language is transported through the layers of a neural network, we will give a definition of semantic information measures which are not scalar quantities as in Shannon information theory, but spaces. Some examples will show the validity of such a definition. A semantic source coding theorem will finally be given.

  • Caire Giuseppe: Advances in multiantenna communications.

Multiuser MIMO has become one of the corner stones of modern wireless communications, and in particular plays a key role in 5G and in Wifi6 (802.11ax).In this talk, I will review the basic principles of MU-MIMO, and recent advances in TDD and FDD systems with baseband (digital) precoding, suited for systems operating below 10MHz. Then, I will discuss the specific problems for systems operating at much higher carrier frequencies and large bandwidths (mmWaves), exploiting hybrid baseband and RF processing.  Finally, I will consider multi-cell systems and cell-free user-centric systems based on distributed remote radio heads (RRHs) and dynamic cluster formation, and some challenges arising in these architectures.

  • Emilio Calvanese Strinati: From 5G to 6G.

This talk promotes the idea that including semantic and goal-oriented aspects in future 6G networks can produce a significant leap forward in terms of system effectiveness and sustainability. Semantic communication goes beyond the common Shannon paradigm of guaranteeing the correct reception of each single transmitted packet, irrespective of the meaning conveyed by the packet. The idea is that, whenever communication occurs to convey meaning or to accomplish a goal, what really matters is the impact that the correct reception/interpretation of a packet is going to have on the goal accomplishment. Focusing on semantic and goal-oriented aspects, and possibly combining them, helps to identify the relevant information, i.e. the information strictly necessary to recover the meaning intended by the transmitter or to accomplish a goal. Combining knowledge representation and reasoning tools with machine learning algorithms paves the way to build semantic learning strategies enabling current machine learning algorithms to achieve better interpretation capabilities and contrast adversarial attacks. 6G semantic networks can bring semantic learning mechanisms at the edge of the network and, at the same time, semantic learning can help 6G networks to improve their efficiency and sustainability.

  • Emilio Calvanese Strinati: 6G Networks: Beyond Shannon Towards Semantic and Goal-Oriented
    Communications.

This talk promotes the idea that including semantic and goal-oriented aspects in future 6G networks can produce a significant leap forward in terms of system effectiveness and sustainability. Semantic communication goes beyond the common Shannon paradigm of guaranteeing the correct reception of each single transmitted packet, irrespective of the meaning conveyed by the packet. The idea is that, whenever communication occurs to convey meaning or to accomplish a goal, what really matters is the impact that the correct reception/interpretation of a packet is going to have on the goal accomplishment. Focusing on semantic and goal-oriented aspects, and possibly combining them, helps to identify the relevant information, i.e. the information strictly necessary to recover the meaning intended by the transmitter or to accomplish a goal. Combining knowledge representation and reasoning tools with machine learning algorithms paves the way to build semantic learning strategies enabling current machine learning algorithms to achieve better interpretation capabilities and contrast adversarial attacks. 6G semantic networks can bring semantic learning mechanisms at the edge of the network and, at the same time, semantic learning can help 6G networks to improve their efficiency and sustainability.

  • Egidio D’Angelo: From the human brain to artificial intelligence

The brain is thought to operate as a complex adaptive system generating an internal representation of the world that is continuously updated through a prediction/action/learning process. This process, coordinated with the management of brain states, is the basis of all brain functions and of intelligence. Current advances in neurophysiology and brain modelling are now allowing a precise representation and simulation of brain functions on multiple scales. Brain modelling is shedding light on the brain functioning principles and is promoting their application to biomedicine and artificial intelligence. In this talk I will address the principles of brain organization and function and the strategies for brain modelling and simulation. I will show solutions to the multiscale problem and I will show applications to virtual brains, neuromorphic computers and closed-loop controllers.

  • Merouane Debbah: Electomagnetic Information Theory: Past present and future.

In this talk, we will discuss this new avenue of research, with the willingness to unify wireless communication theory and electromagnetic theory, something that has never been achieved since the landmark work of Gabor. In particular, it is known that major advances in Wireless Communications are made by re-questioning the model assumptions and we will discuss how the incorporation of neglected physics has a chance of yielding breakthroughs for next generation wireless communication systems.

  • Ioannis Krikidis: Wireless powered communications in the era of 6G.

Conventional energy-constrained wireless systems such as sensor networks are powered by batteries and have limited lifetime. Wireless power transfer (WPT) is a promising technology for energy sustainable networks, where terminals can harvest energy from dedicated electromagnetic radiation through appropriate electronic circuits. The integration of WPT technology into communication networks introduces a fundamental co-existence of information and energy flows; radio-frequency signals are used in order to convey information and/or energy. The efficient management of these two flows through sophisticated networking protocols, signal processing/communication techniques and network architectures, gives rise to a new communication paradigm called wireless powered communications (WPC).  In this talk, we discuss the principles of WPC and we highlight its main network architectures as well as the fundamental trade-off between information and energy transfer. Several examples, which deal with the integration of WPC in modern communication systems, are presented.

  • Josep M. Jornet: Teraherz commuincations

Terahertz (THz)-band (0.1–10 THz) communication is envisioned as a key wireless technology of the next generation of wireless systems. The very large available bandwidth (tens to hundreds of consecutive GHz) and the very small wavelength (sub-millimetric) of THz signals opens the door to Terabit-per-second (Tbps) links for macro- and nano-scale applications as well as unprecedented wireless sensing opportunities, simultaneously. Nevertheless, there are several roadblocks that need to be overcome to tap in the THz band. In this course, the building blocks of THz communication systems will be presented. After briefly reviewing the state of the art in THz device technologies, the key properties of the THz channel will be presented, and their impact on the design of physical layer solutions (including modulation, ultra-massive MIMO and physical layer security) and the link layer (including MAC and neighbor discovery) will be discussed.

  • Massimiliano Mancini: Deep Learning beyond the Closed World.

Deep learning models achieved impressive results in many different fields, such as computer vision, natural language processing, and reinforcement learning. However, these successes heavily rely on large-scale annotated datasets for training, and those cannot capture the infinite variability of the real world. Therefore, deep models may break whenever either the input distribution changes or contains unknown semantic content or both. In the first part of this lecture, we will review recent approaches addressing the domain shift problem (i.e. changes in the input distribution), focusing on the tasks of domain adaptation and generalization. In the second part of the lecture, we will discuss how to expand the output space of a pretrained model, focusing on continual learning. We will finally conclude with a perspective on emerging trends addressing both problems simultaneously.

  • Giovanni Neglia: Distributed Machine Learning Training.

In this lecture, we describe the most common approaches to train machine learning models in a cluster or over the Internet (federated learning). We present open research issues to which the networking and distributed systems communities may contribute. We provide specific examples from our team on improving training efficiency and learning personalized models.

  • Vincenzo Sciancalepore: O-RAN the opportunity bridging 5G to 6G.

Beyond-5G and 6G networks aim to provide flexible compute and connect technologies that should fully support innovative use-cases and unprecedented services via a concrete and sustainable transformation of all existing network designs into smart-connected telecommunication infrastructures. Fortunately, the main network business players have recently worked together to bring up a common concept for the 6G era: Openness. This would require new architectural interfaces thereby opening the telecom industry’s door to hardware and software providers according to the new O-RAN standard. In this context, controlling the propagation environment is of paramount importance: Reconfigurable Intelligent Surfaces as man-made passive surfaces may be seamlessly integrated into the new programmable RAN vision to bring flexibility and innovation into the upcoming network generation.

  • Michele Zorzi: Non-Terrestrial Networks in the 6G Era: Challenges, Opportunities, Technologies, and Trends

Abstract: Many organizations recognize non-terrestrial networks (NTNs) as a key component to provide cost-effective and high-capacity connectivity in future 6th generation (6G) wireless networks. Despite this premise, there are still many questions to be answered for proper network design, including those associated with latency and coverage constraints. In this talk, after reviewing research activities on NTNs, we present the characteristics and enabling technologies of NTNs in the 6G landscape (with a focus on architecture, spectrum, and antenna advancements in the air/space design), and shed light on the challenges in the field that are still open for future research. As a case study, we evaluate the potential of multi-layered hierarchical networks, i.e., the orchestration among different aerial/space platforms, including Unmanned Aerial Vehicles (UAVs), High Altitude Platforms (HAPs), and satellites co-operating at different altitudes, and provide guidelines on the optimal working point(s) for which it is possible to achieve a good compromise between improved system flexibility and network performance, with respect to a baseline standalone deployment. We also discuss the feasibility of configuring UAVs and satellites to operate in the millimeter wave (mmWave) bands, and the research challenges associated with this design.