Jan. 29 - Feb. 02, 2024 Champéry, Switzerland

Theory and Practice of Distributed Computing

CUSO Winter School in Computer Science

WATCH LECTURES

Welcome to the CUSO Winter School 2024 on the Theory and Practice of Distributed Computing!

The school takes place in Champéry, Switzerland, from January 29th to February 2nd, 2024. This school is organized under the Doctoral Program in Computer Science of the CUSO universities. It is primarily targeted at PhD students in computer science. Students from CUSO member universities are reimbursed their participation costs, but other PhD students and academics are welcome to participate as well.

sponsors
sponsors
sponsors
sponsors
sponsors

Overview

Distributed computing concerns networks of nodes that communicate and cooperate with each other, jointly solve problems, exchange and maintain data in order to achieve a desired, network-wide outcome. This school will feature models, problems, data structures, systems, algorithms, and techniques that address fundamental aspects and practical considerations in distributed systems. Particular focus will be given to capacity, coordination, and information propagation in losely connected settings, for instance in mobile or ad-hoc networks, to fundamental protocols for achieving a joint goal or maintaining a common data structure, and to faults and attacks that may prevent the nodes from reaching their common goal, including methods to mitigate their effects.

Goal of the Winter School

In this years' CUSO Winter School we welcome experts to present the State-of-the-Art of current research on theoretical and applied aspects of Distributed Computing. The winter school has two main goals. First, the school will offer an introduction and overview of the research area, illustrating the relevance of the topic in scientific and commercial applications. In addition, it will foster interaction with renowned experts in the field, supporting them to become active and independent researchers on their own. We believe this winter school will address a broad audience of doctoral students and postdoctoral researchers at the CUSO universities, which are home to a number of research groups in related areas.

Venue

The winter school will take place in Champéry, an idyllic alpine resort in Canton de Valais. The venue Hotel Suisse is a family-owned hotel in the center of the village and in walking distance from the train station.

Speakers

About Christian Scheideler

Short Bio: Christian Scheideler has a Diploma in CS and PhD from Paderborn University. After spending a postdoc year at the Weizmann Institute and obtaining his Habilitation in Paderborn, he spent five years as assistant professor at Johns Hopkins University and three years as associate professor at Technical University of Munich. Since 2009 he is a full professor at Paderborn University and has served as department chair from 2013-2017 and 2021-2023. His research interests are distributed algorithms and data structures, randomized algorithms and stochastic processes, and robust distributed systems. He is particularly well-known for his work on overlay networks, robust distributed information systems, and programmable matter.

About Laurent Vanbever

Short Bio: Laurent Vanbever is an associate professor at ETH Zürich where he started as an assistant professor in 2015. Before that, Laurent was a postdoctoral research associate at Princeton University where he worked with Jennifer Rexford. He obtained his PhD degree in Computer Science from the University of Louvain in 2012. His research focuses on making large network infrastructures more manageable, scalable and, secure.

About Patrick Thiran

Short Bio: Patrick Thiran holds an electrical engineering degree from the Université Catholique de Louvain, Louvain-la-Neuve, Belgium, an M.Sc. degree in electrical engineering from the University of California at Berkeley, USA, and he received the PhD degree from EPFL, in 1996. He became an adjunct professor in 1998, an assistant professor in 2002, an associate professor in 2006 and a full professor in 2011. He was with Sprint Advanced Technology Labs in Burlingame, California, in 2000-01. His research interests are in data-driven network science and stochastic models, and in machine learning. He also contributed to the performance analysis of wireless networks and to network calculus, as well to locally coupled neural networks.

About Fabian Kuhn

Short Bio: Fabian Kuhn received the M.Sc. and Ph.D. degrees in computer science from ETH Zurich, Switzerland, in 2001 and 2005, respectively. He afterwards spent time as a Postdoctoral Researcher with Microsoft Research Silicon Valley, Mountain View, CA, USA; ETH Zurich; and the Massachusetts Institute of Technology (MIT), Cambridge, MA, USA. He was an Assistant Professor with the University of Lugano, Lugano, Switzerland, and in 2012, he joined the University of Freiburg, Freiburg, Germany, as a Full Professor. He is interested in a wide range of topics in theoretical computer science, especially in the area of distributed algorithms. In particular, his research interests include the complexity of graph-theoretic problems in a distributed context and the algorithmic foundation of mobile and wireless networks.

About Roger Wattenhofer

Short Bio: Roger Wattenhofer is a professor at ETH Zurich, Switzer­land. He also worked multiple years at Microsoft Research in Redmond, Washington, at Brown University in Providence, Rhode Island, and at Macquarie University in Sydney, Australia. His work received multiple awards, e.g. the Prize for Innovation in Distributed Computing for his work in Distributed Approximation. He published the book “Blockchain Science: Distributed Ledger Technology“, which has been translated to Chinese, Korean and Vietnamese.

Organizers

About Christian Cachin

Christian Cachin is a professor of computer science at the University of Bern (Switzerland), where he has been leading the Cryptology and Data Security Research Group since 2019. Over his career he worked for IBM Research - Zurich during more than 20 years and has also held visiting positions at MIT, EPFL, and Stanford University. With a background in cryptography, he is interested in all aspects of security in distributed systems and especially in cryptographic protocols, consistency, consensus, blockchains, and cloud-computing security. He is known for developing cryptographic protocols, particularly for achieving consensus and for executing distributed cryptographic operations over the Internet. In the area of cloud computing, he has contributed to standards in storage security and developed protocols for key management.

About Pascal Felber

Pascal Felber received his M.Sc. and Ph.D. degrees in Computer Science from the Swiss Federal Institute of Technology. From 1998 to 2002, he has worked at Oracle Corporation and Bell-Labs (Lucent Technologies) in the USA. From 2002 to 2004, he has been an Assistant Professor at Institut EURECOM in France. Since October 2004, he is a Professor of Computer Science at the University of Neuchâtel, Switzerland, working in the field of dependable, concurrent, and distributed computing. He has published over 200 research papers in various journals and conferences.

About Antonio Di Maio

Antonio Di Maio received his Ph.D. degree from the University of Luxembourg in 2020 with a thesis on Routing and Content Dissemination in Software-Defined Vehicular Networks. Since 2020, he has been a Postdoctoral Researcher with the Communications and Distributed Systems Group at the University of Bern, Switzerland, having published 30 articles on topics such as Secure and Privacy-friendly Localization Systems, Secure Federated, Split, and Gossip Learning, Reinforcement-Learning-based Trajectory Prediction, Cooperative Localization, Service Deployment and Slice admission on Mobile Edge Networks, Opportunistic Networking for Floating Content and Services, and Performance Modeling and Evaluation of Mobile Ad Hoc Networks.

About Valerio Schiavoni

Dr. Valerio Schiavoni is a Lecturer at the University of Neuchâtel, Switzerland. His research interests lie at the intersection of systems broadly conceived, security, and data management. He is a former coordinator of the CUSO program in Computer Science.

About Philipp Schneider

Philipp Schneider is a Postdoctoral researcher at the University of Bern and part of Christian Chachins CRYPTO group. He obtained his Masters degree in Computer Science at the Karlsruhe Institute for Technology and his PhD at the University of Freiburg in 2016 and 2022, respectively. His research interests revolve around the theoretical aspects of distributed computing with contributions in the research of the complexity of distributed problems in hybrid communication networks.

Program

The school will take place from Monday January 29th to Friday February 2nd at Hotel Suisse in Champéry, Switzerland. The program is divided into five sessions A-E each of which is subdivided into three blocks. In principle on each day one session is covered, except for the last Session E, which is spread over Thursday and Friday. A detailled time schedule and session overview follows below.


Time Schedule

13:00 - 13:15
Introduction
13:15 - 14:45

Session A1

“Introduction to Hybrid Networks”
14:45 - 15:30
Coffee Break
15:30 - 17:00

Session A2

“Introduction to Hybrid Networks”
17:00 - 17:30
Coffee Break
17:30 - 19:00

Session A3

“Introduction to Hybrid Networks”
19:30
Dinner
Restaurant Alta 1874
08:30 - 10:00

Session B1

"Reasoning about Network-level Distributed Algorithms”
10:00 - 10:30
Coffee Break
10:30 - 12:00

Session B2

"Reasoning about Network-level Distributed Algorithms”
12:00 - 16:30
Leisure Time (lunch on your own)
16:30 - 18:00

Session B3

"Reasoning about Network-level Distributed Algorithms”
18:00 - 18:45
Participant presentations
  • David Lehnherr (Uni BE): From Graphs to Semi-Simplicial Sets
  • Diana Ghinea (ETHZ): Byzantine Agreement: a Quick Overview
  • Pasquale De Rosa (Uni NE): Bias Mitigation in Federated Learning for Edge Computing
  • François-Xavier Wicht (Uni BE): The Multiple Faces of Privacy in Cryptocurrencies
  • Simon Queyrut (Uni NE): Federated Model Extraction Attack Against cGANs
19:30
Dinner
Café du Nord
08:30 - 10:00

Session C1

"Percolation Models and Wireless Networks"
10:00 - 10:30
Coffee Break
10:30 - 12:00

Session C2

"Random Graph Models and Diffusion Source Location"
12:00 - 16:30
Leisure Time (lunch on your own)
16:30 - 18:00

Session C3

"A Short Introduction to Network Calculus"
19:00
Dinner
Hotel Suisse
08:30 - 10:00

Session D1

"Local Algorithms and Symmetry Breaking"
10:00 - 10:30
Coffee Break
10:30 - 12:00

Session D2

"Local Algorithms and Symmetry Breaking"
12:00 - 16:30
Leisure Time (lunch on your own)
16:30 - 17:45

Session D3

"Local Algorithms and Symmetry Breaking"
17:45 - 18:00
Coffee Break
18:00 - 19:15

Session E1

"Learning with Graphs"
19:30
Dinner
At'home
08:30 - 10:00

Session E2

"Learning with Graphs"
10:00 - 10:30
Coffee Break
10:30 - 12:00

Session E3

"Learning with Graphs"
12:00 - 12:30
Closing Out



Session Content

“Introduction to Hybrid Networks”

by Christian Scheideler

  Recording

program

In hybrid networks, nodes have access to two different communication modes: a local mode where (like in traditional networks) communication is only possible between specific pairs of nodes, and a global mode where (like in overlay networks) communication between any pair of nodes is possible. Prominent examples of hybrid networks are datacenter networks that combine high-speed optical communication (the global edges) with a traditional cable network (the local edges) or wireless networks that combine communication via a cellular infrastructure (the global edges) with ad-hoc communication links (the local edges). I will show via some examples like minimum spanning tree and shortest path problems how to best make use of the two communication modes in order to solve the given problems as quickly as possible. Based on that knowledge, the students will have the opportunity to discuss solutions for related problems.

"Reasoning about Network-level Distributed Algorithms”

by Laurent Vanbever

  Recording

program

Managing large-scale networks, like the ones powering Internet Service Providers, Cloud Providers, or data-centers, is very complex, and therefore easy to get it wrong. Recently network verification has emerged as an answer to this complexity by allowing to formally verify network-wide properties such as connectivity. Going one step further, network synthesis aims at automatically designing provably-correct networks from a specification. In my session, I’ll give a overview of network verification and synthesis, the common techniques they rely on, alongside with some of the open research questions that are currently being considered by the community.

"Percolation models and wireless networks"

by Patrick Thiran

  Recording

program

Percolation theory studies phase transitions in random (spatial) graphs. We will review some key results in percolation theory, and show how they are instrumental in deriving tight scaling laws on properties such as connectivity, transport capacity and latency of wireless networks.

"Random graphs models and diffusion source location"

by Patrick Thiran

We survey some recent results on the location of the source of a diffusion (or epidemics) in a network, when the states of only a few sensor nodes in a network can be observed. If propagation delays along the edges are (close to) deterministic values, the smallest number of sensors needed to exactly localize the source is the metric dimension of the network. We compare its value for a static and adaptive sensor placement in Erdös-Renyi random graph models. We comment also on the difference in the high noise regime, when the propagation delays are random.

"A short Introduction to Network Calculus"

by Patrick Thiran

Network Calculus is a set of tools that provide insight in flow problems arising communication networks. It can be viewed as the system theory that applies to applies to deterministic networking. The main difference with the traditional system theory that was so successfully applied to design electronic circuits, is that it is based on min-plus (or max-plus) algebra instead of the usual algebra. We will review the basic concepts of arrival curve, service curve, composition, and illustrate some applications with tight delay and backlog bounds computed with network calculus.

"Local Algorithms and Symmetry Breaking"

by Fabian Kuhn

  Recording

program

The problem of obtaining fast deterministic algorithms for distributed symmetry breaking problems in graphs has long been considered one of the most challenging problems in the area of distributed graph algorithms. Consider for example the distributed coloring problem, where a (computer) network is modeled by an arbitrary graph G = (V,E) and the objective is to compute a vertex coloring of G by running a distributed algorithm on the graph G. It is maybe not surprising that randomization can be a helpful tool to efficiently compute such a coloring. As long as each node v in V can choose among deg(v)+1 different colors, even an almost trivial algorithm in which all nodes keep trying a random available color allows to color all nodes in O(log n) parallel steps. How to obtain a similarly efficient deterministic algorithm is however far less obvious. In fact, for a long time, there has been an exponential gap between the time complexities of the best randomized and the best deterministic distributed algorithms for many basic graph problems. In the last few years, there however has been substantial progress and we now have quite elegant and simple deterministic distributed graph algorithms that are nearly as fast as the classic randomized algorithms for the same tasks. In the first part, I will give an overview on the history of the problem of finding fast deterministic algorithms for distributed symmetry breaking problems. In the second part, we then discuss a concrete, novel distributed derandomization technique that in many cases leads to particularly efficient and simple deterministic distributed algorithms. We will together develop an efficient deterministic distributed graph coloring algorithm from scratch.

“Learning with Graphs”

by Roger Wattenhofer

  Recording

program

In the vast realm of computer science, it may seem that distributed computing and machine learning exist on opposite ends of the spectrum. However, there are many connections between the two domains, both in theory and practice. Recently, machine learning research has become excited about graphs. And when machine learning meets graphs, researchers familiar with distributed algorithms may experience a sense of déjà vu, as many classic distributed computing paradigms are being rediscovered. In my lecture, I will introduce the key concepts in graph machine learning such as underreaching and oversquashing, known for decades in the distributed computing community as local and congest, respectively. Additionally, I will delve into recent breakthroughs and present intriguing open problems that lie ahead in this exciting intersection of fields. My slot will be based on my invited talks at PODC 2023 and OPODIS 2023, but will go well beyond these. Also, participants will be able to have some practical experience with graph learning by working with a Python notebook. While some basic knowledge about machine learning (e.g. backpropagation) is probably advantageous, it is not necessary.

Conclusion

We look back on an exciting week with inspiring talks, discussions, plenty of networking and the wonderful scenery of Champéry. We hope that the school sparked new ideas for your future work, and that you enjoyed the time spent with the organizers, speakers and colleagues.

The group photo is available here. The video recordings can be watched here. The slides will be made available soon.