Program

In recent years, synthetic molecular communications (MC) has emerged as a new field of research in information theory and communication engineering with strong links to several other disciplines, including biology, nanotechnology, and medicine. MC is expected to provide connectivity in environments that are not suitable for conventional communication systems based on electromagnetic (EM) waves, such as the human cardiovascular system, bio-processes, and water pipes, and facilitate novel applications such as the Internet of BioNanoThings (IoBNT), targeted drug delivery, and interfacing with animals and plants.

Although some concepts known from conventional communication systems are applicable in MC, there are also many new aspects and differences. In this tutorial, we will first provide a broad introduction to synthetic MC, reviewing different forms of MC occurring in nature, the propagation of molecules, and potential applications. Subsequently, we will introduce communication-theoretical models for MC channels, discuss phenomena such as degradation and dispersion, and unveil concepts for MC system design. Finally, we will present an overview of state-of-the-art experimental MC systems and provide case studies for applications of MC concepts, including bio-process control and olfactory systems.

Massive multiple-input multiple-output (MIMO) is a key technology for beyond-5G multi-user communication systems that enables high spectral efficiency via fine-grained beamforming. However, all-digital massive MIMO basestation (BS) architectures suffer from excessive system costs, power consumption, and interconnect data rates when implemented naïvely. Moreover, the large bandwidths available at mmWave or terahertz frequencies exacerbate these problems. In this talk, we discuss recent progress on the algorithm, architecture, and hardware implementation levels towards high-throughput and reliable multiuser communication at low power. Concretely, we discuss the following topics: low-resolution baseband processing, beamspace equalization, and jamming-resilient communication.

The scientific community and industry are exploring ways to meet the increasing traffic demand in optical networks at various levels. While technological innovations aim to expand bandwidth and spatial multiplexing through the development of new fibers and optical amplifiers, advancements in modulation techniques, coding, and signal processing are being pursued to maximize spectral efficiency to fully leverage the capacity of existing fiber links. The challenge is further complicated—and made more intriguing—by the nonlinearity of optical fibers, which make the determination of channel capacity an open problem. In this talk, I will introduce the problem of fiber optic channel capacity and discuss recent techniques proposed to evaluate it and to practically approach it through advanced modulation strategies. Specifically, I will focus on a novel method called sequence selection. Originally developed as a tool in information theory to derive capacity bounds, sequence selection also shows promise as a practical and general method for the implementation of efficient modulation schemes, allowing more reliable data transmission over existing fiber networks.

We consider the scenario in which multiple devices track the state of physical/digital processes, quantizes this state, and communicates it to a common receiver through a shared channel in an uncoordinated manner. The receiver aims to estimate the type of the states, i.e., the set of states and their multiplicity in the sequence of states reported by all devices. Such a scenario is relevant in several applications, including over-the-air digital federated learning and multi-target position tracking. We present fundamental information-theoretic bounds to quantify the performance achievable over such a type-based unsourced multiple access channel, as well as practical algorithms to approach these bounds.

Since the seminal work of Shor we know that efficient quantum computers are a threat to the security of public-key cryptography that is currently deployed. A huge amount of work is being put into developing an efficient quantum computer. But even if the advent of such a computer may wait for decades, it is urgent to deploy post-quantum cryptography (PQC), i.e: cryptographic solutions working on our devices which are secure even against « quantum adversaries ». Indeed, an attacker could store encrypted sessions and wait until a quantum computer is available to decrypt. In this context the National Institute of Standard Technology (NIST) has launched in 2017 a call for standardizing public-key PQC schemes. We are currently in the final step of the standardization and most of the selected solutions are basing their security on the hardness of error correcting code problems (but also on hard Euclidean lattice problems). The aforementioned problems come from information theory, which has proven to be an amazing source of hard problems. The most famous is likely the decoding problem: given a code and a target, the goal is to find the closest codeword (in terms of Hamming distance). This problem is considered particularly hard when no structure is assumed for the code, especially when the code is assumed to be random. Surprisingly, since ’78, with the work of McEliece and later Alekhnovich in ’03, we know how to construct cryptographic solutions whose security is based on the difficulty of this problem and which turn out to be secure even against quantum adversaries (at least in our current knowledge). The field studying how to build cryptographic solutions via the hardness of decoding a random code is referred to as code-based cryptography.
The aim of this talk will be to present how we can achieve security via the hardness of the decoding problem of a random code. Then in a last part, we will introduce the basics of quantum computing and explain why it poses a threat to the security of currently deployed cryptographic primitives but not to code-based cryptography.

We introduce the topic of integrated communication and sensing. Then, we show how an in-band full duplex communication radio can easily be extended with radar signal processing, almost for free. When extending the concept to MIMO, massive MIMO and cell-free networks, a range of opportunities arise for multi-static joint communication and sensing, even in scenarios with multipath and other non-idealities.

The rise of machine learning has introduced powerful generative models, including diffusion models, which have achieved remarkable success in areas like image generation. In this talk, we provide a comprehensive overview of diffusion models, explaining their underlying mechanisms and their potential in tackling complex problems. We then explore how these models can be applied to wireless communications to address critical challenges such as channel estimation and reliable data transmission over low-SNR (Signal-to-Noise Ratio) links. By leveraging diffusion models’ ability to incorporate prior knowledge about data distributions, we demonstrate how they can significantly enhance transmission accuracy and efficiency in wireless systems. Experimental results will be presented to showcase the performance improvements achieved using these models, particularly in scenarios where traditional methods struggle. Finally, we will outline key challenges and promising research directions in integrating diffusion models with next-generation wireless technologies.

The integration of radio communication and sensing services in the same network infrastructure will be a key feature of next-generation wireless technology. This emerging paradigm, known as joint communication and sensing (JCAS) or integrated sensing and communication (ISAC), will enable future wireless base-stations to communicate with active devices, and simultaneously sense their surroundings; all using the same spectrum and hardware resources. Realizing provably efficient and reliable JCAS systems, however, strongly hinges on developing an information-theoretic framework through which we can establish fundemental performance limits and trade-offs, and derive insights into the design of optimal JCAS schemes. In this talk, I will discuss recent progress and steps towards an information theory for JCAS systems.

Large-scale MIMO communication (at sub-6GHz, mmWave, and sub-THz bands) is a key enabler for 5G, 6G, and beyond. Scaling MIMO systems, however, is subject to critical challenges, such as the large channel acquisition and beam training overhead and the sensitivity to channel estimation errors (especially at lower frequencies) and blockages (at higher frequencies). These challenges make it difficult for MIMO systems to support applications that have high mobility and strict reliability constraints. In this talk, I will first motivate the use of machine learning and sensory data to address these challenges. Then, I will present a few key machine-learning roles, enabling datasets, and recent hardware proof-of-concept prototypes that demonstrate the machine-learning gains in real-world environments.

With the advancing digitalization of enterprises and society, mobile communication networks are expected to provide the glue to connect physical assets and devices to digital services that support them in their operation. In this trend, also critical applications are increasingly connected via mobile networks. A critical application provides essential functionality to an application domain, and a minimum performance level of connectivity is essential to safeguard its proper operation. This talk will outline key network enablers to provide dependable communication services to critical applications.

Quick welcome by the conference organizers and an quick overview of KIT and the faculty of electrial engineering and information technology by Prof. Dr.-Ing. Eric Sax, dean of the faculty.

The welcome reception will take place at the foyer of the conference venue (Tulla lecture hall) in Building 11.40 at KIT Campus South (Englerstr. 11, 76131 Karlsruhe). Details on how to reach the venue can be found at https://scc2025.net/register-travel.

KIT Campus North is known for large-scale research facilities and scientific experiments in the Helmholtz Association. We are happy to be able to offer excursions on three different topics:

• Infrastructure for high-level research
• Large-scale experiments for science
• (On the trail of) nuclear research

More information about the excursions can be on the official site https://www.cse.kit.edu/english/campus-tours.php.

The banquet will be hosted at the Hoepfner Burghof, which is the restaurant and event location colocated to the brewery Privatbrauerei Hoepfner, one of the largest breweries in the city of Karlsruhe. It is accessible by a short walk (1.3km/20 minutes) from the lecture hall.

This location and the brewery, commonly known as the “Hoepfner Burg” were constructed in the late 1800s after a necessary relocation of the brewery from city center due to size constraints. On November 21st 1899 the first beer was served at the Burghof. The attached beer garden which can host up to 2000 people was constructed concurrently and served locally brewed beer to the population. Previously, at the place where the beer garden is located the brewery stored ice, which was transported from the Alps or at times from Norway, to refrigerate the produced beer. The interior of the Burghof is maintained close to the original design and therefore a rustic and cozy atmosphere is created for the traditional and rustic kitchen. It goes without saying that all locally brewed Hoepfner beer specialties are available at the Burghof.

You can find more information (in German) on the official site https://hoepfner-burghof.com/.