All Years Seminars
Page 1 of 157
Next
[INGI] 2026-12-03 (13:00) : Measuring and Improving Quality of Experience in Cloud Game Streaming
At Nyquist Room - Maxwell a.164
Speaker :
Xiaokun Xu (ICTEAM-INGI)
Abstract : Cloud game streaming moves game rendering from the player’s device to remote cloud servers, enabling high-quality gaming on lightweight hardware. However, its interactive nature makes the player experience particularly sensitive to network conditions such as latency, jitter, and bandwidth variation.
In this talk, I will present my research on measuring, modeling, and improving Quality of Experience (QoE) in cloud gaming. Through controlled user studies, network experiments, and real-world home deployments, I investigate how network and system factors affect responsiveness, visual quality, and gameplay smoothness, and use these observations to develop models for predicting player QoE.
I will also discuss approaches for improving QoE at different layers of the cloud gaming system, including Active Queue Management (AQM) at the network layer, playout buffering at the client, and latency compensation at the server. These results highlight the trade-offs between latency and smoothness and provide practical insights into designing more responsive and robust cloud gaming systems.
Since this is a general seminar for an audience that may not be familiar with the topic, I plan to keep the presentation accessible rather than going too deeply into technical details. I think 30–40 minutes should be sufficient
[INGI] 2026-11-26 (13:00) : Hardware-Software Co-design for Packet-Processing Applications
At Shannon Room - Maxwell (A.105)
Speaker :
Nikita Tyunyayev
Abstract : Network applications are struggling to keep up with the exponential growth of network traffic. To address this, they can rely on the increasingly feature-rich hardware found in modern servers. While powerful, this hardware has its own limitations and challenges. In particular, commodity NICs cannot run stateful packet-processing applications on their own. In this presentation, we will review the capabilities and limitations of modern commodity NICs and show how software can leverage them to accelerate packet-processing applications.
[INGI] 2026-11-12 (13:00) : NVMe-over-QUIC: Secure and Efficient NVMe Disaggregation over IP Networks
At Shannon Room - Maxwell (A.105)
Speaker :
Soyong Kim
Abstract : NVMe-over-Fabrics enables disaggregated storage, but high-performance deployments often rely on specialised lossless RDMA fabrics. This talk presents NVMe-over-QUIC, a secure IP-based transport that does not require such a fabric. Our design maps each NVMe command to an independent QUIC stream, leveraging QUIC’s built-in encryption, multiplexing, and loss recovery while avoiding cross-command head-of-line blocking. I will also describe a UDP-based core-steering architecture that preserves CPU affinity for multi-core scaling. Our evaluation shows up to 4× higher throughput than NVMe/TLS and substantially improved tail latency under packet loss and additional delay. These results show that QUIC can provide an efficient and robust alternative for secure NVMe disaggregation over conventional IP networks.
[ELEN] 2026-10-15 (15:00) : Seminar on Wireless Sensing and Communications
At Nyquist
Section 1:Optimal Sparse Array Design for Sensing and Communications (15:00)
Speaker :
Marius Pesavento (TU Darmstadt)
Abstract : Sparse linear arrays obtained by thinning large uniform linear arrays are widely used in wireless communications and radar sensing because they offer large effective apertures and high angular resolution with fewer sensors. However, thinning also introduces array ambiguities that can induce rank deficiencies in the associated channel or sensing matrix, thereby limiting spatial multiplexing gains and preventing unique multisource parameter estimation in certain scenarios. Conventional sparse-array designs for sensing are often based on local criteria, such as Fisher information matrix (FIM) analysis, which do not fully account for dependencies among sources. More fundamentally, the underlying problem is that of determining the spark of a generalized Vandermonde matrix – an important open question with connections to algebraic coding theory, polynomial approximation, and control theory. In this talk, we present a novel scalable framework for the design and analysis of sparse (thinned) uniform arrays. Specifically, we reformulate the problem as one of designing structured matrices of maximum rank, which in turn yields simple principles for constructing optimal sparse arrays. The resulting approach combines symbolic computation with structured rank optimization. It is simple, scalable, and well suited to the design of very large sparse arrays.
Section 2:Near-field Radio Localization and Mapping with Intelligent Surfaces (16:00)
Speaker :
Anna Guerra (Università di Bologna)
Abstract : The next generation of cellular networks promises to revolutionize wireless communication by creating smart radio environments (SREs) through the use of reconfigurable intelligent surfaces (RISs). These RISs offer significant benefits, including the ability to optimize the radio channel for specific applications, thereby enhancing overall system performance. For instance, RIS elements can be reconfigured to achieve precise localization and sensing, while simultaneously improving communications in cellular networks. Additionally, with the adoption of large aperture arrays and high frequencies, the traditional plane wave approximation, which is valid in the far field, becomes inadequate. In these scenarios, cellular users are likely to be positioned in the near field relative to base stations and RISs. The resulting wavefront curvature allows for the extraction of both distance and angle-of-arrival information, enabling more accurate source localization with fewer deployed devices (single-anchor localization). This webinar will present recent advancements in near-field localization and sensing for next-generation communication systems. It will begin by exploring the localization fundamental localization limits in RIS-aided systems, followed by a review of algorithms designed for imaging and tracking mobile users. The webinar will also show how optimizing the phase profile of RISs can significantly impact localization performance.
[INMA] 2026-10-13 (14:00) : Using randomness to reveal (part of) the mathematics of (deep) learning
At Euler building (room A.002)
Speaker :
Laurent Jacques (UCLouvain)
Abstract : In this talk, we will see how several phenomena in the study of machine and deep learning models, such as their ability to linearly separate two classes of samples or their stability with respect to input variations, can be partly explained by random, untrained parameters. We will draw connections with one-bit compressive sensing and the measure concentration phenomenon.
Page 1 of 157
Next
Seminars
Subscribe (ICS)
INMA Contact Info
Mathematical Engineering (INMA)
L4.05.01
Avenue Georges Lemaître, 4
+32 10 47 80 36
secretaire-inma@uclouvain.be
Mon – Fri 9:00A.M. – 5:00P.M.
JOIN US