Tuesday 3 November
9h00-9h15: Welcome
9h15-10h15: Keynote (chair: Dieter Fiems)
Bruno Gaujal
Reinforcement learning of the optimal control in queues: The exploration-free approach
10h15-10h45: Break
10h45–12h15: Scheduling and load balancing (chair: Urtzi Ayesta)
- Mean field optimal Core Allocation across Malleable jobs
Z. Li, M. Harchol-Balter, B. Berg - Fixed-Point Analysis of a Load Balancing Scheme in Large-Scale Parallel Server Systems with Heterogeneous Capacities and Service Rates
C. Sekita, T. Phung-Duc - Explicit and asymptotic analysis of the mean-field model for Join-Idle-Queue load balancing
B. Van Houdt, A. El Marchouhi
12h15–12h30: SIGMETRICS presentation
12h30–13h30: Lunch
13h30–15h30: Network performance and switching (chair: Sabine Wittevrongel)
- TCP BBR Congestion Control Evaluation in OMNeT++ INET
L. Martino, K. Hielscher, R. German - Latency-Derived Throughput Risk Modeling for Low-Overhead Network Monitoring
J. Bai, Z. Jia - ShortCutting Fast Failover Routes Locally: Efficient eBPF-Implementations and Multiple Link Failures
S. Althoff, M. Weiler, K. Foerster - QPS-ToR: A Parallel Iterative Switching Algorithm for Reconfigurable Optical Datacenter Switching
D. Song, Q. Yu, J. Xu
15h30–16h00: Break
16h00–17h30: Caching (chair: Tuan Phung-Duc)
- Quantized Online Gradient Descent for Constant-Time Caching with Dynamic Regret Guarantees
I. Pinillo Esquivel, S. Alouf, D. Carra, G. Neglia - TTL Approximations for a Family of CLOCK-Like Cache Replacement Policies
Y. Wang, B. Van Houdt, P. Desnoyers - No-Regret Mixing of LRU and LFU with Optimal Switching Cost
Y. Ben Mazziane, X. Zou
17h30-18h30 IFIP WG 7.3 Business meeting
19:00 Social activity
Visit to the Ghent Altarpiece. We will meet at the entrance of the Ghent Altarpiece visiting centre in Kapittelstraat (street on the left side of St Bavo Cathedral) at 19h00 sharp.
Wednesday 4 November
9h00-10h00: Keynote (chair: Michela Meo)
Valeria Cardellini
Orchestrating Compute in the Cloud-Edge Continuum
10h00–10h30 Break
10h30-12h30 Online learning and optimisation (chair: Leana Golubchik)
- Breaking the O(√T) Cumulative Constraint Violation Barrier while Achieving O(√T) Static Regret in Constrained Online Convex Optimization
H. Balasundaram, R. Vaze, K. Mahendran - Diameter-Independent Reinforcement Learning for the Control of State-Dependent Queues
M. Sheldon, G. Casale - Online Adaptive Kernel Mixing for Gaussian Process Decision Making
K. Aravindan, M. Sriram, G. Dasarathy, T. Bodas - Backpropagation-Based Distributed Optimization for Networked Systems with Nonlinear Dynamics
A. Reiffers, D. Menache, N. Shimkin, E. Altman
12h30–13h30 Lunch
13h30-15h00 Distributed systems and resource sharing (chair: Alexandre Reiffers-Masson)
- Distributed Approximations for the Inverse Problem in Multi-Channel CSMA Networks
K. Agham, C. Goenka, S. Peruru, Y. Singh, B. Van Houdt - State Catch-up vs. Availability: A Performance Evaluation of Proactive Quorum Guards
A. Metwally, K. Foerster - The Energy-Data Dilemma: When Self-interest meets Federated Learning
O. Haddaji, O. Brun, B. Prabhu, R. El Azouzi
15h00-15h30: Break
15h30–17h30 Queueing and resource scheduling (chair: Balakrishna Prabhu)
- Pareto-Optimal Scheduling in the Half-batch Multiserver-job Model
Z. Wang, I. Grosof - Queueing analysis of a battery-backed electric vehicle fast charging system
A. Ferragut, S. Bibiloni - Partition-Aware Scheduling for Mobile Heterogeneous Inference Co-Execution
Z. Li, M. Paolieri, L. Golubchik - Monotone Performance Measures in Closed Multiclass Queueing Networks
G. Casale, J. Leake, M. Sheldon
19h30 Conference dinner
Conference dinner location: Pakhuis, Schuurkenstraat 4, 9000 Gent (near Korenmarkt)
Thursday 5 November
Workshop: Analytical and Stochastic Modelling Techniques and Applications
9:00-10:00: Keynote
Sara Alouf
Analytical Modelling and Online Learning for Modern Caching Systems
10:00-10:30 Break
10:30-12:00 Applications
- Occupancy and Charging in Large EV Stations: A Mean-Field Decomposition
Koen De Turck, Tijl Doms, Dieter Fiems - Performance evaluation of Rucio for medical data management in Italy
Marco Gribaudo, Mauro Iacono, Elia Maggioni, Federico Lamperti - Equilibrium joining strategies in a service system with information
about the number of consecutive joining decisions
Antonis Economou, Odysseas Kanavetas, Athanasia Manou, Sotiris Tzinakis
12:00-13:00 Keynote
Giuliano Casale
Stacks, Lists and Permanents: An Ordered Walk through New Product Forms
13:00-14:00 Lunch
14:00-16:30 Queueing theory
- A two-class M/M/1 queueing model with service rate of type-1 customers
depending on the presence of type-2 customers
Matthijs Limpens, Jacques Resing - Delay-Optimal FIFO Residual Service Curves Under General Arrival Curves
Lukas Wildberger, Anja Hamscher, Jens Schmitt - Equilibrium Queueing Strategies in Random Environment
Konstantin Avrachenkov, Uri Yechiali - Greedy failure and a stabilizing partial group discipline on a nested hypergraph
Shu Li, Ana Busic, Jean-Michel Fourneau - Multi-formalism Modeling for Fast Design Space Exploration in ADAS Applications
Mattia Tibaldi, Marco Gribaudo
Workshop: Federated Learning: Recent Advances in Algorithms and Distributed Systems
9:00-10:00 Session 1
- Combinatorial Optimization for Energy-Efficient In-Network Federated Learning Garance Gérard, Patient Ntumba, Safia Kedad-Sidhoum, Amélie Lambert, Nancy Perrot
- Federated Learning with Deadlines in Stackelberg Games Dheeraj Narasimha, Daniela Cuesta, Abdelkrim Alahyane, Matthieu Jonckheere, Balakrishna Prabhu
10:00-10:30 Break
10:30-12:30 Session 2
- The Energy–Data Dilemma: From Subset Sizing to Strategic Federated Clients Oumayma Haddaji, Olivier Brun, Balakrishna J. Prabhu
- Beyond FedAvg: Mapping the Design Space of Personalized Federated Learning Sepanta Farzollahi, Guillaume Fraysse, Rachid El-Azouzi
- Optimization Trade-offs in Asynchronous Federated Learning: A Stochastic Networks Approach Abdelkrim Alahyane, Céline Comte, Matthieu Jonckheere
- Deadline-Aware Workload Adaptation for Straggler Mitigation in Over-the-Air Federated Learning Ahmad Dabaja, Caina Figueiredo Pereira
12:30-13h15 Discussion Exploring New Frontiers in Federated Learning
13:15-14:00 Lunch