Why Interoperability Matters for National Productivity

Happy to share that our paper with my student Mehrdad Salimnejad "Age of Information Versions: a Semantic View of Markov Source Monitoring" has been accepted for publication in the IEEE Transactions on Communications. In our work we introduce two novel performance metrics and propose sampling policies: • Version Innovation Age (VIA): captures how long it has been since a change in the source was successfully delivered. • Age of Incorrect Version (AoIV): measures how outdated the receiver’s information is when it holds an incorrect version of the source. 📄 Read the full paper here: 🔗 https://lnkd.in/d6Cr3Ghs This was in collaboration with Marios Kountouris and Anthony Ephremides. This work has been partially funded by ELLIIT and the ETHER project.

Logical consistency between disagreeing experts can be used to keep us safer when we use their noisy decisions. My latest on this overlooked tool to tame AIs -- submitted to IEEE SaTML 2026 in Berlin. arxiv.org/abs/2510.00821

We are happy to share that our paper, “A New Metric Function for SC-based Polar Decoders: Polarization, Pruning, and Fast Decoders,” has been accepted for publication in IEEE Transactions on Communications. In this work, we show that the variance of SC-based decoder metric values is equal to the polarized varentropy, while our earlier work established that their mean equals the channel capacity—two fundamental parameters that play a central role in the theoretical bounds of finite block-length regimes. Together, these results may open up new directions for analyzing and designing channel codes in short block-length scenarios. We also propose an efficient approximation for computing the varentropy of the BI-AWGN channel. In addition, we present a pruning technique that leverages the polarization of the channel varentropy values. Moreover, we establish a novel connection between polar and PAC codes with convolutional codes, enabling a unified framework for their analysis. Furthermore, we demonstrate that the polarization of the channel cutoff rate highlights the capacity-achieving advantages of polar and PAC codes over convolutional codes, while maintaining the low-complexity decoding benefits of convolutional decoders. Read more here: https://lnkd.in/gy9xq_rt

QMill had the opportunity to participate in the panel discussion “From Algorithm to Advantage” on the final day of the Quantum World Congress. The session was chaired by Outi Keski-Äijö from Business Finland and brought together voices from across the quantum ecosystem. Our CEO Hannu Kauppinen joined the conversation alongside panelists from IQM Quantum Computers, Quanscient, and Haiqu. The discussion focused on the practical path toward quantum advantage — and what it means for industry. As Hannu noted: ▪️ “Quantum advantage is reached when a useful problem – one that has business value for customers – is solved faster on a quantum computer than a classical one. We will arrive at quantum advantage piece by piece.” ▪️ “Quantum advantage is the most important goal for this industry, more so than incremental improvements alone.” ▪️ “Now is the time to start investing into algorithms, not wait for fault-tolerant quantum computing.” At QMill, we are developing quantum algorithms that deliver real-world value for industrial applications. Our focus is on solutions that match customer needs, with performance targets such as 10 000 gate operations at gate fidelities of 99.99% and broad availability of such quantum computing infrastructure. The Quantum World Congress offered a clear snapshot of the global quantum landscape. We met with numerous customers and stakeholders, and it was great to see the Finnish delegation making a strong impact. - Big thank you to Team Finland and the delegation leads for organizing! We also extend our thanks to fellow panel participants: - Outi Keski-Äijö, Business Finland (Chair) - Jouni Flyktman, IQM Quantum Computers - Valtteri Lahtinen, Quanscient - Richard Givhan, Haiqu #QWC2025 #QMill #QuantumAlgorithm #QuantumComputing #QuantumAdvantage #Innovation #DeepTech #QuantumWorldCongress

The research at Information Processing and Transmission Lab proposes a retrieval-augmented generation (RAG)-based framework that combines LLMs with domain-specific knowledge extracted from wireless standards documents, research articles, network-specific data, and other documents. This has been studied for resource allocation in 6G Networks as a potential approach. The paper published as follows: Title: LLM-Based Retrieval-Augmented Generation: A Novel Framework for Resource Optimization in 6G and Beyond Wireless Networks Contributors: Ali Zeeshan, Muhammad Umer, Muhammad A., Aryan Kaushik, Muhammad Ali Jamshed, Haejoon Jung, Syed Ali Hassan Published: IEEE Communications Mag Oct 2025 #6G #LLM #RAG #Research #COMMAG #NUST #SEECS URL: https://lnkd.in/d6DsdhJm

Check this newly published article "Identifying Vital Nodes by Local g-Core on Symmetric Networks" at https://brnw.ch/21wWt9x Authors: Letian Wang The H-index is a widely recognized centrality measure for nodes in symmetric networks, defined as the maximum number of neighbors with degrees equal to or greater than the node’s own degree. However, this metric underestimates the structural influence of “weak nodes”—low-degree nodes connected to high-degree hubs—that often serve as critical connectors in network topology. To address this limitation, we propose the 𝐻𝛼-index, which generalizes the H-index by considering the maximum number of neighbors with degrees at least 𝛼 times the node’s degree, where 𝛼≥1. Based on this refinement, we introduce two novel centrality measures: the g-core and the local g-core, which were derived from iterative applications of the 𝐻𝛼-index to a node’s neighbors... 山东大学 #mdpisymmetry #complexnetworks

Thrilled to share that our paper “A Comparative Study of OpenMP Scheduling Algorithm Selection Strategies” has just been accepted and published in IEEE Access 🎉 📖 You can read the open-access article here: https://lnkd.in/dQKfbMhq This work was a true team effort. We explored both expert-based and reinforcement learning approaches to scheduling algorithm selection in OpenMP, and showed how they can enhance performance. It’s exciting to see this line of research contribute toward self-adaptive scheduling for HPC applications, with potential to go even further through hybrid expert + RL strategies and integration across nodes. #IEEEACCESS #HPC #OpenMP #scheduling #reinforcementlearning #research #teamwork

Day 10 of 21 Days Quantum Computing Challenge: Today I learned about Quantum Parallelism and Interference. Quantum Parallelism is based on Superposition that is the ability of a qubit to exist in a combination of basis states. This allows a quantum computer to evaluate a function for multiple input values at once. However the critical problem is to harness the information as measurement only reads out one value at a time similar to classical computing. This is where interference becomes important. Quantum states can interfere like waves which allow to amplify or filter out information. Entanglement can destroy interference as it act as a measurement. Thus it's important prevent a quantum computer from interacting with its environment #QuCode https://lnkd.in/ecvaGTgt https://lnkd.in/ebxN4jzW

Pleased to share that 6 of our papers have been accepted for presentation at the IEEE Conference on Decision and Control (IEEE CDC 2025). Our papers present distributed optimization and learning, and consensus algorithms (addressing topics such as stepsize selection, quantized communication, dynamic networks, nonlinear MHE, and MPCC problems), operating over networks with possible communication constraints, and open architectures. Grateful to my students and my collaborators: Jiaqi Hu, Xu Du, Nicola Bastianello, Themistoklis Charalambous, Karl Henrik Johansson. Looking forward to engaging discussions this December in IEEE CDC. Paper Links: • Distributed Optimization and Learning for Automated Stepsize Selection with Finite Time Coordination https://lnkd.in/dx9UguMA • Distributed Quantized Average Consensus in Open Multi-Agent Systems with Dynamic Communication Links https://lnkd.in/diy_xmDW • Decentralized Optimization via RC-ALADIN with Efficient Quantized Communication https://lnkd.in/dSHbUdTU • Average Consensus with Dynamic Compression in Bandwidth-Limited Directed Networks https://lnkd.in/dxVU3JvV • A Time Splitting Based Optimization Method for Nonlinear MHE https://lnkd.in/d5d-PXdy • ALADIN-β: A Distributed Optimization Algorithm for Solving MPCC Problems https://lnkd.in/dD8s_wa3 #DistributedOptimization #DistributedLearning #DistributedAlgorithms #IEEE #CDC2025

📈 Most Cited Paper Last Month in IEEE TCSVT Efficient Image Enhancement With a Diffusion-Based Frequency Prior by Qingsen Yan, Tao Hu, Peng Wu, Duwei Dai, Shuhang Gu, Wei Dong, and Yanning Zhang This widely cited paper introduces a diffusion-based frequency prior for efficient image enhancement, offering readers new tools for improving image quality in real-world computer vision and multimedia applications. When authors cite work published in IEEE Transactions on Circuits and Systems for Video Technology, they connect their own research to one of the most influential journals in the field. TCSVT’s bibliometric strength underscores this impact: ✨ Impact Factor: 8.4 ✨ Article Influence® Score: 1.607 ✨ CiteScore: 13.8 Discover why this paper is shaping the conversation in image processing—access the article through the link in the comments 👇