6 papers accepted for IEEE CDC 2025 on distributed optimization and learning.

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

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

Interoperability contributes a large amount to national productivity. I continue to argue for the need to study interoperability as a science and to put forth frameworks. This is an area lacking within academia and practice. Standards organizations should be playing a larger role too. https://lnkd.in/gCy2BzNC

My paper on "Evaluating Conditional handover for 5G networks with dynamic obstacles" has been published in Computer Communications (COMCOM), Volume 240, doi: \url{https://lnkd.in/dMDQ9ms5, 2025. (Publisher: Elsevier, I.F. 4.3, Indexed in: Science Citation Index Expanded) Abstract To enhance seamless connectivity in millimetre wave New Radio networks, Conditional handover has evolved as a promising solution. Unlike A3 handover where handover execution is certain after receiving handover command from the serving access network, in Conditional handover, handover execution is conditional on Reference signal received power measurements from current and target access networks, as well as on handover parameters such as preparation and execution offsets. Presence of dynamic obstacles may block the signal from serving and (or) target access networks, which results in violation of the conditions for handover preparation/execution. Moreover, signal blockage by dynamic obstacles may cause radio link failure, which may cause handover failure as well. Analytic evaluation of Conditional handover in the presence of dynamic obstacles is quite limited in the existing literature. In this work, handover performance of Conditional handover has been analysed in terms of handover latency, handover packet loss and handover failure probability. A Markov model accounting the effect of dynamic obstacles, handover parameters (e.g., execution offset, preparation offset, time-to-preparation and time-to-execution), user velocity and channel fading characteristics has been proposed to characterize handover failure. Results obtained from the proposed analytic model have been validated against simulation results. Our study reveals that optimal configuration of handover parameters is actually conditional on the presence of dynamic obstacles, user velocity and fading characteristics. This study will be helpful for the mobile operators to configure handover parameters for New Radio systems where dynamic obstacles are present.

💡 IEEE Xplore Tip: The IEEE Standards Dictionary has been integrated into IEEE Xplore in a format allowing for contextual discovery and browsing. With over 39,000 standards terms and complete source citations of definitions, the IEEE Standards Dictionary Online is a comprehensive database of standards terminology.  🔗 https://loom.ly/NHY-1lI #IEEE #IEEEXplore #XploreTips #TipsAndTricks #Tips #ResearchTips #PhD

We’re excited to announce the launch of the Protected Research Compute Cluster (PRCC), a high-performance, secure environment for research involving PHI and other sensitive data. PRCC offers customizable workspaces with familiar tools, optimized GPU and storage capacity, and simplified access to Duke Health’s data resources. Replacing PACE, PRCC expands capacity and reliability while enabling global collaboration, all within a NIST 800-53 compliant framework. Together with the Research Computing Cluster (RCC), PRCC forms part of the Duke University School of Medicine Research Enclave, giving researchers flexible, scalable options for secure data analysis. https://lnkd.in/emzX_SvC

We’re working with partners like Rensselaer Polytechnic Institute (RPI) to build the software foundation for quantum-centric supercomputing (QCSC) https://ibm.co/6042Bygh8 Last year, we shared our vision for QCSC architectures (https://ibm.co/6043ByghD). Now, we’re bringing that vision to life with: 🔌 Spank plugins for the Slurm, currently the most popular HPC workload manager in the world: https://ibm.co/6044ByghE 📚 The quantum resource management interface (QRMI), a vendor-agnostic library to control state, run tasks and monitor behavior of quantum computational resources: https://ibm.co/6045Bygh1 Both tools were developed with the help of researchers from RPI’s Future of Computing Institute, now home to the first QCSC environment ever deployed within the walls of a university. Learn more about these tools in this week’s quantum blog linked above to see how you can contribute.

🎉 New paper in IEEE-TAC! Our paper "𝘈 𝘣𝘦𝘩𝘢𝘷𝘪𝘰𝘳𝘢𝘭 𝘢𝘱𝘱𝘳𝘰𝘢𝘤𝘩 𝘧𝘰𝘳 𝘓𝘗𝘝 𝘥𝘢𝘵𝘢-𝘥𝘳𝘪𝘷𝘦𝘯 𝘳𝘦𝘱𝘳𝘦𝘴𝘦𝘯𝘵𝘢𝘵𝘪𝘰𝘯𝘴" has just been published in the IEEE Transactions on Automatic Control 😁! 🔍In this work, we lay the foundations for achieving analysis and controller design of Linear Parameter-Varying (LPV) systems 𝗱𝗶𝗿𝗲𝗰𝘁𝗹𝘆 from data! We use the celebrated behavioral approach (introduced by the legend Jan C. Willems) to construct finite horizon data-driven representations for LPV systems, with which we can directly perform analysis or control design. 🤔𝗦𝗼, 𝘄𝗵𝘆 𝗟𝗣𝗩 𝘀𝘆𝘀𝘁𝗲𝗺𝘀? The LPV framework has been serving as a bridge between LTI systems and the nonlinear realm, providing both rich representation capabilities and convex, intuitive tools. 𝗧𝗵𝗶𝘀 𝗺𝗲𝗮𝗻𝘀 𝘁𝗵𝗮𝘁 𝗼𝘂𝗿 𝗿𝗲𝘀𝘂𝗹𝘁𝘀 𝗺𝗮𝘆 𝘀𝗲𝗿𝘃𝗲 𝗮𝘀 𝗮 𝗳𝘂𝗻𝗱𝗮𝗺𝗲𝗻𝘁𝗮𝗹 𝘀𝘁𝗲𝗽 𝘁𝗼𝘄𝗮𝗿𝗱 𝗱𝗮𝘁𝗮-𝗱𝗿𝗶𝘃𝗲𝗻 𝗮𝗻𝗮𝗹𝘆𝘀𝗶𝘀 𝗮𝗻𝗱 𝗰𝗼𝗻𝘁𝗿𝗼𝗹 𝗼𝗳 𝗻𝗼𝗻𝗹𝗶𝗻𝗲𝗮𝗿 𝘀𝘆𝘀𝘁𝗲𝗺𝘀! I also want to thank my collaborators Ivan Markovsky, Sofie Haesaert, and Roland Toth! You can check out the paper here: https://lnkd.in/eS7mFz82 🤝Happy to chat if you want details, code, discuss possible collaboration, or explore how this could fit into your work! #ControlEngineering #Research #IEEE #Control #Behavior #DataDrivenControl

New Research: Bounded-Complexity Sequential Decoding for PAC Codes with Pareto Distribution Researchers from Northeastern University by Mohsen Moradi and Hessam Mahdavifar have published significant theoretical advances in polarization-adjusted convolutional (PAC) codes, demonstrating that sequential decoding complexity follows a Pareto distribution when rate profiles adhere to polarized cutoff rate constraints. Key Technical Achievements:  📊 Established theoretical upper bounds on computation distribution for PAC code sequential decoding 📊 Developed rate-profile design guidelines based on polarized cutoff rates 📊 Achieved >0.75 dB coding gain at FER 10⁻⁵ compared to state-of-the-art 5G polar and LDPC codes Practical Impact: The designed PAC(1024, 899) code demonstrates exceptional performance while maintaining low decoding complexity. At high SNR values, the average number of visits per decoded bit approaches 1, validating the practical feasibility of the proposed design methodology. This work addresses critical requirements for next-generation wireless systems, where ultra-reliable low-latency communication (URLLC) for short data packets remains a fundamental challenge. The research provides both theoretical foundations and practical implementation guidelines for PAC codes approaching channel capacity with bounded complexity. Published in IEEE Transactions on Information Theory (2025) DOI: 10.1109/TIT.2025.3608563 Read the full paper: https://lnkd.in/gVXEDihm #WirelessCommunications #ChannelCoding #InformationTheory #6G #Research #Engineering #Telecommunications

Thank you to Shannon Wireless for sharing our work! We are excited that our paper “PAC Codes with Bounded-Complexity Sequential Decoding: Pareto Distribution and Code Design” has been accepted for publication in IEEE Transactions on Information Theory. From a practical standpoint, this paper develops a design profile for polar-like codes that relaxes the conventional #polar rate profile. This allows the use of weight-distribution-based constructions to improve the error-correction performance of polar codes while keeping the #decoding complexity linear in the block length. This approach can be a practical way to construct longer polar-like codes. In our numerical results, following the proposed design rule, we froze certain bits of the #RM rate profile to obtain a decoder that achieves performance very close to the theoretical bounds while maintaining low decoding complexity. Convolutional codes decoded via sequential decoding are known to achieve linear complexity in the block length for rates up to the cutoff rate. In this work, we leverage the polarization of the cutoff rate and develop theoretical results that enable approaching channel capacity while keeping the decoding complexity low. Read our paper: https://lnkd.in/gVXEDihm #WirelessCommunications #ChannelCoding #InformationTheory #6G #Research #Engineering #Telecommunications