Category: PhD Students

TNO-ESI and Academic Partners Deliver ASCI PhD Course on Design and Implementation of Real-time Systems

Posted on by bennyakesson

The Netherlands boasts a world-leading high-tech manufacturing industry renowned for constructing distributed real-time systems of continuously growing complexity. These systems must meet stringent timing requirements to ensure the delivery of mission-critical functionalities. To create interest in the high-tech equipment domain and prepare PhD students in Computer Science to address its performance challenges, TNO-ESI has co-created and delivered a one-week PhD course Design and Implementation of Real-time Systems together with academic partners from Eindhoven University of Technology, University of Twente, and University of Amsterdam. The course is given in the context of the Advanced School for Computing and Imaging (ASCI), a Dutch research school for high-quality research and education in computer systems and imaging systems. ASCI encompasses almost all Dutch universities with computer-science departments. The main goals of ESI involvement in this course were to make participants aware of TNO and its role in society and industry and position it as a possible future employer, and creating awareness of TNO-ESIs vision and work in the area of system performance engineering.

The course is focused on providing an overview of selected timing-sensitive applications and the current research landscape on real-time systems and explaining the rationale behind considering real-time requirements in system software design. Through a series of lectures and hands-on labs, the course covers selected topics from scheduling algorithms, priority assignments, resource sharing, resource reservation, together with their implementation in real-time operating systems. It further discusses emerging challenges and practices in an industrial context, based on empirical surveys and experience from TNO-ESIs applied research on telemetry-based system performance engineering for purposes of performance optimization, verification, and diagnostics.

This first instance of the course was given at the Carlton President Hotel in Maarsen, outside Utrecht between June 10 – 14. 15 PhD students from universities all over the Netherlands researching a broad range of topics in computer science participated in the course. TNO-ESI was in the spotlight during the last day of the course. In the morning, I introduced the high-tech equipment domain and its complexity drivers and explained how new model-based engineering methodologies where needed to address them. Next, my colleague Bram van der Sanden presented our view on the field of System Performance Engineering, along with its focus areas and best practices. This was followed by two concrete examples from our system performance research: Kostas Triantafyllidis presented his work on performance analysis and diagnosis with ASML, followed by a presentation by me about performance verification and conformance checking in microservice systems based on our work with Thales.

The course was well-received by the participants and the contents were rated 8.7/10 in the evaluation. We very much enjoyed the experience of creating and delivering this course together with our academic partners. Thank you Kuan-Hsun Chen (leader of the initiative), Mitra Nasri, and Geoffrey Nelissen for the excellent collaboration in organizing this course. Thanks to Kay Heider and Christian Hakert for leading the hands-on exercises. We are also thankful to invited speakers Bram van der Sanden and Kostas Triantafyllidis.

PhD Defense on Governance of Complex Cyber-infrastructure

Posted on by bennyakesson

Today, I chaired a PhD committee at the University of Amsterdam. The PhD candidate Mostafa Mohajeri Parizi defended his dissertation “An Agent-based Approach to the Governance of Complex Cyber-Infrastructures”. The dissertation explores the impact of digitization on society and the need for engineering approaches to ensure that digital systems comply with regulations. Mostafa did his PhD research in the Complex Cyber-Infrastructure (CCI) group under the supervision of TNO colleague Tom van Engers (Principal Scientist ISP DS). The work centers on the use of computational agents and norms to develop tools and methodologies for governing socio-technical systems. The study introduces ASC2, an agent-based programming framework built on the belief-desire-intention model, alongside a scalable multi-agent system environment. It emphasizes the integration of mainstream software development tools into agent-based programming and enhances transparency and decision-making in agents by incorporating context-dependent preferences. Furthermore, the dissertation proposes a modular architecture for integrating norms into multi-agent systems, allowing for the flexible adoption and reasoning of norms without compromising agent autonomy. This is illustrated through two case studies demonstrating the framework’s application in coordinating agent actions and aligning them with encoded laws. The research highlights the potential of agent and norm models in improving the design and policy-making of digital infrastructures.

Mostafa did a good job presenting and defending his work and the committee were impressed by the breadth of the research. We hope the newly minted Dr. Mohajeri Parizi enjoyed the ceremony and the celebration with his friends and family and wish him all the best of luck in his future career.

Advancing Sustainability: Paper Accepted on Estimating Energy Consumption of Applications in the Computing Continuum

Posted on by bennyakesson

I am happy to please that the paper “Estimating the Energy Consumption of Applications in the Computing Continuum with iFogSim” was accepted at the International Workshop on Converged Computing (WOCC). The paper is first-authored by Saaedeh Baneshi and is the first publication to come out of the project Energy Labels for Digital Services. Congratulations Saaedeh!

The paper explains how digital services – applications that often span the entire computing continuum – have become an essential part of our daily lives, but they can have a significant energy cost, raising sustainability concerns. Measuring the energy consumption of such applications is challenging due to the distributed nature of the system and the application. As such, simulation techniques are promising solutions to estimate energy consumption, and several simulators are available for modeling the cloud and fog computing environment. The paper investigates iFogSim’s effectiveness in analyzing the end-to-end energy consumption of applications in the computing continuum through two case studies. We design different scenarios for each case study to map application modules to devices along the continuum, including the Edge-Cloud collaboration architecture, and compare them with the two placement policies native to iFogSim: Cloud-only and Edge-ward policies. We observe iFogSim’s limitations in reporting energy consumption, and improve its ability to report energy consumption from an application’s perspective; this enables additional insight into an application’s energy consumption, thus enhancing the usability of iFogSim in evaluating the end-to-end energy consumption of digital services.

Optimizing Efficiency and Performance: PhD Thesis Defense on Energy- and Time-aware Scheduling for High-Performance Embedded Systems

Posted on by bennyakesson

Yesterday, I participated in the PhD defense committee of Julius Röder, a PhD student in the Parallel Computing Systems group at the University of Amsterdam. The thesis “Energy- and Time-aware Scheduling for Heterogeneous High-Performance Embedded Systems” addresses the relevant problem of optimizing non-functional behavior, such as timing and energy consumption, of heterogeneous high-performance embedded systems. The goal of this optimization Is to reduce energy consumption, thereby also reducing carbon footprint and extending battery-life, as well as ensuring that real-time requirements of applications are satisfied, even at high resource utilizations. To this end, the thesis contributes a discussion on setups used for energy measurements, as well as experiments and a statistical analysis that demonstrate the Importance of sampling frequency on the accuracy of such measurements. The bulk of the thesis proposes heuristic algorithms, both conventional and based on reinforcement learning, for mapping and scheduling applications modelled as directed acyclic graphs (DAG) on heterogeneous platforms. The applications are assumed to be available In different versions, with different non-functional behavior, for the different types of processing elements In the heterogeneous architecture, which enables trade-offs between timing and energy. A key strength of the thesis is that theory is combined with a practical component; the scheduling algorithms are implemented and evaluated on a heterogeneous multi-core systems, where timing and energy behavior are carefully measured and analyzed.

In presence of family, friends, and colleagues, Julius confidently defended his PhD thesis and earned the right to call himself a doctor. Congratulations Julius with this great achievement!

Vacancy for a PhD in Energy Labels for Digital Services

Posted on by bennyakesson

Back in July, I announced that our project proposal “Zero-Waste Computing: Energy Labels for Digital Services” was granted for the Science and Design PhD program at the University of Amsterdam. Now, the Parallel Computing Systems (PCS) group  is looking for a suitable PhD candidate for this project. Among other things, this involves modelling and monitoring to determine how energy is consumed in digital services whose computations are distributed over device, edge, and cloud.

Find a more detailed description of the vacancy, as well as instructions for how to apply here. The application period ends on October 18.

Official Project Kick-off for DSE 2.0

Posted on by bennyakesson

Today was the official project kick-off for the research project “Design Space Exploration 2.0: Towards Optimal Design of Complex, Distributed Cyber Physical Systems”. This project is a part of the Partnership Program Mastering Complexity (MasCot), funded by NWO Domain Applied and Engineering Sciences (AES) together with ESI (TNO). The University of Amsterdam and Leiden University are the academic partners, spearheaded by Andy Pimentel and Todor Stefanov. The carrying industrial partner is ASML, but with Philips, Siemens and ESI as parts of the user committee.

The main goal of the project is to extend existing methods for design-space exploration, often developed for on-chip systems, to cover complex distributed cyber-physical systems (dCPS), such as the lithography machines made by ASML. Designers of such systems need quick answers to so-called “what-if” questions with respect to possible design decisions/choices and their consequences on non-functional properties, such as system performance and cost. This calls for efficient and scalable system level design space exploration (DSE) methods that integrate appropriate application workload and system architectures models, simulation and optimization techniques, as well as supporting tools to facilitate the exploration of a wide range of design decisions. However, such DSE technology for complex dCPS does currently not exist. This projects hence tries to answer the question of how perform efficient and effective DSE for complex, distributed cyber-physical systems.

In today’s kick-off meeting, all stakeholders in the project had an opportunity to introduce themselves and refamiliarize themselves with the project and its goals. The two PhD students who will be working on the project, Marius and Faezeh, from UvA and Leiden, respectively, also gave a brief overview of the work they had done in the first three months of the project, which included a literature review and generation of high-level simulation models for different parameter settings.

I am directly involved in this project through my part-time appointment at UvA. As Marius’ second promotor, I will help him on his journey towards a PhD. I also have an interest in this project as an ESI Research Fellow and part of the MasCot Core Team. In this capacity, I am happy to help linking this project to ESI’s applied research projects, in particular at ASML, to exploit possible synergies, and to stimulate exchanges with other projects in the MasCot program.

Anna Minaeva Successfully Defends Dissertation

Posted on by bennyakesson

Today, Anna Minaeva successfully defended her PhD dissertation entitled “Scalable Scheduling Algorithms for Embedded Systems with Real-Time Requirements” and earned the right to call herself a doctor. The reviewers were pleased with the dissertation and she confidently answered their questions.

The dissertation considers applications with real-time requirements sharing resources, such as memories, cores, and networks, in distributed systems. Scheduling this type of application subject to resource and precedence constraints, among others, while maximizing system performance is a challenging problem. Existing approaches either propose exact solutions that cannot solve industrial-sized instances or propose heuristic algorithms without validating its efficiency with optimal solutions.

The dissertation addresses this problem through a three-stage approach, corresponding to three problems with gradually increasing complexity and accuracy of the model. The four main contributions of are: 1) Comparison of three formalisms to solve the problems optimally, Integer Linear Programming (ILP), Satisfiability Modulo Theory, and Constraint Programming, along with computation time improvements. To increase the scalability of the ILP approach, an optimal approach that wraps the ILP in a branch-and-price framework is presented. 2) For each problem, a scalable and efficient heuristic algorithm is presented that decomposes the problem to decrease its computation time. 3) The efficiency of the optimal and heuristic strategies are quantitatively and qualitatively compared. 4) The practical applicability of the proposed heuristic algorithms and optimal approaches is demonstrated on case studies of real systems in both the automotive and consumer electronics domains.

I wish Anna the best of luck in her future career and hope I will have the opportunity to work with her again.

Hazem Ali Defends Dissertation

Posted on by bennyakesson

Today, Hazem Ali successfully defended his PhD dissertation entitled "Integrating Dataflow and Non-Dataflow Real-time Application Models on Multi-core Platforms" and became a doctor. The main contribution of the thesis is a design flow that integrates applications specified using two different application models, data-flow graphs and periodic real-time task sets, on a shared NoC-based multi-core platform. The committee recognized that the work was building bridges between two different communities in the world of real-time systems and that it was nice that the publications from the PhD work have been nicely integrated into a single design flow.

For those of you that are interested in a short summary of this work, please refer to his publication "Combining Dataflow Applications and Real-time Task Sets on Multi-core Platforms" accepted at the 2017 Workshop on Software and Compilers for Embedded Systems (SCOPES). Lastly, we wish Hazem the best of luck in his future career.

Yonghui Li Defends Dissertation

Posted on by bennyakesson

Today, we celebrate that Yonghui Li successfully defended his PhD dissertation “Design and Formal Analysis of Real-Time Memory Controllers” and became Dr. Li. The thesis defines a dynamically scheduled real-time memory controller architecture, which is implemented as a SystemC simulation model. It then continues by analyzing the worst-case response time and minimum guaranteed bandwidth using three different formal frameworks. The first framework is a mathematical formulation of both the actual and worst-case timing behavior as a set of equations and proofs of their correctness. These equations are also implemented in an open-source tool. The drawback of this kind of mathematical formulation is that it takes a long time to derive and prove correct. The second analysis approach addresses this by shifting the effort of the user from performance analysis to modeling the memory controller as a mode-controlled data-flow graph, which can be analyzed with existing tools. This approach is faster, but only bounds the minimum guaranteed bandwidth and not the worst-case response time. This limitation is overcome by the final approach, which is to model the memory controller using timed automata and bound its worst-case performance using a model checker. So, in summary, one controller architecture and three approaches to analyse its worst-case performance. This work hence gives unique insight into the strengths and weaknesses of different modeling and analysis approaches in terms of accuracy, expressiveness, memory consumption, and computation time.

The defense itself was well-prepared and confident and the committee seemed to really like the work. I am also really pleased with how it came out and I would like to thank Yonghui for the years of hard work that went into creating it. It was a pleasure to work with you during these years and I wish you all the best in your future career.

Sven Goossens Successfully Defended Dissertation

Posted on by bennyakesson

After successfully defending his dissertation “A Reconfigurable Mixed-Time-Criticality SDRAM Controller“, Sven Goossens earned himself a PhD degree and the right to call himself a doctor. The work proposes a pattern-based SDRAM controller targeting mixed-time-criticality systems, i.e. systems where some memory clients need firm worst-case guarantees on bandwidth and latency, while other clients only care about average-case performance. A new memory controller architecture is designed to address this mix of requirements and it is implemented both as a cycle-accurate SystemC simulation model and as synthesizable RTL code for generating FPGA instances. A unique feature of this memory controller is its conservative open-page policy that leaves rows open in the memory banks as long as possible to exploit locality and boost average-case performance, but closes them just in time to avoid reducing the worst-case performance.

The work also parameterizes the concept of memory patterns by allowing the number of banks and the number of bursts per bank to be chosen when the patterns are generated. This allows patterns with different degrees of bank-level parallelism to be created for six different generations of DRAM for any request size, enabling the user to make a trade-off between worst-case bandwidth, worst-case response time, and power consumption. To generate efficient memory patterns, the work proposes an integer linear programming formulation that provides optimal patterns, as well as a near-optimal heuristic that runs in a fraction of the time. In addition to generating predictable memory patterns that provide bounded bandwidth and execution times, composable read and write patterns can be generated with negligible performance loss. These patterns have equal length and can be used to provide complete temporal isolation between memory clients when combined with a non-work-conserving Time-Division Multiplexing (TDM) arbiter in the front-end. The memory patterns are generated offline at design time, but are programmed at run-time when the memory controller is initialized. Lastly, the proposed controller supports run-time reconfiguration of its TDM arbiter, allowing it to be safely reprogrammed when applications dynamically start and stop at run-time without sacrificing the worst-case guarantees of applications that keep running.

I would like to thank Sven for the five years of hard work. It has been a pleasure to work with such a versatile and independent young researcher who seems to be succesful at whatever he attempts, be it design, analysis, writing papers, or hardware/software implementation in more or lesss any language. He has also been an excellent member of the Memory Team and the larger CompSoC Team, never passing on an opportunity to use his skills to support other members of the team. At the end of January, Sven starts his new career with Intrinsic-ID in Eindhoven. We wish him the best of luck with his new job and hope to stay in touch.