Organized by GI/ITG Technical Committee on "Measurement, Modelling and Evaluation of Computing Systems (MMB)"


Reinhard German (FAU Erlangen)


Kai-Steffen Hielscher (FAU Erlangen)
Marco Pruckner (FAU Erlangen)


Markus Siegle (UBw Munich)


Udo Krieger (U Bamberg)


Anatoli Djanatliev (FAU Erlangen)


Thomas Herpel (IEE)


Robert Basmadjian (U Passau)
Marcel Baunach (TU Graz)
Peter Buchholz (TU Dortmund)
Hans Daduna (U Hamburg)
Markus Fidler (U Hannover)
Anna Förster (U Bremen)
Reinhard German (U Erlangen)
Gerhard Haßlinger (Dt. Telekom AG)
Boudewijn Haverkort (EWI/U Twente)
Tobias Hoßfeld (U Dusiburg-Essen)
Holger Hermanns (U Saarbrücken)
Joost-Pieter Katoen (RWTH Aachen)
Peter Kemper (Coll. of William and Mary)
Samuel Kounev (U Würzburg)
Udo Krieger (U Bamberg)
Kai Lampka (Elektrobit Automotive)
Wolfram Lautenschläger (Nokia)
Ralf Lehnert (TU Dresden)
Hermann de Meer (U Passau)
Michael Menth (U Tübingen)
Peter Reichl (U Wien)
Anne Remke (U Münster)
Johannes Riedl (Siemens AG)
Oliver Rose (UBw Munich)
Ramin Sadre (U Catholique de Lovain)
Jens Schmitt (TU Kaiserslautern)
Markus Siegle (UBw Munich)
Helena Szczerbicka (U Hannover)
Andreas Timm-Giel (TU Hamburg)
Dietmar Tutsch (U Wuppertal)
Kurt Tutschku (BTH Karlskrona)
Oliver Waldhorst (HsKA Karlsruhe)
Max Walter (Siemens AG)
Verena Wolf (U Saarbrücken)
Bernd Wolfinger (U Hamburg)
Katinka Wolter (FU Berlin)
Armin Zimmermann (TU Ilmenau)

Tuesday, Wednesday

Time Monday, February 26, 2018
8:30 Start of registration
9:00-10:30 Parallel Tutorials
Joost-Pieter Katoen, Matthias Volk (RWTH Aachen Univ.): A Modern Perspective on Fault Tree Analysis I, Room 01.254-128 Florian Metzger, Tobias Hoßfeld (Univ. Duisburg-Essen): IoT - From Praxis To Theory I,
Room 01.255-128
10:30-11:00 Coffee Break
11:00-12:30 Parallel Tutorials
Joost-Pieter Katoen, Matthias Volk (RWTH Aachen Univ.): A Modern Perspective on Fault Tree Analysis II,
Room 01.254-128
Florian Metzger, Tobias Hoßfeld (Univ. Duisburg-Essen): IoT - From Praxis To Theory II,
Room 01.255-128
Natalia M. Markovich (Russian Academy of Sciences),
Udo R. Krieger (Univ. Bamberg): Data Analysis of Measurements Governed by Immanent Dependences and Heavy-Tailed Distributions,
Room 01.253-128
12:30-13:30 Lunch Break
13:30-13:45 Opening Session
Reinhard German (FAU), Felix Freiling (Head of Computer Science Dep., FAU),
Room H12
13:45-14:45 Invited Talk I
Hans van den Berg (TNO, University of Twente, CWI),
Performance Optimization of 5G Mobile Networks, Room H12
14:45-15:15 Coffee Break
15:15-16:45 Session 1: Analysis of Markov Models, Room H12
Peter Buchholz, Iryna Dohndorf, Dimitri Scheftelowitsch: Time-Based Maintenance Models under Uncertainty
Yuliya Butkova, Ralf Wimmer, Holger Hermanns: Markov Automata on Discount!
Gerhard Hasslinger: Markov Analysis for Optimum Caching as an Alternative to Belady’s Algorithm
16:45-17:15 Coffee Break
17:15-18:15 Session 2: Security & Energy, Room H12
Justyna Chromik, Benedikt Ferling, Anne Remke, Marco Caselli: Intrusion Detection for sequence-based attacks with reduced traffic models
Robert Basmadjian, Yashar Ghiassi-Farrokhfal, Arun Vishwanath: Hidden Storage in Data Centers: Gaining Flexibility Through Cooling Systems
19:30-22:00 Welcome Reception, Wassersaal der Orangerie

Time Tuesday, February 27, 2018
9:00-10:00 Invited Talk II
Hartmut Schmeck (KIT): Future Energy Grids – Challenges and Potential for ICT, Chair: TBD, Room H12
10:00-11:00 Session 3: Tool Presentations, Chair: Markus Siegle, Room H12

Tool Demonstrations, Room: TBD
Freek van den Berg, Jozef Hooman, Boudewijn Haverkort: A Domain-Specific Language and Toolchain for Performance Evaluation based on Measurements
Falko Bause, Peter Buchholz: SLA Tool
Max Kerkers, Justyna Chromik, Anne Remke, Boudewijn Haverkort: A Tool for Generating Automata of IEC60870-5-104 Implementations
Tugrul Dayar: A Software Tool for the Compact Solution of the Chemical Master Equation
Christoph Brandau, Dietmar Tutsch: Logical PetriNet - A Tool to Model Digital Circuit Petri Nets and Transform them into Digital Circuits
Florian Heimgaertner, Thomas Sachs, Michael Menth: ClassCast: A Tool for Class-Based Forecasting
Dimitri Scheftelowitsch: collider – Parallel experiments in silico
Frederik Hauser, Dominik Krauß, Michael Menth: FunSpec4DTMC - A Tool for Modelling Discrete-Time Markov Chains Using Functional Specification
Anna Deitsch, Vitali Schneider: Model-based System Design and Evaluation of Image Processing Architectures with SimTAny Framework
11:00-11:30 Coffee Break
11:30-13:00 Session 4: Special Session on Software Defined Networking, Room H12
Alexej Grigorjew, Stanislav Lange, Thomas Zinner, Phuoc Tran-Gia: Performance Benchmarking of Network Function Chain Placement Algorithms
Sounak Kar, Rhaban Hark, Amr Rizk, Ralf Steinmetz: Towards Optimal Placement of Monitoring Units in Time-varying Networks under Centralized Control
Anh Nguyen-Ngoc, Stanislav Lange, Stefan Geißler, Thomas Zinner, Phuoc Tran-Gia: Impact of Control Plane Delay on FlowMod Processing Times in OpenFlow Switches
13:00-14:00 Lunch Break
14:00-15:30 Session 5: Industrial and Practical Experience & PH.D. track, Chair: Thomas Herpel, Room H12
Ermin Sakic, Vivek Kulkarni, Vasileios Theodorou, Anton Matsiuk, Simon Kuenzer, Nikolaos E Petroulakis, Konstantinos Fysarakis: VirtuWind - An SDN- and NFV-based Architecture for Softwarized Industrial Networks
Jannik Hüls, Anne Remke: A modular environment to test SCADA solutions for wind parks
Thomas Deinlein, Reinhard German, Anatoli Djanatliev: Evaluation of Single-Hop Beaconing with Congestion Control in IEEE WAVE and ETSI ITS-G5
Sebastian Surminski, Christian Moldovan, Tobias Hossfeld: Practical QoE Evaluation of Adaptive Video Streaming
15:30-16:00 Coffee Break
16:00-16:45 Invited Industrial Talk
Frank Keck (ZF Zukunft Mobility)
Autonomous Driving - the "uncrashable" car? What it takes to make self-driving vehicles safe and reliable traffic participants
Room H12
16:45-17:30 Organized bus transfer to Nuremberg
17:30-19:30 Guided tours in the castle of Nuremberg
19:30-22:00 Conference dinner, Bratwurst Röslein, Rathausplatz 6, 90403 Nuremberg
(Walk to the restaurant from the castle ca. 5 min., ca. 400 m)
22:00 Organized bus transfer back to Erlangen

Time Wednesday, February 28, 2018
9:00-10:30 Session 6: Service Scaling, Room H12

Parallel Workshops:

WoNeCa, Room 01.150-128

Room 01.255-128
Esa Hyytiä, Douglas Down, Pasi Lassila, Samuli Aalto: Dynamic Control of Running Servers
André Bauer, Johannes Grohmann, Nikolas Herbst, Samuel Kounev: On the Value of Service Demand Estimation for Auto-Scaling
Alexander Rumyantsev, Polina Zueva, Ksenia Kalinina, Alexander Golovin: Evaluating a Single-Server Queue with Asynchronous Speed Scaling
10:30-11:00 Coffee Break
11:00-12:30 Session 7: Network Performance Modeling, Room H12
Michael Menth, Sebastian Veith: Active Queue Management Based on Congestion Policing (CP-AQM)
Michael Menth, Marcel Mehl, Sebastian Veith: Deficit Round Robin with Limited Deficit Savings (DRR-LDS) for Fairness among TCP Users
Udo Krieger, Krishna Kumar: Modeling the Performance of ARQ Error Control in a LTE Transmission System
12:30-13:30 Lunch Break
13:30-14:30 Session 8: Network Calculus & Quality of Experience, Room H12
Steffen Bondorf, Paul Nikolaus, Jens Schmitt: Catching Corner Cases in Network Calculus – Flow Segregation Can Improve Accuracy
Tobias Hossfeld, Martin Varela, Poul Heegaard, Lea Skorin-Kapov: QoE Analysis of the Setup of Different Internet Services for FIFO Server Systems
14:30-15:30 Session 9: MMB Awards & Closing, Chair: Udo Krieger, Room H12
MMB Master Thesis Award
MMB PH.D. Thesis Award
Best Paper Award MMB


A Modern Perspective on Fault Tree Analysis
Joost-Pieter Katoen, Matthias Volk (RWTH Aachen Univ.)

Abstract: Fault trees are a key technique in safety and reliability engineering. Their application includes aerospace, nuclear power, car, and process engineering industries. Various fault tree extensions exist that increase expressiveness while yielding succinct models. Their analysis is a main bottleneck: techniques do not scale and require substantial manual effort. Formal methods have an enormous potential to solve these issues. We discuss a mixture of techniques rooted in formal methods and tailored to the needs and specifics of fault trees. The result is a fully automated and scalable approach to analyze Dugan's dynamic fault trees. The tutorial will be structured in four blocks:

1. What are DFTs?
- the basics of dynamic fault trees (DFTs)
- uncovering the intricacies of DFTs
- DFT semantics in terms of stochastic Petri nets
2. From DFT to Markov models
- compositional state-space generation
- a plethora of state-space reduction techniques (bisimulation, partial-order reduction, don't care propagation, partial state-space generation, fault tree simplification)
3. DFT analysis by model checking
- model checking Markov models
- from DFT analysis to model checking
4. Experiments
- a software tool (demo)
- industrial applications (railway, aerospace, and automotive)

IoT - From Praxis To Theory
Florian Metzger, Tobias Hoßfeld (Univ. Duisburg-Essen)

Abstract: The prevalence of IoT is driven by industrial requirements and scales, but also by community curiosity and tinkering in participatory crowdsensing endeavours. This tutorial first explores the practical requirements and options of modern IoT appliances and projects, including all aspects of the diverse stack, from PHY to application. With that as base, traffic models can now be derived and evaluated for these IoT topologies that might provide a better fit than traditional approaches.

List of topics:
• Purpose and use cases of IoT (industrial, home automation, community)
• From BAN to LPWAN: an overview of communication standards
• Networking stack overview (from raw PHY to 6LowPAN and CoAP)
• Developments of cellular IoT and the 3GPP ecosystem
• IoT network topologies
• Participatory Crowdsensing with IoT
• Practical Sensing Demo with LoRa
• IoT traffic models: standards, Poisson approximation and bias

Data Analysis of Measurements Governed by Immanent Dependences and Heavy-Tailed Distributions
Natalia M. Markovich (Russian Academy of Sciences), Udo R. Krieger (Univ. Bamberg)

Abstract: Modern tools to measure Internet traffic such as Wireshark or Atheris offer complex opportunities to collect packet data from high speed networks. Advanced statistical methods are required to support an adequate teletraffic analysis of these traces and the evaluation of relevant performance indices, for instance, of captured packet flows stemming from new multimedia services in Internet. They allow us to cope with immanent dependencies and underlying heavy-tailed distributions of interesting features of the traffic such as the bitrates, volumes or lengths of sessions, the inter-arrival times, loss rates and delay distributions of the packet streams or their equivalent bandwidth.

In the tutorial we shall discuss useful statistical techniques to handle the arising strongly correlated or long-range dependent time series and heavy-tailed marginal distributions. The latter features characterize the underlying random variables of the observed data. Advanced procedures to compute the demanded bandwidth of observed streams or the delay-loss profiles of packet flows during a session will be stated. The analysis concepts will be illustrated by real traces arising from some popular Internet applications.

The tutorial shall stimulate the participants to incorporate adaptations of the sketched statistical procedures into open source tools or their own codes according to their personal needs.

Invited Talks

Invited Talk 1 (Monday, 13:45-14:45, Chair: TBD, Room H12)
Performance Optimization of 5G Mobile Networks
Hans van den Berg (TNO, University of Twente, CWI)

Abstract: Research on 5G in Europe is boosted by the 5G PPP consortium consisting of network vendors and operators, system integrators and academia and other research institutes, working closely together with companies from important vertical industries. 5G aims at bringing new, distinctive network and service capabilities fulfilling the needs of the future Internet of Things (IoT). As such it should sustain enormous data volumes and support critical, highly demanding communication services for e.g. self-driving cars, robotics in smart industry, and mobile virtual reality applications. Network ‘softwarization’ through emerging technologies as Software Defined Networking (SDN) and Network Functions Virtualization (NFV) is introduced to provide the flexibility needed to reach the required performance and scalability targets in an efficient way. However, to actually achieve the full potential of future 5G networks huge challenges regarding network management and performance optimization are faced. Big data techniques exploiting data coming from network devices in forms of e.g. device logs and usage histories provide a promising direction to address these challenges. In the talk we will briefly sketch the 5G PPP ambitions, review the aforementioned research challenges and present (ongoing) work on some specific 5G network performance optimization problems.

Curriculum Vitae (Prof. Dr. Hans van den Berg): Hans van den Berg has more than 25 years of experience in ICT research and innovation. His main contributions are in the field of design and performance optimization of (mobile) communication networks and services. He has been active in many national and European research projects and platforms (FP3-FP7, COST, ITEA) and acted as coordinator of the FP7 project SOCRATES (2008-2011) on self-management of 4G mobile networks. He was co-founder and vice-chair of the recently finished COST Action IC1304 “Autonomous Control for a Reliable Internet of Services” (COST ACROSS, 2013-2017). He has published more than 150 refereed papers in international journals and conference proceedings, and is co-editor of several books. In 2017 he received ITC's Arne Jensen Lifetime Achievement Award for his contributions in the area of network performance modeling and analysis. Hans van den Berg started his professional career at the research labs of the Dutch telecom operator KPN in 1990. Since July 2005, Hans van den Berg is principal scientist at TNO and holds a part-time position as full professor within the research group Design and Analysis of Communication Systems at the University of Twente. Since 2016 he is also affiliated with the Centre for Mathematics and Computer Science (CWI) in Amsterdam.

Invited Talk 2 (Tuesday, 9:00-10:00, Chair: TBD, Room H12)
Future Energy Grids – Challenges and Potential for ICT
Hartmut Schmeck (KIT)

Abstract: The energy system is one of the most critical infrastructures of our world. The reliable supply of energy is essential for the adequate operation of almost any process in our private and professional life. Society and industry would suffer enormously, if the steady balance between demand and supply could not be guaranteed. The current transition towards energy from renewable sources is having tremendous effects on this well-established infrastructure. In particular, the restricted capabilities of controlling the supply of electricity from weather-dependent energy sources leads to the need for an essential change in one of the basic principles of the electric power system, which means that it will no longer be feasible to let the power supply follow the demand but there will be a strong need to let the demand follow the supply. This can only be achieved by discovering and exploiting the potential of flexibility of demand and supply in the best possible way. The talk will illustrate how the major challenges of the ongoing energy transition create the need for an adequately designed energy information and control network with distributed intelligence. A fundamental task in the design of this network consists of making the necessary information available to the locations where operating and control decisions have to be taken and to provide appropriate methodology for managing tomorrow’s energy system in the most efficient and most reliable way. In particular, an assessment of the potential contribution of flexibility in demand and supply to guaranteeing the necessary stability and resilience needs appropriate modelling and simulation, based on effective strategies for measuring the current status and behaviour of relevant grid components.

Curriculum Vitae (Dr. Hartmut Schmeck): Hartmut Schmeck studied at the Universities of Kiel (Germany) and Waterloo (Canada). He got his academic degrees (Dipl. Inform., Dr.rer.nat., Dr. habil) at Kiel. Since 1991 he is a Full Professor of Applied Informatics at the Karlsruhe Institute of Technology - KIT. He is (co-)author of more than 140 publications on advanced algorithms and architectures, in particular on nature-inspired methods in optimisation, algorithms for reconfigurable architectures, and on self-organising, adaptive systems applied to energy and traffic systems. He has been program and conference chair for numerous international workshops and conferences (a.o. RAW, ARCS, IFIP BICC 2006, 2008, ATC 2009, ICAC 2011, D-A-CH Energieinformatik) and coordinator of the German priority research program SPP 1183 on “Organic Computing”. As a principal investigator of several cooperative projects in various funding programmes he is pushing the development of intelligent systems in tomorrow's energy systems and for electric mobility, shaping the new discipline of "Energy Informatics", in particular as a director of the FZI Research Center for Information Technology. In 2016, his achievements were honoured with the Heinrich-Hertz Prize from the EnBW foundation.

Invited Industrial Talk (Tuesday, 16:00-16:45, Chair: TBD, Room H12)
Autonomous Driving - the "uncrashable" car? What it takes to make self-driving vehicles safe and reliable traffic participants
Frank Keck (ZF Zukunft Mobility)

Abstract: Autonomous driving is in the spotlight of both scientific research and industrial development. With worldwide constantly growing traffic volumes, the challenging task in putting self-driving vehicles onto the street is to cross the chasm between high system availability and low to zero malfunction rates, even in dense traffic and complex situations on the road. Developing software functions for driver assistance and vehicle safety for autonomously driving cars requires the traditional developments processes and methods to be revised. In this talk, a novel and promising development approach is presented. The combination of use case based requirement specification, algorithm development with machine learning techniques and both simulation based and real-life testing yields an agile yet sound software development framework for autonomous driving functions. Additionally, some thought-provoking impulses are given on how to achieve a high level of system reliability by exploiting the capabilities of virtualization at early development stages.

Curriculum Vitae (Dr. Frank Keck): Frank Keck was born in 1973. In 1998, he received a diploma degree in Physics from the University of Kaiserslautern, Germany. From 1999 to 2002 he was a PhD candidate at the Chair of Theoretical Physics at the University of Kaiserslautern, Germany and was awarded a doctorate in 2002. In 2002 and 2003, he worked at the Physics Department at the University of Kaiserslautern, Germany in a postdoc position. In 2003, he joined the AUDI AG, Ingolstadt, where he was developing car safety algorithms for crash, pedestrian and rollover, occupant classification systems and advanced (foresighted) safety systems until 2009. From 2009 to 2015, he was CEO of Automotive Safety Technologies GmbH, Gaimersheim, Germany and in January 2016, he held the position of the Integral Safety Manager at PSW engineering GmbH, Gaimersheim, Germany. Since February 2016, he is CEO of Zukunft Mobility GmbH – formerly IEE Sensing Germany GmbH – which is a company of ZF Friedrichshafen AG. His experiences and technical interests cover car safety, integral safety systems, advanced driver assistance systems, machine learning, data mining, mathematical and physical modelling.