Visiting Professor at Cochin University (CUSAT), Kerala, India
09/2009 - 2/2011
Freie Universität Berlin (Berlin, Germany) Senior research associate at Computer Systems & Telematics Group, Institute of Computer Science
07/2009 - 08/2009
Senior research associate at Newcastle University, UK
Habilitation in Computer Science at Humboldt-University Berlin
04/2002 - 06/2009
Assistant professor (C1) at Humboldt-University Berlin, Computer Architecture and Communication group, Institute of Computer Science
05/2000 – 08/2000
Visiting researcher at Hewlett-Packard Laboratories, Palo Alto, CA, USA.
PhD in Computer Science at TU Berlin.
07/1997 - 03/2002
DFG project 'Modelling with fluid Petri nets' at TU Berlin, as part of the DFG priority program KONDISK.
Member of the Graduiertenkolleg (graduate course) Communication-based Systems at FU, HU and TU Berlin.
10/1991 – 08/1995
Technische Universität Berlin (Berlin, Germany), studies of Computer Science, Diploma in Computer Science with Statistics.
My Erdös number is 3 (through Miroslaw Malek and Frank Harary).
I am interested in measuring and evaluating the dependability, performance, and security of complex computing systems, with a particular focus on timing behaviour. Within our group we employ a broad range of assessment and evaluation techniques for computing systems and networks, ranging from fault-injection test-beds to simulation and analytical techniques. We develop efficient and accurate modelling and evaluation techniques, applying e.g. Phase-Type distributions in fault-modelling for fault-injection experiments and hybrid discrete-event simulation. We study a large variety of systems, including wireless networks, mobile telephony networks, service-oriented systems, and Computing Clouds.
We conduct extensive statistical analysis of data collected from test beds as well as for medical data.
A tool for fitting phase-type distributions to empirical data.
A tool for fitting MAPs to empirical data, also to correlated data.
Assessment of the service quality of streaming systems, improvement of their fault-tolerance and scalability.
Mobile Application Offloading
Mobile offloading is a collaborative technique for running compute intensive tasks partly on a mobile device and partly on a server, or in the cloud. We investigate fault-tolerant policies for when and what to offload.
We investigate the use and impact of the restart method in distributed environments. Through analysis of stochastic models we aim to confirm the fairness property of the restart algorithm and investigate the timing behaviour of network in which several clients employ restart.
High-Resolution Clock Synchronisation over Packet-Switched Networks
In cooperation with highstreet technologies we analyse timing behaviour of complex packet-switched networks through detailed simulation using the network simulator ns-2.
Efficient Fault-Models for System Evaluation
In cooperation with Prof. Miklos Telek and Prof. Gabor Horvath of the Technical University of Budapest we develop techniques for efficent and accurate stochastic fault-models. We focus on random-number generation from Phase-type and Matrix-Exponential Distributions. This cooperation is supported by the DFG.
We collaborate (past and ongoing) among others with
Rena Bakhshi, Vreje Universiteit Amsterdam, The Netherlands (funded by DFG)
Jean-Michel Fourneau, University of Versailles, France (funded by DAAD)
William Knottenbelt, Imperial College London, UK
Miklos Telek, Technical University Budapest (funded by DFG)
Gabor Horvath, Technical University Budapest (funded by DFG)
I teach courses on Dependable Systems, Model-based Evaluation of Computing Systems, and Distributed Systems in the Master's program and Mathematics and Computer Architecture in the Bachelor's program.
Bachelor- and Masters' students
If you write a thesis with me, please provide me with a copy of the letter indicating your deadline. You may use this latex-template for writing your thesis.
Several topics in the area of evaluating performance, dependability and security evaluation are available upon request. These include experimentation and measurement studies, development of stochastic models, measurements in networks and the Cloud and implementation of algorithms. Modelling related topics are in the area of queueing models with negative customers and phase-type fitting.