S Technische Informatik (Security, Robustness, Reliability)
This seminar is a block seminar, i.e., at the end of the semester, there will be one or two days when all talks are given in a block. During the semester, there will be deadlines for status reports, but no meetings of the complete seminar group.
Students participating will give a talk (20 min presentation + 10 min discussion) and write a report (15-20 pages as a general rule, see templates). Besides that, students will review one or two reports written by other students and prepare questions to be asked after the corresponding talk. This is to stimulate discussions after the talks and make the seminar more worthwhile. This course places special emphasis on dealing competently with academic and primary literature, structure and legibility of the report, and the style of the presentation (50% of final grade).
|Dozent/in||Prof. Dr.-Ing. Jochen Schiller, Prof. Dr. Katinka Wolter|
|Institution||Institute of Computer Science|
Freie Universität Berlin
|Raum||Arnimallee 3 SR 005|
Vordiplom or B.Sc., basic knowlegde in telematics
- Organizational Meeting - 14:00 Uhr (c.t.), room 005, Arnimallee 3
- Read the literature linked below.
- Very last day to choose a topic via e-mail to Jochen Schiller.
- Hand in a preliminary outline and reference list to your advisor via e-mail.
- Final version of your report must be handed in to your advisor and Jochen Schiller via e-mail.
- Final version of your slides must be handed in to your advisor and Jochen Schiller via e-mail.
01.07. (10-18, SR 053):
- The seminar will take place. Attendance is mandatory.
Attention: Students have to meet all deadlines listed in the schedule. Otherwise, she/he will lose the right to take part in the final presentation.
Schedule of final presentations
|10:00-10:30||Compare real military fence monitoring system with FU research||Fabian Schneider||Norman Dziengel|
|10:30-11:00||Data reduction with statistical approaches to ensure data quality in WSNs||Simon Stapelfeld||Norman Dziengel|
|11:00-11:30||How to communicate if all infrastructures break down?Democratic justice?||Daniel Andreas Happ||Jochen Schiller|
|11:30-12:00||A Survey of MANET Security||Martinus Dipobagio||Katinka Wolter|
|12:00-12:30||Are mobile ad-hoc networks really more robust?||Damla Durmaz||Jochen Schiller|
|12:30-13:00||WSNs in critical infrastructures||Ferhat Beyaz||Stephan Adler|
|14:00-14:30||Detecting when interdomain routing goes wrong||Franz Schellenberg||Matthias Wählisch|
|14:30-15:00||Traffic analysis attacks on Skype VoIP calls||Moritz v. H.||Matthias Wählisch|
|15:00-15:30||Is voting a fair way to determine intruders in MANETs?||Maximilian Porsch||Sebastian Trapp|
|15:30-16:00||Radio Interference Detection and Multi-channel media access control in Wireless Sensor Networks||Daniel Lopez Garcia||Marco Ziegert|
|16:30-17:15||Energieeffiziente Netzwerktechniken und deren Auswirkungen||Tomasz Jacewski und Julian Schueler||Katinka Wolter|
|17:15-17:45||Quantification of Security||Lars Ries||Katinka Wolter|
|17:45-18:15||Critical Infrastructures||Stephan Zeisberg||Katinka Wolter|
Advisor: Jochen SchillerHow to communicate if all infrastructures break down?
In case of an earthquake, a large flood etc. classical communication systems like the wired telephone but also infrastructure based mobile communication systems simply break down due to cut wires, power outages, tilted antennas etc. What type of communication system can survive such disaster and thus support emergency response teams? How do they work, what are the limitations?
Assigned to: Daniel Andreas Happ
Security and robustness of VoIP
More and more voice calls use the Internet, i.e. VoIP, including signaling mechanisms (SIP). However, as it is the Internet it is not that complicated to attack VoIP, to create SPIT (Spam over Internet Telephony) etc. The contribution should give an insight into the robustness and vulnerabilities of VoIP including its signaling mechanisms.
Assigned to: Lars Reichelt
Do the new mobile phone networks provide more robust networks? New systems like LTE, LTE advanced and new radio technologies like MIMO, cognitive radio, software defined radio etc. can provide not only higher data rates but can also provide an additional infrastructure, use addition frequencies etc. Do they also contribute to an overall more robust mobile network? Assigned to: Alexander Bach
Are mobile ad-hoc networks really more robust? In theory and in many simulations authors prove that mobile ad-hoc networks are more robust with respect to node failures, link interruptions etc. However, setting up a real mobile ad-hoc network shows that real life challenges many theories: radio propagation is much more complicated, links are asymmetric, heavy fluctuations in link quality cause routing to fail etc. This topic should address important issues of real mobile ad-hoc networks and compare them with theory. Assigned to: Damla Durmaz
Advisor: Matthias WählischHow secure are secure interdomain routing protocols? In the current Internet, inter-domain routing is implemented by BGP. A BGP router advertises IP prefixes that it is responsible for. The security model is based on mutual trust between BGP speakers, which implicates several security problems. On the one hand no one prevents a BGP speaker to announce an arbitrary prefix, on the other hand there is no straightforward way to shield BGP routers from invalid updates. Consequently, BGP extensions have been introduced to overcome these limitations. However, they are not globally deployed. Based on current research you should summarize how secure are secure interdomain routing protocols. A good starting point is a paper by Goldberg et al. recently published at ACM SIGCOMM 2010.
Assigned to: Yves Müller
Detecting when interdomain routing goes wrong The Internet backbone routing is quite vulnerable. For example, in February 2008, one of the most prominent cases occurred when Pakistan Telecom started an unauthorised announcement of an IP prefix that belongs to YouTube. This incorrect update has been globally propagated. Although Pakistan aimed to block the YouTube website only locally, this accident interrupted YouTube traffic around the world. This work should give a summary about current systems that intend to detect BGP routing problems. You should focus on lightweight solutions that do not require knowledge in advance, but detect problems automatically and adapt to the dynamic of the Internet. A good starting point is the paper by Haeberlen et al. published at USENIX NSDI 2009. Assigned to: Franz Schellenberg
Traffic analysis attacks on Skype VoIP calls Skype is a popular software that allows for text, voice and more recently video chat, because it is free and simple. The increased performance of mobile phones and the available Internet connectivity make it quite interesting to use this software on mobiles, as well. The success of Skype nevertheless depends also on its privacy protection functions. This work should present threats that are based on traffic analysis. The insights should be analyzed in the context of lightweight mobile devices (e.g., smartphones). A good starting point is the paper by Zhu and Fu, which is published in Elsevier Computer Communications, December 2010. Assigned to: Moritz v. H.
Advisor: Marco ZiegertRadio Interference Detection and Multi-channel media access control in Wireless Sensor Networks
Due to their limited communication ability and smaller packet sizes Wireless Sensor Networks mainly focus on single frequency solutions compared to wireless ad-hoc networks. We will compare and focus on two approaches to avoid collisions, improve energy efficiency, or to enhance data rates:
- RID-B is based on the fact that the communication range of sensor nodes is different to the interference range and optimizes TDMA protocol for Single-channel environments to avoid packet loss.
- Multi-channel media access control protocols for Wireless Sensor Networks utilize the ability to enhance parallel transmission among neighboring nodes.
Time Synchronization in Wireless Sensor Networks In large environments, it may be useful to capture events not only by their spatial signal propagation but also their exact time of occurrence. This information can enhance distributed event and pattern detection in Wireless Sensor Networks. Due to time-of-flight delays and the requirement of affordable components used by the sensor nodes, the time synchronization in a WSN may allegorize an expensive feature. The aim is to compare different approaches to their required infrastructure, additional energy requirements, component cost and implementation. Assigned to: Annika Gellert
Advisor: Stephan AdlerWSNs in critical infrastructures Today's world is full of autonomous computer systems assisting us in our daily tasks. Some of these tasks are considered 'critical' because a malfunction could have serious consequences such as endangering people's health or causing financially loss. These critical systems can be found in the domain of health care, transportation and finance. Your task is to find out whether wireless sensor networks are used in such systems, how they are secured and what potential tasks could be solved using WSNs. Your adviser will provide you publications to start with.
Assigned to: Ferhat Beyaz
Advisor: Katinka WolterCritical infrastructures (Wolter) – OK for 2 students!
- S. Chiaradonna, P. Lollini, F. Di Giandomenico, On a modeling framework for the analysis of interdependencies in electric power systems, in: EEE/IFIP 37th Int. Conf. on Dependable Systems and Networks (DSN 2007), 2007, pp. 185-195.
- S. Chiaradonna, F. Di Giandomenico, P. Lollini, Assessing the impact of interdependencies in electric power systems, Int. J. System of Systems Engineering, Vol. 1, No. 3, 2009, pp. 367-386
- S. Chiaradonna, P. Lollini, F. Di Giandomenico, Modelling framework of an instance of the electric power system: Functional description and implementation, Tech. Rep. RCL071202 - Version 4, University of Florence, Dip. Sistemi Informatica, RCL group, http://dcl.isti.cnr.it/ Documentation/Papers/Techreports.html (March 2010).
- This work evaluates the availability of the electrical power system under different kinds of disturbances within this large distributed system. The work was done in the context of the EU project CRUTIAL.
Quantification of security
- Starting from the fact that systems are never 100% secure it becomes important to measure and evaluate the degree of security of a system. This requires a very new line of thinking and constitutes a new field of research. The issues being addressed are different system configurations that can be more or less secure as well as characterizing attacker behavior.
- Adversary-Driven State-Based System Security Evaluation. E. LeMay, W. Unkenholz, D. Parks, C. Muehrcke, K. Keefe, and W. H.Sanders. (10VAN02), Proceedings of the 6th International Workshop on Security Measurements and Metrics (MetriSec 2010), Bolzano-Bozen, Italy, Sept. 15, 2010.
- James H. Cowie, David M. Nicol, Andy T. Ogielski, "Modeling the Global Internet," Computing in Science and Engineering, vol. 1, no. 1, pp. 42-50, Jan./Feb. 1999, doi:10.1109/5992.743621
- On inferring autonomous system relationships in the Internet, Lixin Gao; IEEE Transactions on Networking, Vol. 9, Issue 6, Dec. 2001, pp. 733 - 745.
- Power laws and the AS-level Internet topology, Siganos, G.; Faloutsos, M.; Faloutsos, P.; Faloutsos, C.; IEEE Transactions on Networking, Vol. 11, Issue 4, Aug. 2003, pp. 514 - 524.
- A plethora of articles in conferences, journals, magazines
Modelling of network malware
- Simulating realistic network worm traffic for worm warning system design and testing M Liljenstam, DM Nicol, VH Berk, RS Grey - Proceedings of the 2003 ACM workshop on Rapid malcode
- Worm attacks are repeatedly attracting a lot of public attention. This work describes the behavior of worms such as Code Red v2 and Sapphire/Slammer and proposes models to produce worm behavior that also allows to detect worms in a system.
- Efficient simulation of Internet worms, David Nicol, ACM Transactions on Modeling and Computer Simulation (TOMACS)
- Simulating realistic network worm traffic for worm warning system design and testing M Liljenstam, DM Nicol, VH Berk… - Proceedings of the 2003 …, 2003 - portal.acm.org
- Security can be achieved through encryption using a key. The longer a key is being used the more likely decryption is achieved by an attacker. At the same time agreeing on a new key requires computing resources. This work tries to evaluate the cost/benefit balance of rekeying in mobile adhoc networks.
- Performance Analysis of Dynamic Group Communication Systems with Intrusion Detection Integrated with Batch Rekeying in Mobile Ad Hoc Networks Jin-Hee Cho and Ing-Ray Chen and Phu-Gui Feng
Advisor: Marcel Kyas Approximate non-interference, formal verification of security
- Aldini, A. Di Pierro: "Estimating the Maximum Information Leakage" - pdf International Journal of Information Security 7(3):219-242, Springer, June 2008.
- Aldini, A. Di Pierro: "Noninterference and the Most Powerful Probabilistic Adversary" - ps 6th Int. Workshop on Issues in the Theory of Security (WITS'06), Vienna (Austria), March 2006.
- Aldini, M. Bernardo: "A Formal Approach to the Integrated Analysis of Security and QoS" - pdf Journal of Reliability Engineering & System Safety 92(11):1503-1520, Elsevier, November 2007.
Advisor: Norman DziengelCompare real military fence monitoring system with FU research
On construction sites and around other security relevant areas, fences are typically used to protect expensive goods. Wireless Sensor Networks may help to deliver a more flexible system to be attached at the fence itself. Different types of sensors need to be applied on an appropriate sensor board to detect possible intruders. Compare the FU research results with a real world military system that seems to cover our requirements. Research all available sources to get a deep insight into both hardware components, the technical details and the supported features of both systems. Evaluate performance, energy-consumption and state of the art.
Primary literature: please contact advisorAssigned to: Fabian Schneider Influence of Data Quality in Wireless Sensor Networks
Data quality is influenced by a huge amount of factors like sensor fabrication, energy supply, model-design and environmental effects. It is a good idea to discriminate between four components of data quality: accuracy, consistency, timeliness and completeness. These components influence unequally the total data quality depending on the way of communication. Wireless sensor networks (WSNs) underlay restriction like limited energy supply which leads to a restricted computational power, restricted memory and lifetime. Analyze the influence of data quality in Wireless Sensor Networks by taking in advance data gathering and communication.
Primary literature: http://downloads.hindawi.com/journals/wcn/2010/698951.pdfAssigned to: Marc Essien Data reduction with statistical approaches to ensure data quality in WSNs
While optimizing a data stream for wireless communication by compression we reduce data quality due to an inherent loss of information. How far can the data stream be reduced to get an ideal ratio between reduction of data and its representativeness? One option to reduce data in a stream is based on histogram. For this approach we pick random data out of each histogram class and restore the original sample order afterwards. The question is now, how good will this new dataset match the original stream. Techniques for evaluating the losses vs. benefits should be found and compared with the known like Kolmogorov-Smirnov test. Are there other approaches that may leverage the possible data quality?
Primary literature: http://www.inf.ufrgs.br/~buriol/papers/Lins_AINA07.pdfAssigned to: Simon Stapelfeld Consistency driven data quality in WSNs
Wireless Sensor Networks often process and transmit data that have been collected through a varying range of sensors. Depending on the size, fabrication quality, energy supply and environmental roughness the inherent confidence should be known. One option to achieve this goal is by designing and implementing a appropriate model that is able to check the consistency of the collected data. We distinguish numerical consistency, temporal consistency and frequency consistency. Analyze the techniques concerning the usability in WSNs and find other approaches to validate the data consistency of a sensor node. Compare different sensors available at the market and evaluate their symbiotic effect to one another.
Primary literature: Book: Data Quality Management in Wireless Sensor Networks - Kewei ShaAssigned to: Michael Borst
Advisor: Sebastian Trapp Democratic justice? - Is voting a fair way to determine intruders in MANETs?
In mobile ad-hoc networks (MANETs) intrusion detection can achieve better, when the nodes act cooperatively. An approach taken by many researchers is based on voting. The nodes of a MANET vote on the existence and identity of an intruder.
The task of this topic is to compare different voting-based approaches to intrusion detection. It should also be discussed what strengths and weaknesses those approaches, and voting-based intrusion detection in general have.Assigned to: Maximilian Porsch