Towards the future Internet architecture design - Zhonghong Ou

The main design principle and architecture for today?s Internet was created during the 1960?s and 1970?s. Internet has evolved significantly since its birth in terms of usability and amount of traffic. New problems emerge in different direction against the original design. Some researchers argue that the legacy Internet architecture does not fit the current Internet usage any more, thus start the so-called clean slate redesign of Internet architecture. A couple of proposals have already been put forward, e.g. openflow[1], content-centric networking (CCN) [2], and LIPSIN [3]. However, amongst these proposals, there are no certain criteria to evaluate their respective performance and functionality. In this assignment, you are expected to:

1. Study the current mainstream future Internet architecture designs;
2. Find out the perspective, strengths and weaknesses of each proposal;
3. Design a framework to facilitate the evaluation of various proposals;
4. Compare different proposals in the designed framework.
5. Put forward potential improvement of the proposals.

Requirements

• Familiar with C++/Java/C#.
• Experience in Linux/UNIX programming and basic knowledge about networking.
• Fast-learning and self-motivated.
• Good at scientific English writing. Experience in writing survey reports is a plus.

Contact(s): Zhonghong Ou

[1] Nick McKeown, Tom Anderson, Hari Balakrishnan, Guru Parulkar, Larry Peterson, Jennifer Rexford, Scott Shenker, Jonathan Turner (2008) OpenFlow: enabling innovation in campus networks. SIGCOMM Computer Communication Review 38(2):69-74.
[2] Van Jacobson, Diana K. Smetters, James D. Thornton, Michael F. Plass, Nicholas H. Briggs, Rebecca L. Braynard (2009) Networking named content. Proceedings of the 5th international conference on Emerging networking experiments and technologies (CoNEXT '09): 1-12.
[3] Petri Jokela, András Zahemszky, Christian Esteve Rothenberg, Somaya Arianfar, Pekka Nikander (2009) LIPSIN: line speed publish/subscribe inter-networking. Proceedings of the ACM SIGCOMM 2009 conference on Data communication (SIGCOMM '09): 195-206.

Packet Authentication - Ming Li

The existing Internet is still vulnerable to various source spoofing based attacks. In order to provide a strong authentication on the Internet, we propose the packet authentication as a solution. The basic idea is to insert cryptographic signature into packets and each en route entity holding the corresponding keys verify the packets. Invalid packets are dropped immediately. Our target is to implement such kind of packet authentication mechanism in the existing Internet. Some compatibility issues should be solved, such as IP fragmentation, prefix aggregation, interacting with legacy entities (e.g., hosts and routers), etc.

The task is to implement the prototype or simulate it using some simulators (ns2, ns3 or omnet++). Some discussions are needed to determine whether we need to implement the prototype or simulate it. Anyway, it aims to evaluate the feasibility, effectiveness, and performance.

Requirements:
Familiar with one of the simulators, ns2, ns3 and omnet++
Good knowledge of TCP/IP protocols
C/C++ programming experience

Contact: Ming Li, Email: ming.li@tkk.fi, Room: B158 Computer Science Building

Building a medium-sized cloud platform under academic environment - Zhonghong Ou

There have been a handful of commercial products available for public use; amongst them are Amazon EC2/S3, Google AppEngine, and Microsoft Azure. However, as these systems are proprietary, most of the technical details are not available for researchers; there is no way to get the insight of them. To provide academic researchers the opportunity to explore the insight of cloud platforms, researchers have put forward some open source cloud frameworks. Eucalyptus is one of them which claims to provide comparable performance with Amazon EC2/S3 and supports heterogeneous environment typically found in academic settings. The problem of Eucalyptus is that it has not been tested substantially to support its claims. A lot of work needs to be done for that purpose. In this master thesis, you are expected to:
1. Build up a cloud platform based on open-source Eucalyptus framework;
2. Compare the performance of Eucalyptus with Amazon EC2/S3;
3. Substantially evaluate the cross-cluster virtual machine (VM) traffic of Eucalyptus;
4. Study the performance of virtual network overlay made up of VMs;
5. Finding the limitations of Eucalyptus in terms of scalability and try to put forward some solution for that.

Requirements:
Familiar with C++/Java/C#.
Experience in Linux/UNIX programming and basic knowledge about Operating System (OS).
Fast-learning and self-motivated.
Good at scientific English writing. Experience in writing survey reports is a plus.

Contact(s): Zhonghong Ou

References:
[1] Daniel Nurmi, Rich Wolski, Chris Grzegorczyk, Graziano Obertelli, Sunil Soman, Lamia Youseff, Dmitrii Zagorodnov, The Eucalyptus Open-source Cloud-computing System , In Proceedings of Cloud Computing and Its Applications, 2008.
[2] Roy Campbell, Indranil Gupta, Michael Heath, Steven Y. Ko, Michael Kozuch Marcel Kunze, Thomas Kwan, Kevin Lai, Hing Yan Lee, Martha Lyons, Dejan Milojicic, David O'Hallaron, Yeng Chai Soh, Open Cirrus Cloud Computing Testbed: Federated Data Centers for Open Source Systems and Services Research, HotCloud, 2009.
[3] Armbrust, Michael, Fox, Armando, Griffith, Rean, Joseph, Anthony D., Above the Clouds: A Berkeley View of Cloud Computing, February, 10, 2009.

Topology analysis of a DoS resistant network architecture (2 thesis topics) - Tuomas Aura

Research on future Internet architectures has produced new architectures for publish-subscribe routing protocols that promise to not only increase scalability of the network but also protect the end nodes and network infrastructure against denial-of-service attacks. In particular, source routing protocols that encode the route or multicast tree as a Bloom filter in the packet header promise to hide the sender location in such a way that it cannot be flooded with packets. The goal of this thesis project is to analyze proposed Bloom-filter-based routing from the traffic analysis point of view: what can a malicious subscriber or a botnet owner learn about the routes and network topology by looking at some packet headers.The problem is closely linked to both networking and anonymity and traffic analysis research. We have previously identified significant DoS vulnerabilities in these supposedly DoS resistant networks and expect the topology analysis to lead to the discovery of new attacks. While a malicious party may use the results of this analysis for DoS attacks, a honest sender can use the same results to optimize the scalability of multicast transmissions or to reduce congestion created by unicast communication.

This project requires an interest in probability theory and other mathematical models of computer networks. The topic is particularly suitable for a student who wants to try his or her hands at an open research problem with the view of continuing to doctoral studies. Ideally, two students would work simultaneously on independent but closely related aspects of the problem and thus, would be able to exchange ideas.

Online transactions with unreliable communication (2 thesis topics) - Tuomas Aura

This project addresses a fundamental research problem in the security and reliability of electronic wallets and public-transportation tickets. Traditionally, such applications have been built on tamperproof hardware modules that enable offline tracking of the stored value. The offline solutions, however, are fairly inflexible and do not encourage the kinds of service innovation seen in online services. With the availability of ubiquitous wireless network connections, it is possible to transform the stored-value applications to online services, but that will require development of security and reliability models for such services. For example, there is a part of the user population that habitually misuses any system (e.g. those who never buy a metro ticket) and the benefits for such habitual cheaters must be minimized.

The methods to be used in the project are qualitative security analysis as well as analytical models and simulations of system reliability. It is also necessary to learn about the applications of stored-value smartcards to have a realistic picture of the systems. While the project poses open research problems for which there is no guaranteed solution, a prudent systems engineering approach can take us quite far. Ideally, two students would work simultaneously on independent but closely related aspects of the problem and, thus would be able to exchange ideas.

Energy Profiler for Android Phones - Yu Xiao

Android is an open source platform built by Google for the development of devices employing cellular communications. This topic aims at developing a stand-alone application of energy profiling for Android Dev phones. The application is expected to enable the developers to monitor the energy usage in real-time. The energy profiling is based on the power models got from real-life power measurement. The inputs of the power model are the run-time information got from OS and hardware components, while the outputs are the energy cost by each running application or hardware component.

Requirement:
Familiar with Java and C programming. Experience of embedded development is preferred.

Contact(s): Yu Xiao

Power modeling and simulation for Symbian Phones - Yu Xiao

Build system-level power models for Symbian phones, and validate the models using real-life power measurement. Based on the power models, develop a power simulator and integrate it with Symbian emulator. The simulator will enable the developers to test the energy efficiency of applications during development.

Requirement:
Familiar with Symbian C++ programming.

Contact(s): Yu Xiao

Implementing zFilter forwarding functionality in Zigbee protocol stack - Ericsson, Jukka Ylitalo; Aalto, Antti Ylä-Jääski

In this work, the student implements a new sensor network protocol. The target of the work is to implement so-called zFilter protocol at Zigbee network layer (NWK) and to integrate the zFilter functionality tightly with the Medium Access Control (MAC) layer. More precisely, zFilters define a new and novel way to utilize in-packet Bloom-filters for packet forwarding. Zigbee-modules are low-power, low-cost, wireless networking devices running OSI-model based networking stack.

Tutors at Ericsson: Jukka Ylitalo, Petri Jokela

Other: The Thesis work is unpaid, but Ericsson provides a room for the student at Ericsson Research premises to achieve the best possible way for communications between the student and tutors.

More information: jukka.ylitalo@ericsson.com or +358 (0) 400 615 821 Contact in Aalto: Antti Ylä-Jääski

Scheduling TCP flows at Proxy to Improve Mobile Energy Efficiency over WLAN - Mohammad Hoque

It is very common to use a proxy server to increase the energy efficiency at the mobile devices, which are using network dependent applications over WLAN like web browsing, listening music, etc. And, the implementation of the proxy mechanism can also be application dependent. For example, traffic shaping at the proxy for the streaming applications [1] and proxy for web browsing [2]. In the former case, the idea is to send periodic bursts to the mobile client to reduce energy consumption and for the later case to reduce energy consumption by reducing the transmission time.

What to do:
The mobile user may use both kind of applications together and there might be some other mobile users under the same access point. Keeping this in mind there are two tasks:
1. We want to design and develop a proxy which would be able to identify and handle both kind of applications mentioned above (considering both kinds are TCP based) and schedule these different TCP flows for multiple mobile users to improve the energy efficiency at the mobile devices.
2. Measuring the performance of the proxy through active measurement of mobile power consumption and captured traffic analysis.

Requirements:
Programming skills in C and some other tools like wireshark, etc.

Category: Special Assignment/ Masters Thesis.

Contacts: Mohammad Hoque, Matti Siekkinen.

References:
[1] M. Gundlach, S. Doster, D. K. Lowenthal, S. A. Watterson, and S. Chandra, “Dynamic, Power-Aware Scheduling for Mobile Clients Using a Transparent Proxy,� In Proceedings of ICPP 2004.
[2] Anastasi, G., Conti, M., Lapenna, W. Power saving policies for wireless access to TCP/IP networks

9. Mobile Cloud Computing (3 topics) - Antti Ylä-Jääski

According to ABI Research report 2009, "Mobile Cloud Computing, will soon become a disruptive force in the mobile world, and eventually becomes the dominant way in which mobile applications operate." The basic idea of mobile cloud computing is to move both the data storage and the data processing from mobile devices to the cloud. There are already some examples of mobile cloud applications like mobile Gmail, and Google Maps. However, nowadays, most of mobile applications are still using the data storage and processing capacities of mobile devices.

While moving the mobile computing from ground to cloud, there are some key issues that have to be solved such as energy efficiency, security, and services in mobile cloud computing. We would propose three master thesis topics, each focuses on one aspect of mobile cloud computing.

To complete the thesis, students are expected to
1. Survey the state-of-the-art of the research area in academic and industrial communities. The survey should analyze the strengths and weaknesses of previous work, figure out the major challenges, and propose some potential solutions.
2. Choose one solution as case study and implement it on mobile devices and cloud computing platform.
3. Evaluate the solution through real-life measurement or simulation.

You can find the detailed descriptions after the requirements and contacts below

Requirements:
Familiar with C++/Java/C#.
Experience in web service development and mobile programming is a plus.
Fast-learning and good team spirit.
Good at English writing. Experience in writing survey report is a plus.

Contact: Prof Antti Ylä-Jääski.

Here are the detailed descriptions of the topics.

1) Energy Efficiency

Most of previous work has been focusing on the energy consumption of data centers and the techniques to save the energy consumption of processing and networking on data centers. However, few results have been gained from the research on energy efficiency in mobile cloud computing. For mobile users, battery lifetime is a constraint. Cloud-based applications have the potential to move the energy hungry computing to the cloud, whereas the energy cost of network transmission might not decrease but increase due to the communications with the cloud. This topic aims at figuring out the challenges to and the solutions for energy-efficient mobile cloud computing. It is expected to answer the following questions.
1. What are the differences from traditional mobile applications and wired cloud applications from the viewpoint of energy consumption? You could use some cloud-based mobile applications as examples, and summarize their typical power characteristics.
2. Why is energy-efficiency a challenge to mobile cloud computing? Is it caused by some inherent characteristics of mobile cloud computing?
3. How to improve the nergy efficiency in mobile cloud computing? You can present it from different aspects, for example, system architecture, communication mechanism, service design, or other relevant factors. Overview the previous work and analyze their strengths and weaknesses.
4. Is there any solution that could be developed and deployed in our experimental environment? What tools and platforms are needed?
5. How to evaluate the effectiveness of your solutions? Any standard metrics to use?
6. What could be the future research questions in this area?

You could find some information about the practice in energy-efficient cloud computing from ref[1,2,3]. More references about cloud computing could be found at the ending of this document.

2) Security

Cloud computing allows the data storage and processing remotely, which reduces the workload from the cloud users. Meanwhile, "the fact that the cloud users lack control over the cloud raises significant security concerns about the integrity and confidentiality of a user’s workload."[7] There exist risks during the computation including the code execution and flow control, as well as the data transmission over the Internet, such as the transfer between the cloud users and the cloud providers, and between the visualized machines. In addition, mobile services have their specific security issues. When mobile services move from ground to the cloud, there might be new challenges and problems. This topic aims at figuring out the potential security issues in mobile cloud computing and possible solutions. The thesis will cover the following aspects.
1. What are the major security issues in wired cloud computing? Do they also exist in mobile cloud computing?
2. What are the special security issues in mobile cloud services? Are they dependent on the services? Are they different from traditional security issues in mobile applications?
3. What are the solutions proposed in the literature? Analyze the strengths and weaknesses.
4. Analyze the feasibility of developing security into a mobile cloud service, and define the features of this service.
5. Choose a mobile cloud service and propose a security solution for this service, such as Gmail, Google Doc.
6. Implement your solution and evaluate it.

You could find some information about the security issues in cloud computing from ref[4]-[10]. More references about cloud computing could be found at the ending of this document.

3) Mobile Cloud Services

Cloud computing brings new business opportunities and values to IT companies. It is said to make it possible for organizations and consumers to access flexible services over the Internet via a usage-based business model. Many famous IT companies like Google, Microsoft, HP, and IBM have developed their own cloud computing platforms and start gaining huge profit from their cloud services. Currently, most of services are still provided to organization customers. However, mobile customers have become a huge market of services. Moving cloud services to mobile market will become another big thing in mobile technique. This topic will look into the cloud-based mobile services, and figure out the essentials issues on designing, developing, and deploying mobile cloud services. The thesis is expected to solve the following issues.
1. What are the differences among traditional outsourced services, grid computing, and cloud services? You can analyze it from both technique and business aspects.
2. How many types of cloud services are available? What are the differences among them?
3. What are the characteristics of mobile cloud services compared to traditional mobile services? Give some examples.
4. Compare the existing cloud computing platforms, such as Windows Azure and IBM Lotus, and analyze their strengths and weakness in the development of mobile cloud services.
5. What kind of mobile techniques are required for developing or deploying mobile cloud services on mobile devices?
6. What are the other challenges to mobile cloud services? You could analyze it from techniques and business aspects.
7. Analyze the potentials of providing power management and security as mobile cloud services.
8. Analyze the feasibility of building a cloud of mobile devices. If it is feasible, propose some user cases for this scenario and figure out the requirements of techniques. For example, live video edition and delivery.
9. If it is possible, build a simple cloud of several mobile devices.

References:
[1] Kim, K. H., Beloglazov, A., and Buyya, R. 2009. Power-aware provisioning of Cloud resources for real-time services. In Proceedings of the 7th international Workshop on Middleware For Grids, Clouds and E-Science (Urbana Champaign, Illinois, November 30 - December 01, 2009). MGC '09. ACM, New York, NY, 1-6. DOI= http://doi.acm.org/10.1145/1657120.1657121
[2] Sharma, A., Navda, V., Ramjee, R., Padmanabhan, V. N., and Belding, E. M. 2009. Cool-Tether: energy efficient on-the-fly wifi hot-spots using mobile phones. In Proceedings of the 5th international Conference on Emerging Networking Experiments and Technologies (Rome, Italy, December 01 - 04, 2009). CoNEXT '09. ACM, New York, NY, 109-120. DOI= http://doi.acm.org/10.1145/1658939.1658952
[3] Andreas Berl , Erol Gelenbe , Marco Di Girolamo , Giovanni Giuliani , Hermann De Meer , Minh Quan Dang , and Kostas Pentikousis.Energy-Efficient Cloud Computing. The Computer Journal Advance Access published on August 19, 2009, DOI 10.1093/comjnl/bxp080.
Security:
[4] Zhang, X., Schiffman, J., Gibbs, S., Kunjithapatham, A., and Jeong, S. 2009. Securing elastic applications on mobile devices for cloud computing. In Proceedings of the 2009 ACM Workshop on Cloud Computing Security (Chicago, Illinois, USA, November 13 - 13, 2009). CCSW '09. ACM, New York, NY, 127-134. DOI= http://doi.acm.org/10.1145/1655008.1655026
[5] Oberheide, J., Veeraraghavan, K., Cooke, E., Flinn, J., and Jahanian, F. 2008. Virtualized in-cloud security services for mobile devices. In Proceedings of the First Workshop on Virtualization in Mobile Computing (Breckenridge, Colorado, June 17 - 17, 2008). MobiVirt '08. ACM, New York, NY, 31-35. DOI= http://doi.acm.org/10.1145/1622103.1629656
[6] Wong, W., Cheung, D., Kao, B., and Mamoulis, N. 2009. Secure kNN computation on encrypted databases. In Proceedings of the 35th SIGMOD international Conference on Management of Data (Providence, Rhode Island, USA, June 29 - July 02, 2009). C. Binnig and B. Dageville, Eds. SIGMOD '09. ACM, New York, NY, 139-152. DOI= http://doi.acm.org/10.1145/1559845.1559862
[7] Christodorescu, M., Sailer, R., Schales, D. L., Sgandurra, D., and Zamboni, D. 2009. Cloud security is not (just) virtualization security: a short paper. In Proceedings of the 2009 ACM Workshop on Cloud Computing Security (Chicago, Illinois, USA, November 13 - 13, 2009). CCSW '09. ACM, New York, NY, 97-102. DOI= http://doi.acm.org/10.1145/1655008.1655022
[8] Wang, W., Li, Z., Owens, R., and Bhargava, B. 2009. Secure and efficient access to outsourced data. In Proceedings of the 2009 ACM Workshop on Cloud Computing Security (Chicago, Illinois, USA, November 13 - 13, 2009). CCSW '09. ACM, New York, NY, 55-66. DOI= http://doi.acm.org/10.1145/1655008.1655016
[9] Du, J., Wei, W., Gu, X., and Yu, T. 2009. Towards secure dataflow processing in open distributed systems. In Proceedings of the 2009 ACM Workshop on Scalable Trusted Computing (Chicago, Illinois, USA, November 13 - 13, 2009). STC '09. ACM, New York, NY, 67-72. DOI= http://doi.acm.org/10.1145/1655108.1655120
Mobile Cloud Services:
[11] Chow, R., Golle, P., Jakobsson, M., Shi, E., Staddon, J., Masuoka, R., and Molina, J. 2009. Controlling data in the cloud: outsourcing computation without outsourcing control. In Proceedings of the 2009 ACM Workshop on Cloud Computing Security (Chicago, Illinois, USA, November 13 - 13, 2009). CCSW '09. ACM, New York, NY, 85-90. DOI= http://doi.acm.org/10.1145/1655008.1655020
[12] Qingfeng Liu; Xie Jian; Jicheng Hu; Hongchen Zhao; Shanshan Zhang, "An Optimized Solution for Mobile Environment Using Mobile Cloud Computing," Wireless Communications, Networking and Mobile Computing, 2009. WiCom '09. 5th International Conference on , vol., no., pp.1-5, 24-26 Sept. 2009
[13] Bentley, F. R. and Groble, M. 2009. TuVista: meeting the multimedia needs of mobile sports fans. In Proceedings of the Seventeen ACM international Conference on Multimedia (Beijing, China, October 19 - 24, 2009). MM '09. ACM, New York, NY, 471-480. DOI= http://doi.acm.org/10.1145/1631272.1631337
[14] Christensen, J. H. 2009. Using RESTful web-services and cloud computing to create next generation mobile applications. In Proceeding of the 24th ACM SIGPLAN Conference Companion on Object Oriented Programming Systems Languages and Applications (Orlando, Florida, USA, October 25 - 29, 2009). OOPSLA '09. ACM, New York, NY, 627-634. DOI= http://doi.acm.org/10.1145/1639950.1639958
[15] Giurgiu, I., Riva, O., Juric, D., Krivulev, I., and Alonso, G. 2009. Calling the cloud: enabling mobile phones as interfaces to cloud applications. In Proceedings of the 10th ACM/IFIP/USENIX international Conference on Middleware (Urbanna, Illinois, November 30 - December 04, 2009). Middleware Conference. Springer-Verlag New York, New York, NY, 1-20.
[16] Han, S., Hassan, M. M., Yoon, C., and Huh, E. 2009. Efficient service recommendation system for cloud computing market. In Proceedings of the 2nd international Conference on interaction Sciences: information Technology, Culture and Human (Seoul, Korea, November 24 - 26, 2009). ICIS '09, vol. 403. ACM, New York, NY, 839-845. DOI= http://doi.acm.org/10.1145/1655925.1656078
General information:
[17] White Paper. Above the Cloud. http://berkeleyclouds.blogspot.com/
[18] Birman, K., Chockler, G., and van Renesse, R. 2009. Toward a cloud computing research agenda. SIGACT News 40, 2 (Jun. 2009), 68-80. DOI= http://doi.acm.org/10.1145/1556154.1556172
[19] McCann, J. A., Parashar, M., Rana, O., Foschini, L., and Kindberg, T. 2009. Is cloud computing essential to the future of pervasive services?. In Proceedings of the 2009 international Conference on Pervasive Services (London, United Kingdom, July 13 - 17, 2009). ICPS '09. ACM, New York, NY, 147-148. DOI= http://doi.acm.org/10.1145/1568199.1568225
[20] Weiss, A. 2007. Computing in the clouds. netWorker 11, 4 (Dec. 2007), 16-25. DOI= http://doi.acm.org/10.1145/1327512.1327513
[21] Zeng, W., Zhao, Y., Ou, K., and Song, W. 2009. Research on cloud storage architecture and key technologies. In Proceedings of the 2nd international Conference on interaction Sciences: information Technology, Culture and Human (Seoul, Korea, November 24 - 26, 2009). ICIS '09, vol. 403. ACM, New York, NY, 1044-1048. DOI= http://doi.acm.org/10.1145/1655925.1656114

Toolkit for diagnosing home network connectivity - Matti Siekkinen

Did you ever wonder why your download is as slow as it is, even though you have bought a fast Internet subscription? It is common that residential Internet users feel frustrated when their Internet connection performs poorly. Pinpointing the exact cause of the problem is a non-trivial task as there are many possible suspects encompassing local problems such as insufficient resources at the end host (CPU, memory, etc.) or problems with WLAN due to interference and contention, problems at the access network level managed by the ISP (including DNS and DHCP issues), or even at the remote server/peer.

Several tools and methods have been developed for diagnosing certain specific problems. For example, passively diagnosing performance limitations of TCP transmissions have been addressed in [1] and [3]. Diagnosis of WLAN environments has been studied in [2]. In addition, many tools such as NetStumbler have been developed for Wi-Fi monitoring. Active measurement methods such as simple ping are useful in detecting unreachability problems.

Task description:
The student should develop a toolkit (software module) for detailed diagnosis by leveraging already existing tools and analysis methods. The tool should be able to diagnose the following problems:
DNS malfunction
IP address configuration problems (DHCP)
Unreachability of remote host
Local boxes (WLAN router, gateway) are dead
Local host issues (CPU or bandwidth overload)
Detailed diagnosis of active connections
- Local host as bottleneck
- Remote host as bottleneck
- Network as bottleneck (Local network (WLAN issues) and Remote network)

A simple GUI for presenting diagnosis results should be developed. The user should be able to push a button to initiate diagnosis and to stop it. The GUI should show the status of the different diagnosed problems (e.g. a green light indicating that the host has obtained an IP address or that DNS resolving works).

Possible extensions of this work is to make the toolkit collaborative so that diagnosis information could be shared among hosts (e.g. through centralized server or in P2P manner) or even perform analysis through joint efforts (e.g. localizing bottleneck links) to provide more precise diagnostics.

Requirements:
Good knowledge of Internet technologies, programming skills.

Contact(s): Matti Siekkinen, Guillaume Urvoy-Keller

References:
[1] Matti Siekkinen, Guillaume Urvoy-Keller, Ernst W. Biersack, and Denis Collange. A Root Cause Analysis Toolkit for TCP. In Computer Networks. Volume 52, Issue 9, 26 June 2008, Pages 1846-1858.
[2] Cheng, Y., Afanasyev, M., Verkaik, P., Benkö, P., Chiang, J., Snoeren, A. C., Savage, S., and Voelker, G. M. 2007. Automating cross-layer diagnosis of enterprise wireless networks. SIGCOMM Comput. Commun. Rev. 37, 4 (Oct. 2007), 25-36. DOI= http://doi.acm.org/10.1145/1282427.1282384
[3] Hafsaoui, Aymen;Collange, Denis; Urvoy-Keller, Guillaume. Revisiting the performance of short TCP transfers. Networking 2009, 8th IFIP International Conference on Networking, May 11-15, 2009.

Studying the impact of virtualization on network performance - Matti Siekkinen

Virtualization is nowadays a very commonly used technique for consolidating servers in data centers. The key idea is to reduce the number of physical machines and group together servers on the same physical hardware. While virtualization is appealing as it helps saving resources, it might have negative side effects on the performance observed by end users such as abnormal delay variation or throughput degradation. Related studies have been reported very recently in [1,2].

Task description:
The task is two-fold: First the student should design simple and then more realistic use cases for evaluating the impact of virtualization. For instance, while the simple cases might be inspired by the straightforward benchmarks conducted in [1] and [2], different types of web applications (e.g. including dialogues) should be included in the more realistic cases. In the second stage, the tests should be conducted on virtual environments deployed using different kinds of platforms such as VMWare, VirtualBox, Xen, … This analysis of the results from this phase should provide insights in what are the quantitative impacts and how do they vary across different solutions for virtualization.

Requirements: Good knowledge of Internet technologies, programming skills.

Contact(s): Matti Siekkinen, Guillaume Urvoy-Keller

References:
[1] Wang, G. and Ng, T. S. 2010. The impact of virtualization on network performance of amazon EC2 data center. In Proceedings of INFOCOM the 29th Conference on information Communications (San Diego, California, USA, March 14 - 19, 2010).
[2] Barham, P., Dragovic, B., Fraser, K., Hand, S., Harris, T., Ho, A., Neugebauer, R., Pratt, I., and Warfield, A. 2003. Xen and the art of virtualization. In Proceedings of the Nineteenth ACM Symposium on Operating Systems Principles (Bolton Landing, NY, USA, October 19 - 22, 2003). SOSP '03.
[3] Bilal Anwer, Nick Feamster, Ankur Nayak, Ling Liu. Network I/O Fairness in Virtual Machines. In Proceedings of The Second ACM SIGCOMM Workshop on Virtualized Infrastructure Systems and Architectures (VISA 2010).

Delay/Disruption Tolerant Low-Energy Networking with Nokia Instant Community - Matti Siekkinen

Delay or Disruption tolerant networking (DTN) [1] is an emerging paradigm where end-to-end connectivity between communicating end hosts cannot be assumed at all times. Nokia Instant Community [2] is a novel always-on low-power consuming WLAN MANET (Mobile Ad-hoc Network) connectivity solution created by Nokia Research Center. The advanced technology concept enables the exploration of this research area.

There are three possible topics to choose from. These individual tasks are interrelated through the underlying Instant Community solution, and ideally collaboration between these tasks would be possible.
- Explore and identify effective energy saving delay tolerant constructs based on Nokia Instant Community networking [1].
- Study the applicability of the Nokia Instant Community always-on MANET as an underlying communication solution for cloud computing in a local cloud. [3]
- Parameter value optimization and scalability analysis for a truly large and energy efficient Nokia Instant Community MANET [4,5].

The work is carried out by means of simulation with an NS2 tool with existing enhanced energy modeling and UML extension.

Requirements:
Good knowledge of Internet technologies and mobile networks,
Knowledge about NS2 and C++ is a plus.

Position: Special assignment and/or Masters thesis

Contact: Matti Siekkinen

Multimedia Adapter between Mobile Phones and Televisions - Miika Komu

Modern flat-screen televisions are basically small computers and can be connected to the Internet. This opens up new possibilities for the usage of the televisions. For example, a phone could be used as a multimedia storage and used to play music, videos or to even show powerpoint slides on the television. However, it appears there are some challenges with existing devices. In our experiments, a N900 connected to a television set failed to scale the video output to correct size.

The goal of this work is to design and implement software for a small "adapter" device that corrects the media stream to better fit the television. The device can be based on linux or mac platform depending on the student's preferences. The adapter should be independent of the client-side platform (android, maemo, iOS). The adapter is evaluated using a mobile phone.

Requirements:
The student should be familiar with one mobile OS (android, maemo, iOS) or be prepared to sacrifice time to learn it. Previous experience in developing software with Linux or OS-X is highly recommended.

Contacts: Miika Komu, Matti Kemppainen

Reputation metrics for P2P overlay - Andrey Lukyanenko

Reputation is one of the most necessary things in peer-to-peer where whitewashing and free-riding impose big threats for the whole system. Some studies indicates that current BitTorrent architecture functions mainly because of the people's ignorance (of technology, legal issues,and so on), than because of the own desire to contribute to the system. For the future architectures where creators are aware of the contribution stimulus, some form of reputation is required. One of open research questions here is also what metric to use, and how to verify that this exactly metric is able to make participants behave accordingly.

This work is dedicated to the practical development of the reputation metric in P2P networks. The task includes study of P2P network/DHT architectures (such as Chord, Tapestry,...), study of OverSim, a p2p simulator, and study of the possible reputation metrics, and properties of it (locality). The results of the study is supposed to be implementation of the reputations inside of the DHT protocols or over it (in a P2P networks), analyzing it, and listing what they can give to the system, and what they cannot do.

References
[1] Marti, S. and Garcia-Molina, H. 2006. Taxonomy of trust: Categorizing P2P reputation systems. Comput. Netw. 50, 4 (Mar. 2006), 472-484.DOI=http://dx.doi.org/10.1016/j.comnet.2005.07.011
[2] Lukyanenko, A. and Gurtov, A. 2009. Towards behavioral control in multi-player network games. In Proceedings of the First ICST international Conference on Game theory For Networks (Istanbul, Turkey, May 13 - 15, 2009). IEEE Press, Piscataway, NJ, 683-690.

Merging structured and unstructured DHT properties together - Andrey Lukyanenko

DHT is well studied and thought ideas in the networking. The study in the field is limited, but from time to time new inspiring ideas appear in the field. One recent publication [1] revived the topic of Sybil attack problem and the relation of it to DHT. It states that unstructured or not well-structured architectures have higher defense properties against the Sybil attacks [2]. The research topic here would be is it possible to combine good parts of the structured and unstructured DHTs protocols, and how to do it?

References
[1] Viswanath, B., Post, A., Gummadi, K. P., and Mislove, A. 2010. An analysis of social network-based Sybil defenses. In Proceedings of the ACM SIGCOMM 2010 Conference on SIGCOMM (New Delhi, India, August 30 - September 03, 2010). SIGCOMM '10. ACM, New York, NY, 363-374. DOI=http://doi.acm.org/10.1145/1851182.1851226
[2] http://www.springerlink.com/content/3an0ek5gfan3dtx9/

Changing Backoff protocol in real Broadcom device - Andrey Lukyanenko

Broadcom is one of the manufactures who produces wireless cards. Recently Broadcom decided to open own drivers source codes. This gives us a great opportunity to get an access to internal wireless card parameters, which includes Backoff protocols. Open drivers for wireless card itself is not a new feature, however, the Broadcom is who have access to the backoff parameters inside the drivers, Atheros drivers for example does not contain backoff, instead they place it on the chip.

References
[1]https://linux.slashdot.org/story/10/09/09/1925214/Broadcom-Releases-Source-Code-For-Drivers|
[2]http://linux.slashdot.org/comments.pl?sid=10/09/09/1925214
[3]http://www.osnews.com/story/23786/BREAKING_BROADCOM_OPEN_SOURCES_WIRELESS_DRIVERS

MapReduce for mobile cloud computing - Yu Xiao

MapReduce is a programming model for processing and generating large data sets. In mobile cloud computing, when the mobile phones work as information collectors, the processing of data sets can be done by sending all the data to a remote data center or partly processed on mobile phones. MapReduce is possible to be used as a framework for the data processing in such scenario. This thesis is expected to propose a data processing model for mobile cloud computing based on MapReduce.

The model must be implemented and tested on mobile phones such as Nokia N900.

Requirement: Linux, Java/C++

Online provisioning and revocation of cryptographic tokens - Sandeep Tamrakar

Traditional electronic wallets and public-transportation tickets come pre-provisioned with application specific information. In order to update the stored information, e.g. add value in the bus ticket, the user has to go to some service booth or kiosks. Such services may not be available everywhere and around the clock. On the other hand, with the ubiquitous wireless network connections such updates could be made online. Or say, the card could be provisioned and updated online.

Further, different transportation system or companies may choose to use different protocols, information on the card. Integrating such information on a single card is cumbersome and often discouraged. However, with the possibility of online services, the card information can be exchanged online between servers of different transportation companies while the user will not notice any difference, except the travel fair deducted from his ticket when using on different domain.

Managing the life-cycle of the card information becomes easier with the integration of online services. For example, when a card is lost or stolen, the abuse on the card is unavoidable as the service provider has no direct access. The illegitimate holder of the card can continue using the card until the stored information expires. Maintaining and updating blacklist information on each bus terminal may not be feasible in current system. However, with the integration of online services, such abuse can be reduces as the server has direct access to the card information and then can forcefully delete information stored on the card if needed.

The goal of this thesis project is to design and implement a prototype system for online provisioning and revocation of the tickets or other credentials as part of an ongoing research project. Students with experience in software implementation in C, Java and Linux will be able to get started faster.

Many-to-many Discussion Forums with Multiple Userinterface APIs - Sanna Suoranta

The Noppa portal privides course webpages and announcements at Aalto University School of Science and Technology. However, teachers of many courses would like to have two-way discussions with the students and to promote discussion between students. Previously, NNTP based newsgroups were actively used by both students and teachers to discuss study related topics and everyting else. Nowadays, students use many web based discussion forums in their hobbies, but there is no such service for the courses. At Aalto, Optima learning environment allows creating closed discussion forums, but they are seldomly used since they are quite clumsy. Still, the old newsgroups are available but students do not use them anymore. There exist no such system that suites for both students and teachers and allows either open discussions or discussions restricted to students of a course.

The main task of this master thesis is to create API that frees the user to choose his/her favorite communication tool (news reader, web browser, email reader etc) and still be able to use many-to-many discussion forum. The system should allow both open and restricted discussion groups.

Requirements
Familiar with web technologies
Programming skills