Spring 2009 – Current Internet trends

 During the last few years we have experienced that the main developments on the Internet take place in the applications and services areas. Naturally there are developments also in the technical infrastructure domain, but the majority of the most interesting phenomena are towards the content delivery and social networking.

During spring 2009 the seminar on Internetworking will broadly cover the main global trends in the Internet.

1. Services on the Internet

2. Developments in the Web, from Web 2.0 onwards

3. Technology developments in the Internet infrastructure

4. Digital consumers on the Internet

Course Topics 

The concrete list of topics for the student papers will be here after the introductory meeting.

Topics listed 19.1. 16:01. Later additions are in the end of the page

Topics by prof. Tuomas Aura 

Aura, Topic 1: Wireless handoff optimizations

Wi-Fi or 802.11 is the most widely deployed technology for wireless broadband Internet access in densely populated areas. It does, however, suffer from problems related to mobility ad QoS. Users are increasingly mobile, with laptop computers and handheld devices, and they expect quality of service that is sufficient for streaming video or interactive voice. Currently, most 802.11deployments do not meet these requirements. Handoffs between access points belonging to the same wireless network may take seconds. Much of the handoff latency is caused by security mechanisms, such as the 4-way handshake and, in particular, EAP authentication to a remote authentication server. The goal of this seminar paper is to give an up-to-date picture of the handoff optimizations that have been standardized and standard proposals for 802.11 networks. Such optimizations may be implemented either on the 802.11 level or for the 802.1X authentication framework.

IEEE 802.11r-2008 IEEE standard for information technology-telecommunications and information exchange between systems-local and metropolitan area networks-specific requirements part 11: wireless lan medium access control (mac) and physical layer (phy) specifications amendment 2: fast basic service set.

IETF Handover Keying (HOKEY )Working Group,  http://www.ietf.org/html.charters/hokey-charter.html

Aura, Topic 2: Cloud computing with the Azure platform

The Microsoft Azure is one of the several commercial platforms for cloud computing. Its architecture differs considerably from the early cloud-computing models in which the customer creates a virtual-machine image that is instantiated in a data center (see. e.g. Amazon Web Services). In Azure, services are written as .NET applications, with additional mechanisms for scalability and persistent storage in the cloud. Although the applications are run on virtual machines in a data center, that part is not explicitly visible to the developer. The goal of this seminar paper is to explain the essential architectural ideas of the Azure platform, compare it with the VM-image-based approach, and to implement a small sample service (e.g. Hello World). Some prior experience with .NET development (e.g. C# or ASP.NET) or otherwise strong programming background would be helpful.

Microsoft Developer Network Documentation for the Azure platform: http://msdn.microsoft.com/en-us/azure/default.aspx

Topics by Andrei Gurtov

Gurtov Topic1. Game theory in Internet networks and applications

Internet is composed of a large number of users and ISPs, that often behave selfishly when providing services or consuming resources. They try to maximize own profit and minimize costs, which results in negative effects such as hot potato routing or free-riding in P2P systems. Game theory offers a good tool to model selfish player behavior. Using game theoretic modeling, we can design future Internet that can be stable and encourage socially optimal user and ISP behavior. The task of this topic is to survey the state of research on GT in Internet, and outline the currently hot areas.

Selfish Behavior and Stability of the Internet: A Game-Theoretic Analysis of TCP.  Aditya Akella, Srinivasan Seshan (CMU) Richard Karp, Scott Shenker, Christos Papadimitriou (ICSI/UC Berkeley)

http://conferences.sigcomm.org/sigcomm/2002/papers/tcpgame.pdf

Free-Riding and Whitewashing in Peer-to-Peer Systems (2004) by Michal Feldman, Christos Papadimitriou, John Chuang In PINS ’04: Proceedings of the ACM SIGCOMM workshop on Practice and theory of incentives in networked systems

http://www.sigcomm.org/sigcomm2004/workshop_papers/p35-feldman11.pdf

On Selfish Routing in Internet-Like Environments (2003) by Lili Qiu, Yang Richard, Yang Yin, Zhang Scott Shenker in Proceedings of ACM SIGCOMM

http://www-net.cs.yale.edu/publications/QYZS03.pdf

Gurtov topic 2. Adopting Internet protocols to sensor internetworking

The current trend is to adopt IP technology everywhere, including sensor nodes with very limited hardware resources. Since sensor nodes communicate over wireless interfaces, security risks such as eavesdropping and node break-in are quite high. The task of this assignment is to survey security mechanisms for IP protocols in sensor networks, compare symmetric and asymmetric crypto approaches, and study feasibility of running HIP in sensor networks.

Wang, H., Sheng, B., Tan, C. C., and Li, Q. 2008. Comparing Symmetric-key and Public-key Based Security Schemes in Sensor Networks: A Case Study of User Access Control. In Proceedings of the 2008 the 28th international Conference on Distributed Computing Systems - Volume 00 (June 17 - 20, 2008). ICDCS. IEEE Computer Society, Washington, DC,

http://www.cs.wm.edu/~liqun/paper/icdcs08_2.pdf

Durvy, Mathilde and Abeillé, Julien and Wetterwald, Patrick and O'Flynn, Colin and Leverett, Blake and Gnoske, Eric and Vidales, Michael and Mulligan, Geoff and Tsiftes, Nicolas and Finne, Niclas and Dunkels, Adam (2008) Making sensor networks IPv6 ready. In: The Sixth ACM Conference on Networked Embedded Sensor Systems (ACM SenSys 2008), November 2008, Raleigh, North Carolina, USA.

Making TCP/IP Viable for Wireless Sensor Networks

http://www.sics.se/~adam/ewsn2004.pdf 

Topic by Iiro Jantunen 

Jantunen topic 1. Mobile-internet-based healthcare, public safety and welfare services

With rise of the standards of life and advances in medicine, the span of human life has extended and the public services are in danger of being strained by the needs of the aging population. On the other hand, in many parts of the world public services are nonexistent while the traditional way of caring for the elderly at home is not any more feasible for the offspring living in distant cities with small apartments. Mobile-internet-based services provide a possibility to lower the price of getting adequate care at for the elderly at home while preserving their ability to live at their homes - improving their quality of life as they can live in the environment and with the neighbors they know.

Technologies and concepts:

MIMOSA project http://www.mimosa-fp6.com/

MINAmI project http://www.fp6-minami.org/

Continua Health Alliance http://www.continuaalliance.org/home

Jantunen Topic 2. Social networks for improving the quality-of-life of the elderly

With rise of the standards of life and advances in medicine, the span of human life has extended. The traditional way of caring for the elderly at home is not any more feasible for the offspring living in distant cities with small apartments. The elderly would like to meet their offspring, friends, and neighbors on daily basis, but as people move and the elderly get disabilities, the possibilities for daily contacts become scarce. Social networks in the internet may provide the elderly an easy way for daily interaction with their offspring, friends and other people, improving quality of life.

Some article links will be provided. The subject is being studied as a a part of a current Finnish-Chinese research project. There is also a Finnis internet

2.0 company on the field with which the student may interact. 

Topics by Jaakko Kangasharju

Kangasharju Topic 1. Cooperation in Peer-to-peer Content Distribution

Modern peer-to-peer content distribution systems rely on content consumers to also take part in distributing the content, and if a sufficient number of peers refuse to cooperate, the performance of the system will degrade. The goal of this study is to evaluate how significant this problem is in theory and in actual practice, and to study methods of increasing cooperation.

http://portal.acm.org/citation.cfm?id=1080192.1080198&coll=ACM&dl=ACM&CFID=3953356&CFTOKEN=39951226

http://portal.acm.org/citation.cfm?id=1120717.1120723&coll=ACM&dl=ACM&CFID=3953356&CFTOKEN=39951226

Kangasharju Topic 2 Expanding the Web: Is XML Sufficient? 

XML is a ubiquitous data format in the modern Web. With the expansion of the Web into new environments, some believe that XML is too heavy and a binary encoding of XML is needed. Others contend that generic compression and improvements in processing technology are sufficient and no new formats need to be introduced. Study the issue from both sides and draw your own conclusions.

http://portal.acm.org/citation.cfm?id=988672.988719&coll=ACM&dl=ACM&CFID=18384475&CFTOKEN=35648770

http://portal.acm.org/citation.cfm?id=1463788.1463811&coll=ACM&dl=ACM&CFID=17926199&CFTOKEN=11106777

http://www.w3.org/TR/xbc-characterization/

Topics by Miika Komu 

Komu Topic 1: A Survey on Diagnosing Problems in Network Connectivity

Modern laptops and cellular phones are equipped with multiple network access devices which are not used efficiently enough to benefit the  user. The goal of this topic is to survey the state-of-the-art on  existing literature on how a mobile device can detect problems with a  particular network interface and how it can determine automatically when to switch to an alternative interface.

Komu Topic 2: A Survey on Network Application Frameworks 

All modern applications are not using sockets API directly to achieve network connectivity. Instead, they are using network application frameworks, such as Twisted in Python, to access the network. The goal of this topic is to survey the existing literature and implementations on what kind of application frameworks are available and what they are  useful for. 

Topics by Janne Lindqvist

Lindqvist Topic 1: Re-authentication in IEEE 802.11 WLANs

Managed 802.11 networks are supported by backend AAA servers that enable roaming and billing, for example. In the worst case, the WLAN client needs to reauthenticate to the backend server every time it changes the access point.The purpose of this topic is to survey the standard mechanisms of reducing reauthentication delay in wireless networks.

 802.11-preauthentication

IEEE 802.11-2007 IEEE Standard for Information technology-Telecommunications and information exchange between systems-Local and metropolitan area networks-Specific requirements - Part 11: Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specifications

IEEE 802.11r-2008 IEEE standard for information technology-telecommunications and information exchange between systems-local and metropolitan area networks-specific requirements part 11: wireless lan medium access control (mac) and physical layer (phy) specifications amendment 2: fast basic service set (bss)

Lindqvist topic 2: IEEE 802.11 link-layer security

WEP was seriously flawed. The standard 802.11 (formerly 802.11i) is proven to be secure, but there are persistent rumours about the insecurity of WPA/WPA2. It is time to look what is happening in 802.11 link-layer security. This topic is targeted towards a student who is not afraid of formalisms.

Analysis of the 802.11i 4-way handshake

http://doi.acm.org/10.1145/1023646.1023655

Security Analysis and Improvements for IEEE 802.11i - Changhua He, Stanford University; John C. Mitchell, Stanford University

http://www.isoc.org/isoc/conferences/ndss/05/proceedings/papers/NDSS05-1107.pdf

A modular correctness proof of IEEE 802.11i and TLS

http://doi.acm.org/10.1145/1102120.1102124 

Topics by Luo Hong, professor from BUPT, China

Luo Topic 1 :Quality of Information: The quantification metrics of the performances in wireless sensor networks 

Just like QoS (Quality of Service) in traditional networks, which is used to determine the degree of satisfaction of a service to the user, QoI is recently introduced into wireless sensor networks to quantify the degree by which the information is fit-to-use for a purpose. The goal of the study is to identify the main quality attributes of information in sensor networks, analyze different QoI requirements in different type of applications, and make an overview of QoI in structure healthy monitoring system.

[1] C Bisdikian, J Branch, KK Leung, RI Young , A Letter Soup for the Quality of Information in Sensor Networks, IBM Research Report, RC24691(W0811-061) Nov. 2008

[2] S Zahedi, MB Srivastava, C Bisdikian, A Computational Framework for Quality of Information Analysis for Detection-oriented Sensor Networks, 

[3] MA Hossain, PK Atrey, A El Saddik, Modeling Quality of Information in Multi-sensor Surveillance Systems, IEEE 23rd International Conference on Data Engineering Workshop, Apr. 2007 pp: 11-18 

[4] E Gelenbe, L Hey, Quality of information: An empirical approach, 5th IEEE International Conference on Mobile Ad Hoc and Sensor Systems, MASS 2008. Sep. 2008, pp: 730-735 

Luo Topic 2 Trust modeling and management in wireless sensor networks

Wireless sensor networks are data-driven and provide sufficient measurement of phenomena in the sensing field. However, as with almost every data-driven technology, the many benefits come with a significant challenge in data reliability. Are these reported data usable and believable? To solve this problem, some reputation-based strategies are imported for trust evaluation of data. The goal of this research is to survey the schemes of trust modeling and management in sensor networks, analyze their performance, overhead, suitable network structure, and overview the trust management framework for the multi-level cluster-based sensor networks.

[1] S Ganeriwal, LK Balzano, MB Srivastava, Reputation-based framework for high integrity sensor networks, ACM Transactions on Sensor Networks (TOSN), Volume 4, Issue 3, May 2008

[2] Guangjie Han; Deokjai Choi; Wontaek Lim, A Reliable Approach of Establishing Trust for Wireless Sensor Networks, IFIP International Conference on Network and Parallel Computing Workshops, 2007.  Sept. 2007 Pp:232 - 237 

[3] R.A.Shaikh, H Jameel, Sungyoung Lee, S Rajput, Y.J. Song, Trust Management Problem in Distributed Wireless Sensor Networks, 12th IEEE International Conference on Embedded and Real-Time Computing Systems and Applications, 2006. pp: 411 - 414

Topic by Sakari Luukkainen

Luukkainen Topic1. Mobile social networking

Social networking services build online connections between people who share similar interests and has created new ways for the endusers to interact. Currently millions of people use daily these services in the fixed Internet. Current social networking services use autonomous business model, where the endusers both supply and consume content. The service provider then analyses their profiles and collects the revenues from the provision of targeted advertising.

The business aspects related to the social networking solutions supporting mobility are currently of high interest. The existing service providers have started to provide mobile extensions to their solutions and mobile terminal vendors have started to be active in the service business,  too. The mobile platform provides new interesting features for the social networking like micro payments, positioning and increased reachability.

The study should cover following issues:

Review of social networking business related literature

Analysis of what can be learned from the experiences in the fixed Internet

Description of case studies related to mobile social netwrking

Identification of relevant mobile specific service features

Topics by Li Ming 

Ming Topic 1: Mechanisms for extending TCP option space for future Internet

Data Offset (DO) is a required field in every TCP header. It indicates the length of the TCP header of that segment. The DO field has been specified as: "The number of 32 bit words in the TCP Header.  This indicates where the data begins.  The TCP header (even one including options) is an integral number of 32 bits long". It allows for a maximum TCP header length of 60 bytes (15 * 4 bytes).  The required fields in a TCP header occupy a fixed 20 bytes. This only leaves 40 bytes as the maximum amount of space for use by TCP options. But 40 bytes are not enough for some TCP options or the combination of some TCP options. Find out the mechanisms for solve the space limitation of the TCP option in the literature and make a comparison among them to conclude which solution is the best one.

[1]draft-eddy-tcp-loo-01(Extending the Space Available for TCP Options)

[2]draft-kohler-tcpm-extopt-00(Extended Option Space for TCP)

Ming Topic 2: Internet accountability

Currently, the Internet is vulnerable to different kinds of malicious behaviors such as DoS attack, phishing attacks and unsolicited e-mails. If we add the accountability as a fundamental property of the Internet, could the Internet become more efficient and security? Find out the accountability solutions for the Internet, and analyze the benefits they could bring to us and whether it is enough for the future Internet.

[1] Accountable Internet Protocol (AIP). David G. Andersen, Hari Balakrishnan, Nick Feamster, Teemu Koponen, Daekyeong Moon and Scott Shenker. In Proc. SIGCOMM, Aug 2008, Seattle, WA.

Topics by Teemu Muukkonen

Muukkonen topic 1. Country Specific Energy Consumption Estimates of Information and Communication Technology (ICT)

Energy issues are an emerging concern of the ICT industry. A lot of estimates have been made of the energy consumption of ICT equipment. The task for the student is to search and analyze these estimates. The resulting paper will be a synthesis of the published estimates, and a discussion of the estimation methods used, especially their validity and

comparability.

Cremer, C et al., Energy Consumption of Information and Communication Technology (ICT) in Germany up to 2010, Fraunhofer-Institut für Systemtechnik und Innovationsforschung, January 2003,

http://publica.fraunhofer.de/eprints/urn:nbn:de:0011-n-223629.pdf

Koomey, J. G., Estimating Total Power Consumption by Servers in the U.S. and the World. Stanford University, February 2007,

http://hightech.lbl.gov/documents/DATA_CENTERS/svrpwrusecompletefinal.pdf

Muukkonen Topic 2. Information and Communication Technology (ICT) Impact on Other Areas of Life: De-materialising or Re-materialising?

Energy issues are an emerging concern of the ICT industry. Can ICT industry justify its constantly growing energy consumption by enabling other industries to save energy?

Berkout F., Herlin J., De-materialising and re-materialising: digital technologies and the environment, Futures 36 (2004) pp. 903–920, available from ScienceDirect

Gesi, SMART 2020: Enabling the Low Carbon Economy in the Information Age, Report, h http://www.theclimategroup.org/assets/resources/publications/Smart2020Report.pdf

Muukkonen topic 3. Role of x86/x64 Server Virtualization in Data Center Efficiency Development

Energy issues are an emerging concern of the ICT industry. In addition to energy availability and price issues, the data centers are striving with the heat production of the ICT equipment. Server Virtualization can help in reducing energy consumption of the servers. The task for the student is to search and analyze for publications about x86/x64 server virtualization from the energy efficiency point of view. 

Gaudin, S., Case study: Virtualization delivers a cost-saving lesson. InformationWeek, (2006). URL http://www.informationweek.com/

news/software/showArticle.jhtml;?articleID=189400429.

Murugesan S., Harnessing Green IT: Principles and Practices, IT Pro January/February 2008, pp. 24-33

http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=4446673&isnumber=4446663

Topics by Boris Nachaev

Nechaev Topic 1: Cloud computing: new trends in providing Internet-based services and resources

Cloud computing is an emerging paradigm in providing reliable, scalable, QoS-guaranteed computing services to end-users. The managed computing power, storage  and services are delivered on demand to consumers over the Internet. The goal of the study is to  identify main characteristics of cloud computing paradigm, analyze how it is different  from other computing paradigms, and overview existing platforms and services.

[1] Foster, Ian; Zhao, Yong; Raicu, Ioan; Lu, Shiyong, "Cloud Computing 

and Grid Computing 360-Degree Compared," Grid Computing Environments 

Workshop, 2008. GCE '08 , pp.1-10, 12-16 Nov. 2008

[2] Lizhe Wang; Jie Tao; Kunze, M.; Castellanos, A.C.; Kramer, D.; Karl, 

W., "Scientific Cloud Computing: Early Definition and Experience," High 

Performance Computing and Communications, 2008. HPCC '08. 10th IEEE 

International Conference on , pp.825-830, 25-27 Sept. 2008

[3] Buyya, R.; Chee Shin Yeo; Venugopal, S., "Market-Oriented Cloud 

Computing: Vision, Hype, and Reality for Delivering IT Services as 

Computing Utilities," High Performance Computing and Communications, 

2008. HPCC '08. 10th IEEE International Conference on , pp.5-13, 25-27 

Sept. 2008

Nechaev Topic 2: Software as a Service: a novel software licensing business model

Software as a Service (SaaS) is a business model of providing on-demand 

application services across the Internet. Instead of purchasing a perpetual software license, a customer is offered to buy a subscription for an application service. The goal of the research is to compare SaaS and legacy software business models, study the existing SaaS solutions, and analyze for the peculiarities of  the new paradigm influence the software development process.

[1] Sun, Wei; Zhang, Xin; Guo, Chang Jie; Sun, Pei; Su, Hui, "Software 

as a Service: Configuration and Customization Perspectives," Congress on 

Services Part II, 2008. SERVICES-2. IEEE, pp.18-25, 23-26 Sept. 2008

[2] Vidyanand Choudhary, "Software as a Service: Implications for 

Investment in Software Development," System Sciences, 2007. HICSS 2007. 

40th Annual Hawaii International Conference on , pp.209a-209a, Jan. 2007

[3] Dan Ma, "The Business Model of "Software-As-A-Service"," Services 

Computing, 2007. SCC 2007. IEEE International Conference on , 

pp.701-702, 9-13 July 2007

Nechaev Topic 3: Parallelizing data processing with MapReduce

MapReduce is a software framework developed by Google for simplifying distributed computations. The framework libraries are written in many languages which allowed for a big number of applications to be implemented on top of it. The aim of the research is to describe main features of the framework, study existing applications utilizing it and evaluate their performance, and make an overview of possible computational problems that can be  solved using MapReduce.

[1] Jeffrey Dean , Sanjay Ghemawat, "MapReduce: simplified data 

processing on large clusters", Communications of the ACM, v.51 n.1, 

January 2008

[2] Ekanayake, Jaliya; Pallickara, Shrideep; Fox, Geoffrey, "MapReduce 

for Data Intensive Scientific Analyses,"

eScience, 2008. eScience '08. IEEE Fourth International Conference on , 

pp.277-284, 7-12 Dec. 2008

[3] Kim, Kiyoung; Jeon, Kyungho; Han, Hyuck; Kim, Shin-gyu; Jung, 

Hyungsoo; Yeom, Heon Y., "MRBench: A Benchmark for MapReduce Framework," 

Parallel and Distributed Systems, 2008. ICPADS '08. 14th IEEE 

International Conference on , pp.11-18, 8-10 Dec. 2008 

Topics by Long Nguyen 

Topic slides: http://www.tkk.fi/~silver/Long Nguyen.ppt 

Nguyen Topic 1. Searching in P2P Networks: a survey

In this paper the student is expected to do literature survey about searching in P2P network. The paper should cover searching techniques in popular structured P2P networks such as Chord, Pastry, CAN. It?d be excellent if the student can do extra analysis on unstructured P2P networks, Gnutella and FastTrack for example. Comparison and analysis should be done after literature review. For example, the metrics for evaluating seaching techniques can be cost, positive hit rate, responses time.

Nguyen Topic2 Is Information-centric Networking with DTN Possible?

Information networking introduces an interesting communication model in which content providers publish their information into the network and users access information by specifying what information they are interested in; not by connecting to a specific host on the Internet and requesting a particular file stored on that host. DTN?s characteristics, by nature, fit into this model. The question is whether it is possible to build a core network in large scale with DTNs. The topic is quite tough and requires deep analysis. Any good (or crazy) idea from the student is welcome!

Application-aware DTN Routing

http://www.netlab.tkk.fi/~jo/papers/2006-08-dtn-appl-aware-routing-tr.pdf

DTN-based Content Storage and Retrieval

http://www.netlab.hut.fi/~jo/papers/2007-dtn-storage-retrieval.pdf

Topics by Nie Pin 

Nie Topic 1. An Eveluation Framework for Middlewares on Sensor Networks

[1]: "Middleware: Middleware Challenges and Approaches for Wireless Sensor Networks", http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=01621014

[2]: "Middleware Challenges for Wireless Sensor Networks", 

http://www.vs.inf.ethz.ch/res/papers/wsn-middleware.pdf

[3]: "Programming Paradigms and Middleware for Sensor Networks",

http://www.vs.inf.ethz.ch/publ/papers/middleware-kuvs.pdf

Nie Topic 2. Security Solutions on Middlewares for Sensor Networks 

Many middleware solutions have been proposed for sensor networks, but the security resolution is falling behind due to the limitations of the sensors and its application environments. This topic is to do a broad survey for possible security solutions on the middlewares for sensor networks. The result will be carefully classified regarding to the categories of middlewares.

[1]: "Autonomous Middleware Framework for Sensor Networks",

http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=01506385

[2]: "Comprehensible security synthesis for wireless sensor networks",

http://portal.acm.org/citation.cfm?id=1462698.1462702 

Topics by Prabhu Patil 

Patil Topic1. Wireless Sensor Network security - Issues and Challenges

Wireless Sensor networks are deployed to monitor the surroundingsand  keep the end-user informedabout the events witnessed. Wireless sensor network applications include oceanand wildlife monitoring, manufacturing machinery performance monitoring,building safety and earthquake monitoring, and many military applications. Aneven wider spectrum of future applications is likely to follow, including themonitoring of highway traffic, pollution, wildfires, building security, waterquality, and even people’s heart rates. A major benefit of these systems isthat they perform in-network processing to reduce large streams of raw datainto useful aggregated information. As these wireless sensor networks work in avey hostile environment, it is very critical to protect the data.Thus securitymust be an important factor to have in mind when designing the infrastructureand protocols of sensor networks. We need to understand the current issues andstate of the art solutions to deploy secured wireless sensor network.

[1] I. F. Akyildiz, W. Su, Y.Sankarasubramaniam, and E. Cayirci. A survey

on sensor networks. IEEECommunications Magazine, 40(8):102–114,August 2002.

[2] Adrian Perrig, JohnStankovic, and David Wagner. Security in wireless sensor networks COMMUNICATIONS OF THE ACM   June 2004/Vol.47, No. 6

Patil Topic2 Mobile Agentarhitecture for Wireless Sensor Networks

Mobile agents have been proposed for efficient data dissemination insensor networks. The two popular computing models for wireless sensor networksis Client-Server model and Mobile Agent model/ These are  used to facilitate collaboration amongsensor nodes.  In the traditionalclient/server-based computing architecture, data at multiple sources aretransferred to a destination; whereas in the mobile-agent based computingparadigm, a taskspecific executable code traverses the relevant sources togather data. Mobile agents can be used to greatly reduce the communicationcost, especially over low bandwidth links, by moving the processing function tothe data rather than bringing the data to a central processor.

[1] Min Chen , Taekyoung Kwon ,Yanghee Choi. Data Dissemination based on Mobile Agent in Wireless SensorNetworks, Proceedings of the The IEEE Conference on Local Computer Networks30th Anniversary, p.527-529, November15-17, 2005

[2]R. Tynan, C. Muldoon, M.J.O’Grady, & G.M.P. O’Hare. A Mobile Agent Approach to OpportunisticHarvesting in Wireless Sensor Networks.

Proc. of 7th Int. Conf. onAutonomous Agents and Multiagent Systems (AAMAS 2008), Padgham,Parkes, Müller and Parsons (eds.), May,12-16., 2008, Estoril, Portugal

[3]Wang Jietai; Xu Jiadong;Yang Shaojun. Research on Mechanism of Mobile Agent for Wireless SensorNetworks, Wireless Communications, Networking and Mobile Computing, 2007. WiCom 2007

Patil Topic 3. Middleware for wireless sensor networks : Challenges and Approaches 

Wireless sensor networks are being developed for a variety ofapplications. With the continuing advances in network and application design,appropriate middleware is needed to provide both standardized and portablesystem abstractions, and the capability to support and coordinate concurrentapplications on sensor networks. Middleware  bridges  the gapbetween applications and low-level constructs to resolve many wireless sensornetwork issues and enhancing application development.

[1] K Römer, O Kasten, FMattern . Middleware challenges for wireless sensor networks.  ACM SIGMOBILE Mobile Computing andCommunications Review, 2002

[2] Y Yu, B Krishnamachari, VKPrasanna. Issues in Designing Middleware for Wireless Sensor Networks. IEEENetwork, 2004

Topics by Oleg Ponomarev 

Slides: http://www.hiit.fi/~ponomare/ponomarev-5190-topics.ppt 

Ponomarev Topic 1. DNS Cache Poisoning and its prevention

Various flaws in the DNS software are exploited to perform a cache poisoning attack. The attackers aim to replace the target domain entries served by the non-authoritative DNS servers that can lead to various man-in-the middle attacks. The countermeasures include ACLs to deny queries from outside, random query IDs and DNSSEC deployment

2008 DNS Cache Poisoning Vulnerability, Kim Davies, ICANN, Nov 2008

http://www.iana.org/about/presentations/davies-cairo-vulnerability-081103.pdf

DNS Cache Poisoning - The Next Generation,

http://www.secureworks.com/research/articles/cachepoisoning

Ponomarev Topic 2. Are Root Servers Achilles Heal of DNS?

There are number of servers for dot domain (.) redirecting requests for a  particular top-level domain (TLD) to that TLD's nameservers. There are  currently 13 root name servers with many of them distributed using anycast to  multiple sites. Let us review their technical operations and various experince.

Root Server Technical Operations Association, http://www.root-servers.org/

Factsheet - Root server attack on 6 February 2007, ICANN 2007-03-01,

http://www.icann.org/announcements/factsheet-dns-attack-08mar07.pdf

Matsuzaki Yoshinobu, DNS Root Attack on 6th Feb 2007 - End User View, NANOG-40 Meeting, Jun 2007, 

http://www.nanog.org/mtg-0706/Presentations/lightning-maz.pdf 

Ponomarev Topic 3  What is Wrong with DNS?

The DNS is probably the second oldest (after SMTP) service that has not been replaced from the early days of the Internet. For example, telnet has been replaced with SSH and so on. Let us review the known DNS-related issues, such as limited size of UDP message, cache poisoning, bad delegations and lack of network diversity.

EDNS0, RFC2671

DNSSEC, RFC4033

DNS Threat Analysis, NLnet Labs, Mark Santcroos and Olaf M. Kolkman 

http://www.nlnetlabs.nl/downloads/se-consult.pdf

Topics by Sanna Suoranta

Suoranta Topic 1: Identity management in consumer services

Nowadays, Internet community services such as Google, Yahoo, AOL, etc have large clientele and they offer identifying service to other service providers.  What kind of technologies are used in these services? What benefits and drawbacks have these services? 

Suoranta Topic 2: Identity management for mobile devices 

Mobile phones have SIM cards to provide authentication of user for accounting purpose.  However, the SIM card is seldom used for authenticating mobile phone users for Internet services. What kind of solutions exist for mobile device identity management?

Topics by Zhihua Jin

Zhihua Topic1.     Transport Protocols

The traditionaltransport protocols such as TCP have served the Internet well for decades.Though it still works fine nowadays and probably will carry on in near future,it appears more strained to fit into the new environment. Such new environmentincludes, e.g., increased heterogeneity and mobility.

One attemptto improve the performance of transport protocols is to turn it into anadaptive service, which can dynamically configure itself according to thepolicies given by the upper application layer and the context retrieved fromthe underlying network layer.

However,before being able to orchestrate such transport service, the components (qualityattributes such as reliability, flow control, congestion control, etc.) to beorchestrated need to be understand first. The main objective is to identify theatomic quality attributes of the existing protocols and the possible conflictsbetween them, which requires thorough qualitative analysis and novel insightsinto the mechanisms utilized by those protocols.

[1] Bridges etal.: A configurable and extensible transport protocol, In IEEE/ACM TON,2007.

[2] Iren et al.:The transport layer: tutorial and survey. ACM Comput. Surv. 31, 4 (Dec. 1999), 360-404

Zhihua Topic 2. Vertical Handover Mechanisms

Today’s Internet servicesare built on a more and more heterogeneous network infrastructure. Meanwhile,more mobility is also introduced. This is making vertical handover a morecommon routine, for instance, between UMTS and WLAN.

This survey aimsto abstract a common framework to describe the existing typical verticalhandover mechanisms in terms of (though not limited to) their underlyingnetwork service context, supported application type and policy, decision strategy,evaluation criteria and result, and so on. Such framework may be fit into aself-managed component model that performs optimal handover with awareness ofboth underlying context and upper application requirements.

[1] Kassar et al.:An overview of vertical handover decision strategies in heterogeneous wirelessnetworks. Comput. Commun. 31, 10 (Jun. 2008), 2607-2620

[2] Song et al.: Asimulation study of IP-based vertical handoff in wireless convergent networks:Research Articles. Wirel. Commun. Mob. Comput. 6, 5 (Aug. 2006), 629-650

[3] Chakravorty etal.: Performance Issues with Vertical Handovers - Experiences from GPRSCellular and WLAN Hot-spots Integration. In Proceedings of the Second IEEEinternational Conference on Pervasive Computing and Communications (Percom'04)

Newer topics 

Topics by Andrey Lukyanenko

Lukyanenko Topic 1) Survey on reputation systems in p2p networks

Reputation systems are widely being studied these days, however the theoretical research differ from the practice. The practical implementations are simplify and do workaround for some critical theoretical issues. This will be a survey on existing reputation systems or algorithms with additionalanalysis on issues that were done and that were left omitted.

1. Marti, S. and Garcia-Molina, H. 2006. Taxonomy of trust: categorizing P2P reputation systems. Comput. Netw. 50, 4 (Mar. 2006), 472-484.

2. Kamvar, S. D., Schlosser, M. T., and Garcia-Molina, H. 2003. The Eigentrust algorithm for reputation management in P2P networks. In Proceedings of the 12th international Conference on World Wide Web (Budapest, Hungary, May 20 - 24, 2003).

Lukyanenko Topic 2) Can incentives overcome malicious behavior in p2p network?

For the last decade the IT field showed growth of p2p systems. These systems are aimed to be robust against bottlenecks and single-point service places. Incentives are the helping tools to share the system fairly. A set of incentives were proposed, but a number of solutions take out of the consideration malicious behavior of users.

The task of this topic is to find incentive solutions that are aimed to deal with malicious behavior.

1. Dumitriu, D., Knightly, E., Kuzmanovic, A., Stoica, I., and Zwaenepoel, W. 2005. Denial-of-service resilience in peer-to peer file sharing systems. In

Proceedings of the 2005 ACM SIGMETRICS international Conference on Measurement and Modeling of Computer Systems (Banff, Alberta, Canada, June 06 - 10, 2005). SIGMETRICS '05. ACM, New York, NY, 38-49.

2. Dinh, T. T. and Ryan, M. 2008. A Sybil-Resilient Reputation Metric for P2P Applications. In Proceedings of the 2008 international Symposium on Applications and the internet - Volume 00 (July 28 - August 01, 2008). SAINT. IEEE Computer Society, Washington, DC, 193-196.

3. Douceur, J. R. 2002. The Sybil Attack. In Revised Papers From the First

international Workshop on Peer-To-Peer Systems (March 07 - 08, 2002).

Lukyanenko Topic 3  Backoff protocol for TCP?

Backoff protocol is which was developed for Aloha, was used in Ethernet and WI-FI (IEEE802.11). It is aimed for use on shared medium. However, backoff protocol was also implemented on transport layer for TCP communication and now a recent paper of Mondal et al. states that it is not actually necessary to use. You are to study backoff implementations in TCP and different studies on it, and Try answer yourself. Is it realy necessary to use backoff in TCP? 

1. A. Mondal and A.Kuzmanovic “Removing Exponential backoff from TCP”, ACM comp. comm. review, vol. 38. n.5, Oct. 2008.

Topics by prof. Sasu Tarkoma

Tarkoma Topic 1: Google App Engine

Google App Engine is a platform for building and hosting Web applications on dedicated Google servers. The goal of this topic is to present an overview of this system and its APIs (such as the Python runtime, Datastore API, Images API, MemCache API, ..) and then discuss how it relates to current Web technologies and other hosting solutions.

http://code.google.com/appengine/

Tarkoma Topic 2: Amazon Web Services 

The Amazon Web Services are a collection of remote computing services.

The services are invoked with HTTP, using REST and SOAP protocols. The services are billed on usage. The collection includes infrastructure services such as Elastic Compute Cloud, SimpleDB, CloudFront, and Simple Queue Service. In this topic, the aim is to give an overview of the evolution of these services, discuss the underlying technology that allows the infrastructure to scale, and discuss future trends influencing this area.

http://aws.amazon.com/