This thesis provides an introduction to the MMORPG genre, and the challenges of engineering a networking system supporting these games, as well as the state of the art titles that exist on the market today. Further it describes the design and development of a peer-to-peer networking system to support MMORPG games, and basic testing of this system. It focuses on a broad theoretical approach to provide a solid background to understand the options available and choices made in the design of the system. The thesis presents a hybrid peer-to-peer concept that aims to reduce costs of operating an MMORPG, allowing smaller game developers to compete against major titles. It includes the main features that should be enabled in a distributed MMORPG architecture. The concept is designed to be flexible in terms of further development, allowing new features to be added with ease and providing game designers with multiple options. A proof-of-concept demo is implemented in Java, displaying the features through a simple interface, and tests showed that the concept has potential to challenge the client-server solutions that are dominating the market today.

Biometric technology is rapidly evolving in today's society. A large part of the technology has its roots hundreds, or even thousands of years back in time, while other parts are new and futuristic. Research suggest that individuals can be identified by the way they walk, and this kind of biometrics, gait biometrics, is a rather new and definitely intriguing field. However, the technology is far from mature; the performance is not generally competitive to other biometrics, and it has not been thoroughly tested security-wise. This thesis aims to test the security strength of gait biometrics. It will focus on imitation, or mimicking of gait. The bottom line question is whether it is possible to learn to walk like someone else. If this turns out to be easy, it will have a severe effect on the potential of gait as an authentication mechanism in the future. The report is logically twofold. In one part, the reader is brought up to speed on the field of gait biometrics, and a software tool for gait authentication is developed and presented. Second, an experiment is conducted, involving extensive training of test subjects, and using sources of feedback like video and statistical analysis. The data is analyzed by regression, and the goal is to determine whether or not the participants are increasing their mimicking skills, or simply put: if they are learning. The first part of the experiment involves 50 participants that are successfully enrolled using the developed software. The results compete with state of the art gait technology, with an EER of 6.2%. The rest of the experiment is related to mimicking, and the thesis discovers that six out of seven participants seem to have a natural boundary to their performance, a "plateau", forcing them back whenever they attempt to improve further. The location of this plateau predetermines the outcome of an attack; for success it has to lie below the acceptance threshold corresponding to the EER. Exactly one such boundary is identified for almost all participants, but some data also indicate that more than one plateau can exist simultaneously. The final result however, is that a very limited amount of learning is present, not nearly enough to pose a threat to gait biometrics. Gait mimicking is a hard task, and our physiology works against us when we try to adopt specific gait characteristics.

Disk encryption has changed during the last decade from being a mechanism only used by corporate executives and government agencies for their top secret information, to become trivially feasible for everyone to utilize. One of the forces that have been driving this development is the steady flow of new cryptographic primitives such as tweakable narrow- and wide-block ciphers specifically designed for disk encryption implementations. One such tweakable narrow-block cipher is XTS-AES, which is standardized in IEEE Std 1619 and very recently accepted by NIST as an approved mode of operation for AES under FIPS-140. In the first part of this thesis, we study principles and investigate methods used in disk encryption. We present the different implementation types of disk encryption (hardware-based versus software-based, wide-block versus narrow-block, authenticated versus transparent), commonly discussed modes of operations (LRW, XEX, MCB, CMC, EME, XCB), and briefly review some open-source software implementations of disk encryption (TrueCrypt, FreeOTFE, dm-crypt). In the second part of this thesis we provide a thorough examination of XTS-AES, describing both its security and real-world performance. To our knowledge, this is the first scientific work to provide an elaborate description of XTS-AES while also assessing its real-world performance. Our work show that introducing XTS-AES-256 full system disk encryption using TrueCrypt 6.1a on Windows yield a decrease in write and read speed of up to -35% (average for Windows XP, Windows Vista, and Windows 7 Beta). Further, our results also show that disk operations that uses approximately 2% of the CPU resources when no disk encryption is present, takes up to 50% of the CPU resources when full system disk encryption is deployed.

In this thesis the Virtual Ideal Functionality Framework (VIFF) has been used for implementing practical applications based on multiparty computation (MPC). The work started with a theoretical study of secret sharing and multiparty computation. VIFF, its functionality and mode of operation were also thoroughly investigated. In addition, an overview of electronic voting, challenges and typical security requirements have been briefly examined. The background study led to the development of two applications. The first was a small GUI program in which participants on some scientific paper can vote for their choice of 1th, 2th, 3th, etc. author. The second was a secure web voting application accessible from the world wide web.

Setting up large multi-projector arrays today usually come at a cost; manual calibration of each projector requires time. The orientation of the image from each projector must be correctly aligned in six axes to make the final projected output fit the screen. Not all aspects of calibrating projectors are possible to correct, consumer hardware usually only covers two or three of the axes, the remainder are often corrected using clever projector placement. Also, the degree of which it is possible to adjust is also limited, decreasing placement flexibility. As the collaboration surfaces in Hems lab requires a large number of projectors to work seamlessly together, good calibration techniques are required in order to keep setup and maintenance time low, while giving highly accurate calibration results. Abstract By creating a software demonstrator that automates much of the calibration and enables quick and easy setup, I have made possible rapid prototyping, testing and demonstration of multi-projector arrays, both with single and stereoscopic views. As I will prove, this software shows a flexible approach that may be of use, not only to the Caruso lab and future Hems lab, but may also be used in other settings where projector technology up to this date still has not been widely used, by overcoming the calibration and image warping hurdles and limitations. Abstract The software is developed with basis in the OpenCV computer vision library, and implemented in Python. Tests show that calibration time for a single projector may be cut down to a matter of seconds, regardless of the placement of the projector in relation to the screen, whereas traditional calibration often still not reach the same level of accuracy even if taking tens of minutes or require repositioning of the projector to compensate for the lack of adjustment possibilities.

Plan and carry out perception tests of time-varying scene composition, spatial and temporal resolution of scene objects, respectively, as defined in DMP. Base the time-variation on short time intervals of only tenths of milliseconds. Use and extend test setup as in 'The Hems Lab - Perceptual test of scene objects with variable temporal resolution' project, autumn 2008. Find an economic model that can be used to help businesses assess market potential of new technology (software, hardware) and show how this can be used in a DMP setting.

Unauthorized access to computers and theft of proprietary information are two problems leading to large economical losses for organizations around the world. Thousands of laptops often containing vital information are lost at airports every day. Organizations and people in general often believe that the sensitive information is inaccessible because of the login mechanism. Even though we demand that our systems should prevent unauthorized access, we also expect that the access to the operating system can be restored when a password is lost. We require that authorized persons can regain the access to the computer, while unauthorized persons are prevented access. A good solution to reset or recover the Administrator password should exist on all operating systems. This thesis begins with addressing weaknesses in 8 different operating systems. It presents a comprehensive step-by-step guide for already existing procedures and tools that can be used to reset or recover the Administrator password. In total 6 procedures and 10 tools are presented. Because some procedures required a lot of interaction from the user, we decided to automate these and include them in our self made tool named Yet Another Local Password (tool) (YALP). We were able to reset the passwords on all of the 8 operating systems. On some of the operating systems only a few passwords were recovered, and based on that, a more comprehensive password recovery study was desirable. It should be noted that even though Microsoft has introduced a more secure password handling mechanism on newer Windows operating system, many persons and corporations still use the outdated Windows XP operating system. This is partly because Windows Vista has been criticized for its weak performance. An empirical password study was carried out to see what percentage of 30 carefully chosen passwords could be revealed. Some disturbing results were obtained. During a period of 8 hours, 100% of the passwords created on a Windows XP system were revealed. The results from this study show that the use of password as an authentication mechanism for operating systems will not offer sufficient protection in the future, and that other mechanisms have to be considered. Based on results obtained from this thesis, a paper named All in a day's work: Password cracking for the rest of us was submitted to The 14th Nordic Conference on Secure IT Systems, NordSec 2009, in Oslo, Norway. In addition, a poster named Generation of Rainbow Tables was presented at The 8th Annual Meeting on High Performance Computing and Infrastructure, NOTUR2009, in Trondheim, Norway. The poster and the paper are included in Appendix F and Appendix G, respectively.

Semantic Web Services (SWS) are a facility towards full automation of service usage, providing seamless integration of services that are published and accessible on the Web. Based on Semanic Web technology SWS is simply a semantic annotation of the functionalitites and interfaces of Web Services. In the very same way that ontologies and metadata lanaguages will facilitate the integration of static data on the Web, the annotation of services wil help to facilitate the automation of service discovery, service composition, service contracting, and execution.In this thesis we demonstrate how SWS technology can be applied to a network management system (NMS), which can install SNMP managers during run-time in systems running TAPAS platform. Several reasoning applications are made and integrated with the existing system. In addition, we specify a set of Semantic Web Services described using OWL-S, in order to execute these applications.

Reactive systems that provide services to an environment typically interact with numerous users or other components. Session multiplicity enables a component to keep track of these interactions by handling each of them as separate conversations. Session reflection is the ability to look into the state and properties of these conversations at run time, and use that for deciding the actions to be taken when certain events occur. This thesis addresses how to support session multiplicity and reflection during code generation of executable state machines and a runtime support system that can execute them. Using code generation, certain UML composite structures and state machines can be transformed to deployable, executable components automatically. To support these features, an interface for using them is offered to the components that comprise an application and some internal mechanisms have been added to a runtime support system. The interface includes methods for sending messages, creating new sessions and session state machines and retrieving information about run time state machine instances. The internal mechanisms include keeping track of components and sessions, giving intra-component messages priority, creating new session state machine instances and changing the addressing scheme and routing mechanisms to include session state machines. To put the thesis' results into context and to some extent prove that they are good, a proof of concept, multi-player rock-paper-scissors game has been implemented.

This thesis describes the basic theory of multiparty computation (MPC) in addition to a fully functional distributed Rivest-Shamir-Adleman (RSA) protocol for three players implemented in Virtual Ideal Functionality Framework (VIFF) using secure MPC (SMPC). MPC can be used to solve problems where n players, each with a private input xi, wants to compute a function with public output f(x1, x2,…, xn) = y, such that the private inputs remains secret for each player, but the output y is public. A cornerstone in MPC is the concept of secret sharing. In secret sharing, a dealer has a secret and gives each participating player a share of the secret in such a way that a certain number of the players are needed in order to reconstruct the secret. The number of players needed to reconstruct the secret is referred to as the threshold of the scheme. VIFF is a high level programming framework that allows programmers to create applications using SMPC for any number of players in an easy, efficient and secure manner. The distributed RSA solution implemented in VIFF includes distributed key generation, decryption and signature, which are the main functions needed for the distributed RSA scheme, and is based on the distributed RSA algorithm proposed by Dan Boneh and Matthew Franklin in 1997. Four improvements compared to Boneh and Franklin's algorithm are described, two related to the run-time and two related to the security of the algorithm. The run-time improvements are regarding the distributed trial division step and the local trial division on the revealed N, both implemented. The security improvements are related to the way a random number is used to secure a revealed number. The first security improvement is related to the distributed trial division, whereas the second security improvement is regarding the alternative step in the biprimality test. The first security improvement, which is also the more important of the two, has been implemented in this thesis. At last some benchmarks regarding the key generation, decryption and signature process are presented, which indicates that the current implementation is best suited for scenarios where the distributed RSA keys can be generated in advance, whereas the decryption and signature process is fast enough for any type of scenario. The key generation process can become much faster with a few adjustments described at the end of the thesis.