5 January 2011, 2 pm
Security models for key exchange have been around for many years, but only recently have started to include consideration of denial-of-service attacks. This talk will consider security models for client puzzles and in particular introduce a new model to be presented at CT-RSA 2011. The new model incorporates the possibility that an adversary may attack multiple puzzles simultaneously. In addition we will consider the notion of gradual authentication as applied to key exchange and introduce a new mechanism combining client puzzles and digital signatures with fast verification. This is joint work with Juan Gonzalez, Lakshmi Kuppusamy, Jothi Rangasamy and Douglas Stebila.
18 January 2011, 1 pm
The WiMAX standard allows the employment of independent schedulers for the five defined QoS classes. However, the type of scheduling is not defined in the standard to give a freedom to manufacturers to employ their own scheduling policies. We have proposed a greedy-based scheduler to serve a real-time Polling service (rtPS) class. It is known that a Greedy scheduler can serve delay-sensitive applications because it qualitatively shows lower average delay, but it has a low fairness index both in terms of delay and throughput. In addition to that, its average delay in heavy load conditions is not within the required maximum latency. Thus, we propose a combined scheduling and packet dropping algorithm which is shown to bring the average delay below the required maximum latency and also significantly improves the delay fairness when compared to a Greedy scheduler.
25 January 2011, 11 am
Digital environments, such as desktop computers, mobile devices and large surfaces depend on input mechanisms or devices to engage users in interacting with them. These input devices are partly defined by the type of mode they operate under. For example, the mouse is an indirect input device that allows one to position a desktop cursor with relative displacements of the device. A tablet pen facilitates direct input and relies on absolute positioning to control the virtual cursor. Recently, researchers have proposed techniques that allow transitioning between various input modes to harness the inherent potential available in each of these. However most techniques still suffer from additional overheads and other costs associated in switching modes. In this presentation, I summarize our recent results that examine the benefits of blurring the rigid boundaries that have existed between different input modes. With examples taken from three different environments, large surfaces, mobile devices and the desktop computer, I demonstrate that blending input modes can result in significant performance gains in our interactions. I present guidelines for mixing input modes on existing or novel devices and offer suggestions on how these can be leveraged to create novel interactions. I end the talk with a brief description of other HCI projects at the University of Manitoba.
25 January 2011, 2 pm
During a wide variety of tissue remodelling events including embryonic morphogenesis, wound repair and tumour metastases, cohesive groups of cells move together as a unit. For example, epidermal cells surrounding a skin wound collectively migrate into it to seal the defect; similarly, a new blood vessel branch sprouts out by the coordinated movement of the cells constituting it. Understanding the mechanisms of these collective cell migrations is therefore of great importance, both at a fundamental cell biology level, and clinically, because it could reveal a common principle that underpins diverse biological phenomena, and it will be the basis of novel therapeutics aiming at, e.g., acceleration of wound healing or inhibition of cancer metastases. However, it is still largely unknown how the movement of entire cell group is coordinated to achieve proper tissue remodelling. To address this question, we have designed a novel image subtraction method with which we can visualise collective cell mobilisation as a "white wave" propagating in a cell sheet. This method reveals how cells are mobilised and join the stream of collectively migrating cell cohorts, and enables us to analyse the mechanism of collective cell movement. I would also like to discuss how the collective behaviours of cells can be mathematically modelled, and what new predictions can be drawn from this modelling. At the beginning of the seminar, general introductions to the biology of collective cell migration will be given.