|
The group runs a weekly seminar series, covering papers,
projects, and events related with our research. These seminars
are open to anyone on Chalmers or GU who wishes to attend.
Seminars 2009
Current seminar series coordinator: Daniel Cederman.
| 
iCalendar |
RSS Feed |
| Date | Time | Location | Speaker | Refreshment Organizer | Title + Abstract |
| Tue, Sep 22 | 10.00 | Conference Room | Georgios | - |
-
|
| Tue, Sep 15 | 10.00 | Conference Room | Georgios | - |
-
|
| Tue, Sep 8 | 10.00 | Conference Room | Rafia Inam | - |
A* algorithm for Multicore Graphics Processors
|
| Tue, Aug 18 | 10.00 | 5128 | Georgios | - |
A Least-Resistance Path in Reasoning about Unstructured Overlay Networks
Unstructured overlay networks for peer-to-peer applications combined with stochastic algorithms for clustering and resource location are attractive due to low-maintenance costs and inherent fault-tolerance and self-organizing properties. Moreover, there is a relatively large volume of experimental evidence that these methods are efficiency-wise a good alternative to structured methods, which require more sophisticated algorithms for maintenance and fault tolerance. However, currently there is a very limited selection of appropriate tools to use in systematically evaluating performance and other properties of such non-trivial methods.
Based on a well-known association between random walks and resistor networks, and building on a recently pointed-out connection with peer-to-peer networks, we tie-in a set of diverse techniques and metrics of both realms in a unifying framework. Furthermore, we present a basic set of tools to facilitate the analysis of overlay properties and the reasoning about algorithms for peer-to-peer networks. One of the key features of this framework is that it enables us to measure and contrast the local and global impact of algorithmic decisions in peer-to-peer networks. We provide example experimental studies that furthermore demonstrate its capabilities in the overlay network context.
|
| Tue, Jun 16 | 10.00 | Conference Room | Lander Casado | - |
Security in Wireless Sensor Networks
In this thesis we introduce ContikiSec, a secure network layer for wireless sensor networks, designed for the Contiki Operating System. ContikiSec has a configurable design, providing three different security modes: confidentiality-only (ContikiSec-Enc), authentication-only (ContikiSec-Auth) and authenticated-encryption (ContikiSec-AE). ContikiSec has been designed to balance low resource energy consumptions and security. We describe a software-based method to measure the throughput, memory footprint and energy consumption of the considered security algorithms. The design of ContikiSec design was based on performance evaluation of existing security primitives and is part of the contribution of this thesis. Our evaluation was performed in the Modular Sensor Board hardware platform for wireless sensor networks, running Contiki. Contiki is an open source, highly portable, operating system for wireless sensor networks (WSN) that is widely used in WSNs.
|
| Tue, Jun 9 | 10.00 | Conference room | Karl Molin | - |
Measurement and Analysis of the Direct Connect Peer-to-Peer File Sharing Network
Online social networks and peer-to-peer file sharing networks create a digital
mirror of human society, providing insights in social dynamics such as
interaction between entities, structural patterns and flow of information. In
the past such studies were inherently limited due to the vast supply of
information. Today these phenomena can be studied at large scale using
computers to process data from this digital mirror.
Findings from such networks have shown interesting structural properties
shared by both types of systems. In particular, it is often the case that they
show to be scale-free and small-world networks.
By letting ideas and findings from studied peer-to-peer networks guide the
design of novel architectures, improvements on user integrity, usability and
performance have been observed.
This presentation presents a study of the Direct Connect peer-to-peer file
sharing network. We model abstract tools and methods for measuring the network
architecture, and, moreover, custom software tools for data gathering and
analysis from Direct Connect networks are briefly presented.
We look at network topology and properties, statistics on user activities and
geographic distribution, characterization/statistics on data shared and
correlations of users and their shared data.
We verify the scale-free property, small-world network model, strong data
redundancy with clusters of common interest in the set of shared content, high
degree of asymmetry of connections and more.
|
| Wed, May 27 | 11.20 | 6128 | Elad Schiller | - |
Relocation Analysis of Stabilizing MAC Algorithms for Large-Scale Mobile Ad Hoc Networks
|
| Tue, May 12 | 10.00 | Conference Room | Ilhan Uludag | Nhan |
Validating new and existing MAC algorithms for VANET
|
| Tue, Apr 28 | 10.00 | Conference Room | Ning and Gongxi | Elad |
Validating relocation analysis of MAC algorithm for MANET and Evaluating New MAC algorithms for MANET
We will introduce our mobility model, a new MAC algorithm and our test bed.
|
| Tue, Mar 31 | 10.00 | Conference Room | Zhang | - |
Online batch scheduling
We present a transformation from a solution for the batching system that
has two machines with normal speed {2,s,p} on an input I
to a solution for the batching system that has one machine with double
speed {1,s/2,p/2} on the same input such
that the flowtime of the second solution is at most 2 times of that
accumulated in the first solution. This implies that one cannot find an
input sequence such that the total flowtime of the optimal solution for
{1,s/2,p/2} is more than 2 times of the
total flowtime of the optimal solution for {2,s,p}.
Concerning the scheduling two parallel classical machines, it is shown
that we can find a job sequence such that the difference of the makespans
between the solution produced by an algorithm which equally loads the
available jobs when both machines are idle, and the offline optimal
solution aiming at makespan minimization is Ω(log n).
Further, under the condition of jobs with identical processing times we
prove that, if the total flowtime of a solution using job reordering is
F, then there exists another solution that has the same total flowtime
without using job reordering. Finally, a lower bound for the competitive
ratio of online algorithms for a batching system with two identical
machines is given, that is 1+min(s,p)/2(s+p).
|
| Tue, Mar 17 | 10.00 | Conference Room | Nhan | - |
Linearizability consistency in concurrent programming
|
| Wed, Mar 11 | 13.15 | EA | Philippas Tsigas | - |
Making distributed computing more scalable, fault-tolerant and secure
The seminar will be about the research work on distributed computing, that Philippas Tsigas was involved in the last 10 years. First, he will describe their efforts to make it possible to share data between processes with the use of non-conventional synchronization methods, which allow concurrent processes to access data truly concurrently and still without producing unwanted behavior. He will also talk a little about their research efforts to support scalable, fault-tolerant tolerant and secure, interprocess communication in distributed systems like sensor networks.
|
| Tue, Mar 3 | 10.00 | Conference Room | Farnaz Moradi and Asrin Javaheri | Georgios |
Clock synchronization in sensor networks for civil security
Wireless sensor networks are widely deployed in security surveillance applications. Since most of these applications depend on having a common notion of time among the sensors, performing fine-grained clock synchronization is highly desirable. Most of the existing time synchronization approaches mainly focus on improving clock precisions and reducing energy consumption, while ignoring the effect of faults and attacks on system performance. In this thesis we show the importance of fine-grained clock synchronization and implement two of the most well-known synchronization schemes proposed in the literature. We compare these approaches considering the precision, cost and fault tolerance. Our implementations tolerate node failures and adopt newly joint nodes to the network. We also use the synchronized time to avoid message collisions by emulating TDMA-based scheduling in the synchronization protocol. Finally we describe some of the existing attacks against clock synchronization protocols and some of the possible solutions.
|
| Tue, Feb 24 | 10.00 | Conference Room | Daniel Cederman | Zhang |
Introduction to CUDA
|
| Tue, Feb 17 | 10.00 | Conference Room | Georgios Georgiadis | Daniel |
B-matchings: stable marriages for peer-to-peer networks
|
| Tue, Feb 10 | 10.00 | Conference Room | Nhan Nguyen Dang | Andreas |
Consensus, consensus number and the universality of consensus
Can we construct a wait-free linearizable implementation of any concurrent
object from a set of synchronization primitives? The answer yes or no depends
on how powerful the primitives that we have. The power of synchronization
primitive is defined based on the idea of consensus number, which is the
maximum number of threads for which objects of the class can solve an
elementary synchronization problem called consensus. From this consensus number,
one can build a hierarchy of synchronization primitives such that no primitive
at one level can be used for a wait-free or lock-free implementation of
any primitives at higher level. We can prove that one can construct a wait-free
linearizable implementation of any concurrent object in a system of n-threads
from a class of objects if and only if the class has consensus number greater
than or equal to n. This class of objects is called to be universal in n-thread
system. The talk will give the basic understanding about consensus, consensus
number and the universality of consensus.
|
| Tue, Jan 27 | 10.00 | Conference Room | Pierre Leone | Andreas |
Energy-aware routing strategies in wireless sensor networks
Wireless sensor networks are composed of small entities communicating
to each others in order to solve complex global tasks. The limitation
of the amount of energy available to the entities for functioning is
an important incentive to collaboration. In the talk we focus on
energy aware routing strategies and the problem of energy balancing.
Moreover, we discuss the relationships between energy balance routing
strategy and the max flow problem.
We also plan to describe a recently introduced process called "the
interacting urns process" and some applications to network clustering
and pseudo-coordinates computation.
|
 Top of Page
Archived Seminars
Use the following links to get access to the seminar schedules
of previous years:
2009,
2008,
2007,
2006,
2005,
2004,
2003,
2002,
2001,
2000.
Top of Page
|
