Cláudio Fernando Resin Geyer

Toward a General Software Infrastructure for Ubiquitous Computing

This general software architecture is designed to support ubiquitous computings fundamental challenges, helping the community develop and assess middleware and frameworks for this area.

Mobile and ubiquitous computing in an innovative undergraduate course

The increasing use of mobile devices and the dissemination of wireless networks have stimulated mobile and ubiquitous computing research. In this context, education is being considered one of the main application areas. This paper proposes the use of mobile and ubiquitous computing to support and improve learning in a new kind of academic structure called Undergraduate Course of Reference (nicknamed GRefe). The GRefe was proposed in Unisinos, a university located in south of Brazil. Currently, there are four GRefes. These courses are organized in Learning Programs and Learning Projects. They use a practical and multidisciplinary approach to stimulate the learning. We proposed the use of mobile and ubiquitous computing technology to articulate and improve the academic activities of a specific GRefe called Computer Engineering. We believe that GRefe organization simplified and stimulated the use of these technologies in a learning environment.

A load balancing scheme for massively multiplayer online games

In a distributed MMOG (massively multiplayer online game) server architecture, the server nodes may become easily overloaded by the high demand from the players for state updates. Many works propose algorithms to distribute the load on the server nodes, but this load is usually defined as the number of players on each server, what is not an ideal measure. Also, the possible heterogeneity of the system is frequently overlooked. We propose a balancing scheme with two main goals: allocate load on server nodes proportionally to each one’s power and reduce the inter-server communication overhead, considering the load as the occupied bandwidth of each server. Four algorithms were proposed, from which ProGReGA is the best for overhead reduction and ProGReGA-KF is the most suited for reducing player migrations between servers. We also make a review of related works and some comparisons were made, where our approach performed better.

MRSG - A MapReduce simulator over SimGrid

MapReduce is a parallel programming model to process large datasets, and it was inspired by the Map and Reduce primitives from functional languages. Its first implementation was designed to run on large clusters of homogeneous machines. Though, in the last years, the model was ported to different types of environments, such as desktop grid and volunteer computing. To obtain a good performance in these environments, however, it is necessary to adapt some framework mechanisms, such as scheduling and data distribution algorithms. In this paper we present the MRSG simulator, which reproduces the MapReduce work-flow on top of the SimGrid simulation toolkit, and provides an API to implement and evaluate these new algorithms and policies for MapReduce. To evaluate the simulator, we compared its behavior against a real Hadoop MapReduce deployment. The results show an important similarity between the simulated and real executions.

Globaledu - An Architecture to Support Learning in a Pervasive Computing Environment

This paper presents a propose for support learning in a pervasive computing environment. It is composed of educational service agents (ESA) and a pervasive personal pedagogical agent (P3A). ESA will support educational applications using the pervasive environment ISAM. Agents will be accessed through P3A, which shall always be with the learner (the agent moves to devices that the learner is using - desktop, cellular phone, handhelds, etc), assisting the process of global learning through the identification and adaptation of the resources in agreement with the learner’s educational model composed of the learner model, learner context and learning object model.

MRA++: Scheduling and data placement on MapReduce for heterogeneous environments

MapReduce has emerged as a popular programming model in the field of data-intensive computing. This is due to its simplistic design, which provides ease of use for programmers, and its framework implementations such as Hadoop, which have been adopted by large business and technology companies. In this paper we make some improvements to the Hadoop MapReduce framework by introducing algorithms that are suitable for heterogeneous environments. The goal is to efficiently perform data-intensive computing in heterogeneous environments. The need for these adaptations derives from the fact that, following the framework design proposed by Google, Hadoop is optimized to run in large homogeneous clusters. Hence we propose MRA++, a new MapReduce framework design that considers the heterogeneity of nodes during data distribution, task scheduling and job control. MRA++establishes a training task to gather information prior to the data distribution. However, we show that the delay introduced in the setup phase is offset by the effectiveness of the mechanisms and algorithms, that achieve performance gains of more than 70% in 10 Mbps networks.

FreeMMG: a hybrid peer-to-peer and client-server model for massively multiplayer games

A Massively Multiplayer Game (MMG) is an Internet-based distributed game (or multiplayer game) that supports a large amount of simultaneous players interacting, in real-time, in a persistent virtual world. Practically all of todays MMGs employ a client-server distribution approach where the client is trusted only to send interaction requests and receive updates about the current state of the virtual world (position updates, etc). One of the main reasons that drive this choice of a strongly centralized architecture is the protection it offers against player cheating. Examples of successful MMGs that routinely support several thousand simultaneous players include EverQuest and Lineage.

A3: A Novel Interest Management Algorithm for Distributed Simulations of MMOGs

Traditionally, a central server is utilized to provide support to MMOGs (massively multiplayer online games), where the number of participants is in the order of tens of thousands. Much work has been done trying to create a fully peer-to-peer model to support this kind of application, in order to minimize the maintenance cost of its infrastructure, but critical questions remain. Examples of the problems relative to peer-to-peer MMOG support systems are: vulnerability to cheating, overload of the upload links of the peers and difficulty to maintain consistency of the simulation among the participants. In this work, we propose the utilization of geographically distributed lower-cost nodes, working as a distributed game server. The distribution model and some related works are also presented. To address the communication cost imposed to the servers, we specify the A3 algorithm, which is a novel refinement of the interest management technique, significantly reducing the necessary bandwidth. Simulations have been made with ns-2 and their results demonstrate that our approach achieves the least bandwidth utilization, with a 33.10% maximum traffic reduction and 33.58% average traffic reduction, when compared to other algorithms.

Context-Aware Model in a Ubiquitous Learning Environment

The idea of ubiquitous learning is to create a network of devices, people and situation that allows learning experiences to play out. This idea is attractive but it is not easily implemented. The GlobalEdu is a ubiquitous learning architecture integrated in ISAM, a software architecture that manages a large-scale pervasive environment. This paper presents a context-aware model implemented in GlobalEdu. Especially, we present the context management educational service and its interface with a pedagogical agent. A scenario application will be present too

Resource scheduling on grid: handling uncertainty

The heterogeneity and the highly dynamic behavior of the grid computing demands a new class of scheduling strategies. These strategies must not only consider the instantaneous information provided by sensors, but also probabilities of the information received to keep its values in a near future.