This paper gives two contributions; First, it presents an architecture for customized content delivery for Ambient Intelligent Environments. We demonstrate how physical peers made up of a Bluetooth-based network of Java-enabled mobile phones can be used to provide customized content delivery from the web without the need of a dedicated web connection per device. Secondly, we present two algorithms Self-OrganiziNG random walkerS (SONGS) and Peerto-peeR self-organIZ ed tEmporary overlayS (PRIZES), both providing mechanisms of temporary overlay formation in limited connectivity ad-hoc networks. SONGS is an extension of k-random walk algorithm whereas PRIZES is a forest-fire type flooding mechanism. We then show how adding even naive self-organization to these algorithms significantly improves the leftover queries as well as latency in terms of hop-counts.

Over the last decade, clusters have become the fastest growing platforms in high-performance computing. More recently, Grids were emerging as next generation computing platforms for large-scale computation and data intensive problems in industry, academic, and government organizations. Meanwhile, an increasing number of real-time applications running on clusters and Grids have mandatory security requirements in addition to stringent timing constraints. Conventional real-time scheduling algorithms developed for clusters and Grids, however, either disregard applications ’ security needs, and thus expose the applications to security threats, or run applications at inferior security levels without optimizing security performance. In recognition that many applications running on clusters and Grids demand both real-time performance and security, in this dissertation research we investigate the problem of scheduling real-time applications with various security requirements. First, we propose a security middleware model (or SMW for short) from which security-sensitive real-time applications are enabled to exploit a variety of security

This paper is aimed at designing efficient parallel matrix-product algorithms for heterogeneous master-worker platforms. While matrix-product is well-understood for homogeneous 2D-arrays of processors (e.g., Cannon algorithm and ScaLAPACK outer product algorithm), there are three key hypotheses that render our work original and innovative:- Centralized data. We assume that all matrix files originate from, and must be returned to, the master. The master distributes both data and computations to the workers (while in ScaLA-PACK, input and output matrices are initially distributed among participating resources). Typically, our approach is useful in the context of speeding up MATLAB or SCILAB clients running on a server (which acts as the master and initial repository of files).- Heterogeneous star-shaped platforms. We target fully heterogeneous platforms, where computational resources have different computing powers. Also, the workers are connected to the master by links of different capacities. This framework is realistic when deploying the application from the server, which is responsible for enrolling authorized resources.- Limited memory. Because we investigate the parallelization of large problems, we cannot assume that full matrix panels can be stored in the worker memories and re-used for subsequent updates (as in ScaLAPACK). The amount of memory available in each worker is expressed as a given number mi of buffers, where a buffer can store a square block of matrix elements. The size q of these square blocks is chosen so as to harness the power of Level 3 BLAS routines: q = 80 or 100 on most platforms. We have devised efficient algorithms for resource selection (deciding which workers to enroll) and communication ordering (both for input and result messages), and we report a set of numerical experiments on various platforms at École Normale Supérieure de Lyon and the University of Tennessee. However, we point out that in this first version of the report, experiments are limited to homogeneous platforms. 1 1

To date closed form solutions for optimal finish time and job allocation are largely obtained only for network topologies with a single load originating (root) processor. However in large-scale data intensive problems with geographically distributed resources, load is generated from multiple sources. This paper introduces a new divisible load scheduling strategy for single level tree networks with two load originating processors. Solutions for an optimal allocation of fractions of load to nodes in single level tree networks are obtained via linear programming. A unique scheduling strategy that allows one to obtain closed form solutions for the optimal finish time and load allocation for each processor in the network is also presented. The Preprint submitted to Elsevier Science 30 June 2009tradeoff between linear programming and closed form solutions in terms of underlying assumptions is examined. Finally, a performance evaluation of a two source homogeneous single level tree network with concurrent communication strategy is presented.

In previous works we have proposed to use of selforganization based on emergent design as a model for the programming of very large aggregates of heterogeneous computing resources. In our approach, a large scale computation is divided into small independent units of computation, each provided with its own uniform, autonomous behavior; only local information is used by each unit of computation to take all the decisions needed to carry out the computation. One of the challenges of this novel approach is to provide some theoretical foundation that can assist in the rational design of new systems. In this paper is to demonstrate the use of combinatorial techniques for obtaining quantitative analytical models of the organization pattern emerging from a specific type of self-organizing computation. Specifically, in a previous experiment we have demonstrated a computation in which mobile agents organize themselves around an overlay tree, that constantly restructures itself in response to changing node availability and performance levels. In this paper we derive an analytical expression describing how nodes distribute themselves over the tree based on their performance, in a simplified version of the above problem. This result represents an instance of a theoretical tool that can be used to predict global patterns emerging as a result of a selforganizing design, and to establish a direct connection between global features and local behavior parameters.

computational

Meta-scheduling in a Grid is aimed at enabling the efficient sharing of computing resources managed by different local schedulers within a single organization or scattered across several administrative domains. Since current Grid metaschedulers operate in a centralized fashion and thus are single points of failure, we present a distributed meta-scheduler for a service-oriented Grid environment based on peer-topeer (P2P) networking techniques and ant colony optimization algorithms adapted to a P2P network. In the proposed approach, the meta-scheduling process provides automatic load balancing, is completely decentralized, fault tolerant, scalable, and does not require complex administration. Experimental results demonstrate that scheduling decisions are made quickly and lead to a good balance of the computational load.

The amount of genomic data available for study is increasing [1] at a rate similar to that of Moore’s Law [2]. This deluge of data is challenging bioinformaticians to develop newer, faster and better algorithms for analysis and examination of this data. The growing availability of large scale computing grids coupled with high-performance networking [3] is challenging computer scientists to develop better, faster methods of exploiting parallelism in these biological computations and deploying them across computing grids. In this paper, we describe two computations that are required to be run frequently and which require large amounts of computing resource to complete in a reasonable time. The data for these computations are very large and the sequential computational time can exceed thousands of hours. We show the importance and relevance of these computations, the nature of the data and parallelism and we show how we are meeting the challenge of

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: