Abstract: Implementation of a commercial application to a
grid infrastructure introduces new challenges in managing the
quality-of-service (QoS) requirements, most stem from the fact
that negotiation on QoS between the user and the service provider
should strictly be satisfied. An interesting commercial application
with a wide impact on a variety of fields, which can benefit from
the computational grid technologies, is three–dimensional (3-D)
rendering. In order to implement, however, 3-D rendering to a
grid infrastructure, we should develop appropriate scheduling
and resourceallocation mechanisms so that the negotiated (QoS)
requirements are met. Efficient scheduling schemes require
modeling and prediction of rendering workload. In this paper
workload prediction is addressed based on a combined fuzzy
classification and neural network model. Initially, appropriate
descriptors are extracted to represent the synthetic world. The
descriptors are obtained by parsing RIB formatted files, which
provides a general structure for describing computer-generated
images. Fuzzy classification is used for organizing rendering
descriptor so that a reliable representation is accomplished which
increases the prediction accuracy. Neural network performs
workload prediction by modeling the nonlinear input-output
relationship between rendering descriptors and the respective
computational complexity. To increase prediction accuracy, a
constructive algorithm is adopted in this paper to train the neural
network so that network weights and size are simultaneously
estimated. Then, a grid scheduler scheme is proposed to estimate
the queuing order that the tasks should be executed and the
most appopriate processor assignment so that the demanded
QoS are satisfied as much as possible. A fair scheduling policy is
considered as the most appropriate. Experimental results on a real
grid infrastructure are presented to illustrate the efficiency of the
proposed workload prediction — scheduling algorithm compared
to other approaches presented in the literature.
Abstract: In this paper, we propose an efficient non-linear task workload prediction mechanism incorporated with a fair scheduling algorithm
for task allocation and resource management in Grid computing. Workload prediction is accomplished in a Grid middleware approach
using a non-linear model expressed as a series of finite known functional components using concepts of functional analysis. The coefficient
of functional components are obtained using a training set of appropriate samples, the pairs of which are estimated based on
a runtime estimation model relied on a least squares approximation scheme. The advantages of the proposed non-linear task workload
prediction scheme is that (i) it is not constrained by analysis of source code (analytical methods), which is practically impossible to be
implemented in complicated real-life applications or (ii) it does not exploit the variations of the workload statistics as the statistical
approaches does. The predicted task workload is then exploited by a novel scheduling algorithm, enabling a fair Quality of Service oriented
resource management so that some tasks are not favored against others. The algorithm is based on estimating the adjusted fair
completion times of the tasks for task order selection and on an earliest completion time strategy for the grid resource assignment. Experimental
results and comparisons with traditional scheduling approaches as implemented in the framework of European Union funded
research projects GRIA and GRIDLAB grid infrastructures have revealed the outperformance of the proposed method.
Abstract: Grid Infrastructures have been used to solve large scale scientific problems that do not have special requirements on QoS. However, the introduction and success of the Grids in commercial applications as well, entails the provision of QoS mechanisms which will allow for meeting the special requirements of the users-customers. In this paper we present an advanced Grid Architecture which incorporates appropriate mechanisms so as to allow guarantees of the diverse and contradictory usersrsquo QoS requirements. We present a runtime estimation model, which is the heart of any scheduling and resourceallocation algorithm, and we propose a scheme able to predict the runtime of submitted jobs for any given application on any computer by introducing a general prediction model. Experimental results are presented which indicate the robustness and reliability of the proposed architecture. The scheme has been implemented in the framework of GRIA IST project (Grid Resources for Industrial Applications).
Abstract: We study the problem of fair resourceallocation in a simple cooperative multi-agent setting where we have k agents and a set of n objects to be allocated to those agents. Each object is associated with a weight represented by a positive integer or real number. We would like to allocate all objects to the agents so that each object is allocated to only one agent and the weight is distributed fairly. We adopt the fairness index popularized by the networking community as our measure of fairness, and study centralized algorithms for fair resourceallocation. Based on the relationship between our problem and number partitioning, we devise a greedy algorithm for fair resourceallocation that runs in polynomial time but is not guaranteed to find the optimal solution, and a complete anytime algorithm that finds the optimal solution but runs in exponential time. Then we study the phase transition behavior of the complete algorithm. Finally, we demonstrate that the greedy algorithm actually performs very well and returns almost perfectly fair allocations.
Abstract: This Volume contains the 11 papers corresponding to poster and demo presentations
accepted to the 7th ACM/IEEE International Symposium on Modeling,
Analysis and Simulation ofWireless and Mobile Systems (MSWiM 04),
that is held October 4-6, 2004, in Venice, Italy.
MSWiM 2004 (http://www.cs.unibo.it/mswim2004/) is intended to provide
an international forum for original ideas, recent results and achievements on
issues and challenges related to mobile and wireless systems.
A Call for Posters was announced and widely disseminated, soliciting posters
that report on recent original results or on-going research in the area of wireless
and mobile networks. Prospective authors were encouraged to submit interesting
results on all aspects of modeling, analysis and simulation of mobile and
wireless networks and systems. The scope and topics of the Posters Session
were the same as those included in the MSWiM Call for Papers (see above).
Poster presentations were meant to provide authors with early feedback on
their research work and enable them to present their research and exchange
ideas during the Symposium.
All submissions to the call for posters as well as selected papers submitted
to MSWiM 04 were considered and reviewed. The review process resulted in
accepting the set of 11 papers included in this Volume. Accepted posters will
also be on display during the Symposium.
The set of papers in this Proceedings covers a wide range of important topics
in wireless and mobile computing, including channel allocation in wireless
networks, quality of service provisioning in IEEE 802.11 wireless LANs, IP
mobility support, energy conservation, routing in mobile adhoc networks, resource
sharing, wireless access to the WWW, sensor networks etc. The performance
evaluation techniques used include both analysis and simulation.
We hope that the poster papers included in this Volume will facilitate a fruitful
and lively discussion and exchange of interesting and creative ideas during
the Symposium.
We wish to thank the MSWiM Steering Committee Chair Azzedine Boukerche
and the Program Co-Chairs ofMSWiM 04 Carla-Fabiana Chiasserini and
Lorenzo Donatiello for their valuable help in the selection procedure. Also, the
MSWiM 04 Publicity Co-Chairs Luciano Bononi, Helen Karatza and Mirela
Sechi Moretti Annoni Notare for disseminating the Call for Posters.
We wish to warmly thank the Poster Proceedings Chair Ioannis Chatzigiannakis
for carefully doing an excellent job in preparing the Volume you now
hold in your hands.
Abstract: Future Grid Networks should be able to provide Quality of Service (QoS) guarantees
to their users. In this work we examine the way Grid resources should be
configured so as to provide deterministic delay guarantees to Guaranteed Service
(GS) users and fairness to Best Effort (BE) users. The resources are partitioned
in groups that serve GS users only, or BE users only, or both types of users with
different priorities. Furthermore, the GS users are registered to the resources
either statically or dynamically, while both single and multi-Cpu resources are
examined. Finally the proposed resource configurations for providing QoS are
implemented in theGridSim environment and a number simulations are executed.
Our results indicate that the allocation of resources to both types of users, with
different priorities, results in fewer deadlines missed and better resources utilization.
Finally benefits can be derived from the dynamic registration of GS users
to the resources
Abstract: We consider non-cooperative unsplittable congestion games where players share resources, and each player's strategy is pure and consists of a subset of the resources on which it applies a fixed weight. Such games represent unsplittable routing flow games and also job allocation games. The congestion of a resource is the sum of the weights of the players that use it and the player's cost function is the sum of the utilities of the resources on its strategy. The social cost is the total weighted sum of the player's costs. The quality of Nash equilibria is determined by the price of anarchy (PoA) which expresses how much worse is the social outcome in the worst equilibrium versus the optimal coordinated solution. In the literature the predominant work has only been on games with polynomial utility costs, where it has been proven that the price of anarchy is bounded by the degree of the polynomial. However, no results exist on general bounds for non-polynomial utility functions.
Here, we consider general versions of these games in which the utility of each resource is an arbitrary non-decreasing function of the congestion. In particular, we consider a large family of superpolynomial utility functions which are asymptotically larger than any polynomial. We demonstrate that for every such function there exist games for which the price of anarchy is unbounded and increasing with the number of players (even if they have infinitesimal weights) while network resources remain fixed. We give tight lower and upper bounds which show this dependence on the number of players. Furthermore we provide an exact characterization of the PoA of all congestion games whose utility costs are bounded above by a polynomial function. Heretofore such results existed only for games with polynomial cost functions.