Ait-Salaht F.,LIP6 |
Castel-Taleb H.,Orange S.A. |
Fourneau J.-M.,University of Versailles |
Pekergin N.,University Paris Est Creteil
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2017
We generalise to non stationary traffics an approach that we have previously proposed to derive performance bounds of a queue under histogram-based input traffics. We use strong stochastic ordering to derive stochastic bounds on the queue length and the output traffic. These bounds are valid for transient distributions of these measures and also for the steady-state distributions when they exist. We provide some numerical techniques under arrivals modelled by a Switched Batch Bernoulli Process (SBBP). Unlike approximate methods, these bounds can be used to check if the Quality of Service constraints are satisfied or not. Our approach provides a tradeoff between the accuracy of results and the computational complexity and it is much faster than the histogram-based simulation proposed in the literature. © Springer International Publishing AG 2017.
Yu S.-Y.,LIP6 |
Mahanti A.,University of Auckland |
Gong M.,Mount Royal College, Calgary
2016 IEEE 35th International Performance Computing and Communications Conference, IPCCC 2016 | Year: 2016
Measuring quality of Internet access is important because it provides rich information on capability of Internet services and user experience. As Internet services actively evolve, users tend to spend more time and require more network capacity over time. To meet the needs of consumers, there is a wide mix of ISPs and access technologies offered in New Zealand. Benchmarking ISPs in New Zealand for their quality of service enables us to predict what users may experience among different ISPs. In this paper, we present a study of broadband service performance in New Zealand ISPs. Our results will provide accurate information on the quality of service users experience from New Zealand ISPs and helps both ISPs and Government to understand current Internet service market. © 2016 IEEE.
Li Y.,LIP6 |
Constantin C.,LIP6 |
Du Mouza C.,French National Conservatory of Arts and Crafts
ACM International Conference Proceeding Series | Year: 2017
Several distributed frameworks have recently emerged to perform computations on large-scale graphs. However some recent studies have highlighted that vertex-partitioning approaches, e.g. Gi-raph, failed to achieve workload-balanced partitioning for skewed graphs, typically having a heavy-tail degree distribution. While edge-partitioning approaches such as PowerGraph and GraphX provide better balancing and performances for graph computation, they supply a generic framework, independent from the computation. This demonstration presents SGVCut to display our edge partitions designed for random walks-based computation, which is the foundation of many graph algorithms, on skewed graphs. The demonstration scenario introduces SGVCut interface and illustrates the benefits of our approach compare to other partitioning strategies for different settings and algorithms. © 2017 ACM.
News Article | October 28, 2016
MENLO PARK, CA--(Marketwired - October 24, 2016) - ONOS Project, the rapidly growing open source community advancing the software-defined networking (SDN) OS for service providers with high availability, scale and performance and the right abstractions to create apps and services, today announced BII, Canonical, NCTU and UPMC have joined as collaborators. To build on the continued project momentum, more than 215 ONOS® developers focused on coding and testing will converge in Paris from November 2-4 for the project's first large-scale developer summit, ONOS Build 2016. "ONOS Build 2016 will offer the unique opportunity for ONOS users and contributors across the globe to meet, align, plan and hack together in-person," said Bill Snow, VP of engineering for ON.Lab. "Open to members and non-members alike, attendees will have the chance to talk directly with the ONOS core architects and Technical Steering Team, collaborate on ideas that will directly impact the future of ONOS, and promote their work among the community." The ONOS Build 2016 agenda includes a mix of sharing the ONOS roadmap, social activities for the community to bond, and hacking activities to help deliver specific features in the pipeline. The event will feature keynotes and industry panelists from ONOS community partners and collaborators Ciena, Create-Net, DT, GARR, Huawei, NOKIA Bell Labs, NTT Communications, ON.Lab, Radisys and UPMC. Topics for discussion will include basics of the ONOS architecture, northbound and southbound protocols, deployments, application development, performance and testing, community planning and the Ambassador program. Click here to learn more about the confirmed speakers. A separate Community Showcase track will be devoted to members and ONOS Brigade teams presenting updates on interesting new features developed by and for the ONOS community through open source collaboration. For example, the Dynamic Configuration Brigade has focused on building a vendor-agnostic driver into ONOS that allows automatic discovery and activation of NETCONF-enabled devices and services into the network as long as they expose a Yang Model. Click here to read more about the ONOS Brigades, and here to view the complete ONOS Build 2016 agenda. To obtain an ONOS Build ticket, complete the online registration form. ONOS Build 2016 is free for students and ONOS contributors. General admission costs $400 USD. ONOS is grateful for its generous sponsors who are helping organize the event. These include Platinum sponsors Ciena, Fujitsu, and Huawei, Gold sponsors Kisti and Radisys, and Silver sponsors ADARA Networks and Gandi. If interested in sponsoring ONOS Build 2016 or future Build events, please contact firstname.lastname@example.org. More about the new ONOS collaborators: BII: BII Group Holdings Ltd. centered on the core of technical research and development, seated on the basis of testing service, and navigated by internationalization and marketization, is a global open platform for internet infrastructure technologies with a focus on IPv6, Domain Name System (DNS), SDN and Internet of Things (IoT). Canonical: Canonical is the company behind Ubuntu, the leading OS for container, cloud, scale-out and hyperscale computing. 65% of large-scale OpenStack deployments are on Ubuntu, using both KVM and the pure-container LXD hypervisor for the world's fastest private clouds. Canonical provides enterprise support and services for commercial users of Ubuntu. NCTU: National Chiao Tung University (NCTU) is a prestigious university known for its special strengths in the areas of computer science and electronics engineering. Its highly-reputable computer science department, which is the largest one in Taiwan, will participate in the ONOS Project with a focus on SDN-IP. UPMC: The University Pierre and Marie Curie LIP6 computer science research laboratory is dedicated to application modeling and testing, as well as implementation and validation through academic and industry partnerships. Currently, LIP6 is strategizing data consistency for distributed network control-plane systems, and enhancing SDN southbound interfaces. Redefining network economics, ONOS provides the only SDN control plane that can support both disruptive and incremental SDN for service providers and enterprises seeking to virtualize and optimize their networks to keep agile pace with the explosion of mobile devices, video and big data applications. The rapidly growing and diverse ONOS community comprises a core engineering team at ON.Lab, along with developers from service providers, vendors and Research and Educational Networks spanning across industries. Whether an individual or an organization, as an open source project all are encouraged to get involved with the growing ONOS community and help contribute to the project today. ONOS® is the open source SDN networking operating system for Service Provider networks architected for high performance, scale and availability. The ONOS ecosystem comprises ON.Lab, organizations that are funding and contributing to the ONOS initiative, and individual contributors. These organizations include AT&T, China Unicom, Comcast, Google, NTT Communications Corp., SK Telecom Co. Ltd., Verizon, Ciena Corporation, Cisco Systems, Inc., Ericsson, Fujitsu Ltd., Huawei Technologies Co. Ltd., Intel Corporation, NEC Corporation, Nokia, Radisys and Samsung. See the full list of members, including ONOS' collaborators, and learn how you can get involved with ONOS at onosproject.org. ONOS is an independently funded software project hosted by The Linux Foundation, the nonprofit advancing professional open source management for mass collaboration to fuel innovation across industries and ecosystems.
Crespelle C.,LIP6 |
Todinca I.,University of Orléans
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2010
The minimal interval completion problem consists in adding edges to an arbitrary graph so that the resulting graph is an interval graph; the objective is to add an inclusion minimal set of edges, which means that no proper subset of the added edges can result in an interval graph when added to the original graph. We give an O(n2)-time algorithm to obtain a minimal interval completion of an arbitrary graph. This improves the previous O(nm) time bound for the problem and lower this bound for the first time below the best known bound for minimal chordal completion. © 2010 Springer-Verlag.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2013
The aim of this paper is to provide a unifying axiomatic justification for a class of qualitative decision models comprising among others optimistic/pessimistic qualitative utilities, binary possibilistic utility, likelihood-based utility, Spohn's disbelief function-based utility. All those criteria that are instances of Algebraic Expected Utility have been shown to be counterparts of Expected Utility thanks to a unifying axiomatization in a von Neumann-Morgenstern setting when non probabilistic decomposable uncertainty measures are used. Those criteria are based on (⊕,⊗) operators, counterpart of (+, ×) used by Expected Utility, where ⊕ is an idempotent operator and ⊗ is a triangular norm. The axiomatization is lead in the Savage setting which is a more general setting than that of von Neumann-Morgenstern as here we do not assume that the uncertainty representation of the decision-maker is known. © 2013 Springer-Verlag.
Bodini O.,LIP6 |
Bodini O.,University Paris Diderot |
Gardy D.,University of Versailles |
Gittenberger B.,Institute of Discrete Mathematics and Geometry
8th Workshop on Analytic Algorithmics and Combinatorics 2011, ANALCO 2011 | Year: 2011
We aim at the asymptotic enumeration of lambda-terms of a given size where the order of nesting of abstractions is bounded whereas the size is tending to infinity. This is done by means of a generating function approach and singularity analysis. The generating functions appear to be composed of nested square roots which exhibit unexpected phenomena. We derive the asymptotic number of such lambda-terms and it turns out that the order depends on the bound of the height. Furthermore, we present some observations when generating such lambda randomly and explain why powerful tools for random generation, such as Boltzmann samplers, face serious difficulties in generating lambda-terms. © Copyright (2011) by SIAM: Society for Industrial and Applied Mathematics. All rights reserved.
Carpentier G.,IRCAM |
Assayag G.,IRCAM |
Journal of Heuristics | Year: 2010
In this paper a computational approach of musical orchestration is presented. We consider orchestration as the search of relevant sound combinations within large instruments sample databases and propose two cooperating metaheuristics to solve this problem. Orchestration is seen here as a particular case of finding optimal constrained multisets on a large ensemble with respect to several objectives. We suggest a generic and easily extendible formalization of orchestration as a constrained multiobjective search towards a target timbre, in which several perceptual dimensions are jointly optimized. We introduce Orchidée, a time-efficient evolutionary orchestration algorithm that allows the discovery of optimal solutions and favors the exploration of non-intuitive sound mixtures. We also define a formal framework for global constraints specification and introduce the innovative CDCSolver repair metaheuristic, thanks to which the search is led towards regions fulfilling a set of musical-related requirements. Evaluation of our approach on a wide set of real orchestration problems is also provided. © Springer Science+Business Media, LLC 2009.
Frontiers in Artificial Intelligence and Applications | Year: 2012
Setting the values of rewards in Markov decision processes (MDP) may be a difficult task. In this paper, we consider two ordinal decision models for MDPs where only an order is known over rewards. The first one, which has been proposed recently in MDPs , defines preferences with respect to a reference point. The second model, which can been viewed as the dual approach of the first one, is based on quantiles. Based on the first decision model, we give a new interpretation of rewards in standard MDPs, which sheds some interesting light on the preference system used in standard MDPs. The second model based on quantile optimization is a new approach in MDPs with ordinal rewards. Although quantile-based optimality is state-dependent, we prove that an optimal stationary deterministic policy exists for a given initial state. Finally, we propose solution methods based on linear programming for optimizing quantiles. © 2012 The Author(s).
Galand L.,University of Paris Dauphine |
Lesca J.,LIP6 |
IJCAI International Joint Conference on Artificial Intelligence | Year: 2013
Multiobjective Dynamic Programming (MODP) is a general problem solving method used to determine the set of Pareto-optimal solutions in optimization problems involving discrete decision variables and multiple objectives. It applies to combinatorial problems in which Pareto-optimality of a solution extends to all its sub-solutions (Bellman principle). In this paper we focus on the determination of the preferred tradeoffs in the Pareto set where preference is measured by a Choquet integral. This model provides high descriptive possibilities but the associated preferences generally do not meet the Bellman principle, thus preventing any straightforward adaptation of MODP. To overcome this difficulty, we introduce here a general family of dominance rules enabling an early pruning of some Pareto-optimal sub-solutions that cannot lead to a Choquet optimum. Within this family, we identify the most efficient dominance rules and show how they can be incorporated into a MODP algorithm. Then we report numerical tests showing the actual efficiency of this approach to find Choquet-optimal tradeoffs in multiobjective knapsack problems.