Search

The Autonomous System (AS) topology of the Internet (up to 61k ASs) is growing at a rate of about 10% per year. The Border Gateway Protocol (BGP) starts to show its limits in terms of the number of routing table entries it can dynamically process and control. Due to the increasing routing information processing and storage, the same trend is observed for routing model simulators such as DRMSim specialized in large-scale simulations of routing models. Therefore, DRMSim needs enhancements to support the current size of the Internet topology and its evolution (up to 100k ASs). To this end, this paper proposes a feasibility study of the extension of DRMSim so as to support the Distributed Parallel Discrete Event paradigm. We first detail the possible distribution models and their associated communication overhead. Then, we analyze this overhead by executing BGP on a partitioned topology according to different scenarios. Finally, we conclude on the feasibility of such a simulator by computing the expected additional time required by a distributed simulation of BGP compared to its sequential simulation., Comment: 26th ACM/IEEE/SCS Workshop on Principles of Advanced and Distributed Simulation (2012)

Most studies on path-vector routing stability have been conducted empirically by means of ad-hoc analysis of BGP data traces. None of them consider prior specification of an analytic method including the use of stability measurement metrics for the systematic analysis of BGP traces and associated meta-processing for determining the local state of the routing system. In this paper, we define a set of metrics that characterize the local stability properties of path-vector routing such as BGP (Border Gateway Protocol). By means of these stability metrics, we propose a method to analyze the effects of BGP policy- and protocol-induced instability on local routers., Comment: 14\`emes Rencontres Francophones sur les Aspects Algorithmiques des T\'el\'ecommunications (AlgoTel), La Grande Motte : France (2012)

Most studies on path-vector routing stability have been conducted empirically by means of ad-hoc analysis of BGP data traces. None of them consider prior specification of an analytic method including the use of stability measurement metrics for the systematic analysis of BGP traces and associated meta-processing for determining the local state of the routing system. In this paper, we define a set of metrics that characterize the local stability properties of path-vector routing such as BGP (Border Gateway Protocol). By means of these stability metrics, we propose a method to analyze the effects of BGP policy- and protocol-induced instability on local routers., 14\`emes Rencontres Francophones sur les Aspects Algorithmiques des T\'el\'ecommunications (AlgoTel), La Grande Motte : France (2012)

© 7th International Conference on Learning Representations, ICLR 2019. All Rights Reserved. In weakly-supervised temporal action localization, previous works have failed to locate dense and integral regions for each entire action due to the overestimation of the most salient regions. To alleviate this issue, we propose a marginalized average attentional network (MAAN) to suppress the dominant response of the most salient regions in a principled manner. The MAAN employs a novel marginalized average aggregation (MAA) module and learns a set of latent discriminative probabilities in an end-to-end fashion. MAA samples multiple subsets from the video snippet features according to a set of latent discriminative probabilities and takes the expectation over all the averaged subset features. Theoretically, we prove that the MAA module with learned latent discriminative probabilities successfully reduces the difference in responses between the most salient regions and the others. Therefore, MAAN is able to generate better class activation sequences and identify dense and integral action regions in the videos. Moreover, we propose a fast algorithm to reduce the complexity of constructing MAA from O(2T) to O(T2). Extensive experiments on two large-scale video datasets show that our MAAN achieves a superior performance on weakly-supervised temporal action localization.

The distributed monitoring problem refers to the placement and configuration of passive monitoring points to jointly realize a task of monitoring traffic flows. Given a monitoring task, the objective consists in minimizing the total monitoring cost to realize this task. We formulate this problem as a mixed-integer program. This formulation can also be dualized to determine the gain obtained when varying the number of monitoring points (i.e., the installation cost) and the fraction of monitored traffic (i.e., the configuration cost). As traffic flows can follow different paths depending on the routing strategy, we compare the resulting cost and gain when they are routed along the min-cost path, the paths obtained by solving the min-cost multicommodity flow and the multicommodity capacity network design problem.

info:eu-repo/semantics/published

International audience

International audience

International audience

The cooperative monitoring problem refers to the placement and configuration of passive monitoring points (or monitors) to jointly realize a task of monitoring time-varying traffic flows. Given a traffic monitoring task, e.g., monitor a traffic flow percentage, the objective consists in minimizing the total monitoring cost to realize this task. We formulate the corresponding problem as a mixed-integer linear program (MILP). This formulation can also be dualized to determine the gain obtained when increasing the number of monitoring points (i.e., the monitoring installation cost) and the fraction of monitored traffic (i.e., the monitoring configuration cost). As traffic flows are subject to uncertainty, we extend the MIP formulations developed in [1] to their robust counter-part when the uncertainty in traffic demands is modeled by means of box+polyhedral and box+ellipsoidal perturbation sets.

In communication networks, the routing decision process (distributed and online) remains decoupled from the network design process, i.e., resource installation and allocation planning process (centralized and offline). To reconcile both processes, we ambition to design a distributed optimization technique aware of distributed nature of the routing process by decomposing the optimization problem along same dimensions as (distributed) routing decision process. For this purpose, we generalize the capacitated multi-commodity capacitated fixed charge network design (MCND) class of problems by including different types of fixed costs (installation and maintenance costs) and variable costs (routing costs) but also variable traffic demands over multiple periods. However, conventional integer programming methods can typically solve only small to medium size instances. As an alternative, we propose a Lagrangian approach for computing a lower bound by relaxing the flow conservation constraints such that the Lagrangian subproblem itself decomposes by node. Though this approach yields one subproblem per network node, solving the Lagrangian dual by means of the bundle method remains a complex computational tasks. However, the approach is more robust than any LP solvers and it always returns some solutions. Instead, we proved that CPLEX, which uses the Dual Simplex algorithm, is not able to provide a solution for large instances.

Background DAVID syndrome is a rare condition combining anterior pituitary hormone deficiency with common variable immunodeficiency. NFKB2 mutations have recently been identified in patients with ACTH and variable immunodeficiency. A similar mutation was previously found in Nfkb2 in the immunodeficient Lym1 mouse strain, but the effect of the mutation on endocrine function was not evaluated. Methods We ascertained six unrelated DAVID syndrome families. We performed whole exome and traditional Sanger sequencing to search for causal genes. Lym1 mice were examined for endocrine developmental anomalies. Results Mutations in the NFKB2 gene were identified in three of our families through whole exome sequencing, and in a fourth by direct Sanger sequencing. De novo origin of the mutations could be demonstrated in three of the families. All mutations lie near the C-terminus of the protein-coding region, near signals required for processing of NFΚB2 protein by the alternative pathway. Two of the probands had anatomical pituitary anomalies, and one had growth and thyroid hormone as well as ACTH deficiency; these findings have not been previously reported. Two children of one of the probands carried the mutation and have to date exhibited only an immune phenotype. No mutations were found near the C-terminus of NFKB2 in the remaining two probands; whole exome sequencing has been performed for one of these. Lym1 mice, carrying a similar Nfkb2 C-terminal mutation, showed normal pituitary anatomy and expression of proopiomelanocortin (POMC). Conclusions We confirm previous findings that mutations near the C-terminus of NFKB2 cause combined endocrine and immunodeficiencies. De novo status of the mutations was confirmed in all cases for which both parents were available. The mutations are consistent with a dominant gain-of-function effect, generating an unprocessed NFKB2 super-repressor protein. We expand the potential phenotype of such NFKB2 mutations to include additional pituitary hormone deficiencies as well as anatomical pituitary anomalies. The lack of an observable endocrine phenotype in Lym1 mice suggests that the endocrine component of DAVID syndrome is either not due to a direct role of NFKB pathways on pituitary development, or else that human and mouse pituitary development differ in its requirements for NFKB pathway function. Electronic supplementary material The online version of this article (doi:10.1186/s12881-014-0139-9) contains supplementary material, which is available to authorized users.

The combined network design and (distributed)traffic routing problem can be formulated as a large-scale multi-period mixed integer optimization problem. This problem com-bines network design decisions and routing decisions, with time-dependent demands. A compact formulation for this problem,based on the aggregation of flows by destination, was proposedin [1]. They observed that the resolution on realistic instancesbecomes intractable and unscalable with state-of-the-art solversdue to the weak linear programming bound provided by theformulation. In this paper, we consider an extended formulationwhere flows are decomposed by origin-destination pairs, whilekeeping the requirement of destination-based routing. The qualityof the extended formulation and the computational time neededto solve it are evaluated on a representative set of networktopologies and traffic demands, and compared to results obtainedwith the base formulation of [1]. The extended formulation canthen be considered for more efficient resolution methods involvingdecomposition and cutting planes approaches., info:eu-repo/semantics/published

The capacitated Fixed-Charge Network Design (FCND) problem deals with the simultaneous optimization of capacity installation and routing of traffic, where a fixed cost is paid for opening a link and a linear routing cost is paid for sending traffic flow on a link. The routing decision must be performed such that flows remain bounded by the installed capacities. This problem appears as a particular case of the combined network design and traffic flow routing problem with time-dependent demands as developed in this paper. This general problem can be formulated as a multi-period mixed integer optimization problem. A compact formulation based on the aggregation of flows by destination and an extended formulation, where flows are decomposed by origin-destination pairs while keeping the requirement of destination-based routing, have been proposed in [1] and [2], respectively. As these computational improvements enable to progressively overcome the limit of solving such problem when the network size and number of periods increases, we propose an extended model that accounts for different patterns of failures affecting installed arcs over time. The latter lead to consider time-variable link maintenance cost function as part of our generalized formulation. We also analyze the quality of the results for the proposed formulation with different link maintenance cost functions with the objective to derive the best arc replacement strategy.

Language of publication: en; International audience; no abstract

International audience; The expansion of Internet topology, which comprises thousands of Autonomous Systems (AS), has resulted in a number of important research challenges. The Border Gateway Protocol (BGP), which is used to make core routing decisions on the Internet, starts to show its limitations in terms of the number of routing table entries it can store locally, update in a timely fashion and dynamically exchange. Because it is impractical to deploy newly designed routing protocols on the Internet a large-scale, simulation is an unavoidable step to validate their properties. However, the increasing routing information processing (CPU) and storage (memory) introduces similar challenges for the simulation of state-full routing. For this purpose, we introduce DRMSim a Dynamic Routing Model simulator of routing models on large-scale networks.

International audience; Compact routing schemes address the fundamental tradeoff between the memory space required to store the routing table entries and the length of the routing paths they produce. This paper introduces a dynamic compact routing scheme that allows the distribution of traffic from any source to any set of leaf nodes along a point-to-multipoint routing path referred to as multicast (traffic) distribution tree. By means of the proposed scheme, a multicast distribution tree can dynamically evolve according to the arrival of leaf-initiated join/leave requests. We evaluate by means of simulation the performance of the proposed scheme and analyze the tradeoffs between the stretch of the produced routing paths, the size and the number of routing table entries, and the communication cost.

International audience; The expansion of the Internet routing system results in a number of research challenges, in particular, the Border Gateway Protocol (BGP) starts to show its limits a.o. in terms of the number of routing table entries it can dynamically process and control. Dynamic routing protocols showing better scaling properties are thus under investigation. However, because deploying under-development routing protocols on the Internet is not practicable at a large-scale (due to the size of the Internet topology), simulation is an unavoidable step to validate the properties of a newly proposed routing scheme. Unfortunately, the simulation of inter-domain routing protocols over large networks (order of tens of thousands of nodes) poses real challenges due to the limited memory and computational power that computers impose. This paper presents the Dynamic Routing Model simulator DRMsim which addresses the specific problem of large-scale simulations of (inter-domain) routing models on large networks. The motivation for developing a new simulator lies in the limitation of existing simulation tools in terms of the number of nodes they can handle and in the models they propose.

Part 5: Mixed Reality and 3D Worlds; International audience; Globalization and increasing collaboration between remote teams drive the need for teleconference systems. However, currently no videoconferencing system matches the face-to-face experience for a business meeting with many participants in a flexible and affordable manner. In search for a better solution, we created a Virtual Meeting Room (VMR) application that visualizes key events detected using computer vision (e.g., participant entering the meeting room, talking, presenting) in a 3D virtual environment. The goal was to provide a good sense of overview to users when many meeting participants - represented by 3D avatars - from remote locations join a teleconference. In this paper, a technical overview of the working prototype - built using 3D game technology - is presented. Also, feedback from multiple user tests performed during the development of the prototype is discussed and presented as a set of recommendations. From the technical perspective, we found that existing 3D game technology is mature, affordable and contains the features needed to build the VMR application. From the users’ and experts’ feedback, we conclude that the VMR has merits as a teleconferencing support tool accompanying a video stream that conveys more detailed non-verbal communication of the active speaker.