Michael Schapira | The Hebrew University of Jerusalem (original) (raw)
Uploads
Papers by Michael Schapira
IEEE INFOCOM 2018 - IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS), Apr 1, 2018
Acm Transactions on Algorithms, Jun 1, 2014
Proceedings of the Thirty Eighth Annual Acm Symposium on Theory of Computing, May 21, 2006
Mathematics of Operations Research, Jan 27, 2010
Siam Journal on Computing, Nov 1, 2011
Proceedings of the 10th Usenix Conference on Networked Systems Design and Implementation, Apr 2, 2013
ABSTRACT We typically think of network architectures as having two basic components: a data plane... more ABSTRACT We typically think of network architectures as having two basic components: a data plane responsible for forwarding packets at line-speed, and a control plane that instantiates the forwarding state the data plane needs. With this separation of concerns, ensuring connectivity is the responsibility of the control plane. However, the control plane typically operates at timescales several orders of magnitude slower than the data plane, which means that failure recovery will always be slow compared to data plane forwarding rates. In this paper we propose moving the responsibility for connectivity to the data plane. Our design, called Data-Driven Connectivity (DDC) ensures routing connectivity via data plane mechanisms. We believe this new separation of concerns -- basic connectivity on the data plane, optimal paths on the control plane -- will allow networks to provide a much higher degree of availability, while still providing flexible routing control.
Acm Sigcomm Computer Communication Review, Oct 22, 2011
Distributed Computing, Feb 25, 2011
We use ideas from distributed computing to study dynamic environments in which computational node... more We use ideas from distributed computing to study dynamic environments in which computational nodes, or decision makers, follow adaptive heuristics (Hart 2005), i.e., simple and unsophisticated rules of behavior, e.g., repeatedly "best replying" to others' actions, and minimizing "regret", that have been extensively studied in game theory and economics. We explore when convergence of such simple dynamics to an equilibrium is guaranteed in asynchronous computational environments, where nodes can act at any time. Our research agenda, distributed computing with adaptive heuristics, lies on the borderline of computer science (including distributed computing and learning) and game theory (including game dynamics and adaptive heuristics). We exhibit a general non-termination result for a broad class of heuristics with bounded recall---that is, simple rules of behavior that depend only on recent history of interaction between nodes. We consider implications of our result across a wide variety of interesting and timely applications: game theory, circuit design, social networks, routing and congestion control. We also study the computational and communication complexity of asynchronous dynamics and present some basic observations regarding the effects of asynchrony on no-regret dynamics. We believe that our work opens a new avenue for research in both distributed computing and game theory.
Lecture Notes in Computer Science, 2015
ABSTRACT Deterministic constructions of expander graphs have been an important topic of research ... more ABSTRACT Deterministic constructions of expander graphs have been an important topic of research in computer science and mathematics, with many well-studied constructions of infinite families of expanders. In some applications, though, an infinite family is not enough: we need expanders which are "close" to each other. We study the following question: Construct an an infinite sequence of expanders G0,G1,dotsG_0,G_1,\dotsG0,G1,dots, such that for every two consecutive graphs GiG_iGi and Gi+1G_{i+1}Gi+1, Gi+1G_{i+1}Gi+1 can be obtained from GiG_iGi by adding a single vertex and inserting/removing a small number of edges, which we call the expansion cost of transitioning from GiG_iGi to Gi+1G_{i+1}Gi+1. This question is very natural, e.g., in the context of datacenter networks, where the vertices represent racks of servers, and the expansion cost captures the amount of rewiring needed when adding another rack to the network. We present an explicit construction of ddd-regular expanders with expansion cost at most 5d/25d/25d/2, for any dgeq6d\geq 6dgeq6. Our construction leverages the notion of a "2-lift" of a graph. This operation was first analyzed by Bilu and Linial, who repeatedly applied 2-lifts to construct an infinite family of expanders which double in size from one expander to the next. Our construction can be viewed as a way to "interpolate" between Bilu-Linial expanders with low expansion cost while preserving good edge expansion throughout. While our main motivation is centralized (datacenter networks), we also get the best-known distributed expander construction in the "self-healing" model.
IEEE INFOCOM 2018 - IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS), Apr 1, 2018
Acm Transactions on Algorithms, Jun 1, 2014
Proceedings of the Thirty Eighth Annual Acm Symposium on Theory of Computing, May 21, 2006
Mathematics of Operations Research, Jan 27, 2010
Siam Journal on Computing, Nov 1, 2011
Proceedings of the 10th Usenix Conference on Networked Systems Design and Implementation, Apr 2, 2013
ABSTRACT We typically think of network architectures as having two basic components: a data plane... more ABSTRACT We typically think of network architectures as having two basic components: a data plane responsible for forwarding packets at line-speed, and a control plane that instantiates the forwarding state the data plane needs. With this separation of concerns, ensuring connectivity is the responsibility of the control plane. However, the control plane typically operates at timescales several orders of magnitude slower than the data plane, which means that failure recovery will always be slow compared to data plane forwarding rates. In this paper we propose moving the responsibility for connectivity to the data plane. Our design, called Data-Driven Connectivity (DDC) ensures routing connectivity via data plane mechanisms. We believe this new separation of concerns -- basic connectivity on the data plane, optimal paths on the control plane -- will allow networks to provide a much higher degree of availability, while still providing flexible routing control.
Acm Sigcomm Computer Communication Review, Oct 22, 2011
Distributed Computing, Feb 25, 2011
We use ideas from distributed computing to study dynamic environments in which computational node... more We use ideas from distributed computing to study dynamic environments in which computational nodes, or decision makers, follow adaptive heuristics (Hart 2005), i.e., simple and unsophisticated rules of behavior, e.g., repeatedly "best replying" to others' actions, and minimizing "regret", that have been extensively studied in game theory and economics. We explore when convergence of such simple dynamics to an equilibrium is guaranteed in asynchronous computational environments, where nodes can act at any time. Our research agenda, distributed computing with adaptive heuristics, lies on the borderline of computer science (including distributed computing and learning) and game theory (including game dynamics and adaptive heuristics). We exhibit a general non-termination result for a broad class of heuristics with bounded recall---that is, simple rules of behavior that depend only on recent history of interaction between nodes. We consider implications of our result across a wide variety of interesting and timely applications: game theory, circuit design, social networks, routing and congestion control. We also study the computational and communication complexity of asynchronous dynamics and present some basic observations regarding the effects of asynchrony on no-regret dynamics. We believe that our work opens a new avenue for research in both distributed computing and game theory.
Lecture Notes in Computer Science, 2015
ABSTRACT Deterministic constructions of expander graphs have been an important topic of research ... more ABSTRACT Deterministic constructions of expander graphs have been an important topic of research in computer science and mathematics, with many well-studied constructions of infinite families of expanders. In some applications, though, an infinite family is not enough: we need expanders which are "close" to each other. We study the following question: Construct an an infinite sequence of expanders G0,G1,dotsG_0,G_1,\dotsG0,G1,dots, such that for every two consecutive graphs GiG_iGi and Gi+1G_{i+1}Gi+1, Gi+1G_{i+1}Gi+1 can be obtained from GiG_iGi by adding a single vertex and inserting/removing a small number of edges, which we call the expansion cost of transitioning from GiG_iGi to Gi+1G_{i+1}Gi+1. This question is very natural, e.g., in the context of datacenter networks, where the vertices represent racks of servers, and the expansion cost captures the amount of rewiring needed when adding another rack to the network. We present an explicit construction of ddd-regular expanders with expansion cost at most 5d/25d/25d/2, for any dgeq6d\geq 6dgeq6. Our construction leverages the notion of a "2-lift" of a graph. This operation was first analyzed by Bilu and Linial, who repeatedly applied 2-lifts to construct an infinite family of expanders which double in size from one expander to the next. Our construction can be viewed as a way to "interpolate" between Bilu-Linial expanders with low expansion cost while preserving good edge expansion throughout. While our main motivation is centralized (datacenter networks), we also get the best-known distributed expander construction in the "self-healing" model.