TY - GEN

T1 - Self-adjusting grid networks to minimize expected path length

AU - Avin, Chen

AU - Borokhovich, Michael

AU - Haeupler, Bernhard

AU - Lotker, Zvi

PY - 2013/12/1

Y1 - 2013/12/1

N2 - Given a network infrastructure (e.g., data-center or on-chip-network) and a distribution on the source-destination requests, the expected path (route) length is an important measure for the performance, efficiency and power consumption of the network. In this work we initiate a study on self-adjusting networks : networks that use local-distributed mechanisms to adjust the position of the nodes (e.g., virtual machines) in the network to best fit the route requests distribution. Finding the optimal placement of nodes is defined as the minimum expected path length (MEPL) problem. This is a generalization of the minimum linear arrangement (MLA) problem where the network infrastructure is a line and the computation is done centrally. In contrast to previous work, we study the distributed version and give efficient and simple approximation algorithms for interesting and practically relevant special cases of the problem. In particular, we consider grid networks in which the distribution of requests is a symmetric product distribution. In this setting, we show that a simple greedy policy of position switching between neighboring nodes to locally minimize an objective function, achieves good approximation ratios. We are able to prove this result using the useful notions of expected rank of the distribution and the expected distance to the center of the graph.

AB - Given a network infrastructure (e.g., data-center or on-chip-network) and a distribution on the source-destination requests, the expected path (route) length is an important measure for the performance, efficiency and power consumption of the network. In this work we initiate a study on self-adjusting networks : networks that use local-distributed mechanisms to adjust the position of the nodes (e.g., virtual machines) in the network to best fit the route requests distribution. Finding the optimal placement of nodes is defined as the minimum expected path length (MEPL) problem. This is a generalization of the minimum linear arrangement (MLA) problem where the network infrastructure is a line and the computation is done centrally. In contrast to previous work, we study the distributed version and give efficient and simple approximation algorithms for interesting and practically relevant special cases of the problem. In particular, we consider grid networks in which the distribution of requests is a symmetric product distribution. In this setting, we show that a simple greedy policy of position switching between neighboring nodes to locally minimize an objective function, achieves good approximation ratios. We are able to prove this result using the useful notions of expected rank of the distribution and the expected distance to the center of the graph.

UR - http://www.scopus.com/inward/record.url?scp=84893025529&partnerID=8YFLogxK

U2 - https://doi.org/10.1007/978-3-319-03578-9_4

DO - https://doi.org/10.1007/978-3-319-03578-9_4

M3 - Conference contribution

SN - 9783319035772

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 36

EP - 54

BT - Structural Information and Communication Complexity - 20th International Colloquium, SIROCCO 2013, Revised Selected Papers

T2 - 20th International Colloquium on Structural Information and Communication Complexity, SIROCCO 2013

Y2 - 1 July 2013 through 3 July 2013

ER -