FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W222100618> ?p ?o ?g. }

• W222100618 abstract "Simulation is the most flexible means for assessments of quality of service in complex, tightly coupled distributed systems such as telecommunication systems. A major problem with simulation is that it is very inefficient when the quality measures depend on the occurrence of rare events (e.g. cell losses or system failure). Importance sampling (IS) is a speed-up simulation technique that has been successfully applied to increase the simulation efficiency. The basic idea is to provoke the rare events to occur more often and adjust for the bias in observations afterwards. However, the effect of IS is observed to be very sensitive to the choice of the simulation parameters. Much work are therefore invested to obtain optimal parameters, or at least good heuristics for the setting of these. In queuing systems, the main effort has been on single-dimensional problems such as single queue/single customer type, and some achievements are made using results from large deviation theory. Unfortunately, these results are not directly applicable to multi-dimensional problems, i.e. systems with several queues and/or customer types. The failure biasing techniques from IS in dependability simulations can be applied to multi-dimensional problems. However, failure biasing needs some heuristics for choosing good parameters (bias) to be as efficient as the strategies based on large deviation theory. This paper proposes an adaptive optimisation strategy where the large deviation results are combined with (failure distance) ideas from dependability simulations. This strategy is applicable to problems with a multi-dimensional state space where a large number of different resource limitations exists. This is a major extension compared to previous strategies which related to single resource limitations, and mainly one source type). The adaptive optimisation as been successfully applied to a large network example including 11 links and 10 different traffic sources. The estimated blocking probabilities for all source and all resource types show good agreement compared with exact values obtained by Iversens convolution method. Additionally, investigations of the sensitivity to the number of resources and the number of dimensions are carried out. 1. Norwegian University of Science and Technology, Department of Telematics, N-7034 Trondheim; Telephone: +47 73 59 2890, Fax: +47 73 59 6973, E-mail: Poul.Heegaard@item.ntnu.no Adaptive optimisation of importance sampling for multi-dimensional state space models with irregular resource boundaries" @default.
• W222100618 created "2016-06-24" @default.
• W222100618 creator A5032464896 @default.
• W222100618 date "2007-01-01" @default.
• W222100618 modified "2023-09-27" @default.
• W222100618 title "Adaptive optimisation of importance sampling for multi-dimensional state space models with irregular resource boundaries" @default.
• W222100618 cites W1599415737 @default.
• W222100618 cites W1972716520 @default.
• W222100618 cites W1976655797 @default.
• W222100618 cites W2017036530 @default.
• W222100618 cites W2070057864 @default.
• W222100618 cites W2093673791 @default.
• W222100618 cites W2146416787 @default.
• W222100618 cites W2151958841 @default.
• W222100618 cites W3009870883 @default.
• W222100618 cites W125717091 @default.
• W222100618 cites W6906867 @default.
• W222100618 hasPublicationYear "2007" @default.
• W222100618 type Work @default.
• W222100618 sameAs 222100618 @default.
• W222100618 citedByCount "1" @default.
• W222100618 crossrefType "journal-article" @default.
• W222100618 hasAuthorship W222100618A5032464896 @default.
• W222100618 hasConcept C105795698 @default.
• W222100618 hasConcept C106131492 @default.
• W222100618 hasConcept C111472728 @default.
• W222100618 hasConcept C115903868 @default.
• W222100618 hasConcept C126255220 @default.
• W222100618 hasConcept C127705205 @default.
• W222100618 hasConcept C138885662 @default.
• W222100618 hasConcept C140779682 @default.
• W222100618 hasConcept C160403385 @default.
• W222100618 hasConcept C199360897 @default.
• W222100618 hasConcept C22684755 @default.
• W222100618 hasConcept C2779530757 @default.
• W222100618 hasConcept C31258907 @default.
• W222100618 hasConcept C31972630 @default.
• W222100618 hasConcept C33923547 @default.
• W222100618 hasConcept C41008148 @default.
• W222100618 hasConcept C72434380 @default.
• W222100618 hasConcept C77019957 @default.
• W222100618 hasConceptScore W222100618C105795698 @default.
• W222100618 hasConceptScore W222100618C106131492 @default.
• W222100618 hasConceptScore W222100618C111472728 @default.
• W222100618 hasConceptScore W222100618C115903868 @default.
• W222100618 hasConceptScore W222100618C126255220 @default.
• W222100618 hasConceptScore W222100618C127705205 @default.
• W222100618 hasConceptScore W222100618C138885662 @default.
• W222100618 hasConceptScore W222100618C140779682 @default.
• W222100618 hasConceptScore W222100618C160403385 @default.
• W222100618 hasConceptScore W222100618C199360897 @default.
• W222100618 hasConceptScore W222100618C22684755 @default.
• W222100618 hasConceptScore W222100618C2779530757 @default.
• W222100618 hasConceptScore W222100618C31258907 @default.
• W222100618 hasConceptScore W222100618C31972630 @default.
• W222100618 hasConceptScore W222100618C33923547 @default.
• W222100618 hasConceptScore W222100618C41008148 @default.
• W222100618 hasConceptScore W222100618C72434380 @default.
• W222100618 hasConceptScore W222100618C77019957 @default.
• W222100618 hasLocation W2221006181 @default.
• W222100618 hasOpenAccess W222100618 @default.
• W222100618 hasPrimaryLocation W2221006181 @default.
• W222100618 hasRelatedWork W1992879234 @default.
• W222100618 hasRelatedWork W2055029865 @default.
• W222100618 hasRelatedWork W206052733 @default.
• W222100618 hasRelatedWork W2185897934 @default.
• W222100618 hasRelatedWork W2341744983 @default.
• W222100618 hasRelatedWork W2594876019 @default.
• W222100618 hasRelatedWork W2807186 @default.
• W222100618 hasRelatedWork W2903366716 @default.
• W222100618 hasRelatedWork W2936484317 @default.
• W222100618 hasRelatedWork W2946228343 @default.
• W222100618 hasRelatedWork W2965054383 @default.
• W222100618 hasRelatedWork W2972376672 @default.
• W222100618 hasRelatedWork W3086169459 @default.
• W222100618 hasRelatedWork W3089207924 @default.
• W222100618 hasRelatedWork W3090859680 @default.
• W222100618 hasRelatedWork W3110019368 @default.
• W222100618 hasRelatedWork W3127665452 @default.
• W222100618 hasRelatedWork W3178384524 @default.
• W222100618 hasRelatedWork W642711208 @default.
• W222100618 hasRelatedWork W992484401 @default.
• W222100618 isParatext "false" @default.
• W222100618 isRetracted "false" @default.
• W222100618 magId "222100618" @default.
• W222100618 workType "article" @default.