Estimating small cell-loss ratios in ATM switches via importance sampling

L’Ecuyer, Pierre; Champoux, Yanick

doi:10.1145/379525.379528

Cited by 14 publications

(4 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Another priority queue called twol-amalgamated priority queue [49] uses three efficient Henriksen's queues, skew heap, and splay tree, to achieve efficient operations. The large number of events can also be minimized using techniques that manipulate the statistical properties of the model to reduce the size of events [50].…”

Section: Related Workmentioning

confidence: 99%

Performance Measurement of Dynamic Structure DEVS for Large-Scale Cellular Space Models

Sun

2009

SIMULATION

View full text Add to dashboard Cite

Dynamic structure Discrete Event System Specification (DSDEVS) is an advanced modeling formalism that allows DEVS models and their couplings to be dynamically changed. The modeling power and advantages of DSDEVS have been well studied, however, the performance aspect is generally overlooked. This paper provides a comprehensive performance measurement of DSDEVS for a large-scale cellular space model. We consider the modeling and simulation layers for performance analysis, and carry out performance measurement based on a token ring model and a fire spread model. The results show that DS modeling can improve simulation performance for large-scale cellular space models, due to the fact that it makes the simulation focus only on those active models, and thus is more efficient than when the entire cellular space is loaded. On the other hand, the DS overhead cannot be ignored and can become significant and even dominant when a large number of cells are dynamically added/deleted.

show abstract

Section: Related Workmentioning

confidence: 99%

Performance Measurement of Dynamic Structure DEVS for Large-Scale Cellular Space Models

Sun

2009

SIMULATION

View full text Add to dashboard Cite

show abstract

“…single server queues, under restrictive assumptions regarding the underlying arrival and service processes (e.g., having light-tail distributions [26], [29]). Some queueing networks of certian topologies have also been considered (see, e.g., [8], [23], [22]). The overflow event of interest is usually that of an individual buffer or that of the total network population.…”

Section: Introductionmentioning

confidence: 99%

Efficient heuristics for the simulation of population overflow in series and parallel queues

Nicola¹,

Zaburnenko²

2006

Proceedings of the 1st International Conference on Performance Evaluation Methodolgies and Tools - Valuetools '06

View full text Add to dashboard Cite

In this paper we propose state-dependent importance sampling heuristics to estimate the probability of population overflow in Markovian networks of series and parallel queues. These heuristics capture state-dependence along the boundaries (when one or more queues are empty) which is critical for the asymptotic optimality of the change of measure. The approach does not require difficult (and often intractable) mathematical analysis or costly optimization involved in adaptive importance sampling methodologies. Experimental results on tandem and parallel networks with a moderate number of nodes yield asymptotically efficient estimates (often with bounded relative error) where no other state-independent importance sampling techniques are known to be efficient. Insight drawn from simulating basic networks in this paper promises the applicability of the proposed methodology to larger networks with more general topologies.

show abstract

“…An example is the use of importance sampling if the metric of concern is the small cell-loss ratio (e.g., less than 10 −8 ) in an ATM switch [L'Ecuyer and Champoux 2001]. However, such techniques are specific only to certain models of concern and cannot be generalized for most simulations.…”

Section: Introductionmentioning

confidence: 99%

Twol-amalgamated priority queues

Goh

Thng

2004

ACM J. Exp. Algorithmics

View full text Add to dashboard Cite

Priority queues are essential function blocks in numerous applications such as discrete event simulations. This paper describes and exemplifies the ease of obtaining high performance priority queues using a two-tier list-based structure. This new implementation, called the Twol structure, is amalgamated with three priority queues, namely, the Henriksen's queue, splay tree and skew heap, to enhance the efficiency of these basal priority queue structures. Using a model that combines traditional average case and amortized complexity analysis, Twol-amalgamated priority queues that maintain N active events are theoretically proven to offer O(1) expected amortized complexity under reasonable assumptions. They are also demonstrated empirically to offer stable near O(1) performance for widely varying priority increment distributions and for queue sizes ranging from 10 to 10 million. Extensive empirical results show that the Twol-amalgamated priority queues consistently outperform those basal structures (i.e., without the Twol structure) with an average speedup of about three to five times on widely different hardware architectures. These results provide testimony that the Twol-amalgamated priority queues are suitable for implementation in sizeable application scenarios such as, but not limited to, large-scale discrete event simulation.

show abstract

Estimating small cell-loss ratios in ATM switches via importance sampling

Cited by 14 publications

References 32 publications

Performance Measurement of Dynamic Structure DEVS for Large-Scale Cellular Space Models

Performance Measurement of Dynamic Structure DEVS for Large-Scale Cellular Space Models

Efficient heuristics for the simulation of population overflow in series and parallel queues

Twol-amalgamated priority queues

Contact Info

Product

Resources

About