2009
DOI: 10.1016/j.peva.2009.01.002
|View full text |Cite
|
Sign up to set email alerts
|

A unified model for synchronous and asynchronous FDL buffers allowing closed-form solution

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
31
0

Year Published

2009
2009
2016
2016

Publication Types

Select...
5
2

Relationship

3
4

Authors

Journals

citations
Cited by 24 publications
(31 citation statements)
references
References 11 publications
0
31
0
Order By: Relevance
“…A possible explanation is that when the load is 'low' (60 %), the probability of finding a horizon index n = 8 is smaller than for higher load (80 %), thus reducing the number of instances fillable voids can be created and the LP is reduced. The load for which this inversion is observed may be linked to the critical load of 69.3 % for which the infinite system without void creation becomes unstable (ρ = ln(2), see [4]). …”
Section: Performance Resultsmentioning
confidence: 99%
“…A possible explanation is that when the load is 'low' (60 %), the probability of finding a horizon index n = 8 is smaller than for higher load (80 %), thus reducing the number of instances fillable voids can be created and the LP is reduced. The load for which this inversion is observed may be linked to the critical load of 69.3 % for which the infinite system without void creation becomes unstable (ρ = ln(2), see [4]). …”
Section: Performance Resultsmentioning
confidence: 99%
“…The length of the arriving packets is assumed a fixed length B equal to the granularity, B = D. While matching the granularity and the packet size is a natural choice in view of performance (see [9]), it also enables to devise an intuitive void-creating scheduling algorithm, as argued below. Algorithms with a similar scheduling strategy for other values of D of course can be thought of, but are considered out of the scope of this paper.…”
Section: A Assumptionsmentioning
confidence: 99%
“…, N . Since this setting is known to be optimal in many (but not all) cases (for discussion, see [9]), it is also assumed in this work.…”
Section: Quantization Modelmentioning
confidence: 99%
“…This assumption allows to derive a simple characterization of the equivalent load that is independent of the actual number of delay lines. As discussed in [9], the most efficient way to tackle the analysis is by focusing on the evolution of assigned waiting times. In the infinite FDL buffer system, we associate waiting time W k with the kth packet, and define it as the time between the acceptance of packet k, and the start of its transmission.…”
Section: Quantizationmentioning
confidence: 99%
See 1 more Smart Citation