ACM Home Page
Please provide us with feedback. Feedback
Performance bounds in communication networks with variable-rate links
Full text PdfPdf (1,000 KB)
Source Applications, Technologies, Architectures, and Protocols for Computer Communication archive
Proceedings of the conference on Applications, technologies, architectures, and protocols for computer communication table of contents
Cambridge, Massachusetts, United States
Pages: 126 - 136  
Year of Publication: 1995
ISBN:0-89791-711-1
Also published in ...
Author
Kam Lee  Carnegie Mellon University, 5000 Forbes Ave., Pittsburgh, PA
Sponsor
SIGCOMM: ACM Special Interest Group on Data Communication
Publisher
ACM  New York, NY, USA
Bibliometrics
Downloads (6 Weeks): 2,   Downloads (12 Months): 26,   Citation Count: 7
Additional Information:

abstract   references   cited by   index terms  

Tools and Actions: Request Permissions Request Permissions    Review this Article  
DOI Bookmark: Use this link to bookmark this Article: http://doi.acm.org/10.1145/217382.217422
What is a DOI?

ABSTRACT

In most network models for quality of service support, the communication links interconnecting the switches and gateways are assumed to have fixed bandwidth and zero error rate. This assumption of steadiness, especially in a heterogeneous internet-working environment, might be invalid owing to subnetwork multiple-access mechanism, link-level flow/error control, and user mobility. Techniques are presented in this paper to characterize and analyze work-conserving communication nodes with varying output rate. In the deterministic approach, the notion of "fluctuation constraint," analogous to the "burstiness constraint" for traffic characterization, is introduced to characterize the node. In the statistical approach, the variable-rate output is modelled as an "exponentially bounded fluctuation" process in a way similar to the "exponentially bounded burstiness" method for traffic modelling. Based on these concepts, deterministic and statistical bounds on queue size and packet delay in isolated variable-rate communication server-nodes are derived, including cases of single-input and multiple-input under first-come-first-serve queueing. Queue size bounds are shown to be useful for buffer requirement and packet loss probability estimation at individual nodes. Our formulations also facilitate the computation of end-to-end performance bounds across a feedforward network of variable-rate server-nodes. Several numerical examples of interest are given in the discussion.


REFERENCES

Note: OCR errors may be found in this Reference List extracted from the full text article. ACM has opted to expose the complete List rather than only correct and linked references.

 
1
A. Banerjea and S. Keshav, "Queueing Delays in Rated Controlled ATM networks," Proc. INFOCOM '93, pp. 547- 556, Mar. 1993.
 
2
 
3
C. S. Chang, "Stability, Queue Length and Delay, Part I: Deterministic Queueing Networks." IEEE Trans. Auto. Control, vol. 39, no. 5, pp. 913-931, May 1994.
 
4
R.L. Cruz, "A Calculus for Network Delay, Part I: Network Elements in isolation," IEEE Trans. Inform. Theory, vol. 37, no. 1, pp. 114-131, Jan. 1991.
 
5
R. L. Cruz, "A Calculus for Network Delay, Part II: Network Analysis," IEEE Trans. Inform. Theory, vol. 37, no. 1, pp. 132-141, Jan. 1991.
 
6
R. L. Cruz, "Service Burstiness and Dynamic Burstiness Measures: A Framework," J. of High Speed Networks, vol. 1, no. 2, pp. 105-1127, 1992.
 
7
R.L. Cruz and H. N. Liu, "End-to-End Queueing Delay in ATM Networks," J. of High Speed Networks, vol. 3, no. 4, pp. 413-428, 1994.
 
8
R.L. Cruz, "Quality of Service Guarantees in Virtual Circmt Switched Networks," to appear in IEEE JSAC, 1995.
 
9
F.P. Kelly, Reversibility and Stochastic Networks, New York: Wiley, 1979.
 
10
 
11
L. Kleinrock, Queueing Systems, Vol. 2: Computer Applications, New York: Wiley, 1976.
12
 
13
 
14
 
15
 
16
O. Yaron and M. Sidi, "Generalized Processor Sharing Networks with Exponentially Bounded Burstiness Arrivals," Proc. IEEE INFORM '94, pp. 5b.4.1-5b.4.7, June 1994.
17
18