Home // ICNS 2012, The Eighth International Conference on Networking and Services // View article

A Novel Probabilistic Deadline Scheduling Mechanism for DCCP

Authors:
Daniel Wilson
Mike Dixon
Terry Koziniec

Keywords: DCCP; CCID3; Congestion control

Abstract:
This paper introduces a novel cross layer probability based deadline scheduling mechanism designed specifically for real time Data Congestion Control Protocol (DCCP) flows. Scheduling in this mechanism is determined based on the probability of a data packet being received within its useful lifespan. In order to predict this probability, DCCP is modified to access routing table information used by CISCO Systems Inc.’s Enhanced Interior Gateway Routing Protocol (EIGRP), to estimate the approximate forward path delay period. Once the packet’s probability of arriving within its useful lifetime is determined, the scheduling algorithm then places the packet into one of three predefined queues to ensure all packets that are received are given the highest chance of being delivered within their useful lifespan period. In addition to describing the design of the mechanism, this paper will also present proof of concept modelling carried out to quantify the effectiveness of the mechanism. The results presented in this paper show the mechanism described is able to predict the time a packet will need to traverse the network using EIGRP’s metrics with greater than 90 percent accuracy (on average) in the tested topologies. The results will also show the mechanism is stable and able to operate in medium sized networks with marginal overhead.

Pages: 87 to 96

Copyright: Copyright (c) IARIA, 2012

Publication date: March 25, 2012

Published in: conference

ISSN: 2308-4006

ISBN: 978-1-61208-186-1

Location: St. Maarten, The Netherlands Antilles

Dates: from March 25, 2012 to March 30, 2012