Optimum performance model of ARQ protocol under adaptive modulation scheme.

Li, Yue and Nguyen, Huan X. and Xing, Weixi (2008) Optimum performance model of ARQ protocol under adaptive modulation scheme. In: Communications, 2008. ICC '08. IEEE International Conference [proceedings]. IEEE, pp. 225-229. ISBN 9781424420759

Full text is not in this repository.


In this paper, a novel analytic framework of cross-layer design between physical and network layers is devised to enable an efficient communication link under various channel qualities. Based on the principle of maximum entropy (ME), the performance modelling and evaluation of a wireless channel with bursty multiple traffics and automatic repeat request (ARQ) traffic flows subject to intelligent buffer management scheme (IBMS) is investigated. As the packet error rate varies depending on channel quality, better performance of IBMS can be achieved under adaptive modulation (AM). In this context, an open queue- ing network model (QNM) is proposed to a GE/GE/1/N/HoL/CBS (T) queueing system with finite capacity, threshold T, head of line (HoL) and complete buffer sharing (CBS) to model the wireless channel with IBMS corresponding to ARQ protocol under various wireless channel qualities. Subject to appropriate GE-type queueing and delay theoretic mean value constraints, ME analytic solutions are characterized and new closed form expressions for the state and blocking probability distributions are determined. Typical numerical experiments are included to verify the credibility of the ME solutions at 95% confidence intervals and to compare wireless network performance with ARQ traffic handling schemes.

Item Type: Book Section
Research Areas: A. > School of Science and Technology > Computer and Communications Engineering
A. > School of Science and Technology > Computer Science > SensoLab group
Item ID: 7687
Useful Links:
Depositing User: Dr Huan Xuan Nguyen
Date Deposited: 19 Apr 2011 12:57
Last Modified: 13 Oct 2016 14:22
URI: http://eprints.mdx.ac.uk/id/eprint/7687

Actions (login required)

Edit Item Edit Item