The Community for Technology Leaders
2009 IEEE International Conference on Wireless and Mobile Computing, Networking and Communications (2009)
Marrakech, Morocco
Oct. 12, 2009 to Oct. 14, 2009
ISBN: 978-0-7695-3841-9
pp: 331-337
ABSTRACT
Wireless channels exhibit complex signal strength fluctuation over time due to signal propagation effects interacting with a dynamic environment. For accurate simulation of wireless networking protocols, it is important to model wireless channels accurately; without accurate models, simulation results have been shown to be significantly different from measured results. Accurate statistical models of the state of a particular channel can be constructed based on obtained experimental measurements. However, this approach cannot be easily exploited for dynamic simulations (e.g., those with mobility), because it is impossible to pre-measure all of the possible channels that arise during the simulation; for example, in a mobile ad hoc network the nodes move and a very large number of channels is encountered in the lifetime of the simulation. In this paper, we propose the use of interpolation between measured channels to produce models for other channels that have not been measured, significantly increasing the flexibility of this approach. Instead of obtaining measurement traces for every link, our method estimates the essential parameters for a new link by interpolation from nearby measured links. Thus, the flexibility of measurement based models is increased dramatically, providing a modeling approach that is both accurate and flexible. We validate the approach using representative experiments.
INDEX TERMS
Ad-hoc network, channel model simulation, Markov model
CITATION

S. Han and N. B. Abu-Ghazaleh, "Estimated Measurement-Based Markov Models: Towards Flexible and Accurate Modeling of Wireless Channels," 2009 IEEE International Conference on Wireless and Mobile Computing, Networking and Communications(WIMOB), Marrakech, Morocco, 2009, pp. 331-337.
doi:10.1109/WiMob.2009.63
156 ms
(Ver 3.3 (11022016))