Realistic Modeling of Wireless Network Environments
Final technical rept. Aug 2010-Sep 2014
CARNEGIE-MELLON UNIV PITTSBURGH PA
Pagination or Media Count:
Wireless networking has become a part of our everyday life, but it is a challenging networking technology. The reason is that wireless channel properties, and thus link quality, depend strongly on the physical environment. This creates a challenge for both simulation and emulation platforms, since they must accurately model the impact of many physical layer factors on signal propagation in order to achieve a high degree of realism. This project developed a number of techniques to improve the accuracy of developing realistic and efficient environment-specific channel models. We looked at how the accuracy of both geometric and stochastic channel models can be improved by generating model inputs that accurately reflect a specific environment. Our evaluation uses two important properties, channel dynamics and correlation between nearby channels, in a challenging wireless environment, namely vehicular networks. We also made a number of improvements to an emulation-based wireless testbed to improve channel model accuracy and transferred a copy of the testbed to LTS as the basis for research collaboration.
- Computer Systems