Models and methodologies for realistic propagation simulation for urban mesh networks

By providing connectivity to mobile users, mesh networks are poised to become a major extension of the Internet. More than 300 cities and towns have plans to deploy mesh networks, and several dozen cities have already deployed these networks. These mesh networks are meant to enhance city and emergency services communication as well as to provide city-wide, low-cost, ubiquitous Internet access for residents and visitors. Such networks promise to bring dramatic changes to data accessibility and hence have a major impact on society.; Simulation is the integral step in the validation of mesh networking protocols. However, currently, most simulations use the trivial disk propagation model (i.e., the signal propagates exactly R meters and no further) or a highly idealized propagation model such as the free-space or the two-ray model. Due to the presence of buildings, propagation in urban environments is far more complicated than the propagation presented by these simple models. Consequently, channel variability, which is a key aspect of wireless networking, is not well modeled in today's simulations. The result is that many insights gained from the simplistic environment do not necessarily hold in the urban environment, casting doubt on the applicability of the conclusions drawn from these simulations. While the reasons that realistic propagation models are typically not included into network simulations is not well documented, it seems that one of the main reasons include the belief that propagation modeling is computationally intractable and that propagation cannot be realistically modeled due to small-scale fading and delay spread.; This dissertation focuses on developing models for realistic propagation simulation and the implications of these models on simulation of urban mesh networks. The dissertation also provides measurements and models for time-varying nature of propagation and characterization of the effect of interference on packet success probability.; To this effect a realistic propagation simulator has been developed. This propagation simulator can estimate channel gain from any given point to any other given point in the simulated region (e.g., city downtown). In addition to estimating channel gain, other key metrics such as delay spread, angle of arrival and departure can be computed. It is shown that the simulation time scales linearly with the number of processors and also this computation needs to be carried out only once for a simulated region.; Channel variability that results in temporally varying topologies is one of the biggest challenges faced by wireless networks. This dissertation presents measurements and models of propagation between stationary transmitters and receivers in a dynamic environment. A diffusion-based stochastic model is presented that can be used to accurately model the time-varying propagation.; Propagation model only provides the signal strength and other propagation metrics. However, realistic simulation requires the packet success probability (PSP). The signal to Interference plus noise ratio impacts the probability with which the receiver will be able to successfully decode the signal.; This dissertation aims at characterizing the effect of varying interference and signal strengths on the packet success probability. As a result of this characterization a model for bit-rate selection is provided. This dissertation provides a framework for realistic simulation of urban mesh networks. The models developed are available online and are compatible with widely adopted discrete event packet simulators such as NS2 and QualNet.