On the modeling of mobile ad hoc networks

This thesis examines the effects of node mobility and signal fading on the performance of mobile ad hoc networks. Chapter 3 presents novel expressions for the statistical properties of the received signal-to-noise ratio (SNR) and the separation distance in mobile ad hoc networks (MANETs). A full mathematical description of the squared distance process is provided, including its distribution and autocorrelation function. Next, the stochastic differential equation of the link SNR is derived. Closed-form expressions for the distribution and bivariate distribution of the link SNR are given in the absence and presence of signal fading. Finally, the connectivity probability is calculated for both scenarios. Chapter 4 develops an entropy-based metric that quantifies link uncertainty and evaluates its stability by taking full advantage of the correlation between the link's current and future state. This formulation is based on a stationary Markov chain representation of link connectivity whose source of randomness is the separation distance between the nodes or the variations in the propagation channel. The simulation results demonstrate that the proposed metric thoroughly captures the link dynamics and accurately reflects changes in the state of a wireless connection. Chapter 5 combines the theoretical modeling of the communication link developed in Chapter 3 with the topological uncertainty analysis carried out in Chapter 4 to conclude that the primary source of uncertainty on the link state is random mobility. It then proposes a new routing decision strategy for MANETs based on the entropy-based metric presented in Chapter 4. The key idea of this strategy is to identify and select the route with the longest lifetime between a source and a destination in an environment where different paths are available. Identifying the set of most stable paths, i.e., those that are expected to be active for a longer time, can potentially reduce the number of route discoveries and overhead throughout the network. The performance evaluation of the proposed approach is performed via simulation and analysis.