| With the recent advances in the development of affordable and portable wireless communication and computation devices, wireless ad hoc networks have been receiving considerable research interest in recent years. A wireless ad hoc network is a collection of wireless terminals (usually referred to as nodes) that can communicate without the need for a fixed backbone or infrastructure. A communication session between two nodes is achieved through a direct link, if the two nodes are within the transmission range of each other. Otherwise, intermediate nodes are used to establish a multihop connection between the source and destination. These characteristics make a wireless ad hoc network an ideal choice wherever a communication infrastructure is not present or not viable, e.g. in battlefield and emergency disaster relief. While the architecture of ad hoc networks certainly has many advantages, it also introduces several new challenges which stem from the inherent characteristics of such a design. Managing the scarce resources, mainly link bandwidths and node powers, while satisfying the user service requirements, is crucial for efficient operation of wireless ad hoc networks.; Our research is by no means the first attempt to approach such a problem. However, earlier work has been either mostly based on simulations and no mathematical modeling has been used, has not taken into account the cross layer interaction in the analysis, has been based on simplified regular pattern for the node distribution, or has considered a simple non-realistic radio propagation model. Moreover, earlier research efforts have been tackling the case of single service networks only.; In this work, we formulate and present a framework for optimum allocation of such limited resources in a multiservice interference limited wireless ad hoc networks with quality of service (QoS) constraints. We start by developing an analytical model showing the interaction among the power levels, supported transmission rates, and number of nodes in the network and the effects of these variables on the network performance. First, we consider a single service ad hoc network where all the nodes choose the same power level for all their transmissions. Next, we extend the analysis to multiservice networks, where nodes use service dependent power levels for their transmissions. We extend the analysis further by focusing on how the scarce resources of power and bandwidth are managed. To this objective, and building on the mathematical formulation of the QoS attributes, we consider an optimization problem with the objective of maximizing the aggregate network throughput subject to QoS constraints set by the supported services. Then, we turn our objective to minimizing the total transmission power subject to similar constraints. In both cases, we end up with a nonlinear constrained optimization problem. We tackle the problems and obtain solutions using well developed numerical techniques. |
