Capacity analysis of wireless multi-hop ad hoc access networks with directional antenna

Abstract

The future of wireless networks points towards an integration of multi-hop ad hoc connections. The flexibility of the scheme enables a wide range of applications. The implementations, however, are still very limited, as many technical issues remain unsolved. There have been many research works addressing those challenging issues of the wireless multi-hop ad hoc networks, including mobility, connectivity, routing, scheduling, and media access control. Nevertheless, the inherit problems of wireless networks is radio frequency resource scarcity. One of the tools that could improve the resource usage is directional antenna. It can lead to better resource-usage and capacity. The overall effects, however, have not been entirely discovered. The aim of this dissertation is to disclose the physical constraints that govern the capacity of the multi-hop ad hoc communications equipped with directional antenna. The results shed the light on the capacity maximization of multi-hop ad hoc access networks. The proposed analytical framework integrates the ability to reduce spatial interference of directional antenna. The novelty of the proposed formula is in the usage of the vector representations. Enabled by the cone-plus-ball antenna model, the cumulative interference is found to be conveniently expressed by the concept of equivalent interferers. For verification purpose, the derived formula is also numerically compared to Monte Carlo simulations of realistic antenna patterns, which shows good agreements. Based on the polar coordinate system, the optimal conditions for each dimension are herein derived. The optimal condition for the angle dimension is described as the minimum separable condition around a gateway. The optimal condition for the distance dimension is found as the optimal relay selection condition. The results provide valuable insights to the protocol design.