Directional wireless communications networks (DWNs) are expected to

become a workhorse of the military, as they provide great network capacity in hostile

areas where omnidirectional RF systems can put their users in harm's way. These

networks will also be able to adapt to new missions, change topologies, use different

communications technologies, yet still reliably serve all their terminal users. DWNs

also have the potential to greatly expand the capacity of civilian and commercial

wireless communication. The inherently narrow beams present in these types of

systems require a means of steering them, acquiring the links, and tracking to

maintain connectivity. This area of technological challenges encompasses all the

issues of pointing, acquisition, and tracking (PAT).

iii

The two main technologies for DWNs are Free-Space Optical (FSO) and

millimeter wave RF (mmW). FSO offers tremendous bandwidths, long ranges, and

uses existing fiber-based technologies. However, it suffers from severe turbulence

effects when passing through long (>kms) atmospheric paths, and can be severely

affected by obscuration. MmW systems do not suffer from atmospheric effects

nearly as much, use much more sensitive coherent receivers, and have wider beam

divergences allowing for easier pointing. They do, however, suffer from a lack of

available small-sized power amplifiers, complicated RF infrastructure that must be

steered with a platform, and the requirement that all acquisition and tracking be done

with the data beam, as opposed to FSO which uses a beacon laser for acquisition and

a fast steering mirror for tracking.

This thesis analyzes the many considerations required for designing and

implementing a FSO PAT system, and extends this work to the rapidly expanding

area of mmW DWN systems. Different types of beam acquisition methods are

simulated and tested, and the tradeoffs between various design specifications are

analyzed and simulated to give insight into how to best implement a transceiver

platform.

An experimental test-bed of six FSO platforms is also designed and constructed

to test some of these concepts, along with the implementation of a three-node biconnected

network. Finally, experiments have been conducted to assess the

performance of fixed infrastructure routing hardware when operating with a

physically reconfigurable RF network.