Wirless Mast Place Application for the Scottish Highlands
View/ Open
NelsonResearchPaper2008.doc (1.399Mb)
NelsonTechnicalPaper2008.doc (5.221Mb)
Date
2008Item status
Restricted AccessAuthor
Nelson, Thomas
Metadata
Abstract
This application attempts to provide wireless service to the maximum number of people with a minimal number of masts. This aim creates a challenge for most algorithms as they are not designed to take into account both such objectives. An Evolutionary Algorithm (EA) is a technique inspired by biological evolution used to approximate solutions for this problem. This research uses two such algorithms to solve the radio network design problem (RND) and adds additional constraints to make it more practically applicable to the physical world. The challenge one faces when applying the RND problem to remote locations is that it fails to incorporate mast-to-mast interaction. In order to provide broadband signals to under-serviced areas, masts must be linked together, thereby permitting signals to be propagated from a single source. In an attempt to overcome this issue and enable mast interaction, terrain information was incorporated into an EA thus allowing a line-of-sight verification. The end result of such an analysis was to provide the user with possible mast sites based on their input parameters for geographical features, such as roads, water, and land use. However, these modified algorithms often resulted in conflicting requirements. It is therefore concluded that this approach to solving the mast-to-mast interaction was not consistently able to produce a result and requires the loosening of constraints and further refinements.