Due to the volume of data to be managed, and the computational complexity of many
spatial operations, GIS systems tend to face a number of performance issues. One
common approach to cope with the large volumes of data and high CPU demands is
to partition the system and deploy it over either a parallel or distributed system.
This thesis seeks to address a number of performance issues faced by parallel and
distributed GIS system.
Contributions of this thesis include:
Creation of new class of approximations, which allow for improved
performance of filtration phase, of spatial join algorithms.
Two new approaches to perform polygon amalgamation with significantly
reduced resource demands.
Creation of load balancing algorithms to reduce the impact of data skew
during spatial query processing in a parallel environment.
Creation of data allocation strategies that minimise the introduction of data
skew during spatial query processing, and reduce the need for load balancing.