Go to Paper
Return to GeoComputation 99 Index

A Distributed Approach For Planning Radio Communications

KIDNER, David (dbkidner@glam.ac.uk), FITZELL, Ian and AL NUAMI, Miqdad, University of Glamorgan, School of Computing; RALLINGS, Phil and WARE, Andrew, Division of Mathematics and Computing, Pontypridd, Rhondda Cynon Taff, Wales, CF37 1DL, U.K.

Key Words: digital terrain modelling, 3-D modelling, parallel and distributed computing, radio propagation

This paper focuses on integrating detailed 3-D topographic features with high-resolution digital terrain models, radio path propagation algorithms with respect to terrain and topographic features, and a parallel implementation of these algorithms across a distributed network of Pentium PCs to predict broadcast coverages over a 'cluttered' field-of-view

Radio communications planning has been a prominent application area of digital terrain modeling for the last 30 years. Path loss models describe the signal attenuation between the transmitter and receiver as a function of the propagation distance, and other parameters related to the terrain profile and its surface features. Many of the problems of radio propagation are akin to the problems of intervisibility analysis, i.e., greater accuracy can be obtained with higher resolution digital terrain models (DTMs), such as LiDAR-derived DTMs, and detailed 3-D surface features or "clutter." Clutter can include buildings and other relevant properties including roof structure, building materials and elevations; vegetation and shrubbery, including individual trees; and other 2-D surface features that have distinctive propagation properties, such as roads and water bodies; however, the complexity of these 3-D data models and propagation algorithms means increased computational workloads. This is particularly significant for determining broadcast coverages in the field-of-view. One solution is to parallelise the application. The need for specialised hardware is no longer an obstacle to producing parallel solutions, as recent advances in modern operating systems allied with PC networks are available in most organisations. The real problem is identifying the best strategy for assigning the workloads to machines, which minimises both processor redundancy and communication overheads. The paper discusses the issues for developing a distributed GIS for radio propagation modelling and presents a domain decomposition strategy that optimises performance.