The Model
Image Gallery
Contact Us

The Model

The completed model involved the integration of several sub-processes. The model and the sub-processes were written in Avenue, the scripting language of ArcView. The sub-processes were controlled by a master control script, also written in Avenue. The stages of the model are described below.

Stage 1: Tower Locations

Tower locations were determined by using the locations of the individual antennae of the various cellular providers in the GVRD. Total tower capacities were determined by the integration of the individual capacities of the antennae at each cell site.

Stage 2: Incorporating Disaster Scenarios

Nhematis was used to evaulate the potential impacts of earthquakes in the GVRD. Output MMI maps from Nhematis were combined with the towers and threshold information to determine which towers would be 'destroyed'. The remaining towers were then used for continuation of the model

Click here for a larger image

Stage 3: Diurnal Migration Patterns

Traffic flow data was combined with population data to model the movement of people to and from urban centers during the day and night, respectively (view the script). The following diagrams are based on 1.2 people travelling per car with the 7:30 to 8:30 rush representing one-third of the total morning population movement. Numerical differences are significant but the differences shown on the maps are subtle.

Click here for a larger image Click here for a larger image

Stage 4: Cell Phone Usage

This step involved the determination of the actual percent of people that own cell phones and the percent of those that will actually use their phones immediately after an earthquake occurs. These parameters were input as variables, allowing several different scenarios to be run.

Stage 5: Modelling Tower Access Areas

Tower access areas are defined as the area around each cell tower within which calls will be directed to that tower. The tower access areas were modelled by using a Thiessen polygon approach (view the script). The Thiessen polygons were modified according to tower ranges to allow for the mixing of towers of many different ranges.

Click here for a larger image

Stage 6: Determining Cellular Service Supply and Demand

Tower access areas were integrated with the cell phone user population distribution. Individual users in each area were assigned to towers, resulting in increased use of the network and a decreased amount of remaining cell phone users in each area.

Stage 7: Determining Saturated Towers and Remaining Demand

Subtracting the demand for each tower from the amount of users attempting to access the tower, provides the number of users remaining to get on the network and the number of towers that are 'saturated' with calls. Saturated towers were removed and new population amounts were calculated.

Stage 8: Reiteration Procedure without Saturated Towers

A series of iterations through stages 4-7 were run until all no further changes were evident in either the remaining tower capacities or the number of users attempting to access a cell tower.

After Stage 8 is complete, the model produces a map of locations where cell phone users cannot access the network. These users also include emergency response personnel, who may be relying on cellular communication for coordinating emergency response efforts.

back to top         

Simon Fraser University Homepage
This site produced by the CELL-GIS Project Group   --   March 22, 2000
Web site layout and graphic design by Allan Glen