A Group Mobility Model for MANETs Using Terrain Maps
Due to the lack of money some month ago - now sounds funny - I skipped a conference where I wanted to present an approach to model group mobility of nodes. The idea is to use a form of terrain map (normally used to generate landscape models) as a basis to emulate real world motion characteristics similar. The focus are groups and their motion characterstic within the group. The terrain map is interpreted as what I called penalty map: the object wish is to evade areas with higher penalties to move from start to end. The intra-group mobility is minted through a "virtual magnet". Nodes are placed in a position around a virtual center of the nodes. Nodes are appointed to stay there but the terrain map will also shift nodes to a more penalty free zone. For example: within a valley all nodes will come closer to each other because the hill have a higher penalty. In a flat environemt all nodes will stay at the pre-defined relative position because all penalty factors are equal.
The previous section is a summary of the algorithm. See