|Speaker :||Dalia Popescu|
|Location:||LINCS Seminars room|
Confinement of human settlement in areas limited in size is the foundation of the long-standing Central Place Theory which assumes the existence of regular spatial patterns in regional human organization. For example, a department has rural areas, with low density of population, and urban areas, with high density of population, namely cities and towns. Similarly, cities reflect a statistical self-similarity or hierarchy of clusters. Towns are split in neighborhoods; each neighborhood is organized in quarters then blocks separated by streets. Blocks are made of buildings that are themselves split in apartments and so on. Vehicles and devices are deployed where human activities occur and therefore inherit the self-similarity of the environment where they are deployed.
This talk will present our macro level model for the topology of vehicular communication entities in cities by exploiting self-similarity. The model is called hyperfractal and presents two variations: the hyperfractal model for density of vehicles and the hyperfractal model for road-side units (RSU). The model exploits the self-similarity of the density map of the vehicular communications entities and avoids the extremes of regularity and randomness.
Furthermore, we introduce a method for computing the hyperfractal dimension on cities. Using this fitting procedure applied to open data sets, we prove the validity of the model and show how it can be extended to cities that do not follow a regular hierarchical pattern.
Next, we study the information propagation speed of a broadcast in an urban delay tolerant network which is disconnected at all time, i.e., where end-to-end multihop paths may not exist (requiring a store-carry-and-forward routing model). We prove bounds on the average broadcast time in a hyperfractal setup and show that the performance is due in part to an interesting self-similar phenomenon, that we denote as information teleportation, that arises as a consequence of the topology and allows an acceleration of the broadcast time.