SYED Mohammad Imran

The research community suffers from the lack of datasets reflecting the mobility of users in space and time. Existing mobility traces available to the research community are limited. Interesting datasets have been collected by private companies (such as telecom operators), but these traces are very rarely made available to researchers. In other contexts such as for the mobility of vehicles (taxis, buses), GPS collection can be found and exploited to retrieve the mobility pattern, but collected information are limited to a subset of vehicles. Although the amount of collected information cannot provide with pertinent cross layer information (Strength of a signal, inter-contact times, etc.).
In this thesis, the candidate will focus on developing measurement strategies to collect mobility traces in some target areas without having to deploy any software at the user devices. There are two main ways of obtaining mobility data from real users. The first one relies on active measurement techniques, where measurement components are deployed at end- devices. The second approach, the one we will focus on during the thesis, is to rely on passive measurements. The main idea is to rely on an infrastructure composed of sniffers that analyzes the wireless traffic generated by the nodes to estimate their spatial and temporal displacements in the target region. The objective of such a non-intrusive strategy is to identify changing trends of movements and spatial occupancy without the need to identify the nodes explicitly.